Homology modeling of Homo sapiens lipoic acid synthase: Substrate docking and insights on its binding mode

Author’s Accepted Manuscript Homology modeling of Homo sapiens Lipoic acid Synthase: substrate docking and insights on its binding mode Ezhilarasi Krishnamoorthy, Sameer Hassan, Luke Elizabeth Hanna, Indira Padmalayam, Rama Rajaram, Vijay Viswanathan www.elsevier.com/locate/yjtbi

PII: DOI: Reference:

S0022-5193(16)30287-9 http://dx.doi.org/10.1016/j.jtbi.2016.09.005 YJTBI8813

To appear in: Journal of Theoretical Biology Received date: 11 January 2016 Revised date: 22 June 2016 Accepted date: 5 September 2016 Cite this article as: Ezhilarasi Krishnamoorthy, Sameer Hassan, Luke Elizabeth Hanna, Indira Padmalayam, Rama Rajaram and Vijay Viswanathan, Homology modeling of Homo sapiens Lipoic acid Synthase: substrate docking and insights on its binding mode, Journal of Theoretical Biology, http://dx.doi.org/10.1016/j.jtbi.2016.09.005 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Homology modeling of Homo sapiens Lipoic acid Synthase: substrate docking and insights on its binding mode Ezhilarasi Krishnamoorthya, Sameer Hassanb, Luke Elizabeth Hannab, Indira Padmalayamc, Rama Rajaramd, Vijay Viswanathana* a

Department of Biochemistry and Molecular Genetics, Prof. M. Viswanathan Diabetes Research Centre and M.V.

Hospital for Diabetes (WHO Collaborating Centre for Research, Education and Training in Diabetes), No. 4, West Madha Church Road, Royapuram, Chennai 600013, Tamil Nadu. India b

Department of Biomedical Informatics, National Institute for Research in Tuberculosis (Indian Council of Medical

Research), No. 1, Mayor Sathiyamoorthy Road, Chetpet, Chennai 600031, Tamil Nadu. India c

Drug Discovery Division, Southern Research Institute, Birmingham, USA

d

Department of Biochemistry, Central Leather Research Institute, Adyar, Chennai, Tamil Nadu. India

*Address for correspondence: Dr.Vijay Viswanathan, MD, PhD, FICP, FRCP (London & Glasgow) M. V. Hospital for Diabetes and Prof. M. Viswanathan Diabetes Research Centre. (WHO Collaborating Centre for Research, Education and Training in Diabetes) No. 4, Main Road, Royapuram, Chennai 600 013. INDIA. Tel.: +91 44 2595 49 13-15; fax: +91 44 2595 49 19. E-mail: [email protected]. Abstract: Lipoic acid synthase (LIAS) is an iron-sulfur cluster mitochondrial enzyme which catalyzes the final step in the de novo pathway for the biosynthesis of lipoic acid, a potent antioxidant. Recently there has been significant interest in its role in metabolic diseases and its deficiency in LIAS expression has been linked to conditions such as diabetes, atherosclerosis and neonatal-onset epilepsy, suggesting a strong inverse correlation between LIAS reduction and disease status. In this study we use a bioinformatics approach to predict its structure, which would be helpful to understanding its role. A homology model for LIAS protein was generated using X – ray crystallographic structure of Thermosynechococcus elongatus BP-1

2

(PDB ID: 4U0P). The predicted structure has 93 % of the residues in the most favour region of Ramachandran plot. The active site of LIAS protein was mapped and docked with S-Adenosyl Methionine (SAM) using GOLD software. The LIAS - SAM complex was further refined using molecular dynamics simulation within the subsite 1 and subsite 3 of the active site. To the best of our knowledge, this is the first study to report a reliable homology model of LIAS protein. This study will facilitate a better understanding mode of action of the enzyme-substrate complex for future studies in designing drugs that can target LIAS protein. Keywords: Lipoic acid synthase, Lipoic acid, Thermosynechococcus elongatus, Homology model, SAdenosyl Methionine.

1. Introduction Lipoic acid synthase (LIAS) is an iron-sulfur cluster mitochondrial enzyme responsible for the endogenous synthesis of Lipoic acid (LA) [1]. LA is an essential cofactor of mitochondrial enzyme complexes involved in oxidative metabolism [2]. Previously, the synthesis of LA was unknown in mammals, but the discovery of mouse homolog of LIAS validated the existence of LIAS synthesizing mechanism in mammals [3]. Recently two studies have reported the importance of LIAS in human subjects. A functional defect in the maturation of LIAS resulted in a significant reduction in mitochondrial LA levels due to impairment in the activity of LIAS, which led to decreased activities of the LA-requiring complexes such as pyruvate dehydrogenase complex (PDHC) and α-ketoglutarate dehydrogenase (KGDH) [4]. In another study, a polymorphism in LIAS was shown to cause significant reduction in mitochondrial LA levels, and was found to be associated with several abnormalities. A key abnormality was impairment in mitochondrial metabolism due to decreased PDHC activity and reduced oxidation of pyruvate [5]. Expression of LIAS has also been reported to be reduced in many animal models of metabolic diseases. LIAS was found to be significantly down regulated (70%) in skeletal and adipose tissues in animal models of Type 2 Diabetes mellitus (T2DM) and obesity [6]. Similarly, in the model of accelerated nephopathy (Ins2Akita/+) and apoE deficient (apoE-/-) diabetic mice, the

3

expression of LIAS was found to be decreased by 30% compared to wild type mice [7]. In vitro gene silencing studies also revealed that knockdown of LIAS has detrimental effects on various aspects of cellular function [6]. One of them is the disruption of the antioxidant defense system. The deficiency of LIAS also caused reduced mitochondrial lipoic acid levels [6]. In keeping with the in vitro observation, genetic reduction of LIAS expression in Ins2 Akita/+ mice (a type 1 diabetes model) caused a marked elevation of oxidative stress in the kidney of the mice and accelerated the onset and development of diabetic nephropathy [8]. Similarly, crossing apoE-/mice with LIAS heterozygous null mice resulted in the enhancement of atherosclerosis [9], suggesting a strong inverse correlation between LIAS reduction and disease status. Recent study also shows that pharmacological increase in LIAS and LA levels had many beneficial metabolic effects in rodent models of obesity and T2DM such as decrease in body weight gain, improvement in plasma glucose and lipid profile. These experiments provide strong in vivo evidence for the pathological benefits of correcting of LIAS deficiency in obesity and diabetes. Further, they validate the crucial role played by LIAS in the maintenance of mitochondrial redox balance, and metabolic consequences of LIAS deficiency [10]. Since the molecular function of a protein is tightly coupled to its function, we sought to determine the structure of the LIAS protein to better understand its functional role. There is currently no 3D structure of the mammalian LIAS protein available in the PDB database. It is also believed that as homologous proteins evolve, the basic structure often remains more conserved than their sequence [11]. Identifying similarity between protein structures can therefore yield valuable clues to their function [12] and also can reveal the structural plasticity of the fold within a family [13]. LIAS is a member of the radical-SAM (RS) superfamily, enzymes that use an electron donated from a [4Fe−4S] + cluster to cleave S-adenosyl-L-methionine (SAM) reductively to Lmethionine and a 5′-deoxyadenosyl 5′-radical (5′-dA•). 5′-dA• is a potent oxidant used to abstract hydrogen atoms (H•) both from protein and small-molecule substrates, many of which contain inert target C−H bonds [11]. The [4Fe−4S] cluster is ligated by three Cys residues that reside typically in a CX3CX2C motif, which serves as a signature sequence for RS enzymes [12].The LIAS reaction proceeds as two distinct half-reactions. In the first half-reaction, SAM bound to

4

the [4Fe−4S] cluster ligated by Cys residues in the CX3CX2C motif (RS cluster) is reductively cleaved to a 5′-dA•, which is used to abstract an H• from C6 of the octanoyl chain. The resulting carbon-centered radical attacks an activated form of sulfide, believed to be a bridging μ3-sulfido ion of the [4Fe−4S] cluster ligated by Cys residues in the CX4CX5C motif (auxiliary cluster). In the second half-reaction, another equivalent of 5′- dA• is similarly generated, but it abstracts an H• from C8 of the octanoyl chain. The carbon-centered radical at C8 attacks a second form of activated sulfide, presumably from the same [4Fe−4S] cluster upon addition of two protons; the lipoic acid is formed in its two-electron reduced state [13]. In this study, we aimed to pridct the 3D structure of LIAS –SAM complex using a combination of computational modeling and docking. Further, the predicted complex structure was refined using molecular dynamics simulation. 2. Material and Methods 2.1. Homology modeling The protein sequence of LIAS was downloaded from the UNIPROT database (http://www.uniprot.org/downloads) with accession number O43766. In order to identify homologous sequences with known 3D structures and find related protein structures as templates, a protein–protein BLAST [14] was performed against the NCBI protein data bank (PDB) [15] database using LIAS as a query sequence. Based on the BLAST results, the template structure was downloaded from the RCSB Protein Data Bank (http://www.rcsb.org/pdb/). The target sequence was then aligned to the template sequence using Discovery Studio 2.1 and the Align 123 program (Accelrys Software, Inc., San Diego, CA). Homology model of LIAS was generated by submitting the sequence alignments to DS 2.1 and the homology model building Modeler 9v4 program [16]. The 3D models were ranked based on probability density function (PDF) total energy and evaluated by Verify Protein (Profiles-3D) [17], PROCHECK [18] and ProSa-web [19]. 2.2. Active site analysis The active site was identified and reconfirmed by superimposition and multiple structure based sequence alignments of LIAS with the selected crystal structure of Biotin Synthase, an S-

5

adenosylmethionine-dependent radical enzyme (PDB code: 1R30) in complex with SAdenosylMethionine (SAM). Based on the alignment, the binding site of SAM in 1R30 was then mapped on to the LIAS protein. There are many different servers available for predicting the active site for the protein of interest. However, Knowledge based method where structurally related proteins bound with ligand are used to map the binding site on to the uncharacterized protein sequence [20]. As a result, the predicted site was chosen as the most favorable binding site for docking with the substrate.

2.3. Molecular Docking Molecular docking analysis was performed using SAM molecule as ligand with LIAS protein to study its mode of interaction within the binding site. GOLD v5.2(Genetic Optimisation for Ligand Docking) is a widely used protein–ligand docking programme [21]. GOLD uses a genetic algorithm methodology for protein ligand docking that allows the full ligand and partial protein flexibility. The homology model of LIAS protein used as a receptor for docking was added with hydrogen atoms. The SAM molecule was corrected with Auto Edit Ligand option GOLD program. The binding site mapped from section 2.3 was used as input for GOLD calculation. Default genetic algorithm (GA) settings that ensure 100% search efficiency were used for docking. A total of 20 poses of docked molecule were generated. An early termination of the number of GA runs was allowed when the RMSDs of the top three GA solutions were within 1.5 Å. The best pose of the docked ligand was selected based on CHEMPLP score. 2.4. Molecular dynamic simulation Molecular Dynamics (MD) simulations were performed on the LIAS apo and LIAS – SAM complex obtained from docking studies using AMBER12 [22]. The Amber ff99SB force field was applied to the protein structure. The parameters for SAM were produced from the Generalized Amber Force Field (GAFF) and the missing parameters are calculated using ANTECHAMBER using the AM1-bcc model. The topology and parameter files for LIAS – SAM complex and LIAS apo structures were generated using ANTECHAMBER tools and tLEAP program [23]. Both strucures of apo-LIAS protein and the LIAS – SAM complex were

6

surrounded by rectangular periodic boxes of TIP3P water molecules with a margin of 10.0 Å. The system was neutralized by addition of Na+ ions. Prior to MD, 1500 steps of steepestdescent minimization and 1000 steps of conjugated gradient minimization were applied to the solvent and the entire model system, respectively. The entire system was heated from 0 to 300 K gradually over 200 ps (picoseconds) using NVT (constant volume and normal temperature) conditions and then simulations were carried out under NPT conditions at 1 atm pressure for another 200 ps. MD simulations were finally performed for 10 ns at 1 atm and 300 K under the NPT ensemble with a time step of 2 fs (femtoseconds). The temperature was controlled using Langevin dynamics and simulations were run using periodic boundary conditions. The PTRAJ module of AMBER package was used for analyzing the trajectories from MD simulations. 3. Results and Discussion 3.1. Sequence analysis and homology modeling The purpose of homology modeling is to design a 3D structure of a protein by using previously resolved structures of homologous protein. Results of the

sequence similarity

revealed that LIAS belongs to the RADICAL SAM super family. There are recent reports which state that the 3D structures of two proteins could be similar if they share a sequence identity of more than 25% [24,25]. Comparing the LIAS sequence against the PDB database using BlastP revealed similarity of mamalian LIAS with Lipoyl Synthase from Thermosynechococcus elongatus BP-1. The target sequence showed 44% identity and 59% similarity with the sequence of Thermosynechococcus elongatus BP-1 (PDB code: 4U0P). The crystal structure of Lipoyl Synthase from Thermosynechococcus elongatus BP-1 incomplex with S-adenosyl Homocysteine at a resolution of 1.62 Å was selected as a template for generating the three dimensional model for LIAS protein. The target sequence was amenable to alignment from redidues 76 -357 of the selected template, while residues from 1 – 75 were not amenable to alignment. The aligned region was then manually checked for the conserved regions and the final alignment was used for predicting the model structure of LIAS protein. Five model structures were generated using a modeling module in Discovery Studio. Based on the PDF total energy, the best model was selected from the five models (Fig.1). The

7

final model was validated using PROCHECK and ProSa-web. The Ramachandran plot revealed the distribution of the backbone dihedral angle of all amino acid residues of LIAS protein as follows: 93 % were in the most favored regions, 5.7 % were in additional allowed regions, 0.8 % were in generously allowed regions and 0.4 % in the disallowed region (Fig. 2). Further, the ProSa - web analysis further indicated that the generated LIAS model has a Z-score of -9.17, which is well within the range of scores typically found in native proteins of similar size. The LIAS model was further evaluated by superimposing it over the selected template (PDB code: 4U0P). The RMSD value obtained was 0.29Å based on the backbone atoms (Fig. 3). The overall analysis revealed that the predicted model of LIAS protein as satisfactory and was therefore considered for the rest of the study.

Fig.1. The 3D structure of LIAS protein of Homo Sapiens. The alpha helix is represented in cyan, the beta strand is shown in magenta, the loop is represented in orange colour.

8

Fig. 2. Ramachandran plot for LIAS protein obtained by PROCHECK analysis.

Fig.3. Structural superimposition of LIAS protein with the template 4U0P. LIAS is shown in green and 4U0P in cyan colour.

9

The binding site of the substrate S-AdenosylMethionine (SAM) molecule in LIAS protein was identified and mapped on the LIAS sequence by superimposing the model structure with crystal structure of Biotin Synthase, an S- adenosylmethionine-dependent radical enzyme (PDB code: 1R30) in complex with SAM (Fig. 4A). The identitifed active site of LIAS protein was used as the binding site for docking studies with the substrate (Fig. 4B).

10

Fig. 4 (a). Structure based sequence alignment of LIAS model with the crystal structures of 4U0P, and 1R30. The mapped SAM binding site is highlighted in black color. (b). Active site superimposition of crystal structures, 4U0P (green) and 1R30 (light blue) with the mapped active site of LIAS (pink). 3.2. Binding of SAM with LIAS model using molecular docking In order to identify the interaction of SAM ligand within the active site of LIAS model, GOLD software was used as the docking tool. Based on the GOLD score, the best pose among the 20 poses generated was analyzed by superimposing with the crystal structure, 1R30. The docked pose of SAM that is logically closer to the conformation of the bound SAM in the crystal structure, 1R30 was examined (Fig. 5a). The pose with a GOLD score of 67 Kcal/mole was

11

selected as the best conformation for further analysis. Wei Hong et al [26] reported for human protein arginine methyltransferase 1 that the SAM binding site can be divided into two sub-sites, sub-site 1 and sub site 3. Sub-site 1 accommodates the adenine moiety and the sub-site 3 accommodates the homocysteine moiety. While comparing the binding pose of the docked SAM within the binding pocket of LIAS protein, a very similar arrangement is observed as reported by Wei Hong et al [26], wherein the adenine moiety of SAM is in the sub-site 1 and the homocysteine is in sub-site 3 (Fig. 5b). Based on the interaction analysis of the selected docked pose of SAM molecule, it is observed that six residues within the active site of LIAS protein form hydrogen bonds with SAM molecule. The adenine moiety of SAM forms three hydrogen bonds; two with MET235 and one with PHE68. The adenosine ribose moiety forms one hydrogen bond with SER276. The carboxylate group of SAM homocysteine moiety forms a total of two hydrogen bonds; one with ASN162 and the other with GLU164(Fig. 5b).

Fig. 5 (a) Superimposition of the LIAS protein docked with SAM and the crystal structure of 1R30 (b) Interaction of LIAS active site residues with docked SAM molecule. The red and

12

green dotted lines represents hydrogen acceptor and donor interaction. The red and blue shaded circles represent negative and positive ionisable interactions. 3.3. Refinement of docked LIAS-SAM complex using molecular simulation Predicting the binding pose of ligands has been successful using molecular docking tools. However, estimating the affinity of the docked ligands has been unsuccessful [27,28].Treating the protein receptor as a rigid molecule and not allowing to adjust conformation during docking is one of the major reasons for the failure. Homology modeling, molecular docking, and molecular dynamics simulations have aided in revealing the binding mode of SAM. Molecular simulation method is widely used to analyze the conformational changes that occur with protein - ligand interaction [29,30], dynamics of protein [31,32] and protein folding [33,34]. In the current study, docking was performed with LIAS protein by allowing flexibility at the LIAS active site. In order to study the flexibility of the receptor protein and conformational changes upon the binding to the SAM molecule both the apo structure of LIAS and the selected docked pose of SAM with LIAS protein were further refined using explicit solvent MD simulation. To study the stability and conformational changes of the structures of apo and LIAS – SAM complex during molecular simulation, the root-mean-square deviation (RMSD) for the backbone atoms were plotted against the initial LIAS homology model (Fig. 6). From Figure 6, it can be noted that both the systems have a RMSD of 1.5 – 2.5Å during 10 ns of simulation. The RMSD of the LIAS apo structure is observed to gradually increase during the first 2 ns; after this there was no further increase in RMSD and the LIAS apo structure reaches equilibrium. However, the LIAS-SAM complex, has a RMSD fluctuation, where as a steady increase in RMSD from 1 ns to 4ns and after 7 ns there was a gradual dip in the RMSD up to 9 ns following which the system was observed to achieve equillibrium. This conformational change could be attributed to the rearrangement of the SAM molecule and the LIAS protein during simulation. Superimposing the LIAS structures before and after MD simulation shows a backbone RMSD difference of 1.95 Å (Fig. 7A). It is also observed that the active site residues of LIAS protein and the ligand after 10 ns of simulation rearranges its conformation and orientation (Fig. 7A and 7B). This indicates that the flexibilities incorporated into the structure and ligand by simulation helps to attaina better correlation between the experiment and the simulation [35]. The interaction of SAM within the

13

binding site of LIAS protein after simulation is shown in Fig. 8A and B. The adenine moiety of SAM forms three hydrogen bonds; two with MET235 and one withSER70. The adenosine ribose moiety forms one hydrogen bond with SER276. The carboxylate group of SAM homocysteine moiety forms a total of two hydrogen bonds; one with ASP106 and the other with SER104 (Fig. 8A). Analyzing the interaction of SAM with the binding site residues of LIAS protein before and after simulation, hydrogen bond interaction of residues MET235 and SER276 with the adenine and adenosine ribose moieties are retained. The root mean square fluctuation (RMSF) of the backbone atoms of the residues in LIAS apo and LIAS – SAM complex was calculated to study the flexibility of the residues in these structures. The residues with high RMSF values signify more flexibility whereas low RMSF values signify limited movement during the 10 ns simulation. The flexibility differences for the residues in LIAS apo and LIAS-SAM complex structures are shown in Fig. 9. It is observed that the residues 14- 40 and 70 – 80 in LIAS apo structure exhibit more flexibility than in the LIASSAM complex. The active site residues of LIAS protein that bind to SAM molecule exhibit a small degree of movement with a RMSF less than 1 nm and revealed small differences when compared to that of the LIAS apo structure. This indicates that the binding of SAM within the active site of LIAS protein induces minor conformational changes.

14

Fig.6. The backbone RMSD are shown for the LIAS apo and LIAS – SAM complex with reference to the initial homology model structure. The line in black colour indicates the LIAS apo structure and the line in red indicates the LIAS – SAM complex.

15

Fig.7. (a) Superimposition of LIAS-SAM complex before (green) and after (cyan) MD simulation. The bound SAM within the binding site for before and after MD is shown in yellow and blue colour respectively. (b) Superimposition of binding site residues of LIAS protein before (green) and after (cyan) MD simulation. The amino acids are labeled with their three letter code along with the residue number.

Fig. 8. (a) Interaction of LIAS active site residues with docked SAM molecule after molecular simulation. The red and green dotted lines represent hydrogen acceptor and donor interaction. (b) Molecular surface representation of SAM molecule within the active site. Red color represents

16

an electronegative and blue color represents electropositive region.

Fig. 9. The RMSF plot showing the fluctuation of the backbone atoms of both LIAS apo and LIAS – SAM complex. The black line indicates the LIAS apo and the red line indicates LIAS – SAM complex. Nowadays, the LIAS protein was shown to be involved in the potential application of medicine. In animal model partial deletion of radical SAM domain in LIAS leads to defect in the development of forebrain in the mouse [36]. Recent studies show that pharmacological induction of LIAS offers a novel peripheral approach for management of obesity and associated metabolic abnormalities. And also another study in 2014 by Tsou etal shows that LA and LIAS could reverse the myofibroblast differentiation of Systemic sclerosis dermal fibroblasts and found to be effective treatment in Systemic sclerosis[37,38]. The disorder of non-ketotic hyperglycinaemia is associated with LIAS mutation which would be important information for physicians evaluating patients with abnormalities in glycine as this will affect the genetic causation and genetic counselling, and provide prognostic information on the expected phenotypic course[39]. Taken together, deficiency or mutations in LIAS are associated with inherited disorders of amino acid and energy metabolism in human.

17

4. Conclusions To our knowledge, this is the first study to report a reliable homology model of LIAS protein. The binding pocket and key residues involved in binding to SAM molecule were identified. Molecular dynamics simulation analysis of both the LIAS apostructure and with SAM complex was performed to study the dynamic behavior of LIAS protein, and further, this study will facilitate a better understanding the mode of action of the enzyme-substrate complex and would be of help for future studies in designing drugs that can target LIAS protein. List of abbrevations LIAS- Lipoic acid synthase LA- Lipoic acid ORF - Open reading frame KGDH - α-Ketoglutarate dehydrogenase PDHC- Pyruvate dehydrogenase complex Ins2Akita/+

- Accelerated nephropathy

apoE-/- - Apo E deficient T2DM- Type 2 Diabetes mellitus GA- Genetic algorithm MD- Molecular Dynamics RMSD-Root Mean Square Deviation RMSF- The root mean square fluctuation Ps - Picoseconds Fs - Femtoseconds NVT - Constant volume and normal temperature SAM - S- Adenosyl Methionine

18

Competing interests None declared. Authors' contributions E.K — as a PhD student,was involved in the entire study, the design, and interpretation of the data; S.H—participated in the design, coordination, interpretation of data and helped to draft the manuscript; I.P — contributed in the design, revision and editing the manuscript for critical content. R.R — contributed in the design, revision and editing the manuscript for critical content. L.E.H — contributed in the design, revision and editing the manuscript for critical content. V.V— contributed in the design, revision and editing the manuscript for critical content. All authors read and approved the final manuscript. Acknowledgements The authors thank Dr. Soumya Swaminathan and Dr. Hemanth Kumar of NIRT for their support.

Reference: [1] R.M.Cicchillo, S.J.Book, Mechanistic investigations of lipoic acid biosynthesis in Escherichia coli: both sulfur atoms in lipoic acid are contributed by the same lipoyl synthase polypeptide, J Am Chem Soc.127 (2005) 2860-2861. [2] L.J.Reed, From lipoic acid to multi-enzyme complexes, Protein Sci.7 (1998) 220-224. [3] T.Morikawa, R.Yasuno, H.Wada, Do mammalian cellssynthesize lipoic acid? Identification of a mouse cDNA encoding alipoic acid synthase located in mitochondria FEBS Lett.498 (2001) 16-21. [4] A.Navarro-Sastre, F. Tort, O.Stehling, M.A. Uzarska, J.A.Arranz, M. Del Toro, M.T. Labayru, J.Landa, A.Font, J. Garcia-Villoria, B.Merinero, M.Ugarte, L.G. GutierrezSolana, J.Campistol, A. Garcia-Cazorla, J. Vaquerizo, E. Riudor, P.Briones, O. Elpeleg,

19

A. Ribes, R. Lill, A Fatal Mitochondrial Disease Is Associated with Defective NFU1 Function in the Maturation of a Subset of Mitochondrial Fe-S Proteins, Am.J.Hum. Genet. 89(2011) 656-667. [5] J.A.Mayr, F.A. Zimmermann, C. Fauth, C. Bergheim, D.Meierhofer, D. Radmayr, J. Zschocke, J. Koch, W. Sperl, Lipoic acid synthetase deficiency causes neonatal-onset epilepsy, defective mitochondrial energy metabolism, and glycine elevation, Am. J. Hum. Genet. 89(2011) 792-797. [6] Padmalayam, I.; Hasham, S.; Saxena, U.; Pillarisetti, S. Lipoic Acid Synthase (LASY): A Novel Role in Inflammation, Mitochondrial Function and Insulin Resistance. Diabetes, 2009, 58(3), 600-608. [7] X.Yi, V.Nickeleit, L.R. James, N. Maeda, α-Lipoic acid protects diabetic apolipoprotein E-deficient mice from nephropathy, J.Diabetes Complications.25 (2011) 193-201. [8] X.Yi, L. Xu, S. Hiller, H.S. Kim, V. Nickeleit, L.R. James, N.Maeda, Reduced Expression of Lipoic Acid Synthase Accelerates Diabetic Nephropathy, J. Am. Soc. Nephrol. 23(2011) 103-111. [9] X.Yi, L. Xu, K. Kim, H.S. Kim,N.Maeda, Genetic reduction of lipoic acid synthase expression modestly increases atherosclerosisin male, but not in female, apolipoprotein Edeficient mice,Atherosclerosis.211(2010) 424-430. [10] I. Padmalayam, Targeting mitochondrial oxidative stress through lipoic acid synthase: a novel strategy to manage the diabetic cardiovascular disease, Cardiovasc Hematol Agents Med Chem. 10(2012) 223-33. [11] Roach, P. L. (2011) Radicals from S-adenosylmethionine and their application to biosynthesis. Curr. Opin. Chem. Biol. 15, 267−275. [12] Sofia, H. J., Chen, G., Hetzler, B. G., Reyes-Spindola, J. F., and Miller, N. E. (2001) Radical SAM, a novel protein superfamily linking unresolved steps in familiar biosynthetic pathways with radical mechanisms: functional characterization using new analysis and information visualization methods. Nucleic Acids Res. 29, 1097−1106.

20

[13] Cicchillo, R. M., and Booker, S. J. (2005) Mechanistic investigations of lipoic acid biosynthesis in Escherichia coli: both sulfur atoms in lipoic acid are contributed by the same lipoyl synthase polypeptide. J. Am. Chem. Soc. 127, 2860−2861. [14] S.F. Altschul, T.L. Madden, A.A. Schäffer, J. Zhang, Z. Zhang, W. Miller, D.J. Lipman,Gapped BLAST and PSI-BLAST: a new generation of protein database searchprograms, Nucleic Acids Res. 25 (1997) 3389–3402. [15] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig, I.N.Shindyalov, P.E. Bourne, The Protein Data Bank, Nucleic Acids Res. 28 (2000)235– 242. [16] A. Sali, L. Potterton, F. Yuan, H. van Vlijmen, M. Karplus, Evaluation of comparative protein modeling by MODELLER, Proteins. 23(1995) 318-26. [17] Discovery Studio Profile-3D User Guide, 2000 (Biosym/MSI: San Diego, CA). [18] R.A. Laskowski, D.S. Moss, J.M. Thornton, Main-chain bond lengths and bond angles in protein structures, J. Mol. Biol. 231 (1993) 1049–1067. [19] M.J. Sippl, Recognition of errors in three-dimensional structures of proteins, Proteins. 17 (1993) 355–362. [20] D. Ghersi, R. Sanchez, Beyond structural genomics: computational approaches for the identification of ligand binding sites in protein structures, J Struct Funct Genomics 12(2011)109–117. [21] G. Wu, D.H. Robertson, C.L.Brooks, M. Vieth, Detailed analysis of grid-based molecular docking: a case study of CDOCKER-A CHARMm-based MD docking algorithm, J. Comput. Chem. 24 (2003) 1549–1562. [22] D.A.Pearlman, D.A. Case, J.W.Caldwell, W.S. Ross, T.E. Cheatham III,S. DeBolt, D. Ferguson, G. Seibel, P. Kollman, AMBER, a package of computer programs for applying molecular mechanics, normal mode analysis, molecular dynamics, and free energy calculations to simulate the structural and energetic properties of molecules. Comput. Phys. Commun. 91(1995) 1–41.

21

[23] D.A. Case, T.E. Cheatham III, T. Darden, H. Gohlke, R. Luo, K.M.Merz, A.Onufriev, C. Simmerling, B. Wang, R.Woods, The Amber biomolecular simulation programs,J. Comput. Chem. 26(2005)1668–1688. [24] A.S.Yang, B. Honig, An integrated approach to the analysis and modeling of protein sequences and structures. III. A comparative study of sequence conservation in protein structural families using multiple structural alignments, J Mol Biol. 301(2000) 691-711. [25] B. Rost, Twilight zone of protein sequence alignments, Protein Eng. 12(1999) 85-94. [26] W. Hong, J. Li, C.A.Laughton, L.F. Yap, I.C. Paterson , H.Wang, Investigating the binding preferences of small molecule inhibitors of human protein arginine methyltransferase 1 using molecular modeling, J Mol Graph Model. 51(2014)193-202. [27] S.F. Sousa, P.A. Fernandes,M.J. Ramos, Protein-ligand docking: Current status and future challenges, Proteins.65(2006)15-26. [28] T. Cheng, Q. Li , Z. Zhou, Y.Wang, S. Bryant, Structure-Based Virtual Screening for Drug Discovery: a Problem-Centric Review, AAPS J. 12(2012)133-141. [29] W. Li, J. Shen, G. Liu, Y.Tang ,T. Hoshino, Exploring coumarin egress channels in human cytochrome p450 2a6 by random acceleration and steered molecular dynamics simulations, Proteins. 79( 2011) 271-281. [30] J. Shen, W. Li, G. Liu, Y. Tang,H. Jiang, Computational Insights into the Mechanism of Ligand

Unbinding

and

Selectivity

of

Estrogen

Receptors,

J

Phys

ChemB

113(2009)10436-10444. [31] R. Nygaard, Y. Zou, R.O. Dror, T.J. Mildorf, D.H. Arlow, A. Manglik, A.C. Pan C.W. Liu, J. Fung, M.P. Bokoch, F.S. Thian, T.S. Kobilka, D.E Shaw, R.Mueller, R. Prosser, B. Kolbilka, The Dynamic Process of ²2-Adrenergic Receptor Activation, Cell 152(2013) 532-542.

22

[32] A.C. Kruse, J. Hu, A.C. Pan, D.H. Arlow, D.M. Rosenbaum, E. Rosemond, H.F. Green, T. Liu , P.S.Chae, R.O.Dror, D.E. Shaw, W.I.Weis, J. Wess, B.K. Kolbilka, Structure and dynamics of the M3 muscarinic acetylcholine receptor, Nature.482(2012) 552-556. [33] K. Lindorff-Larsen, S. Piana, R.O.Dror, D.E. Shaw, How Fast-Folding Proteins Fold. Science, 334(2011) 517-520. [34] S. Piana, K. Lindorff-Larsen, D.E. Shaw, Atomic-level description of ubiquitin folding, Proc Natl Acad Sci. 110(2013) 5915-5920. [35] G.Tiwari G, D. Mohanty D, An in silico analysis of the binding modes and bindingaffinities of small molecule modulators of PDZ-peptide interactions, PLoS One. 8(2013) e71340. [36] X. Zhou, K. V. Anderson, Development of Head Organizer of the Mouse Embryo Depends on a High Level of Mitochondrial Metabolism, Dev Biol. 344(2010) 185–195. [37] I. Padmalayam, R. Hamm, D. Srinivas Reddy, D. Bhuniya, A. Vines, S. Seenu, R. Chakrabarti, U. Saxena, S. Pillarisetti, Identification of a Pharmacological Inducer of Lipoic Acid Synthase that Impacts Mitochondrial Function: Metabolic Benefits and Body Weight Loss without Changes in Caloric Intake, J Diabetes Metab 11(2012) 2155-6156. [38] P.S.Tsou, B.Balogh, A.J. Pinney, G. Zakhem, A. Lozier, M.A. Amin, W.A. Stinson, E. Schiopu, D. Khanna, D.A. Fox, A.E Koch, Lipoic acid plays a role in scleroderma: insights obtained from scleroderma dermal fibroblasts. Arthritis Res Ther. 16(2014) 411. [39] P.R. Baker, M. W. Friederich, M.A. Swanson, T. Shaikh, K. Bhattacharya, G.H. Scharer, J. Aicher, G. Creadon-Swindell, E. Geiger, K.N. MacLean, W.T. Lee, C. Deshpande, M.L. Freckmann, L.Y. Shih, M. Wasserstein, M.B. Rasmussen, A.M. Lund, P. Procopis, J.M. Cameron, B.H. Robinson, G.K. Brown, R.M. Brown, A.G. Compton, C.L. Dieckmann, R. Collard, C.R. Coughlin, E. Spector, M.F. Wempe, J.L. Van Hove, Variant non ketotic hyperglycinemia is caused by mutations in LIAS, BOLA3 and the novel gene GLRX5. Brain. 137(2014)366-379.

23

Highlights:   

This is the first study to report a reliable homology model of LIAS protein. The binding pocket and key residues involved in binding to SAM molecule were identified. This study will facilitate a better understanding the mode of action of the enzymesubstrate complex and would be of help for future studies in designing drugs that can target LIAS protein.