In Silico Structure Analysis of Type 2 Diabetes Associated Toll Like Receptor-4 (TLR4)

INTRODUCTION

Type 2 Diabetes Mellitus (T2DM), previously known as non-insulin-dependent diabetes (NIDDM), is a complex heterogeneous group of metabolic conditions characterized by increased levels of blood glucose due to impairment in insulin action and/or insulin secretion (Das & Elbein, 2006). Toll-like receptor 4 is a protein (M. wt. 95kDa) that in humans is encoded by the TLR4 gene located on the chromosome no. 9. Polymorphisms in genes that encode proteins related to the innate immune system, such as the toll-like receptors (TLRs), could influence the immune response as well as the development of T2DM (Bagarolli, et. al., 2010). TLRs are evolutionary conserved Pattern-Recognition Receptors (PRRs) that play a key role in the activation of innate immune response by recognizing highly conserved pathogen-associated molecular patterns (PAMPs), such as the Lipopolysaccharide (LPS) component of gram-negative bacteria (Meylan, et. al., 2006, Arancibia, et. al., 2007). Human orthologs of TLR4 are known to comprise at least 10 members (Zarember & Godowski, 2002). Each TLR family member recognizes a specific pathogen component and upon activation, triggers a signaling cascade leading to the production of inflammatory cytokines, releasing of antimicrobial peptides and activation of the adaptive immune response (Akira, et. al., 2001, Takeda & Akira, 2005). TLR4 recognizes LPS as a ligand (Takeda & Akira, 2005) and is expressed in macrophages, airway epithelia, adipose tissue, skeletal muscle, pancreas and vascular endothelial and smooth muscle cells (Shi, et. al., 2006). It also interacts with endogenous ligands such as free-fatty acids, heat shock proteins 60 and 70, fibrinogen and fibronectin, which are elevated in T2DM patients (Shi, et. al., 2006, Sasu, et. al., 2001, Smiley, et. al., 2001). Homology modelling refers to the modelling of protein 3D structure using a known experimentally determined structure of homologous protein as a template. Homology modelling provides structural information, that is important to understand of protein function, dynamics, interactions with ligands (Barry, et. al., 1994). In the present study, we used protein homology modelling

MATERIAL METHODS:

The protein sequence of human TLR4 (O00206.2) was retrieved from GeneBank-NCBI (www.ncbi.nlm.nih.gov) in the FASTA format. The sequence Alignment was performed via BlastP against Protein Data Base (PDB) to search the best template for TLR4 protein modelling. Homology modelling of human TLR4 protein was performed by MODELLER using crystal structure of homologous template human TLR4-human MD-2-E.coli LPS Ra complex (3FXI) and crystal structure of homologous template mouse TLR4/md-2/Lipid IVa complex (3VQ1). The PBD file of human TLR4 (O00206.2) was generated by using PyMOL.

The SAVES v5.0 server provides a visual analysis of the quality of a putative crystal structure of protein. Verify 3D expects this crystal structure to be submitted in PDB format. RAMPAGE program was used for visualizing and assessing the Ramachandran plot of human TLR4 (O00206.2) and its homologous template. The model was selected on the basis of various factors such as overall G-factor, number of residues in core that fall in generously allowed and disallowed region in Ramachandran plot (figure 1,2,3). The model was further analyzed by QMEAN (Benkert, et. al., 2009, Novotny, et. al., 1998) and ProSA (Wiederstein & Sippl, 2007). ProSA was used for the display of Z-score and energy plots.

Building of protein Model: To build this model, BLAST was done with the maximum E-value allowed for template being 5e-62. Sequence alignment of human TLR4 (O00206.2) protein by using the BLAST revealed sequence homology with TLR4 Human Md-2-E.Coli (3FXI) 100% identity and crystal structure of Mouse TLR4/md-2 (3VQ1) 62% identity, which was selected as template for the model building of TLR4 (O00206.2) protein. Total 604 residues (72% of query sequence) have been modelled with 100% and 62% confidence by the highest scoring template. Model Reputation: The human TLR4 (O00206.2) model showed good stereo chemical property in terms of overall G-factor value of -0.23. G-factor value indicating that geometry of the model corresponds to the probability conformation with 81.2% residue in the most favoured region of the Ramachandran plot showing high accuracy of model predicted (Prajapat, et. al., 2014). The number of residues in allowed region is 18.1%, generally allowed region 0.5% and disallowed region 0.2%.(figure 1a,1b). A similar approach was also used for homologous templates TLR4 Human Md-2- allowed region and none of the residue in the disallowed region. (figure 2a,2b). The homologous templates of Mouse TLR4/md-2 (3VQ1) has Ramachandran plot in which 77.8% residue in the most favoured region, 20.9% residue in the allowed region, 0.7% residue in the generally allowed region and 0.5% residue in the disallowed region (figure 3a,3b). The Ramachandran plot has 90% residue in most favoured region has a good quality (Morris, et. al., 1992) therefore the target model of TLR4 has consider in good quality model due to 81.2% residue in the most favoured region.

The verified 3D high score 0.81 for human TLR4 (O00206.2) indicate that environmental profile of the model is good (figure 4). In the verified 3D plot, 100% of the residue have an averaged 3D-1D score >=0.2. The profile score above zero in the verified 3D graph is corresponds to an acceptable environment of the model (Bowie, et. al., 1991, Luthy, et. al., 1992) The Errat is a program used in protein structure determined by crystallography. The two line are protein for which the calculated error value falls below the 95% rejection limit. Good high resolution structures generally produce values around 95% or higher. For lower resolutions (2.5 to 3A) the average overall quality factor is around 91%. The overall quality factor for human TLR4 (O00206.2) was 83.137 (figure 5). Model Validation: ProSA tool was used for the potential errors detection of protein structure based on energy plot and display in three- dimensional manner. ProSA model of human TLR4 (O00206.2), human TLR4- human MD-2-E.coli LPS Ra complex (3FXI) and mouse TLR4/MD- 2/lipid lVa complex (3VQ1) protein for potential errors (figure 6,7and 8). The ProSA web z-scores of protein chains in PDB determined by X-ray crystallography (light blue) or NMR spectroscopy (dark blue) with respect to their length. The plot shows only chains with less than 1,000 residues and a z-score of 10.The z-scores of human TLR4 (O00206.2), human TLR4- human MD-2-E.coli LPS Ra complex (3FXI) and mouse TLR4/MD- 2/lipid lVa complex (3VQ1) are highlighted as large dots. The human TLR4 (O00206.2) ProSA Z-score was -7.82 for (Figure 6),for template human TLR4- human MD-2-E.coli LPS Ra complex (3FXI) ProSA Z-score was -7.81 (figure 7) and for template mouse TLR4/MD- 2/lipid lVa complex (3VQ1) ProSA Z-score -7.99 that indicates the overall model quality of target and template (Figure 8) measures the deviation of the total energy of the structure with respect to an energy distribution derived from random conformations (Teilum, et. al., 2005).

The QMEAN Z- score of human TLR4 (O00206.2) -2.53 (Figure 9), template human TLR4- human MD-2-E.coli LPS Ra complex (3FXI) QMEAN Z-score is -3.06 (Figure 10) and for template mouse TLR4/MD- 2/lipid lVa complex (3VQ1) QMEAN Z-score is -1.10 (figure 11). which very closed to the value of 0 and this shows the fine quality for the models [15; 21]. A comparison between the normalized QMEAN score (0.40) and protein size in non-redundant set of PDB structures in the plot revealed different set of Z-values for different parameters such as C-beta interactions (-0.37), interactions between all atoms (-2.98), solvation (-0.87), torsion (-2.51), SSE agreement (-1.44) and ACC agreement (-0.64) (Figure 9; Table 1).

The human TLR4(O00206.2)protein model was obtained through homology modelling and the main interaction are similar to those observed for homologous template human TLR4- human MD-2-E.coli LPS Ra complex (3FXI) and mouse TLR4/MD- 2/lipid lVa complex (3VQ1). The human TLR4 (O00206.2) model showed overall G-factor value of -0.23 with 81.2% residue in the most favoured region Ramachandran plot and its homologous templates human TLR4- human MD-2-E.coli LPS Ra complex (3FXI) 69.1% residue in the most favoured region and mouse TLR4/MD- 2/lipid lVa complex (3VQ1) have 77.8% residue in the most favoured region that was -7.99. That indicates overall model quality and measures the deviation of total structural energy with respect to an energy distribution derived from random conformations. We hope that, these results will be useful for the designing of inhibitors for TLR4 (O00206.2) and understanding of the mechanism of inhibition at the molecular level.

[1] Das S.K. & Elbein S.C. (2006). The genetic basis of type 2 diabetes. Cellscience 2, pp. 100-131. [2] Bagarolli R.A., Saad M.J., Saad S.T. (2010). Toll-like receptor 4 and inducible nitric oxide synthase gene polymorphisms are associated with Type 2 diabetes. J Diabetes Complications 24, pp. 192-198. [3] Meylan E., Tschopp J., Karin M. (2006). Intracellular pattern recognition receptors in the host response. Nature 442 (7098), pp. 39-44. [4] Arancibia S.A., Beltrán C.J., Aguirre I.M., Silva P., Peralta A.L., Malinarich F., Hermoso MA (2007). Toll-like receptors are key participants in innate immune responses. Biol Res 40 (2), pp. 97- 112. [5] Zarember K.A. & Godowski P.J. (2002). Tissue expression of human Toll-like receptors and differential regulation of Toll-like receptor mRNAs in leukocytes in response to microbes, their products, and cytokines. J Immunol 168 (2), pp. 554-561. [6] Akira S., Takeda K., Kaisho T. (2001). Toll-like receptors: critical proteins linking innate and acquired immunity. Nat Immunol 2 (8), pp. 675-680. [7] Takeda K. & Akira S. (2005). Toll-like receptors in innate immunity. Int Immunol 17 (1), pp. 1-14. [8] Shi H., Kokoeva M.V., Inouye K., Tzameli I., Yin H., Flier J.S. (2006). TLR4 links innate immunity and fatty acid-induced insulin resistance. J Clin Invest 116 (11), pp. 3015- 3025. [9] Sasu S., LaVerda D., Qureshi N., Golenbock D.T., Beasley D. (2001). Chlamydia pneumoniae and chlamydial heat shock protein 60 stimulate proliferation of human vascular smooth muscle cells via toll-like receptor 4 and

[10] Smiley S.T., King J.A., Hancock W.W. (2001). Fibrinogen stimulates macrophage chemokine secretion through toll-like receptor 4. J Immunol 167 (5), pp. 2887-2894. [11] Barry M.M., Mol C.D., Anderson W.F., Lee, J.S. (1994). Sequencing and modeling of anti-DNA immunoglobulin Fv domains. Comparison with crystal structures, J. Biol. Chem. 1994; 269: pp. 3623–3632. [12] Lima A.H., Souza P.R.M., Alencar N., Lameira J., Govender T., Kruger H.G., Maguire G.E.M., Alves C.N. (2012). Molecular modeling of T. rangeli, T. brucei gambiense, and T. evansi Sialidases in complex with the DANA inhibitor. Chem. Biol. Drug Des. 2012; 80: pp. 114-120. [13] Benkert P., Künzli M., Schwede T. (2009). QMEAN server for protein model quality estimation. Nucleic Acids Res. 2009; 37: W510-4. doi: 10.1093/nar/gkp322. [14] Novotny W.F., Girard T.J., Miletich J.P., Broze G.J.J. (1998). Platelets secrete a coagulation inhibitor functionally and antigenically similar to the lipoprotein associated coagulation inhibitor. Blood. 1998; 72(6): pp. 2020-2025. [15] Wiederstein M., Sippl M.J. (2007). Pro SA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Research. 35: pp. 407-410. [16] Prajapat R., Marwal A., Gaur R.K. (2014). Recognition of Errors in the Refinement and validation of three-dimensional structures of AC1 proteins of begomovirus strains by using ProSA-Web. J. of Viruses. ID 752656, .doi.org/10.1155/2014/752656. [17] Morris A.L., Mac Arthur M.W., Hutchinson E.G. and Thornton J.M. (1992). Stereo chemical quality of protein structure coordinates. Proteins. 12: pp. 345-364. [18] Bowie J.U., Luthy R., Eisenberg D. (1991). A method to identify protein sequences that fold into a known three-dimensional structure. Science. 253: pp. 164-170. [19] Luthy R., Bowie J.U., Eisenberg D. (1992). Assessment of protein models with three-dimensional profiles. Nature. 356: pp. 83-85. 2005; 351(4): pp. 810–823. [21] Zhang Z., Schwartz S., Wagner L., Miller W. (2000). A greedy algorithm for aligning DNA sequences. J. Comput. Biol. 2000; 7: pp. 203-214.

Department of Biochemistry, Kurukshetra University, Kurukshetra, Haryana, India