Computational approaches for the characterization of the dipeptidyl peptidase iv inhibition: applications to drug discovery, drug design and binding site similarity
- Autores: María José Ojeda Montes
- Directores de la Tesis: Gerard Pujadas Anguiano (dir. tes.), Santiago Garcia Vallvé (codir. tes.), Cristina Valls Bautista (codir. tes.)
- Lectura: En la Universitat Rovira i Virgili ( España ) en 2017
- Idioma: español
- Tribunal Calificador de la Tesis: Lluís Arola Ferrer (presid.), Albert Antolin Hernandez (secret.), Sonsoles Martín-Santamaría (voc.)
- Programa de doctorado: Programa de Doctorado en Nutrición y Metabolismo por la Universidad Rovira i Virgili
- Materias:
- Enlaces
- Tesis en acceso abierto en: TDX hdl.handle.net
- Resumen
The inhibition of dipeptidyl peptidase-IV (DPP-IV) has emerged over the last decade as one of the most effective treatments for type II diabetes mellitus with low risk of hypoglycemia and weight gain. The success of this target is reflected on the launch of eleven gliptins on the market in different countries since 2006. DPP-IV inhibitors avoid the degradation of glucagon-like peptide-1 (GLP-1) and gastric inhibitory polypeptide (GIP), thus amplifying the incretin effect (i.e., increasing insulin release from the β-cells and inhibiting glucagon secretion). Moreover, beyond glycemic control, the inhibition of this enzyme has an effect on cardiovascular and renal outcomes. Consequently, DPP-IV has been extensively studied and there is a considerable amount of information related to the protein structure and its inhibition, including a large number of crystal structures of the human DPP-IV and several structure-activity relationship (SAR) studies. In recent years, different virtual screening (VS) protocols have used this accurate information available for explaining how ligands interact with the DPP-IV binding site and to mine large databases of small molecules searching for new DPP-IV inhibitors. Nevertheless, most of the DPP-IV inhibitors identified have inhibitory bioactivities in the μM range and no measurement of their selectivity was performed over related enzymes like DPP8 and DPP9.
The present doctoral thesis has been therefore focused on: (a) the characterization of DPP-IV inhibition in order to suggest how virtual screening protocols may be improved either to favor the identification of potent and selective DPP-IV inhibitors or to look for new lead molecules; (b) the design of a computational strategy suitable for identifying new lead compounds with very low (or no) similarity to known actives in purchasable databases; (c) the demonstration that, at least partly, the described antidiabetic effect of different Ephedra species extracts is the result of the DPP-IV inhibitory bioactivity by ephedrine and the ephedrine-derivatives found in these extracts and (d) the analysis of the physico-chemical features shared by the DPP-IV and β2-adrenergic receptors binding sites and their comparison in order to evaluate if small molecules with dual bioactivity as DPP-IV inhibitors and β-blockers are possible.
First, critical insights into the DPP-IV inhibition were established from congeneric series of SAR studies for highlighting key activity changes. Thus, we disclosed the essential features that a ligand requires to interact with DPP-IV (Manuscript 1). As a result we were able to draw eight simple rules when searching for potent and selective DPP-IV inhibitors through VS: (1) a positively charged donor group (preferably a primary or secondary amine) that can establish salt bridges and/or hydrogen bonds with the N-terminal recognition region; (2) a group (preferably an aromatic ring) that hydrophobically contacts with the S1 pocket; (3) a negatively charge group (with a partial or net charge), an acceptor group or a phenyl ring with a halogen substituent in ortho position that can establish an electrostatic interaction with Arg125; (4) an aromatic ring near Phe357 to form additional π-π interactions to further increase activity and selectivity; (5) a negatively charge group (with a partial or net charge) that can establish an electrostatic interaction with Arg358 to further increase activity and selectivity; (6) an aromatic ring near Tyr547 to form additional π-π interactions to further increase activity and selectivity; (7) a negatively charge group (with a partial or net charge) that can establish an electrostatic interaction with Lys554; and (8) do not reach the S2’ pocket since this may result in a decrease activity. Among these interactions, the Glu205, Glu206 and Tyr662 residues form the N-terminal recognition region and the S1 pocket are considered to be the most important anchor points for getting basal DPP-IV inhibition.
Then, we have applied a VS workflow following protocols similar to those used by pharmaceutical industries in drug discovery, in order to focus on the most promising candidates from small molecules databases and remove those compounds that do not possess the required features. Thus, we have successfully developed a VS workflow identifying new scaffolds for DPP-IV inhibition with basal bioactivity and no similarity to known actives which therefore allow that the synthetic effort to focus solely on improving the core structure (Manuscript 2). The different sequential filters used included ADME, protein-ligand docking, pharmacophore screening and electrostatic and shape similarity analyses. It is noteworthy that a fingerprint similarity analysis was also performed at the beginning of the VS in order to find those structures that could contribute to new scaffolds and that were significantly different from co-crystallized inhibitors. The VS protocol for identifying these lead DPP-IV inhibitors was validated by an in silico analysis with the help of a set of actives and decoys, and by an in vitro analysis using an enzymatic assay. From the five compounds tested, ZINC02751967 and ZINC03823281, at 500 μM, reduced a 25.4% and a 7.6% the DPP-IV enzymatic activity, respectively. Although the bioactivity value is significantly lower than most DPP-IV inhibitors, these results showed that the VS workflow is able to identify novel scaffolds not structurally related to any known molecule that inhibits DPP-IV in purchasable databases of small molecules. Since the experimental measure of the IC50 value for ZINC02751967 showed that this compound only presents basal activity as DPP-IV inhibitor, different substituents were attached to ZINC02751967 with the purpose of improving the binding affinity and its selectivity according to the most important interactions described for DPP-IV. Thus, lead optimization of ZINC02751967 was addressed in order to improve the occupation of the hydrophobic S1 pocket and to reach the S2 extensive subsite. The resulting derivatives with proper ADME properties were finally selected.
A previous in silico study of our group (Guasch et al., 2012) predicted that ephedrine and five ephedrine-derivatives alkaloids inhibit DPP-IV, which could explain, at least partially, the described antidiabetic effect of Ephedra extracts. Our experimental results confirmed that all six molecules are DPP-IV inhibitors, with an IC50 ranging from 124 μM for ephedrine to 28.890 mM for N-methylpseudoephedrine (Manuscript 3). However, more important than their bioactivity values is the fact that these alkaloids are able to simultaneously inhibit DPP-IV and activate the adrenergic receptors. Unfortunately, several cardiovascular adverse effects are the result of the adrenergic receptor activation by ephedrine and ephedrine-derivatives and, as a consequence, their use has been severely restricted as food supplements by the EFSA and the FDA. Therefore, a second goal of this work was to use norephedrine as a lead compound for improving affinity and selectivity for DPP-IV while avoiding the adverse effects associated to norephedrine. Thus, norephedrine-derivatives reached the S2 extensive subsite and interacted with Tyr54. The resulting derivatives with proper ADME properties were finally selected.
