Fingerprint-Based Machine Learning Approach to Identify Potent and Selective 5-HT2BR Ligands

Abstract

The identification of subtype-selective GPCR (G-protein coupled receptor) ligands is a challenging task. In this study, we developed a computational protocol to find compounds with 5-HT2BR versus 5-HT1BR selectivity. Our approach employs the hierarchical combination of machine learning methods, docking, and multiple scoring methods. First, we applied machine learning tools to filter a large database of druglike compounds by the new Neighbouring Substructures Fingerprint (NSFP). This two-dimensional fingerprint contains information on the connectivity of the substructural features of a compound. Preselected subsets of the database were then subjected to docking calculations. The main indicators of compounds’ selectivity were their different interactions with the secondary binding pockets of both target proteins, while binding modes within the orthosteric binding pocket were preserved. The combined methodology of ligand-based and structure-based methods was validated prospectively, resulting in the identification of hits with nanomolar affinity and ten-fold to ten thousand-fold selectivities.

Biased Ligands Differentially Shape the Conformation of the Extracellular Loop Region in 5-HT2B Receptors

“…This phenomenon is also observed in the structural difference between the ergotamine-bound forms of 5HT 1B R and the 5HT 2B R; the shorter distance between TM5 and TM6 in the 5HT 2B R appear to play a role in the stronger bias of ergotamine at this receptor. Investigations into ligand selectivity for 5HT 2B R over 5HT 1B R found that selectivity is determined by non-conserved amino acid residues in the extracellular secondary binding pocket involving ECL2, as previously shown for other aminergic receptors [ 34 , 35 ].…”

Structure, Function, and Pharmaceutical Ligands of 5-Hydroxytryptamine 2B Receptor

“…Among them, compound 37 (Table 3) showed high binding affinity (K i = 23 nM) and antagonist activity for the 5-HT 2B R, with 12-fold binding selectivity and 170-fold functional selectivity for the 5-HT 2B R over the 5-HT 2C R [150]. In 2018, Rataj et al combined a fingerprint-based machine learning approach and molecular docking that led to the identification of compound 38 (Table 3) [151]. Notably, compound 38 showed potent binding affinity (K i = 0.3 nM) and >10,000-fold selectivity over other five tested 5-HTRs [151].…”

“…In 2018, Rataj et al combined a fingerprint-based machine learning approach and molecular docking that led to the identification of compound 38 (Table 3) [151]. Notably, compound 38 showed potent binding affinity (K i = 0.3 nM) and >10,000-fold selectivity over other five tested 5-HTRs [151]. In addition, in 2018, as a proof-of-concept of halogen bonding in designing 5-HT 2B R ligands, a series of halogensubstituted analogs of doxepin (K i = 25.3 nM for the 5-HT 2B R) were synthesized.…”

“…Notably, co-crystal ligands fit well in this pharmacophore model (Figure 11b). Most of the 5-HT2BR ligands bear positively charged nitrogens or polar NH K i = 23 nM 38 [151] Pharmaceuticals 2021, 14, x FOR PEER REVIEW 22 of 32 the 5-HT2R affinity and simultaneously reduce the affinity to adenosine receptors (ARs) [150]. A SAR study on a series of adenosine derivatives assisted them in identifying several antagonists of the 5-HT2B/5-HT2C receptors with selectivity over ARs.…”

Frontiers in Computational Chemistry for Drug Discovery

“…GPCRs are also the subject examined in the contribution by Bojarski, Keserü, and coworkers [ 8 ]. In this study, they report a computational protocol to find compounds with selectivity between two receptors: 5-HT 2B versus 5-HT 1B .…”