- Email: [email protected]
Regulatory networks, genes and glycerophospholipid biosynthesis pathway in schistosomiasis: A systems biology view for pharmacological intervention
Gene 550 (2014) 214–222
Contents lists available at ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
Regulatory networks, genes and glycerophospholipid biosynthesis pathway in schistosomiasis: A systems biology view for pharmacological intervention Sonali Shinde, Milsee Mol, Shailza Singh ⁎ National Centre for Cell Science, NCCS Complex, Ganeshkhind, Pune University Campus, Pune 411007, India
a r t i c l e
i n f o
Article history: Received 16 April 2014 Received in revised form 5 August 2014 Accepted 13 August 2014 Available online 19 August 2014 Keywords: Metabolic network Schistosomiasis Glycerophospholipid biosynthesis Cytidyltransferase Inhibitor designing
a b s t r a c t Understanding network topology through embracing the global dynamical regulation of genes in an active state space rather than traditional one-gene–one trait approach facilitates the rational drug development process. Schistosomiasis, a neglected tropical disease, has glycerophospholipids as abundant molecules present on its surface. Lack of effective clinical solutions to treat pathogens encourages us to carry out systems-level studies that could contribute to the development of an effective therapy. Development of a strategy for identifying drug targets by combined genome-scale metabolic network and essentiality analyses through in silico approaches provides tantalizing opportunity to investigate the role of protein/substrate metabolism. A genome-scale metabolic network model reconstruction represents choline–phosphate cytidyltransferase as the rate limiting enzyme and regulates the rate of phosphatidylcholine (PC) biosynthesis. The uptake of choline was regulated by choline concentration, promoting the regulation of phosphocholine synthesis. In Schistosoma, the change in developmental stage could result from the availability of choline, hampering its developmental cycle. There are no structural reports for this protein. In order to inhibit the activity of choline–phosphate cytidyltransferase (CCT), it was modeled by homology modeling using 1COZ as the template from Bacillus subtilis. The transition-state stabilization and catalytic residues were mapped as ‘HXGH’ and ‘RTEGISTT’ motif. CCT catalyzes the formation of CDP-choline from phosphocholine in which nucleotidyltransferase adds CTP to phosphocholine. The presence of phosphocholine permits the parasite to survive in an immunologically hostile environment. This feature endeavors development of an inhibitor specific for cytidyltransferase in Schistosoma. Flavonolignans were used to inhibit this activity in which hydnowightin showed the highest affinity as compared to miltefosine. © 2014 Elsevier B.V. All rights reserved.
1. Introduction A prerequisite for understanding any disease is acquiring knowledge of its component interactions between the host–parasite systems. This is possible with the available sequence and omics' data obtained through various high throughput platforms. The volumes of data generated can be analyzed to identify individual component interactions that make up a system, which can be analyzed within a mathematical framework to identify possible targets for new drug discovery (Kitano, 2002).
Abbreviations: PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine; GPC, glyceryl phosphorylcholine; CTP, cytidine triphosphate; CCT, choline–phosphate cytidyltransferase; CK, choline kinase; ADP, adenine diphosphate; DAG, diacylglycerol; AAG, alkyl-acylglycerol; IL, interleukin; DC, dendritic cells; Lyso-PC, lysophosphatidylcholine; FBA, flux balance analysis; GEM, genome-scale model; FBA, flux based analysis; MD, Molecular Dynamic; SPC, Simple Point Charge; NVT, Number of moles, Volume, Temperature; NPT, Number of moles, Pressure, Temperature; PME, Particle Mesh Ewald; RMSD, root mean square deviations; RMSF, root mean square fluctuations; LGA, Lamarckian Genetic Algorithm; PE, potential energy. ⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (S. Singh).
http://dx.doi.org/10.1016/j.gene.2014.08.031 0378-1119/© 2014 Elsevier B.V. All rights reserved.
Under this backdrop our work represents abstracting the complex lipid biosynthetic process in schistosomiasis for novel drug target identification. Schistosomiasis is a tropical disease caused by trematode belonging to the species Schistosoma. Control of schistosomiasis is challenging as more than 600 million people in 76 developing countries are at risk and with 200 million people already affected by this deadly disease (Berriman et al., 2009). The intestinal form of the disease is caused by Schistosoma mansoni. The availability of genomic data of the intestinal form of the parasite opened new avenues to understand disease novel pathophysiology that can be harnessed to progression of the disease (Useh, 2013). The parasite enters human, a definitive host (snail: Biomphalaria glabrata, being the intermediate host) by dissolving the skin barrier and migrates towards the liver and lungs to lay eggs. These eggs are lodged within the blood capillaries which is a major cause of pathology. Praziquantel (Biltricide®) is the effective monotherapy to control schistosomiasis but reports of drug resistance and tolerance have limited its use (Doenhoff et al., 2008). Designing new drug becomes difficult due to the rapid proliferation and changing surface variants of the parasite. One of the important surface components involved in renewal of membrane complex is associated with lipid
S. Shinde et al. / Gene 550 (2014) 214–222
metabolism (Ramakrishnan et al., 2012). Impairment reduces these pathogens' ability to proceed in its infectious life cycle and therefore could be a possible source of drug target (Marechal, 2011). Dissection of lipid metabolism using lipidomics in S. mansoni can provide cues on the alteration in the lipid pool of parasites that is associated with drug resistance (Lewis et al., 1996). Metabolic profile of S. mansoni infected mice liver indicates lower levels of glucose and glycogen but higher levels of choline metabolites, (phosphatidylcholine (PC), glyceryl phosphorylcholine (GPC), and alanine) than in controls suggesting that Schistosoma can suppress the metabolic response of the host (Wu et al., 2010). Many evidence suggests that lipids are essential in the life cycle of the parasite, and its regulation mechanisms are largely unknown (Loukas and Maizels, 2000). Glycerophospholipids are one such class of lipids in Schistosoma which accounts for PC (25%), phosphatidylserine (PS—15%) and phosphatidylethanolamine (PE— 8%). Fatty acids are differentially incorporated into the various phospholipid classes, principally into PC and, to a lesser extent, into PE, lyso-PC, and PS (Furlong and Caulfield, 1989). They show immune modulation properties e.g. PS induces dendritic cells (DC) to polarize IL-4/IL10 producing T-cells. Whereas lysophosphatidylserine (lyso-PS) specifies DC to induce IL-10 secreting regulatory T cells, thus swaying the immune system away from a protective Th1 immune response (Hewitson et al., 2009a, 2009b). Lysophosphatidylcholine (Lyso-PC) is known to adhere, immobilize and lyse the red blood cell to alter the cell plasma membrane, thereby neutralizing the attacking host cells. It promotes membrane fusion with the host resulting in the acquisition of host membrane components by the parasite. The presence of lysoPC may alter antigen presentation and immune recognition of the parasite antigens (Golan et al., 1986). These immune modulating lipids confer on the parasite the major adaptation mechanism, allowing survival in an immunologically hostile environment within the host (Schmid-Hempel, 2009). Also PCs are essential in the development of Schistosoma, as large amount is needed for a rapid membrane renewal and to renew its outer bilayer surface. The synthesis of PC in S. mansoni adults occurs via the Kennedy pathway (de novo synthesis of PC) consisting of three enzymatic steps shown in Fig. 1 (Young and Podesta, 1982). In protozoan parasites like Trypanosoma brucei, Plasmodium falciparum Kennedy pathways have already been validated as a drug
215
target (Ancelin et al., 2003; Cui and Houweling, 2002; Gibellini and Smith, 2010). In order to understand the highly redundant phospholipid pathway and to establish a mechanism to identify newer drug targets a kinetic model was developed in S. mansoni. Lipid mining connects the lipid profiles with the metabolic pathways during different stages of disease progression. This may allow deciphering the complexities and regulatory mechanisms involved in maintaining lipid homeostasis in diseased condition. Potential drug targets can also be identified that may be important regulatory hubs in the lipid biosynthesis. Designing small molecule inhibitors against these targets may be valuable ‘schistosomicides’. Our current understanding of Kennedy pathway in Schistosoma development and pathogenesis could help to develop anti-schistosomal therapies targeting lipid metabolic pathways. We have attempted to delineate lipid metabolism specifically PC biosynthetic pathway in parasite and identify key regulatory enzymes. These were screened against natural compounds like flavonolignan derivatives as potential lead compounds for possible drug design. 2. Material and Methods 2.1. Metabolic Network Analysis To address the current challenge of metabolic network biology and understand how organisms adapt to their environment to maintain optimal growth in changing environmental conditions the mathematical model was optimized for PC biomass production. The built model represents a schematic way to unravel the molecular components that underlie cellular processes with the available concentration of metabolites and kinetic data from KEGG database (Kanehisa and Goto, 2000) and Biocyc database (Caspi et al., 2010). The metabolic network was designed using CellDesigner v4.3 (Funahashi et al., 2008; Kitano et al., 2005). The detail work flow is shown in Fig. 2. Reactions in the metabolic network are connected via gene–protein-reaction (GPR) mechanism where the cofactors are considered to be independent of time and thus are not directly considered for the study. The boundary conditions of genes are constant. The reactions in the network are under the control of several kinetic laws like Hill–Hinze, convenience kinetics, and Michaelis–Menten (MM). Numerical simulation of the built model identifies essential genes/ metabolites vital for cell growth under the imposed constraint. The sensitivity of the flux and the topology of the network are studied by summarizing the stoichiometric matrix using a Copasi tool (Hoops et al., 2006). The stoichiometry of the reaction network is described mathematically as a stoichiometric matrix with rows representing reactants and columns corresponding to the reaction deduced by flux balance analysis (FBA) and ExPa (Bell and Palsson, 2005). In ExPa, the topology of the metabolic network is summarized as a stoichiometric coefficient matrix (Sij) in which, the rate of change of concentration of a metabolite is the sum of metabolite flow to and from the metabolite represented as Sij b 0 when consumed and Sij N 0 when produced. An Optim tool box of Matlab version 7.12.0.35 (R2011a) finds the minimum and maximum objective parameters required to get the optimal solution of a linear equation determined by the feasible boundary region (Lee et al., 2005). 2.2. Sequence Retrieval, Model Building and Evaluation
Fig. 1. Biosynthesis of phosphatidylcholine.
The target identified from network analysis was modeled by homology modeling protocol to get an insight into the functional characteristic of the protein. The sequence of choline–phosphate cytidyltransferase (CCT) was retrieved from Uniprot database (Uniprot ID: G4VD67_SCHMA) and blast against RCSB PDB (The Research Collaboratory for Structural Bioinformatics Protein Data Bank) to obtain a template for protein modeling. The PDB ID: 1COZ which is a prototypical glycerol-3-phosphate cytidyltransferase (GCT) from B. subtilis was used as a template to model CCT. Fifty models of CCT were generated by modeler v9.10
216
S. Shinde et al. / Gene 550 (2014) 214–222
Fig. 2. An overview of the methodology.
(Eswar et al., 2006), and top 5 models were submitted to structure analysis and verification (SAVEs server (Laskowski et al., 1993) and PROSA web server (Wiederstein and Sippl, 2007)). Subsequently one of the models with less disallowed residues in Ramachandran plot was minimized to remove stereochemical clashes explained in Section 2.4. 2.3. Inhibitor Design Inhibitor designing or screening against S. mansoni was carried out using bioactive compounds like shagoal, curcumin and flavonoids. Among flavonoids when the lignan group was attached (flavonolignans) showed better binding activity, hence discussed further. Flavonoids are a group of polyphenolic metabolite compounds that are ubiquitous in fruits, vegetables and several beverages. They are also termed as antioxidants that protect against the damaging effect of reactive oxygen species (Dai and Mumper, 2010). Several biological activities like metal chelation, antioxidation, cell signaling, normal cell cycle regulation, and decreased inflammation are known to be associated with them (Flora, 2009). The 2D structures of four flavonolignans tested against CCT are built with MavinSketchv5.10 (MavinSketch, 2012) (Fig. 3).
Van der Waals interactions of 10 Å. The simulations were further checked for the root mean square deviations (RMSD) and root mean square fluctuations (RMSF). The protein model obtained after simulation was processed (removed water molecules and Na+ ion) to carry out molecular docking study. The inhibitors were docked with the protein using AutoDock Tools (version 1.5.6 rc2). The docking simulations were performed with AutoDock Vina (Trott and Olson, 2010) with the search space of 0.572 Å. Binding pocket of CCT was investigated for the interaction with the dimension of 30 × 34 × 28 Å common for all inhibitors under study. The inhibitor with affinity coordinates was converted to topology files using the ffgmx force field (GROMOS87) by a PRODRG2 server (Schüttelkopf and van Aalten, 2004). The topology (.top) and gromacs file (.gro) file were manually edited to match the number of molecules and further subjected for 4 ns MD simulation with the same conditions as used for protein MD.
2.4. Molecular Dynamics Simulation (MD), Molecular Docking and Post-Docking Simulation The modeled protein was minimized and MD simulation studies were done using GROMACS (GROningen MAchine for Chemical Simulations) software. The protein molecule was relaxed in GROMOS 53a6 force field for refinement, to remove artifacts from the modeled structure (Oostenbrink et al., 2004;Pronk et al., 2013). Explicit SPC216 water molecule and sodium counter ions were used to neutralize the system [the solvation volume of (64 nm)3 with ~ 13824 atoms] and 4 ns stochastic simulation (time step of 2 fs) was performed. The loss of secondary structure was prevented by a harmonic potential (force 10 kcal/mol). Berendsen pressure coupling and tcouple were used for simulation with constant pressure of 1 atm and temperature (300 K). Electrostatic interactions were calculated using the particle-mesh Ewald algorithm with a spacing of 1.0 Å and a cutoff radius for the
Fig. 3. The 2D structures of screened compounds selected as drug candidates against schistosomiasis built with MavinSketchv5.10.
S. Shinde et al. / Gene 550 (2014) 214–222
3. Result and Discussion 3.1. Metabolic Network Analysis The pathway comprises 58 reactions with 82 species and three compartments (nucleus, ER and cytoplasm). In order to get an overview of the eventual links between different pathways the source of the pathway was Glycerone-P from Glycerolipid and O-acetylcholine of Kennedy pathway. The end products of the system are PC, PE, PS 1-acyl-snglycero-3-phosphocholine, lysophosphatidylcholine transferase and phosphatidyl 1D-myo-inositol. The model consists of biochemical reactions where the number of influx is 2 i.e. m = 2 and the number of exchange fluxes is 6. Sensitivity analysis of (cytidine 5′-diphosphocholine) CDP-choline pathway shows that the carrier mediated choline entry
217
into the parasite showed 5.140 as the positive flux and CCT reaction has the largest coefficients in the pathway distinguishing it to be a unique rate-limiting step. The first reaction is catalyzed by choline kinase (CK) by an ATP-dependent phosphorylation of choline, forming phosphocholine, and ADP. In further steps, CCT uses phosphocholine and cytosine triphosphate (CTP) to form high-energy donor CDPcholine with the release of pyrophosphate, this is considered to be the rate-limiting step and hence an important step of the Kennedy pathway. The CDP-choline: 1,2-diacylglycerol cholinephosphotransferase (CPT) advances the final reaction of the pathway, using CDP-choline and a lipid anchor diacylglycerol (DAG) or alkyl–acylglycerol (AAG) to form PC, PE and CMP as byproducts. In S. mansoni CCT is specific for phosphocholine, catalyzing the primary regulatory step of choline incorporation to PC. This is highly stage dependent and is reported in all
Fig. 4. Reconstructed glycerophospholipid pathway of the oval compartment depicts nucleus and the square compartment is the endoplasmic reticulum. The dotted arrow shows Hill–Hinze equation which converts a gene to a product (enzyme, protein) with boundary condition set to true, the dark black line represents Michaelis–Menton equation.
218
S. Shinde et al. / Gene 550 (2014) 214–222
four stages of the parasite (Ancelin et al., 1987) essential to its survival. In the case of humans two pathways namely the Kennedy and demethylation pathway exist for PC biosynthesis whereas in S. mansoni it is only the Kennedy pathway that synthesizes PC. The model shown (Fig. 4) is a typical genome-scale metabolic network model (GEM) which accounts for all known metabolic genes in a glycerophospholipid pathway with their chemical reactions providing mechanistic predictions of how each gene affects the metabolic network function. The metabolites enter the system through boundary pseudoreactions (b) further metabolized in internal reactions (i) with a flux V. Reaction flux is limited by thermodynamics, the velocity of the forward and reverse enzymecatalyzed reactions described as lower and upper boundaries (0 to 10) for the internal fluxes and the external fluxes for unknown biological reactions. The lower and upper limits were set to ±10 to 0 for those metabolites whose fluxes were not known. The known fluxes were set as per the percentage found in the parasite (for e.g. the value 0.25 indicates the percentage of PE viz. 25% and similar with the other molecules). The biological constraints on Glycerone-P was (− 10 to 0), acetylcholine (− 10 to 0), PE (0.25 to 10), lyso-PC (− 10 to 0), acyl-sn-glycero-PI (0.13 to 10), PS (0.15 to 10), betaine (0.1 to 10), choline (0.12 to 10) and PC (0.28 to 10). The optimal metabolic state was accounted using ExPa, 18 reactions were taken into account as both reversible and irreversible with 6 external fluxes, after obtaining the stoichiometry and sensitivity for the same. A detailed understanding of cellular process involves gene regulation, linked to the metabolic pathway in a gene–protein reaction (GPR). The function is optimized with linear programming to maximize a particular objective under a given set of conditions used in FBA to understand metabolite usage and phenotype. This identifies which of the important metabolites when removed or inhibited alters the biomass production (in our case PC, PS and PE). The wild type GEM pathway showed negative influx suggesting the usability of the molecules in production of PC, PS and PE with the maximum 0.28 of PC. The simultaneous knockout of choline, PC and PI shows negative effect on snglycerol 3P and Glycerone 3P. The first enzyme of glycerophospholipid pathway glycerol-3-phosphate dehydrogenase was found to be the drug target in Leishmania mexicana and T. brucei suggesting the utility of this pathway in kinetoplastids (Suresh et al., 2000). PC is seen to have a positive control on 1-acyl-sn-glycero-3 phosphocholine and CDP-choline metabolite with a strong negative control on itself. 1Acyl-sn-glycero-3 phosphocholine is negatively controlled by itself and positively controlled by PC as shown in Supplementary file, Table 3. 3.2. Homology Model for Structure Based Drug Design Metabolic network analysis study identifies CCT as a drug target which alters choline production. CCT exists as a homodimer, in nearly all the cell that expresses it, but the monomeric form is said to be more active and hence is the topic of interest. Four domains are reported in the CCT viz. nuclear targeting domain, catalytic activity, membrane binding and a phosphorylating domain. CCT is regulated in the cells by lipids, reversible phosphorylation, and at the level of mRNA. As seen in Fig. 5a, CTP activates CCT and converts it to CDP and pyrophosphate. CCT regulates the overall biosynthesis of PC, and the activity of CCT is modulated in response to the need of the cell for PC. The sequence alignment shows that ~ 30% of the residues in GCT from B. subtilis are identical to the residues in the catalytic domains of the larger members of the GCT family. 126 residues were modeled out of the total 354 amino acids (aa) with 1–122 aa with E-value of 2.4e−17 that belongs to the CTP_transferase_2 family (Punta et al., 2012) and the rest were avoided due to low-complexity regions (Fig. 5b, c). The level of sequence identity implies that GCT may provide a good template for the catalytic domains to model larger eukaryotic CCT and ECT. The three stretches of conserved residues observed in the alignment are ‘DXXHXGH’, ‘RTXGISTT’ and ‘RYVDEVI’ this is the signature pattern in all CCT. The monomer structure of CCT appears as a typical Rossmann
fold with five β sheets and six α helices. The signature motifs are visible in the loop and in the sheet region Fig. 5d. Further the verification of the structure practiced by Ramachandran plot verifies 3D and Z-score. 91.2% residues (114 amino acids) were observed in core region, and 2 residues (Asp100, Ser106) in generously allowed regions. The rest of the regions were in the favorable regions. The 3D–1D average score indicated that 86% residues have an average score of b 0.2. The G-factor results suggested some disorders in residues in dihedral which were corrected in MD simulation. The local structure and Z-score were within the desired range of −5.7 of X-ray and NMR experiments (Supplementary Fig. 1). 3.3. Inhibitor Design and Molecular Docking The Ramachandran plot of this minimized model showed that 100% of the residues are located in the allowed regions for further analysis. The reference drug used for docking is miltefosine, it is a phospholipid analog initially used against cancer but currently used against leishmaniasis and schistosomiasis. It acts on phospholipid group of alkylphosphocholine present on the surface of the organism. CCT-miltefosine guided to identify the interaction details with the four inhibitors at the active site Fig. 6a. Herbacetin (3,5,7,8 tetrahydroxy-2-4-hydroxyphenyl chromen-4one), a flavonolignan, is known to induce apoptosis in cancer cells (Seo et al., 2011). Reactive oxygen species (ROS) produced by the macrophages is the primary pathway of immune attack against the parasite antioxidants secreted by glutathione peroxidase and thioredoxin perioxidase which are responsible for inducing activated macrophages (Hewitson et al., 2009a, 2009b). The 2-4-hydroxyphenyl chromen ring of herbacetin interacts with CCT and 7,8 hydroxyl group of flavonol. It interacts with Ala9, Gln20, Gly46, Asp92, Thr120, Thr119, and Ser118 in CCT. Gln20, Thr120, Thr119, and Ser118 are the important amino acids/ motifs that have nucleotransferase activity that transfers the cofactor CTP required for phosphocholine catalysis. Thr119, 120 forms H-bond with chromen ring and hydroxyl group of threonine with bond distances of 3.30 Å and 2.95 Å. The 2-amino group of Thr117 interacts with the hydroxyl group of chromen with a distance of 3.09 Å. The catalytic residue Ser118 forms H-bond with 5-hydroxyl residue of herbacetin with a distance of 2.84 Å. The overall affinity score of herbacetin –CCT complex was 7.6 kcal/mol shown in Fig. 6b. Hydnocarpin (IUPAC: 2-[(2R, 3R)-2-(3,4-dihydroxyphenyl)-3(hydroxymethyl)-2,3-dihydro-1, 4-benzodioxin-6-yl]-5,7-dihyroxy4H-chromen-4-one) has strong hypolipidomic effect and antiinflammatory effect primarily in Brucea javanica, Hydnocarpus wightiana, Verbascum sinaiticum, and certain Berberis species (Sharma and Hall, 1991). It is known to be a potent inhibitor with free radical scavenging activity that inhibits Staphylococcus aureus multidrug resistant efflux pump (Guz and Stermitz, 2000). Hydnocarpin and CCT interact with the affinity score of −9.6 kcal/mol forming H-bonds with Asp40, Ser36, Asp71, His14, His17, Thr119 and several Van der Waals bonds with Ser118, Ala72, Thr77, Trp34, Asp7, Gly8, Tyr44, etc. Imidazole ring of His14, 17 interacts with the terminal hydroxyl group forming H-bond with 2.85 and 2.77 Å (Fig. 6c). 3.3.1. Hydnowightin It is a 2-[(2S)-2-{4-hydroxy-2[(E)-2-(3-hydroxyphenyl) ethenyl] phenyl}-2,3-dihydro-1,4-benzodioxin-6-yl]-4H-chromen-4-one compound. Of the flavonolignans mentioned all have the antineoplastic and anti-inflammatory activity. Hydnowightin can be isolated from H. wightiana seeds. They competitively inhibit the binding of CTP with CCT and show the highest affinity of 10.8 kcal/mol (Fig. 6d). Hymenaea palustris leaves exhibit antineoplastic activity and antiinflammatory activity with the presence of Palstatin (IUPAC: 5,7-dihydroxy-2-[8-hydroxy-2-(hydroxymethyl)-3-(3,4,5-trihydoxyphenyl)1,4-benzodioxin-6-yl]-4H-chromen-4-one), flavonolignans in them (Pettit et al., 2003). The benzodoxin ring of Palstatin is most active
S. Shinde et al. / Gene 550 (2014) 214–222
219
Fig. 5. (a) Conversion of phosphocholine to CDP choline with the aid of choline cytidyltransferase, (b) ClustalW alignment of the CCT target and template showing notations for CTP binding site. Identical residues (*), conserved substitutions (:) and semi-conserved substitutions (.) are shown. (c) Homology model of choline cytidyltransferase. (d) The conserved signature motif highlighted is catalytic residues.
than chromen ring and binds CCT with affinity of −9.9 kcal/mol. Several H-bonding hooks are formed with His14, His17, Thr119 and Thr120, the major players for nucleotidyltransferase (Fig. 6e).
3.4. Molecular Simulation and Post-Docking Simulation The homology model subjected for minimization showed the potential energy, the temperature fluctuations and the pressure in the desired range of 300 K, 1 atm pressure respectively. Protein attained stability at 4 ns of simulation observed in RMSD, to measure the structural distance between coordinate sets and RMSF graphs measure conformational variance on single residue than a single value. The post-docking simulation showed that the compounds were located in the active site cavity with higher affinity. Time-dependent Cα RMS deviation (RMSD) provides the global drift of the protein model during the simulation period. The RMSD drift of Cα atoms from the initial protein structure was determined (Fig. 7a). The drift observed for the structure reaches a plateau after ~0.25 ns of simulation and is equal to 0.5 nm.
4. Discussion S. mansoni and its four different stages ensure that metabolism changes multiple times. The present repertoires of drugs are limited for S. mansoni. We utilized the predictive power of ‘systems biology’ to test a hypothesis and derive signals for interactions with and without the knowledge of rate laws, to ensure which enzyme controls the flux and how well does the system cope with variations in demand. GEM was used to study the regulation that controls metabolism and biomass production essential for the survival of the parasite. The biosynthesis of phospholipids is essential in the development as change in the membrane rheology could affect multiple biological events that occur at the cellular membrane level like transporting metabolites, signal transduction events, and interaction of membrane components (Hashimoto, 2001). The built model clarifies the relative importance of various reactions in a metabolic pathway. The modeling strategy serves to elucidate the regulatory mechanisms governing the metabolism of PC for surface protection as well as the mechanism of action of drugs on the membrane biosynthetic pathway and eventual mechanism of resistance.
220
S. Shinde et al. / Gene 550 (2014) 214–222
Fig. 6. Interaction details of the inhibitors chosen for the study at the active site.
Impairment of lipid biosynthesis thus appears as a potential strategy of choice to fight against these parasites. The ability to alter the lipid composition was addressed with the use of naturally screened compounds. Altogether with the findings we show the system that poised to sense the metabolic response and metabolic imbalances, illustrating extensive communication in glycerophospholipid metabolic network. The current model posts the stress on the bilayer curvature which is sensed by CCT governing the degree of membrane association and providing a mechanism for positive and negative regulations. CCT is the exclusively studied target for which homology model and inhibitors were designed of which hydnowightin. Henceforth, systems biology approaches may unveil CCT as an important drug target for pharmacological intervention against schistosomiasis.
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2014.08.031.
Acknowledgement The authors would like to thank the Department of Biotechnology, Government of India (BT/PR3140/BID/7/379/2011), for funding the work. Ms. Sonali Shinde acknowledges the junior research fellowship from DBT, Government of India. We are thankful to Dr. Shekhar C. Mande, Director of NCCS, for supporting the Bioinformatics and High Performance Computing Facility (BHPCF) at NCCS, Pune, India. Softwares available at BHPCF are acknowledged too.
S. Shinde et al. / Gene 550 (2014) 214–222
221
Fig. 7. (a) RMSD shows stability at 4 ns simulation. (b) RMSF shows fluctuation at certain residues found to be identified as catalytic residues. (c) Simulation behavior of CCT and different competitive inhibitors checked.
References Ancelin, M.L., Torpier, G., Vial, H.J., Capron, A., 1987. Choline incorporation by Schistosoma mansoni: distribution of choline metabolites during development and after sexual differentiation. J. Parasitol. 73 (3), 530–535. Ancelin, M.L., Calas, M., Bonhoure, A., Herbute, S., Vial, H.J., 2003. In vivo antimalarial activities of mono- and bis quaternary ammonium salts interfering with Plasmodium phospholipid metabolism. Antimicrob. Agents Chemother. 47, 2598–2605. Bell, S.L., Palsson, B.Ø., 2005. Expa: a program for calculating extreme pathways in biochemical reaction networks. Bioinforma. Appl. Notes 21, 1739–1740. Berriman, M., et al., 2009. The genome of the blood fluke Schistosoma mansoni. Nature 460, 352–358. Caspi, R., Altman, T., Dale, J.M., Dreher, K., Fulcher, C.A., Gilham, F., Kaipa, P., Karthikeyan, A.S., Kothari, A., Krummenacker, M., Latendresse, M., Mueller, L.A., Paley, S., Popescu, L., Pujar, A., Shearer, A.G., Zhang, P., Karp, P.D., 2010. The MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res. 38, D473–D479. Cui, Z., Houweling, M., 2002. Phosphatidylcholine and cell death. Biochim. Biophys. Acta 1585, 87–96. Dai, J., Mumper, R.J., 2010. Plant phenolics: extraction, analysis and their antioxidant and anticancer properties. Molecules 15, 7313–7352. Doenhoff, M.J., Cioli, D., Utzinger, J., 2008. Praziquantel: mechanism of action, resistance and new derivatives for Schistosomiasis. Curr. Opin. Infect. Dis. 21, 659–667. Eswar, N., Marti-Renom, M.A., Webb, B., Madhusudhan, M.S., Eramian, D., Shen, M., Pieper, U., Sali, A., 2006. Comparative protein structure modeling with MODELLER. Current Protocols in Bioinformatics. Supplement 15. John Wiley & Sons, Inc., pp. 5. 6.1–5.6.30. Flora, S.J., 2009. Structural, chemical and biological aspects of antioxidants for strategies against metal and metalloid exposure. Oxid. Med. Cell Longev. 4, 191–206. Funahashi, A., Matsuoka, Y., Jouraku, A., Morohashi, M., Kikuchi, N., Kitano, H., 2008. Cell designer 3.5: a versatile modeling tool for biochemical networks. Proc. IEEE 96 (8), 1254–1265. Furlong, S.T., Caulfield, J.P., 1989. Schistosoma mansoni: synthesis and release of phospholipids, lysophospholipids, and neutral lipids by schistosomula. Exp. Parasitol. 69 (1), 65–77. Gibellini, F., Smith, T.K., 2010. The Kennedy pathway—de novo synthesis of phosphatidylethanolamine and phosphatidylcholine. IUBMB Life 62 (6), 414–428. Golan, D.E., Brown, C.S., Catherine, M.L.C., Furlong, S.T., Caulfield, J.P., 1986. Schistosomula of Schistosoma mansoni use lysophosphatidylcholine to lyse adherent human red blood cells and immobilize red cell membrane components. J. Cell Biol. 103, 819–828.
Guz, N.R., Stermitz, F.R., 2000. Synthesis and structures of regioisomeric hydnocarpintype flavonolignans. J. Nat. Prod. 63 (8), 1140–1145. Hewitson, J.P., Grainger, J.R., Maizels, R.M., 2009a. Helminth immunoregulation: the role of parasite secreted proteins in modulating host immunity. Mol. Biochem. Parasitol. 167 (1–9), 1–11. Hewitson, J.P., Grainger, J.R., Maizels, R.M., 2009b. Helminth immunoregulation: the role of parasite secreted proteins in modulating host immunity. Mol. Biochem. Parasitol. 167, 1–11. Hoops, S., Sahle, S., Gauges, R., Lee, C., Pahle, J., Simus, N., Singhal, M., Xu, L., Mendes, P., Kummer, U., 2006. COPASI: a COmplex PAthway SImulator. Bioinformatics 22, 3067–3074. Kanehisa, M., Goto, S., 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucl. Acids Res. 28, 27–30. Kitano, H., 2002. Systems biology: a brief overview. Science 295 (5560), 1662–1664. Kitano, H., et al., 2005. Using process diagram for the graphical representation of biological networks. Nat. Biotechnol. 23 (8), 961–966. Laskowski, R.A., MacArthur, M.W., Moss, D.S., Thornton, J.M., 1993. PROCHECK — a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, 283–291. Lee, S.J., Lee, D.Y., Kim, T.Y., Kim, B.H., Lee, J., Lee, S.Y., 2005. Metabolic engineering of Escherichia coli for enhanced production of succinic acid, based on genome comparison and in silico gene knockout simulation. Environ. Microbiol. 71, 7880–7887. Lewis, R.N., Pohle, W., McElhaney, R.N., 1996. The interfacial structure of phospholipid bilayers: differential scanning calorimetry and Fourier transform infrared spectroscopic studies of 1,2-dipalmitoyl-sn-glycero-3-phosphorylcholine and its dialkyl and acyl–alkyl analogs. Biophys. J. 70 (6), 2736–2746. Loukas, A., Maizels, R.M., 2000. Helminth C-type lectins and host–parasite interactions. Parasitology 16, 333–340. Marechal, Eric, Riou, Mickael, Kerboeuf, Dominique, Beugnet, Frederic, Chaminade, Pierre, Loiseau, Philippe M., 2011. Membrane lipidomics for the discovery of new antiparasitic drug targets. Trends Parasitol. 27 (11), 496–504. MavinSketch 5.10.1 ChemAxon Ltd. 1998–2012. http://www.chemaxon.com. Oostenbrink, C., Villa, A., Mark, A.E., van Gunsteren, W.F., 2004. A Biomolecular Force Field Based on the Free Enthalpy of Hydration and Solvation: The GROMOS Force-Field Parameter Sets 53A5 and 53A6. pp. 1656–1676. Pettit, G.R., Meng, Y., Stevenson, C.A., Doubek, D.L., Knight, J.C., Cichacz, Z., Pettit, R.K., Chapuis, J.C., Schmidt, J.M., 2003. Isolation and structure of palstatin from the Amazon tree Hymeneae palustris (1). J. Nat. Prod. 66 (2), 259–262. Pronk, Sander, Páll, Szilard, Schulz, Roland, Larsson, Per, Bjelkmar, Pär, Apostolov, Rossen, Shirts, Michael R., Smith, Jeremy C., Kasson, Peter M., van der Spoel, David, Hess, Berk,
222
S. Shinde et al. / Gene 550 (2014) 214–222
Lindahl, Erik, 2013. GROMACS 4.5: a high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics 29 (7), 845–854. Punta, M., Coggill, P.C., Eberhardt, R.Y., Mistry, J., Tate, J., Boursnell, C., Pang, N., Forslund, K., G., Clements, J., Heger, A., Holm, L., Sonnhammer, E.L., Eddy, S.R., Bateman, A., Finn, R.D., 2012. The Pfam protein families database. Nucleic Acids Res. 40, D290–D301. Ramakrishnan, S., Docampo, M.D., MacRae, J.I., Pujol, F.M., Brooks, C.F., van Dooren, G.G., Kalervo Hiltunen, J., Kastaniotis, A.J., McConville, M.J., Striepen, B., 2012. Apicoplast and endoplasmic reticulum cooperate in fatty acid biosynthesis in apicomplexan parasite Toxoplasma gondii. J. Biol. Chem. 287, 4957–4971. Schmid-Hempel, P., 2009. Immune defense, parasite evasion strategies and their relevance for ‘macroscopic phenomena’ such as virulence. Philos. Trans. R. Soc. B Biol. Sci. 364, 85–98. Schüttelkopf, A.W., van Aalten, D.M.F., 2004. PRODRG: a tool for high-throughput crystallography of protein–ligand complexes. Acta Crystallogr. D60, 1355–1363. Seo, M., Kim, J.-D., Neau, D., Sehgal, I., Lee, Y.-H., 2011. Structure-based development of small molecule PFKFB3 inhibitors: a framework for potential cancer therapeutic agents targeting the Warburg effect. PLoS One 6 (9), e24179. http://dx.doi.org/10. 1371/journal.pone.0024179.
Sharma, D.K., Hall, I.H., 1991. Hypolipidemic, anti-inflammatory, and antineoplastic activity and cytotoxicity of flavonolignans isolated from Hydnocarpus wightiana seeds. J. Nat. Prod. 54 (5), 1298–1302. Suresh, S., Turley, S., Opperdoes, F.R., Michels, P.A., Hol, W.G., 2000. A potential target enzyme for trypanocidal drugs revealed by the crystal structure of NAD-dependent glycerol-3-phosphate dehydrogenase from Leishmania mexicana. Structure 15, 541–552. Trott, O., Olson, A.J., 2010. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading. J. Comput. Chem. 31, 455–461. Useh, M.F., 2013. Parasitic Diseases — Schistosomiasis. InTech http://dx.doi.org/10.5772/ 53553. Wiederstein, M., Sippl, M.J., 2012. ProSA-Web: Interactive Web Service for the Recognition of Errors in Three-Dimensional Structures of Proteins. Nucleic Acids Res. 35 (Web Server issue) W407–10. Wu, J., Xu, W., Ming, Z., Dong, H., Tang, H., et al., 2010. Metabolic changes reveal the development of schistosomiasis in mice. PLoS Negl. Trop. Dis. 4 (8), e807. Young, B.W., Podesta, R.B., 1982. Major phospholipids and phosphatidylcholine synthesis in adult Schistosoma mansoni. Mol. Biochem. Parasitol. 5 (3), 165–172.
Contents lists available at ScienceDirect
Gene journal homepage: www.elsevier.com/locate/gene
Regulatory networks, genes and glycerophospholipid biosynthesis pathway in schistosomiasis: A systems biology view for pharmacological intervention Sonali Shinde, Milsee Mol, Shailza Singh ⁎ National Centre for Cell Science, NCCS Complex, Ganeshkhind, Pune University Campus, Pune 411007, India
a r t i c l e
i n f o
Article history: Received 16 April 2014 Received in revised form 5 August 2014 Accepted 13 August 2014 Available online 19 August 2014 Keywords: Metabolic network Schistosomiasis Glycerophospholipid biosynthesis Cytidyltransferase Inhibitor designing
a b s t r a c t Understanding network topology through embracing the global dynamical regulation of genes in an active state space rather than traditional one-gene–one trait approach facilitates the rational drug development process. Schistosomiasis, a neglected tropical disease, has glycerophospholipids as abundant molecules present on its surface. Lack of effective clinical solutions to treat pathogens encourages us to carry out systems-level studies that could contribute to the development of an effective therapy. Development of a strategy for identifying drug targets by combined genome-scale metabolic network and essentiality analyses through in silico approaches provides tantalizing opportunity to investigate the role of protein/substrate metabolism. A genome-scale metabolic network model reconstruction represents choline–phosphate cytidyltransferase as the rate limiting enzyme and regulates the rate of phosphatidylcholine (PC) biosynthesis. The uptake of choline was regulated by choline concentration, promoting the regulation of phosphocholine synthesis. In Schistosoma, the change in developmental stage could result from the availability of choline, hampering its developmental cycle. There are no structural reports for this protein. In order to inhibit the activity of choline–phosphate cytidyltransferase (CCT), it was modeled by homology modeling using 1COZ as the template from Bacillus subtilis. The transition-state stabilization and catalytic residues were mapped as ‘HXGH’ and ‘RTEGISTT’ motif. CCT catalyzes the formation of CDP-choline from phosphocholine in which nucleotidyltransferase adds CTP to phosphocholine. The presence of phosphocholine permits the parasite to survive in an immunologically hostile environment. This feature endeavors development of an inhibitor specific for cytidyltransferase in Schistosoma. Flavonolignans were used to inhibit this activity in which hydnowightin showed the highest affinity as compared to miltefosine. © 2014 Elsevier B.V. All rights reserved.
1. Introduction A prerequisite for understanding any disease is acquiring knowledge of its component interactions between the host–parasite systems. This is possible with the available sequence and omics' data obtained through various high throughput platforms. The volumes of data generated can be analyzed to identify individual component interactions that make up a system, which can be analyzed within a mathematical framework to identify possible targets for new drug discovery (Kitano, 2002).
Abbreviations: PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine; GPC, glyceryl phosphorylcholine; CTP, cytidine triphosphate; CCT, choline–phosphate cytidyltransferase; CK, choline kinase; ADP, adenine diphosphate; DAG, diacylglycerol; AAG, alkyl-acylglycerol; IL, interleukin; DC, dendritic cells; Lyso-PC, lysophosphatidylcholine; FBA, flux balance analysis; GEM, genome-scale model; FBA, flux based analysis; MD, Molecular Dynamic; SPC, Simple Point Charge; NVT, Number of moles, Volume, Temperature; NPT, Number of moles, Pressure, Temperature; PME, Particle Mesh Ewald; RMSD, root mean square deviations; RMSF, root mean square fluctuations; LGA, Lamarckian Genetic Algorithm; PE, potential energy. ⁎ Corresponding author. E-mail addresses: [email protected], [email protected] (S. Singh).
http://dx.doi.org/10.1016/j.gene.2014.08.031 0378-1119/© 2014 Elsevier B.V. All rights reserved.
Under this backdrop our work represents abstracting the complex lipid biosynthetic process in schistosomiasis for novel drug target identification. Schistosomiasis is a tropical disease caused by trematode belonging to the species Schistosoma. Control of schistosomiasis is challenging as more than 600 million people in 76 developing countries are at risk and with 200 million people already affected by this deadly disease (Berriman et al., 2009). The intestinal form of the disease is caused by Schistosoma mansoni. The availability of genomic data of the intestinal form of the parasite opened new avenues to understand disease novel pathophysiology that can be harnessed to progression of the disease (Useh, 2013). The parasite enters human, a definitive host (snail: Biomphalaria glabrata, being the intermediate host) by dissolving the skin barrier and migrates towards the liver and lungs to lay eggs. These eggs are lodged within the blood capillaries which is a major cause of pathology. Praziquantel (Biltricide®) is the effective monotherapy to control schistosomiasis but reports of drug resistance and tolerance have limited its use (Doenhoff et al., 2008). Designing new drug becomes difficult due to the rapid proliferation and changing surface variants of the parasite. One of the important surface components involved in renewal of membrane complex is associated with lipid
S. Shinde et al. / Gene 550 (2014) 214–222
metabolism (Ramakrishnan et al., 2012). Impairment reduces these pathogens' ability to proceed in its infectious life cycle and therefore could be a possible source of drug target (Marechal, 2011). Dissection of lipid metabolism using lipidomics in S. mansoni can provide cues on the alteration in the lipid pool of parasites that is associated with drug resistance (Lewis et al., 1996). Metabolic profile of S. mansoni infected mice liver indicates lower levels of glucose and glycogen but higher levels of choline metabolites, (phosphatidylcholine (PC), glyceryl phosphorylcholine (GPC), and alanine) than in controls suggesting that Schistosoma can suppress the metabolic response of the host (Wu et al., 2010). Many evidence suggests that lipids are essential in the life cycle of the parasite, and its regulation mechanisms are largely unknown (Loukas and Maizels, 2000). Glycerophospholipids are one such class of lipids in Schistosoma which accounts for PC (25%), phosphatidylserine (PS—15%) and phosphatidylethanolamine (PE— 8%). Fatty acids are differentially incorporated into the various phospholipid classes, principally into PC and, to a lesser extent, into PE, lyso-PC, and PS (Furlong and Caulfield, 1989). They show immune modulation properties e.g. PS induces dendritic cells (DC) to polarize IL-4/IL10 producing T-cells. Whereas lysophosphatidylserine (lyso-PS) specifies DC to induce IL-10 secreting regulatory T cells, thus swaying the immune system away from a protective Th1 immune response (Hewitson et al., 2009a, 2009b). Lysophosphatidylcholine (Lyso-PC) is known to adhere, immobilize and lyse the red blood cell to alter the cell plasma membrane, thereby neutralizing the attacking host cells. It promotes membrane fusion with the host resulting in the acquisition of host membrane components by the parasite. The presence of lysoPC may alter antigen presentation and immune recognition of the parasite antigens (Golan et al., 1986). These immune modulating lipids confer on the parasite the major adaptation mechanism, allowing survival in an immunologically hostile environment within the host (Schmid-Hempel, 2009). Also PCs are essential in the development of Schistosoma, as large amount is needed for a rapid membrane renewal and to renew its outer bilayer surface. The synthesis of PC in S. mansoni adults occurs via the Kennedy pathway (de novo synthesis of PC) consisting of three enzymatic steps shown in Fig. 1 (Young and Podesta, 1982). In protozoan parasites like Trypanosoma brucei, Plasmodium falciparum Kennedy pathways have already been validated as a drug
215
target (Ancelin et al., 2003; Cui and Houweling, 2002; Gibellini and Smith, 2010). In order to understand the highly redundant phospholipid pathway and to establish a mechanism to identify newer drug targets a kinetic model was developed in S. mansoni. Lipid mining connects the lipid profiles with the metabolic pathways during different stages of disease progression. This may allow deciphering the complexities and regulatory mechanisms involved in maintaining lipid homeostasis in diseased condition. Potential drug targets can also be identified that may be important regulatory hubs in the lipid biosynthesis. Designing small molecule inhibitors against these targets may be valuable ‘schistosomicides’. Our current understanding of Kennedy pathway in Schistosoma development and pathogenesis could help to develop anti-schistosomal therapies targeting lipid metabolic pathways. We have attempted to delineate lipid metabolism specifically PC biosynthetic pathway in parasite and identify key regulatory enzymes. These were screened against natural compounds like flavonolignan derivatives as potential lead compounds for possible drug design. 2. Material and Methods 2.1. Metabolic Network Analysis To address the current challenge of metabolic network biology and understand how organisms adapt to their environment to maintain optimal growth in changing environmental conditions the mathematical model was optimized for PC biomass production. The built model represents a schematic way to unravel the molecular components that underlie cellular processes with the available concentration of metabolites and kinetic data from KEGG database (Kanehisa and Goto, 2000) and Biocyc database (Caspi et al., 2010). The metabolic network was designed using CellDesigner v4.3 (Funahashi et al., 2008; Kitano et al., 2005). The detail work flow is shown in Fig. 2. Reactions in the metabolic network are connected via gene–protein-reaction (GPR) mechanism where the cofactors are considered to be independent of time and thus are not directly considered for the study. The boundary conditions of genes are constant. The reactions in the network are under the control of several kinetic laws like Hill–Hinze, convenience kinetics, and Michaelis–Menten (MM). Numerical simulation of the built model identifies essential genes/ metabolites vital for cell growth under the imposed constraint. The sensitivity of the flux and the topology of the network are studied by summarizing the stoichiometric matrix using a Copasi tool (Hoops et al., 2006). The stoichiometry of the reaction network is described mathematically as a stoichiometric matrix with rows representing reactants and columns corresponding to the reaction deduced by flux balance analysis (FBA) and ExPa (Bell and Palsson, 2005). In ExPa, the topology of the metabolic network is summarized as a stoichiometric coefficient matrix (Sij) in which, the rate of change of concentration of a metabolite is the sum of metabolite flow to and from the metabolite represented as Sij b 0 when consumed and Sij N 0 when produced. An Optim tool box of Matlab version 7.12.0.35 (R2011a) finds the minimum and maximum objective parameters required to get the optimal solution of a linear equation determined by the feasible boundary region (Lee et al., 2005). 2.2. Sequence Retrieval, Model Building and Evaluation
Fig. 1. Biosynthesis of phosphatidylcholine.
The target identified from network analysis was modeled by homology modeling protocol to get an insight into the functional characteristic of the protein. The sequence of choline–phosphate cytidyltransferase (CCT) was retrieved from Uniprot database (Uniprot ID: G4VD67_SCHMA) and blast against RCSB PDB (The Research Collaboratory for Structural Bioinformatics Protein Data Bank) to obtain a template for protein modeling. The PDB ID: 1COZ which is a prototypical glycerol-3-phosphate cytidyltransferase (GCT) from B. subtilis was used as a template to model CCT. Fifty models of CCT were generated by modeler v9.10
216
S. Shinde et al. / Gene 550 (2014) 214–222
Fig. 2. An overview of the methodology.
(Eswar et al., 2006), and top 5 models were submitted to structure analysis and verification (SAVEs server (Laskowski et al., 1993) and PROSA web server (Wiederstein and Sippl, 2007)). Subsequently one of the models with less disallowed residues in Ramachandran plot was minimized to remove stereochemical clashes explained in Section 2.4. 2.3. Inhibitor Design Inhibitor designing or screening against S. mansoni was carried out using bioactive compounds like shagoal, curcumin and flavonoids. Among flavonoids when the lignan group was attached (flavonolignans) showed better binding activity, hence discussed further. Flavonoids are a group of polyphenolic metabolite compounds that are ubiquitous in fruits, vegetables and several beverages. They are also termed as antioxidants that protect against the damaging effect of reactive oxygen species (Dai and Mumper, 2010). Several biological activities like metal chelation, antioxidation, cell signaling, normal cell cycle regulation, and decreased inflammation are known to be associated with them (Flora, 2009). The 2D structures of four flavonolignans tested against CCT are built with MavinSketchv5.10 (MavinSketch, 2012) (Fig. 3).
Van der Waals interactions of 10 Å. The simulations were further checked for the root mean square deviations (RMSD) and root mean square fluctuations (RMSF). The protein model obtained after simulation was processed (removed water molecules and Na+ ion) to carry out molecular docking study. The inhibitors were docked with the protein using AutoDock Tools (version 1.5.6 rc2). The docking simulations were performed with AutoDock Vina (Trott and Olson, 2010) with the search space of 0.572 Å. Binding pocket of CCT was investigated for the interaction with the dimension of 30 × 34 × 28 Å common for all inhibitors under study. The inhibitor with affinity coordinates was converted to topology files using the ffgmx force field (GROMOS87) by a PRODRG2 server (Schüttelkopf and van Aalten, 2004). The topology (.top) and gromacs file (.gro) file were manually edited to match the number of molecules and further subjected for 4 ns MD simulation with the same conditions as used for protein MD.
2.4. Molecular Dynamics Simulation (MD), Molecular Docking and Post-Docking Simulation The modeled protein was minimized and MD simulation studies were done using GROMACS (GROningen MAchine for Chemical Simulations) software. The protein molecule was relaxed in GROMOS 53a6 force field for refinement, to remove artifacts from the modeled structure (Oostenbrink et al., 2004;Pronk et al., 2013). Explicit SPC216 water molecule and sodium counter ions were used to neutralize the system [the solvation volume of (64 nm)3 with ~ 13824 atoms] and 4 ns stochastic simulation (time step of 2 fs) was performed. The loss of secondary structure was prevented by a harmonic potential (force 10 kcal/mol). Berendsen pressure coupling and tcouple were used for simulation with constant pressure of 1 atm and temperature (300 K). Electrostatic interactions were calculated using the particle-mesh Ewald algorithm with a spacing of 1.0 Å and a cutoff radius for the
Fig. 3. The 2D structures of screened compounds selected as drug candidates against schistosomiasis built with MavinSketchv5.10.
S. Shinde et al. / Gene 550 (2014) 214–222
3. Result and Discussion 3.1. Metabolic Network Analysis The pathway comprises 58 reactions with 82 species and three compartments (nucleus, ER and cytoplasm). In order to get an overview of the eventual links between different pathways the source of the pathway was Glycerone-P from Glycerolipid and O-acetylcholine of Kennedy pathway. The end products of the system are PC, PE, PS 1-acyl-snglycero-3-phosphocholine, lysophosphatidylcholine transferase and phosphatidyl 1D-myo-inositol. The model consists of biochemical reactions where the number of influx is 2 i.e. m = 2 and the number of exchange fluxes is 6. Sensitivity analysis of (cytidine 5′-diphosphocholine) CDP-choline pathway shows that the carrier mediated choline entry
217
into the parasite showed 5.140 as the positive flux and CCT reaction has the largest coefficients in the pathway distinguishing it to be a unique rate-limiting step. The first reaction is catalyzed by choline kinase (CK) by an ATP-dependent phosphorylation of choline, forming phosphocholine, and ADP. In further steps, CCT uses phosphocholine and cytosine triphosphate (CTP) to form high-energy donor CDPcholine with the release of pyrophosphate, this is considered to be the rate-limiting step and hence an important step of the Kennedy pathway. The CDP-choline: 1,2-diacylglycerol cholinephosphotransferase (CPT) advances the final reaction of the pathway, using CDP-choline and a lipid anchor diacylglycerol (DAG) or alkyl–acylglycerol (AAG) to form PC, PE and CMP as byproducts. In S. mansoni CCT is specific for phosphocholine, catalyzing the primary regulatory step of choline incorporation to PC. This is highly stage dependent and is reported in all
Fig. 4. Reconstructed glycerophospholipid pathway of the oval compartment depicts nucleus and the square compartment is the endoplasmic reticulum. The dotted arrow shows Hill–Hinze equation which converts a gene to a product (enzyme, protein) with boundary condition set to true, the dark black line represents Michaelis–Menton equation.
218
S. Shinde et al. / Gene 550 (2014) 214–222
four stages of the parasite (Ancelin et al., 1987) essential to its survival. In the case of humans two pathways namely the Kennedy and demethylation pathway exist for PC biosynthesis whereas in S. mansoni it is only the Kennedy pathway that synthesizes PC. The model shown (Fig. 4) is a typical genome-scale metabolic network model (GEM) which accounts for all known metabolic genes in a glycerophospholipid pathway with their chemical reactions providing mechanistic predictions of how each gene affects the metabolic network function. The metabolites enter the system through boundary pseudoreactions (b) further metabolized in internal reactions (i) with a flux V. Reaction flux is limited by thermodynamics, the velocity of the forward and reverse enzymecatalyzed reactions described as lower and upper boundaries (0 to 10) for the internal fluxes and the external fluxes for unknown biological reactions. The lower and upper limits were set to ±10 to 0 for those metabolites whose fluxes were not known. The known fluxes were set as per the percentage found in the parasite (for e.g. the value 0.25 indicates the percentage of PE viz. 25% and similar with the other molecules). The biological constraints on Glycerone-P was (− 10 to 0), acetylcholine (− 10 to 0), PE (0.25 to 10), lyso-PC (− 10 to 0), acyl-sn-glycero-PI (0.13 to 10), PS (0.15 to 10), betaine (0.1 to 10), choline (0.12 to 10) and PC (0.28 to 10). The optimal metabolic state was accounted using ExPa, 18 reactions were taken into account as both reversible and irreversible with 6 external fluxes, after obtaining the stoichiometry and sensitivity for the same. A detailed understanding of cellular process involves gene regulation, linked to the metabolic pathway in a gene–protein reaction (GPR). The function is optimized with linear programming to maximize a particular objective under a given set of conditions used in FBA to understand metabolite usage and phenotype. This identifies which of the important metabolites when removed or inhibited alters the biomass production (in our case PC, PS and PE). The wild type GEM pathway showed negative influx suggesting the usability of the molecules in production of PC, PS and PE with the maximum 0.28 of PC. The simultaneous knockout of choline, PC and PI shows negative effect on snglycerol 3P and Glycerone 3P. The first enzyme of glycerophospholipid pathway glycerol-3-phosphate dehydrogenase was found to be the drug target in Leishmania mexicana and T. brucei suggesting the utility of this pathway in kinetoplastids (Suresh et al., 2000). PC is seen to have a positive control on 1-acyl-sn-glycero-3 phosphocholine and CDP-choline metabolite with a strong negative control on itself. 1Acyl-sn-glycero-3 phosphocholine is negatively controlled by itself and positively controlled by PC as shown in Supplementary file, Table 3. 3.2. Homology Model for Structure Based Drug Design Metabolic network analysis study identifies CCT as a drug target which alters choline production. CCT exists as a homodimer, in nearly all the cell that expresses it, but the monomeric form is said to be more active and hence is the topic of interest. Four domains are reported in the CCT viz. nuclear targeting domain, catalytic activity, membrane binding and a phosphorylating domain. CCT is regulated in the cells by lipids, reversible phosphorylation, and at the level of mRNA. As seen in Fig. 5a, CTP activates CCT and converts it to CDP and pyrophosphate. CCT regulates the overall biosynthesis of PC, and the activity of CCT is modulated in response to the need of the cell for PC. The sequence alignment shows that ~ 30% of the residues in GCT from B. subtilis are identical to the residues in the catalytic domains of the larger members of the GCT family. 126 residues were modeled out of the total 354 amino acids (aa) with 1–122 aa with E-value of 2.4e−17 that belongs to the CTP_transferase_2 family (Punta et al., 2012) and the rest were avoided due to low-complexity regions (Fig. 5b, c). The level of sequence identity implies that GCT may provide a good template for the catalytic domains to model larger eukaryotic CCT and ECT. The three stretches of conserved residues observed in the alignment are ‘DXXHXGH’, ‘RTXGISTT’ and ‘RYVDEVI’ this is the signature pattern in all CCT. The monomer structure of CCT appears as a typical Rossmann
fold with five β sheets and six α helices. The signature motifs are visible in the loop and in the sheet region Fig. 5d. Further the verification of the structure practiced by Ramachandran plot verifies 3D and Z-score. 91.2% residues (114 amino acids) were observed in core region, and 2 residues (Asp100, Ser106) in generously allowed regions. The rest of the regions were in the favorable regions. The 3D–1D average score indicated that 86% residues have an average score of b 0.2. The G-factor results suggested some disorders in residues in dihedral which were corrected in MD simulation. The local structure and Z-score were within the desired range of −5.7 of X-ray and NMR experiments (Supplementary Fig. 1). 3.3. Inhibitor Design and Molecular Docking The Ramachandran plot of this minimized model showed that 100% of the residues are located in the allowed regions for further analysis. The reference drug used for docking is miltefosine, it is a phospholipid analog initially used against cancer but currently used against leishmaniasis and schistosomiasis. It acts on phospholipid group of alkylphosphocholine present on the surface of the organism. CCT-miltefosine guided to identify the interaction details with the four inhibitors at the active site Fig. 6a. Herbacetin (3,5,7,8 tetrahydroxy-2-4-hydroxyphenyl chromen-4one), a flavonolignan, is known to induce apoptosis in cancer cells (Seo et al., 2011). Reactive oxygen species (ROS) produced by the macrophages is the primary pathway of immune attack against the parasite antioxidants secreted by glutathione peroxidase and thioredoxin perioxidase which are responsible for inducing activated macrophages (Hewitson et al., 2009a, 2009b). The 2-4-hydroxyphenyl chromen ring of herbacetin interacts with CCT and 7,8 hydroxyl group of flavonol. It interacts with Ala9, Gln20, Gly46, Asp92, Thr120, Thr119, and Ser118 in CCT. Gln20, Thr120, Thr119, and Ser118 are the important amino acids/ motifs that have nucleotransferase activity that transfers the cofactor CTP required for phosphocholine catalysis. Thr119, 120 forms H-bond with chromen ring and hydroxyl group of threonine with bond distances of 3.30 Å and 2.95 Å. The 2-amino group of Thr117 interacts with the hydroxyl group of chromen with a distance of 3.09 Å. The catalytic residue Ser118 forms H-bond with 5-hydroxyl residue of herbacetin with a distance of 2.84 Å. The overall affinity score of herbacetin –CCT complex was 7.6 kcal/mol shown in Fig. 6b. Hydnocarpin (IUPAC: 2-[(2R, 3R)-2-(3,4-dihydroxyphenyl)-3(hydroxymethyl)-2,3-dihydro-1, 4-benzodioxin-6-yl]-5,7-dihyroxy4H-chromen-4-one) has strong hypolipidomic effect and antiinflammatory effect primarily in Brucea javanica, Hydnocarpus wightiana, Verbascum sinaiticum, and certain Berberis species (Sharma and Hall, 1991). It is known to be a potent inhibitor with free radical scavenging activity that inhibits Staphylococcus aureus multidrug resistant efflux pump (Guz and Stermitz, 2000). Hydnocarpin and CCT interact with the affinity score of −9.6 kcal/mol forming H-bonds with Asp40, Ser36, Asp71, His14, His17, Thr119 and several Van der Waals bonds with Ser118, Ala72, Thr77, Trp34, Asp7, Gly8, Tyr44, etc. Imidazole ring of His14, 17 interacts with the terminal hydroxyl group forming H-bond with 2.85 and 2.77 Å (Fig. 6c). 3.3.1. Hydnowightin It is a 2-[(2S)-2-{4-hydroxy-2[(E)-2-(3-hydroxyphenyl) ethenyl] phenyl}-2,3-dihydro-1,4-benzodioxin-6-yl]-4H-chromen-4-one compound. Of the flavonolignans mentioned all have the antineoplastic and anti-inflammatory activity. Hydnowightin can be isolated from H. wightiana seeds. They competitively inhibit the binding of CTP with CCT and show the highest affinity of 10.8 kcal/mol (Fig. 6d). Hymenaea palustris leaves exhibit antineoplastic activity and antiinflammatory activity with the presence of Palstatin (IUPAC: 5,7-dihydroxy-2-[8-hydroxy-2-(hydroxymethyl)-3-(3,4,5-trihydoxyphenyl)1,4-benzodioxin-6-yl]-4H-chromen-4-one), flavonolignans in them (Pettit et al., 2003). The benzodoxin ring of Palstatin is most active
S. Shinde et al. / Gene 550 (2014) 214–222
219
Fig. 5. (a) Conversion of phosphocholine to CDP choline with the aid of choline cytidyltransferase, (b) ClustalW alignment of the CCT target and template showing notations for CTP binding site. Identical residues (*), conserved substitutions (:) and semi-conserved substitutions (.) are shown. (c) Homology model of choline cytidyltransferase. (d) The conserved signature motif highlighted is catalytic residues.
than chromen ring and binds CCT with affinity of −9.9 kcal/mol. Several H-bonding hooks are formed with His14, His17, Thr119 and Thr120, the major players for nucleotidyltransferase (Fig. 6e).
3.4. Molecular Simulation and Post-Docking Simulation The homology model subjected for minimization showed the potential energy, the temperature fluctuations and the pressure in the desired range of 300 K, 1 atm pressure respectively. Protein attained stability at 4 ns of simulation observed in RMSD, to measure the structural distance between coordinate sets and RMSF graphs measure conformational variance on single residue than a single value. The post-docking simulation showed that the compounds were located in the active site cavity with higher affinity. Time-dependent Cα RMS deviation (RMSD) provides the global drift of the protein model during the simulation period. The RMSD drift of Cα atoms from the initial protein structure was determined (Fig. 7a). The drift observed for the structure reaches a plateau after ~0.25 ns of simulation and is equal to 0.5 nm.
4. Discussion S. mansoni and its four different stages ensure that metabolism changes multiple times. The present repertoires of drugs are limited for S. mansoni. We utilized the predictive power of ‘systems biology’ to test a hypothesis and derive signals for interactions with and without the knowledge of rate laws, to ensure which enzyme controls the flux and how well does the system cope with variations in demand. GEM was used to study the regulation that controls metabolism and biomass production essential for the survival of the parasite. The biosynthesis of phospholipids is essential in the development as change in the membrane rheology could affect multiple biological events that occur at the cellular membrane level like transporting metabolites, signal transduction events, and interaction of membrane components (Hashimoto, 2001). The built model clarifies the relative importance of various reactions in a metabolic pathway. The modeling strategy serves to elucidate the regulatory mechanisms governing the metabolism of PC for surface protection as well as the mechanism of action of drugs on the membrane biosynthetic pathway and eventual mechanism of resistance.
220
S. Shinde et al. / Gene 550 (2014) 214–222
Fig. 6. Interaction details of the inhibitors chosen for the study at the active site.
Impairment of lipid biosynthesis thus appears as a potential strategy of choice to fight against these parasites. The ability to alter the lipid composition was addressed with the use of naturally screened compounds. Altogether with the findings we show the system that poised to sense the metabolic response and metabolic imbalances, illustrating extensive communication in glycerophospholipid metabolic network. The current model posts the stress on the bilayer curvature which is sensed by CCT governing the degree of membrane association and providing a mechanism for positive and negative regulations. CCT is the exclusively studied target for which homology model and inhibitors were designed of which hydnowightin. Henceforth, systems biology approaches may unveil CCT as an important drug target for pharmacological intervention against schistosomiasis.
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.gene.2014.08.031.
Acknowledgement The authors would like to thank the Department of Biotechnology, Government of India (BT/PR3140/BID/7/379/2011), for funding the work. Ms. Sonali Shinde acknowledges the junior research fellowship from DBT, Government of India. We are thankful to Dr. Shekhar C. Mande, Director of NCCS, for supporting the Bioinformatics and High Performance Computing Facility (BHPCF) at NCCS, Pune, India. Softwares available at BHPCF are acknowledged too.
S. Shinde et al. / Gene 550 (2014) 214–222
221
Fig. 7. (a) RMSD shows stability at 4 ns simulation. (b) RMSF shows fluctuation at certain residues found to be identified as catalytic residues. (c) Simulation behavior of CCT and different competitive inhibitors checked.
References Ancelin, M.L., Torpier, G., Vial, H.J., Capron, A., 1987. Choline incorporation by Schistosoma mansoni: distribution of choline metabolites during development and after sexual differentiation. J. Parasitol. 73 (3), 530–535. Ancelin, M.L., Calas, M., Bonhoure, A., Herbute, S., Vial, H.J., 2003. In vivo antimalarial activities of mono- and bis quaternary ammonium salts interfering with Plasmodium phospholipid metabolism. Antimicrob. Agents Chemother. 47, 2598–2605. Bell, S.L., Palsson, B.Ø., 2005. Expa: a program for calculating extreme pathways in biochemical reaction networks. Bioinforma. Appl. Notes 21, 1739–1740. Berriman, M., et al., 2009. The genome of the blood fluke Schistosoma mansoni. Nature 460, 352–358. Caspi, R., Altman, T., Dale, J.M., Dreher, K., Fulcher, C.A., Gilham, F., Kaipa, P., Karthikeyan, A.S., Kothari, A., Krummenacker, M., Latendresse, M., Mueller, L.A., Paley, S., Popescu, L., Pujar, A., Shearer, A.G., Zhang, P., Karp, P.D., 2010. The MetaCyc database of metabolic pathways and enzymes and the BioCyc collection of pathway/genome databases. Nucleic Acids Res. 38, D473–D479. Cui, Z., Houweling, M., 2002. Phosphatidylcholine and cell death. Biochim. Biophys. Acta 1585, 87–96. Dai, J., Mumper, R.J., 2010. Plant phenolics: extraction, analysis and their antioxidant and anticancer properties. Molecules 15, 7313–7352. Doenhoff, M.J., Cioli, D., Utzinger, J., 2008. Praziquantel: mechanism of action, resistance and new derivatives for Schistosomiasis. Curr. Opin. Infect. Dis. 21, 659–667. Eswar, N., Marti-Renom, M.A., Webb, B., Madhusudhan, M.S., Eramian, D., Shen, M., Pieper, U., Sali, A., 2006. Comparative protein structure modeling with MODELLER. Current Protocols in Bioinformatics. Supplement 15. John Wiley & Sons, Inc., pp. 5. 6.1–5.6.30. Flora, S.J., 2009. Structural, chemical and biological aspects of antioxidants for strategies against metal and metalloid exposure. Oxid. Med. Cell Longev. 4, 191–206. Funahashi, A., Matsuoka, Y., Jouraku, A., Morohashi, M., Kikuchi, N., Kitano, H., 2008. Cell designer 3.5: a versatile modeling tool for biochemical networks. Proc. IEEE 96 (8), 1254–1265. Furlong, S.T., Caulfield, J.P., 1989. Schistosoma mansoni: synthesis and release of phospholipids, lysophospholipids, and neutral lipids by schistosomula. Exp. Parasitol. 69 (1), 65–77. Gibellini, F., Smith, T.K., 2010. The Kennedy pathway—de novo synthesis of phosphatidylethanolamine and phosphatidylcholine. IUBMB Life 62 (6), 414–428. Golan, D.E., Brown, C.S., Catherine, M.L.C., Furlong, S.T., Caulfield, J.P., 1986. Schistosomula of Schistosoma mansoni use lysophosphatidylcholine to lyse adherent human red blood cells and immobilize red cell membrane components. J. Cell Biol. 103, 819–828.
Guz, N.R., Stermitz, F.R., 2000. Synthesis and structures of regioisomeric hydnocarpintype flavonolignans. J. Nat. Prod. 63 (8), 1140–1145. Hewitson, J.P., Grainger, J.R., Maizels, R.M., 2009a. Helminth immunoregulation: the role of parasite secreted proteins in modulating host immunity. Mol. Biochem. Parasitol. 167 (1–9), 1–11. Hewitson, J.P., Grainger, J.R., Maizels, R.M., 2009b. Helminth immunoregulation: the role of parasite secreted proteins in modulating host immunity. Mol. Biochem. Parasitol. 167, 1–11. Hoops, S., Sahle, S., Gauges, R., Lee, C., Pahle, J., Simus, N., Singhal, M., Xu, L., Mendes, P., Kummer, U., 2006. COPASI: a COmplex PAthway SImulator. Bioinformatics 22, 3067–3074. Kanehisa, M., Goto, S., 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucl. Acids Res. 28, 27–30. Kitano, H., 2002. Systems biology: a brief overview. Science 295 (5560), 1662–1664. Kitano, H., et al., 2005. Using process diagram for the graphical representation of biological networks. Nat. Biotechnol. 23 (8), 961–966. Laskowski, R.A., MacArthur, M.W., Moss, D.S., Thornton, J.M., 1993. PROCHECK — a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, 283–291. Lee, S.J., Lee, D.Y., Kim, T.Y., Kim, B.H., Lee, J., Lee, S.Y., 2005. Metabolic engineering of Escherichia coli for enhanced production of succinic acid, based on genome comparison and in silico gene knockout simulation. Environ. Microbiol. 71, 7880–7887. Lewis, R.N., Pohle, W., McElhaney, R.N., 1996. The interfacial structure of phospholipid bilayers: differential scanning calorimetry and Fourier transform infrared spectroscopic studies of 1,2-dipalmitoyl-sn-glycero-3-phosphorylcholine and its dialkyl and acyl–alkyl analogs. Biophys. J. 70 (6), 2736–2746. Loukas, A., Maizels, R.M., 2000. Helminth C-type lectins and host–parasite interactions. Parasitology 16, 333–340. Marechal, Eric, Riou, Mickael, Kerboeuf, Dominique, Beugnet, Frederic, Chaminade, Pierre, Loiseau, Philippe M., 2011. Membrane lipidomics for the discovery of new antiparasitic drug targets. Trends Parasitol. 27 (11), 496–504. MavinSketch 5.10.1 ChemAxon Ltd. 1998–2012. http://www.chemaxon.com. Oostenbrink, C., Villa, A., Mark, A.E., van Gunsteren, W.F., 2004. A Biomolecular Force Field Based on the Free Enthalpy of Hydration and Solvation: The GROMOS Force-Field Parameter Sets 53A5 and 53A6. pp. 1656–1676. Pettit, G.R., Meng, Y., Stevenson, C.A., Doubek, D.L., Knight, J.C., Cichacz, Z., Pettit, R.K., Chapuis, J.C., Schmidt, J.M., 2003. Isolation and structure of palstatin from the Amazon tree Hymeneae palustris (1). J. Nat. Prod. 66 (2), 259–262. Pronk, Sander, Páll, Szilard, Schulz, Roland, Larsson, Per, Bjelkmar, Pär, Apostolov, Rossen, Shirts, Michael R., Smith, Jeremy C., Kasson, Peter M., van der Spoel, David, Hess, Berk,
222
S. Shinde et al. / Gene 550 (2014) 214–222
Lindahl, Erik, 2013. GROMACS 4.5: a high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics 29 (7), 845–854. Punta, M., Coggill, P.C., Eberhardt, R.Y., Mistry, J., Tate, J., Boursnell, C., Pang, N., Forslund, K., G., Clements, J., Heger, A., Holm, L., Sonnhammer, E.L., Eddy, S.R., Bateman, A., Finn, R.D., 2012. The Pfam protein families database. Nucleic Acids Res. 40, D290–D301. Ramakrishnan, S., Docampo, M.D., MacRae, J.I., Pujol, F.M., Brooks, C.F., van Dooren, G.G., Kalervo Hiltunen, J., Kastaniotis, A.J., McConville, M.J., Striepen, B., 2012. Apicoplast and endoplasmic reticulum cooperate in fatty acid biosynthesis in apicomplexan parasite Toxoplasma gondii. J. Biol. Chem. 287, 4957–4971. Schmid-Hempel, P., 2009. Immune defense, parasite evasion strategies and their relevance for ‘macroscopic phenomena’ such as virulence. Philos. Trans. R. Soc. B Biol. Sci. 364, 85–98. Schüttelkopf, A.W., van Aalten, D.M.F., 2004. PRODRG: a tool for high-throughput crystallography of protein–ligand complexes. Acta Crystallogr. D60, 1355–1363. Seo, M., Kim, J.-D., Neau, D., Sehgal, I., Lee, Y.-H., 2011. Structure-based development of small molecule PFKFB3 inhibitors: a framework for potential cancer therapeutic agents targeting the Warburg effect. PLoS One 6 (9), e24179. http://dx.doi.org/10. 1371/journal.pone.0024179.
Sharma, D.K., Hall, I.H., 1991. Hypolipidemic, anti-inflammatory, and antineoplastic activity and cytotoxicity of flavonolignans isolated from Hydnocarpus wightiana seeds. J. Nat. Prod. 54 (5), 1298–1302. Suresh, S., Turley, S., Opperdoes, F.R., Michels, P.A., Hol, W.G., 2000. A potential target enzyme for trypanocidal drugs revealed by the crystal structure of NAD-dependent glycerol-3-phosphate dehydrogenase from Leishmania mexicana. Structure 15, 541–552. Trott, O., Olson, A.J., 2010. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading. J. Comput. Chem. 31, 455–461. Useh, M.F., 2013. Parasitic Diseases — Schistosomiasis. InTech http://dx.doi.org/10.5772/ 53553. Wiederstein, M., Sippl, M.J., 2012. ProSA-Web: Interactive Web Service for the Recognition of Errors in Three-Dimensional Structures of Proteins. Nucleic Acids Res. 35 (Web Server issue) W407–10. Wu, J., Xu, W., Ming, Z., Dong, H., Tang, H., et al., 2010. Metabolic changes reveal the development of schistosomiasis in mice. PLoS Negl. Trop. Dis. 4 (8), e807. Young, B.W., Podesta, R.B., 1982. Major phospholipids and phosphatidylcholine synthesis in adult Schistosoma mansoni. Mol. Biochem. Parasitol. 5 (3), 165–172.