Rational Designing, Synthesis and Biological Evaluation of Novel Anti-diabetic Agents

Abstract

The present thesis work entitled “Rational Designing, Synthesis and Biological Evaluation of Novel Anti-diabetic Agents” is divided into 9 chapters as described below in brief. Chapter 1 describes the introduction about type II diabetes, its prevalence, conventional and novel treatments. Among the novel targets, dipeptidyl peptidase-4 (DPP-4) has been discussed in depth along with the marketed DPP-4 inhibitors. Novel longer acting molecules in the late development phase were also highlighted. Lastly, the roll of computer aided drug design in the current drug design and discovery has been briefly described. Quantitative structure–activity relationship (QSAR) and Pharmacophore modelling are the important tools of drug design. It also describes the aim and objectives of present work. The aim of this research is to design, synthesis and evaluate the new heterocyclic compounds as novel DPP-4 inhibitor hits and so as anti-diabetic agents. Chapter 2 gives detailed literature review of DPP-4 inhibitors. It discusses the chemistry and medicinal aspects of diverse heterocyclic scaffolds reported under peptidomimetic and nonpeptidomimetic class of DPP-4 inhibitors. Chapter 3 describes the in-silico design of novel DPP-4 inhibitor hits from various ligand and structure based drug design approaches. Initially, ligand based 3D-QSAR study of previously reported quinoline and isoquinoline derivatives was carried out to identify the important structural features of ligands responsible for the inhibitory activity. Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA) models were built for a set of 36 reported quinoline and isoquinoline based inhibitors using different alignment methods using SybylX1.2. The distill based alignment gave the best CoMFA and CoMSIA models with statistical significance and good predictive abilities. The q2 value of CoMFA model was found 0.803 while for CoMSIA model it was 0.826. The activities of the training set predicted by CoMFA/CoMSIA models fit well with the actual activities (r2 = 0.991 for the CoMFA; r2 = 0.983 for the CoMSIA). The predicted r2 (r2 pred) values of CoMFA and CoMSIA models were found 0.874 and 0.847 respectively which indicated that both models were good in predictions. On the basis of the CoMFA/CoMSIA model contour maps, significant regions for steric, electrostatic, hydrophobic, H-bond interactions were identified which contribute to bioactivity. The obtained results can be used as a guideline to design and predict new potent DPP-4 inhibitors, which could be an effective way to find novel leads for the development of antidiabetic drug. To further enrich designing, ligand and structure based pharmacophore modelling was performed through identifying specific regions on the structurally diverse DPP-4 inhibitor ligands beyond the quinoline/isoquinoline scaffold. The main aim behind this study was scaffold hopping from quinoline/isoquinoline to novel fused heterocycle which has not been explored so far for DPP-4 inhibition. Ligand based pharmacophore A was generated using HipHop module of Discovery Studio 2.1 while structure based pharmacophore B was constructed from LigandScout 3.0. Both models consist of five features. When compared with pharmacophore-B, pharmacophore-A contains two HBAs, two HYs and one PI features whereas on the other hand pharmacophore-B contains one HBA, one HBD, one HY, one RA and one PI features. Both pharmacophore models were validated for their ability in distinguishing active compounds from inactive compounds. Pharmacophore A outperformed than pharmacophore B with GH score of 0.865 and ROC-AUC of 80%. Further, validated pharmacophores were used for virtual screening of various databases such as Maybridge Hit Finder collection, Chemdiv and Asinex to identify the novel scaffolds/hits. The retrieved hit molecules were further filtered by several properties like maximum fit value, Lipinki rule of five, other drug like properties filter and high docking scores. Finally those hits that showed interactions with important amino acid residues were considered and reported as novel virtual hits to modify and design new DPP-4 inhibitors. Finally, new DPP-4 inhibitor hits were designed from the various conclusions drawn from 3DQSAR studies and pharmacophore modelling as well as looking into the literature review of DPP-4 inhibitors described in chapter 2. Total 72 novel derivatives were designed with various substitutions and all were screened against validated pharmacophore A and molecular docking. It was found that, out of all designed molecules, total 17 compounds showed acceptable insilico results comparable to marketed drugs, linagliptin and sitagliptin so chosen for actual synthesis. The P1 and P2 fragments of all the designed molecules got fit into the complimentary target pockets, S1 and S2 respectively of DPP-4 active site. Their in-silico ADMET properties were also predicted using OSIRIS property explorer and found acceptable in terms of drug score and no toxicities. Chapter 4 discusses synthesis and characterization of total 17 designed molecules under a series of triazine derivatives. All molecules were characterized by FT-IR, Mass (ESI-MS), 1H-NMR and 13C-NMR. Single crystal X-Ray crystallographic analysis of one of the compound 5p was also done to enrich the structure elucidation data. Purity of all final compounds were determined using HPLC chromatography and was found >95% in each compound. Also, experimental logP was determined for each molecule by traditional shake flask method and compared with predicted clogP values from the software. Both values were found almost near to each other for every compound. Chapter 5 describes the in-vitro and in-vivo pharmacological evaluation of 17 synthesized molecules for targeted DPP-4 enzyme inhibition and so as an anti-diabetic agents. Initial invitro screening of all molecules was carried out in DPP-4 enzyme inhibition assay. Compound 5q, and 5c, gave 53% and 48% DPP-4 inhibition respectively at 100 μM concentration so were chosen for IC50 determination. They were found 28.05 μM and 166.4 μM for compound 5q and 5c respectively. From the results of in-vitro screening, structure activity relationship (SAR) was discussed. Selectivity assay was performed against DPP-8 and DPP-9. Both compounds 5q and 5c were found more selective for DPP-4 compared to DPP-8 and DPP-9. Also they were checked for the cytotoxicity on Vero cell line by in-vitro MTT assay and found safe and noncytotoxic to the normal Vero cells. During in-vivo OGTT in C57BL/6 mice, both compounds 5q and 5c produced reduction in blood glucose levels. Glucose-induced blood glucose excursion was found significantly inhibited by 19.63%, 33.59% and 47.14% by compound 5q in a dose-dependent manner from 5mg/kg, 10 mg/kg to 20 mg/kg doses, respectively. Compound 5c was found to produce moderate reduction in blood glucose by 12.01%, 19.21%, and 23.72% at a dose of 5mg/kg, 20 mg/kg and 50 mg/kg respectively. The dose dependent effect was significant and potent in compound 5q than in compound 5c. Chapter 6 describes the summary and future prospects of the present thesis work.