A validated molecular docking study of lipid–protein interactions

A Validated Molecular Docking Study of Lipid–Protein Interactions Rajyalakshmi Gaddipati College of Engineering and Science Victoria University Submitted in fulfillment of the requirements of the degree of Doctor of Philosophy September, 2015 To my mother, father, husband, mother-in-law and children, with all my love and respect Abstract The interaction of proteins with lipids is an important aspect of research as it plays a main role in various biological responses such as metabolic pathways, signal transduction and in drug discovery Proteins that take part in the treatment of different diseases act as drug targets and hence research is ongoing to find new series of ligands of medicinally significant proteins Few such proteins, peroxisome proliferator activated receptors (PPARs), retinoid receptors, cannabinoid receptors (CB1 and CB2), lipoxygenase (LOX), cyclooxygenases (COXs) were selected for the author’s study due to their therapeutic role to act as pharmacological targets The existing ligands for these protein targets are causing some side effects For example, thiazolidinediones are the currently used ligands for PPARs Thiazolidinediones bind to PPARs and used in the treatment of diabetes However, this treatment results in obesity Similarly, the use of Nonsteroidal Anti-inflammatory Drugs (NSAIDs) like aspirin and ibuprofen lead to stomach or gastrointestinal ulcers, heartburn, headache and dizziness Hence, a set of ligands which have a significant role in the treatment of diseases were selected and compared for their binding affinities towards the design of a new series of drugs In order to find the new series of ligands of the above proteins, three groups of lipid ligands— tocotrienols (α, β, γ and δ tocotrienols), omega fatty acids (Docosahexaenoic acid (DHA), Eicosapentaenoic acid (EPA)) and endocannabinoids (anandamide and 2-arachidonyl glycerol) —were tested for their ability to bind to PPARs, CBs and COX-2 Two molecular docking programs, AutoDock and Glide, were used to study the above lipid-protein interactions The stability of docked complexes was tested through molecular dynamic simulations Further, the in silico results were validated with in vitro experimental results The three groups of lipid ligands were provided same conditions in both in vitro and in silico experiments Still, omega fatty acids have shown strong interactions with PPARs and retinoid receptors This is because of the ligand binding cavity of PPARs and retinoid receptors that accommodates polyunsaturated fatty acids better than the other ligands Among the fatty acids, omega fatty acids possess most potent immunomodulatory activities and among omega fatty acids DHA and EPA are biologically more potent Furthermore, DHA and EPA have anti-inflammatory and cancer preventing properties COX-2 also has shown strong binding interactions with DHA in both virtual and wet laboratory experiments compared to the other ligands Next to omega fatty acids, endocannabinoids have exhibited strong affinity with COX-2 Tocotrienols did not show favorable binding interactions with cyclooxygenases due to their orientation and structure which failed to fit into the binding pocket of cyclooxygenases The ligand binding cavity of COX-2 is larger than COX-1 and hence COX-2 has shown strong binding interactions with the ligands compared to COX-1 Endocannabinoids have shown strong binding interactions with both cannabinoid receptors compared to the other two groups of lipid ligands A web-based validated tool, Lipro Interact was developed with the results of all the above lipid-protein interactions The purpose of Lipro Interact was to provide the author’s study of 80 lipid-protein interactions for global use Lipro Interact provided the detailed information on the binding affinities of each lipid-protein interaction along with the microscopic atomic interactions, bond distances and ligand binding sites The advantage of Lipro Interact is that all the lipid-protein interacting studies included were downloadable in image form Further, Lipro Interact allows the users to download the PDB files of the above lipid-protein interactions Future versions of Lipro Interact can calculate the binding affinity for any pair of protein and ligand Student Declaration “I, Rajyalakshmi Gaddipati, declare that the PhD thesis entitled “The Study of Lipid-Protein Interactions towards the Design of Lipro Interact- A Validated Web based Tool” is no more than 100,000 words in length including quotes and exclusive of tables, figures, appendices, bibliography, references and footnotes This thesis contains no material that has been submitted previously, in whole or in part, for the award of any other academic degree or diploma Except where otherwise indicated, this thesis is my own work” Signature: Rajyalakshmi G Date: 03/09/2015 Acknowledgements First, I would like to thank my Master and God for everything I would like to thank my husband who was with me all the times providing his support My full respect and thanks to my mother and mother-in-law who helped me in looking after my kids during my PhD studies I would like to thank my daughters who did not ever disturb me while I am studying I would like to extend my deep thanks and gratitude to all the people who contributed to enrich my knowledge and improve my competencies I am grateful to my principal supervisor Dr Gitesh Raikundalia who guided me throughout my project and provided me with his valuable comments and advises Further, I would like to thank my co-supervisor Dr Michael Mathai without whose encouragement and support, this research would not have been completed I would like to thank Dr John Orbell for his valuable guidance provided at the beginning stages of my project I would like to extend my deep thanks to Dr Mike Kuiper for providing his suggestions required for the entire project I would like to thank Dr Elizabeth Yuriev for her help in guiding me in molecular docking techniques I would like to extend my thanks to Dr Phil Beart for giving me permission to work with Howard Florey Laboratories I would like to thank Dr Linda Lau for supervising my project for the period of my wet laboratory experiments at Howard Florey Institute I would like to thank Victoria University for giving me the scholarship for my project I would like to thank Howard Florey Institute for allowing me to conduct my wet laboratory experiment in Beart lab Finally, I would like to thank all my peers, for all the fun we have had together in the last four years List of Publications Gaddipati, R.S., Raikundalia, G.K., & Mathai, M.L (2012) Towards the Design of PPAR Based drugs Using Tocotrienols as Natural Ligands Paper presented at the International Conference on Engineering and Science, Beijing Gaddipati, R.S., Raikundalia, G.K., & Mathai, M.L (2014) Dual and selective lipid inhibitors of cyclooxygenases and lipoxygenase: a molecular docking study Medicinal Chemistry Research, 23 (7), 3389-3402 Gaddipati, R.S, Raikundalia, G.K., & Mathai, M.L (2014) Comparison of AutoDock and Glide towards the Discovery of PPAR Agonists International Journal of Bioscience, Biochemistry and Bioinformatics, (2), 100-105 Table of Contents Chapter : Thesis overview 21 1.1 Introduction 21 1.2 Research Question 25 1.2.1 Gaps and Limitations of Previous Research 27 1.3 Aims of this Research 30 1.4 Originality and Uniqueness of the Research 32 1.5 Significance of the Research 34 1.6 Organization of the Thesis 36 Chapter : Literature Review 39 2.2 Ligand-Receptor Interactions 41 2.3 Lipid Ligands 42 2.3.1 Tocotrienols 43 2.3.2 Omega Fatty Acids 48 2.3.3 Endocannabinoids 52 2.4 Target Proteins 56 2.4.1 Therapeutic Proteins and Their Significance 58 2.4.2 Mutations in Target Proteins 61 2.4.2.1 Mutations in PPAR-α 63 2.4.2.2 Mutations in PPAR-γ isoform1 63 2.4.2.3 Mutations in PPAR-γ isoform2 64 2.4.2.4 Mutations in SXR 64 2.4.2.5 Mutations in RXR-α 64 2.4.2.6 Mutations in RXR-γ 65 2.4.2.7 Mutations in RAR-α 65 2.4.2.8 Mutations in FXR 65 2.4.3 Pathways of Target Proteins 66 2.4.4 Active Sites of Target Proteins 68 2.5 Bioinformatics Tools to Assess the Lipid-Protein Interactions 71 2.5.1 Lipid Structural Databases and Tools 72 2.5.2 Protein Structural Databases 74 2.5.3 Bioinformatics Tools to Study the Active Site of Proteins 76 2.5.4 Molecular Docking Tools for the Study of Lipid-Protein Interactions 78 2.5.5 Molecular Dynamic Simulations 82 2.5.6 Experimental Validation 83 2.6 Novelty and Uniqueness of Contribution 84 2.7 Conclusion 86 Chapter 3: Bioinformatic and Biochemical Methods to Create Lipro Interact Software 88 3.1 Introduction 88 3.2 Methodology Framework 89 3.2.1 Biochemical Component 91 3.2.2 Lipro Interact 91 3.3 Selection of Molecular Docking Tool 92 3.4 Molecular Docking Studies using AutoDock 93 3.4.1 Ligand Preparation 94 3.4.2 Protein Preparation 95 3.4.3 Receptor Grid Generation 99 3.4.4 Docking with AutoDock 100 3.5 Molecular Docking Studies using Glide 104 3.5.1 Ligand Preparation 105 3.5.2 Protein Preparation 106 3.5.3 Receptor Grid Generation 107 3.5.4 Docking Studies 108 3.5.5 Validation of Docking Results 108 3.6 Molecular Dynamic Simulations 109 3.6.1 Theory of MD Simulations 109 3.6.2 System Building 111 3.6.3 Minimization 113 3.6.4 Molecular Dynamics 115 3.7 Scintillation Proximity Assay 115 3.7.1 Preparation of SPA System 116 3.7.2 Normalization and Standardization of MicroBeta Trilux 116 3.7.3 Saturation Binding Assay 117 3.7.4 Competitive Binding Assay 119 3.8 Developing Lipro Interact 120 3.9 Conclusion 121 10 Appendices if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("Binding Energy - Glide Score") == 0) { Image1.Visible = false; Label4.Visible = true; Label5.Visible = true; Label5.Text = bdrow[0][2].ToString(); btndownload.Visible = false; btndwnpdb.Visible = false; } //code to be executed if 4A-AutoDock is selected in Protein- Ligand Interaction if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("4AAutoDock") == 0) { Label4.Visible = false; Label5.Visible = false; Image1.Visible = true; Image1.ImageUrl = "~/SW-input files/4A-AutoDock/" bdrow[0][3].ToString(); ViewState["FileName"] = bdrow[0][3].ToString(); + //check file exists if (!System.IO.File.Exists(Server.MapPath(Image1.ImageUrl))) { Label5.Visible = true; Label5.Text = DropDownList1.SelectedValue.ToString() + " did not produce considerable binding poses with " + DropDownList2.SelectedValue.ToString(); btndownload.Visible = false; btndwnpdb.Visible = false; Image1.Visible = false; return; } else { Label5.Visible = false; btndownload.Visible = true; btndwnpdb.Visible = false; //divref2.Visible = true //divref1.Visible = false; } } //modifed 19/7 based on drop down value //code to be executed if 4A-Glide is selected in Protein- Ligand Interaction 435 Appendices if Glide") == 0) { (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("4ALabel4.Visible = false; Label5.Visible = false; Image1.Visible = true; Image1.ImageUrl = "~/SW-input files/4A-Glide/" + bdrow[0][5].ToString(); ViewState["FileName"] = bdrow[0][5].ToString(); //check file exists if (!System.IO.File.Exists(Server.MapPath(Image1.ImageUrl))) { Label5.Visible = true; Label5.Text = DropDownList1.SelectedValue.ToString() + " did not produce considerable binding poses with " + DropDownList2.SelectedValue.ToString(); btndownload.Visible = false; btndwnpdb.Visible = false; Image1.Visible = false; return; } else { Label5.Visible = false; btndownload.Visible = true; btndwnpdb.Visible = false; //divref2.Visible = true //divref1.Visible = false; } } //if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("Binding_Pocket") == 0) //code to be executed if Bonded_Interactions is selected in Protein- Ligand Interaction if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("Bonded_Interactions") == 0) { Label4.Visible = false; Label5.Visible = false; Image1.Visible = true; Image1.ImageUrl = "~/SW-input files/Bonded_Interactions/" bdrow[0][3].ToString(); ViewState["FileName"] = bdrow[0][3].ToString(); //check file exists if (!System.IO.File.Exists(Server.MapPath(Image1.ImageUrl))) { Label5.Visible = true; + 436 Appendices Label5.Text = DropDownList1.SelectedValue.ToString() + " Did not produce considerable binding poses with " + DropDownList2.SelectedValue.ToString(); btndownload.Visible = false; btndwnpdb.Visible = false; Image1.Visible = false; return; } else { Label5.Visible = false; btndownload.Visible = true; btndwnpdb.Visible = false; //divref2.Visible = true //divref1.Visible = false; } } //Code to be executed if Comparing_AutoDock_with_Glide is selected in Protein- Ligand Interaction if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("Comparing_AutoDock _with_Glide") == 0) { Label4.Visible = false; Label5.Visible = false; Image1.Visible = true; Image1.ImageUrl = "~/SW-input files/Comparing_AutoDock_with_Glide/" + bdrow[0][7].ToString(); ViewState["FileName"] = bdrow[0][7].ToString(); //check file exists if (!System.IO.File.Exists(Server.MapPath(Image1.ImageUrl))) { Label5.Visible = true; Label5.Text = DropDownList1.SelectedValue.ToString() + " Did not produce considerable binding poses with " + DropDownList2.SelectedValue.ToString(); btndownload.Visible = false; btndwnpdb.Visible = false; Image1.Visible = false; return; } else { Label5.Visible = false; btndownload.Visible = true; btndwnpdb.Visible = false; //divref2.Visible = true //divref1.Visible = false; } } 437 Appendices //if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("Molecular_Dynamic_S imulation") == 0) //{ // Label4.Visible = false; // Label5.Visible = false; // Image1.Visible = true; // Image1.ImageUrl = "~/SW-input files/Molecular Dynamic Simulation/" + bdrow[0][7].ToString(); // ViewState["FileName"] = bdrow[0][7].ToString(); // //check file exists // if (!System.IO.File.Exists(Server.MapPath(Image1.ImageUrl))) // { // Label5.Visible = true; // Label5.Text = DropDownList1.SelectedValue.ToString() + " did not produce considerable binding poses with " + DropDownList2.SelectedValue.ToString(); // btndownload.Visible = false; // btndwnpdb.Visible = false; // Image1.Visible = false; // return; // } // else // { // Label5.Visible = false; // btndownload.Visible = true; // btndwnpdb.Visible = false; // } //} //code to be executed if PDB_Files is selected in Binding Ligand Interation if (DropDownList3.SelectedItem.Value.ToString().Trim().CompareTo("PDB_Files") == 0) { Label4.Visible = false; Label5.Visible = false; filePath = Server.MapPath("~/SW-input files/PDB_Files/"); fileName = bdrow[0][8].ToString(); //check file exists if (!System.IO.File.Exists(filePath + fileName)) { Label5.Visible = true; Label5.Text = DropDownList1.SelectedValue.ToString() + " Did not produce considerable binding poses with " + DropDownList2.SelectedValue.ToString(); btndownload.Visible = false; Image1.Visible = false; return; 438 Appendices } else { ViewState["PDBFilePath"] = filePath; ViewState["PDBFileName"] = fileName; } Label5.Visible = false; btndownload.Visible = false; Image1.Visible = false; btndwnpdb.Visible = true; } } } else { Label4.Visible = false; Label5.Visible = false; Image1.Visible = false; } } catch (Exception ex) { Response.Write(ex.Message); } } protected void btndownload_Click(object sender, EventArgs e) { //code to be executed if Download Button is clicked On clicking Lipro Interact will download respective image file try { byte[] bytes = System.IO.File.ReadAllBytes(Server.MapPath(Image1.ImageUrl)); Response.Clear(); //Response.ContentType = "image/png"; Response.Cache.SetCacheability(HttpCacheability.Private); Response.Expires = -1; Response.Buffer = true; Response.AddHeader("Content-Disposition", string.Format("{0};FileName=\"{1}\"", "attachment", ViewState["FileName"].ToString())); Response.BinaryWrite(bytes); Response.End(); } catch (Exception ex) 439 Appendices {} } protected void btndwnpdb_Click(object sender, EventArgs e) { //code to be executed if Download PDB File Button is clicked.On clicking Lipro Interact will download respective PDBfile try { byte[] bytes = System.IO.File.ReadAllBytes(ViewState["PDBFilePath"].ToString() + ViewState["PDBFileName"].ToString()); Response.Clear(); Response.ContentType = "application/pdb"; Response.Cache.SetCacheability(HttpCacheability.Private); Response.Expires = -1; Response.Buffer = true; Response.AddHeader("Content-Disposition", string.Format("{0};FileName=\"{1}\"", "attachment", ViewState["PDBFileName"].ToString())); Response.BinaryWrite(bytes); Response.End(); } catch (Exception ex) {} } protected void Button1_Click(object sender, EventArgs e) { Response.Redirect("Experiment.aspx"); } protected void Button2_Click(object sender, EventArgs e) { Response.Redirect("Welcome.aspx"); } } Third Page-Design //Design page for Lipro Interact Software for the Binding Interactions of Lipid Ligands [...]... chromanol ring A previous study performed by Aggarwal, et al., suggested that tocotrienols have positive health effects on bone health, brain health, blood sugar metabolism and cancer (Aggarwal et al., 2010)  Omega 3 fatty acids are the natural ligands of PPARs because of the presence of polar head group and a hydrophobic tail in their structures (Sheu et al., 2005) Omega 3 fatty acids prevent cancer... 57 Table 2.2 Computational tools and databases of lipids 73 Table 2.3 Computational tools and databases of proteins 75 Table 2.4 Computational tools and databases for the study of ligand binding site of proteins 77 Table 2.5 Computational tools and databases for the study of lipid-protein interaction 80 Table 3.1 Target proteins and lipid ligands 91 Table 3.2... particular ligand as an effective drug candidate, the action of ligand has to be studied in vivo Further the mechanism of action of ligand and the effect of ligand on a particular biological reaction has to be analyzed The metabolism of ligand-protein interaction has to be studied Lipids are a group of organic compounds that are insoluble in water and soluble in organic solvents Lipids, commonly known as... dimensional structure of proteins and ligands are now available Hence, the atomic interactions between ligand and protein are studied in close proximity There are different bioinformatic techniques available to study the ligand-receptor interactions To name a few are virtual screening, molecular docking and molecular dynamic simulations (MD simulations) Comparing the virtual results with wet laboratory... to initiate a biological response, which is termed as efficacy (Strange, 2008) Agonist is a small molecule (ligand) which is usually a chemical binds to a receptor and activates the receptor resulting in a biological response Antagonist is also a ligand that blocks the action of a receptor An agonist can be a full agonist, partial agonist or inverse agonist Full agonists are the compounds that bind... experimental results However, Padgett et al.’s study is different from the author’s in the choice of ligands Padgett’s study has concluded that during docking, anadamide adapted certain conformations which are analogous to arachidonic acid, the substrate of COX-2 Moreover, the study was limited to the binding of 28 Chapter 1: Thesis Overview anandamide and its analogues only with CB1 The author’s study. .. design of a new series of drugs starts with the finding of new molecular targets The strong binding affinity and the microscopic atomic interactions are important to analyze if a particular ligand molecule can be a potential drug candidate At the same time virtual computational results need to be validated with the wet laboratory experimental results, because the virtual results alone are not sufficient... and EPA and the differences in the mechanism of DHA and EPA involving PPAR-γ (Oster et al., 2010) The author’s study has filled this gap of a study that compares both microscopic atomic interactions with wet laboratory experimental validations Furthermore, the author’s study is extended to some other lipid ligands and proteins along with DHA and PPARs Stone et al., have examined the binding of PPAR-γ... these drugs have some side effects (discussed in detail in Chapters 4-6) and so the current study was designed to find new series of ligands that can act as drug candidates and cause few or no side effects The aim is to study the potential of each ligand to be an effective drug candidate 3 There are various molecular docking techniques available today to study the ligandreceptor interactions However,... selecting a suitable docking technique remains as a 31 Chapter 1: Thesis Overview challenge Hence, a comparison study was conducted between two molecular docking techniques The study was aimed to use molecular dynamic simulations to support the molecular docking results 4 Virtual results need an experimental validation Hence, the study conducted a wet laboratory experiment Then, an evaluation study determined ... brain health, blood sugar metabolism and cancer (Aggarwal et al., 2010)  Omega fatty acids are the natural ligands of PPARs because of the presence of polar head group and a hydrophobic tail... (Sierra et al., 2007) The anti-diabetic drugs-TZDs use PPAR-γ as a molecular target (Malapaka et al., 2012) However, TZDs cause side effects like obesity and cardiovascular diseases (Malapaka et al.,... interactions are important to analyze if a particular ligand molecule can be a potential drug candidate At the same time virtual computational results need to be validated with the wet laboratory