Enamine’s Lipid GPCR Library has been designed in collaboration with NQuiX. Over 6 400 compounds have been selected from off-the-shelf collection or specifically synthesized for this project from Enamine REAL screening Collection. They were designed to target the family of eight endothelial differentiation gene (EDG) receptors (S1P1–5 and LPA1–3). The compounds may also be appropriate for screening at GPR3, GPR6 and GPR12 orphan receptors given some TM bundle binding site similarity and/or GPR23, GPR92 and P2Y5 given their more recent classification as additional, albeit distinct, LPA receptors.

Library design

The compounds have been selected using a combination of ligand-based methods using chemical fingerprints, 2D pharmacophores and 3D shape/feature matches (Figure 1). The training active compound set included 1 393 compounds (870 acidic and 523 non-acidic), collected across 73 papers (Figure 2). They represent a combination of compounds for expanding SAR around known chemotypes and scaffold hops seeking novelty. The current set of compounds covers the non-acidic classes of ligand, with acidic compounds being designed via a different procedure and added to the library separately.

Fig. 1.Dedicated design of the compounds included in the Library (general scheme).

Fig. 2. Distribution of the compounds in the training set.

Examples of the molecules

Z1576067638

Z740728124

Z641956124

Z1303449800

Z1523110947

Z1303440017

Z1576459756

Although modulating GPCRs with orthosteric selective drug candidates has numerous success stories, in many cases, allosteric GPCR ligands are more effective for these biological targets. The small molecule allosteric modulation of GPCRs can promote a conformational change in the receptor that often has no noticeable effects, but can have a strong impact on the signaling pathway in the presence of an orthosteric endogenous ligand. On the other hand, GPCRs actuated by intractable stimuli can be modulated allosterically, opening a new opportunity for these currently unassailable targets. Two allosteric GPCRs modulators that have been introduced in the clinic (Cinacalcet and Maraviroc) are excellent evidence of prospects of investigation in this area. Another benefit of these ligands is their relatively low molecular weight and hence higher potential for oral bioavailability, in contrast to most orthosteric ligands. Moreover, allosteric sites are generally less conserved enabling development of actives with greater subtype selectivity than drugs targeting the conserved binding site.

Enamine’s Allosteric GPCR Library is a utile starting point for the drug discovery in the field of allosteric GPCR modulators.

Library design

Both ligand- and structure-based methods were used in design of the library. Ligand-based approach relied on using 3D pharmacophore search to the reference set of known ligands with high activity values (over 200 compounds from ChEMBL and BindingDB). The main emphasis has been done on allosteric modulators of class B GPCRs that are well established in drug discovery. Two pharmacophore models were created using the reference compounds sets and then validated with the training sets of actives and non-active ligands (Figure 1). Finally, over 6 200 small molecules were selected into Pharmacophore-based allosteric GPCR subset starting from Enamine MedChem set (710 K).

Fig. 1. Pharmacophore superposition of a reference active compound (in green) and a hit compound from the Library (in grey).

Structure-based approach was applied to human Glucagon receptor (GCGR) and human Corticotropin-releasing factor 1 receptor (CRF1R). The corresponding protein structures recorded in 5EE7 and 4K5Y PDB entries were optimized and used for the docking calculation with Glide (Schrödinger software). Compounds with at least 50% score-based efficacy as compared to the co-crystallized ligands were selected as hits and included in the targeted sets: h-GCGR set – 200 compounds; h-CRF1R set: 7 200 compounds (Figures 2 and 3).

Fig. 2. Binding mode of a hit (in red) obtained from the docking and the co-crystallized ligand (in green) of human Glucagon receptor.
 

Fig. 3. Superposition of a hit compound obtained after docking of CRF1R allosteric site with Enamine MedChem screening compounds (in green) and the co-crystallized ligand CP-376395 (in orange).

Examples of the molecules

Z2052759665

Z1256586621

Z2052158874

Z1299484172

Z2234781168

Z1192156186

Z1572741745

Z2326994277

The most straightforward approach to the design of kinase inhibitors relies on targeting ATP pocket. In fact, all compounds of this class approved by FDA (Copanlisib, Neratinib, Brigatinib, and Ribociclib) demonstrate this binding mode. The standard kinase interaction pattern consists of a hydrogen bond acceptor for the hinge region, heteroaromatic core with various substituents, and a second hydrogen acceptor for the conserved Lysine residue. This type of Inhibitors can be considered as ATP mimetics in the sense that they make interactions similar to that of ATP. Hydrogen bonds with Hinge Region formed by the adenosine moiety of ATP are crucial for effective binding (Figure 1, left). Analysis of a broad number of the kinase-inhibitor interactions has shown that similar hydrogen bonding patter is necessary for high inhibitory potency.

Fig. 1. ATP binding mode with hinge region (left) and examples of core structures that can mimic that mode (right).

Library design

Starting from the results of our previous study (Anticancer Agents Med. Chem. 2007, 7, 171–188) and detailed structure analysis of known and most potent kinase inhibitors, we have developed a series of unique structure filters aimed to identify potential inhibitors targeting ATP pocket. Molecular fragments able to form at least two hydrogen bonds with hinge region (Figure 1, right) have been analyzed, and appropriate topological models were developed for searching of directed inhibitors. The screening models were evaluated with a reference set of known kinase inhibitors (over 2 000 molecules with high inhibitory activity). Validated topological models were applied to over 3.2 M Enamine stock collection to produce Kinase Hinge Binders Libarary. The resulting set of compounds was evaluated with main emphasis on novel chemotypes. MedChem filters including PAINS and Ro5 PhysChem restrictions were applied as a prerequisite.

Examples of the molecules

Z1082776814

Z4521999208

Z1757787998

In response to high demand for new and small carboxylic acid we have created set of carefully selected specific molecules that strictly respond to FBDD requirements. Our carboxylic acid fragments are mainly consist from the compounds derived in the last three years from previously established at Enamine in-house program for design and producing diverse carboxylic acids to enlarge core Building Blocks Collection and diversification of decorating reagents.

Mainly this type of fragments has great potential for hit finding on new and difficult targets, e.g. protein-protein interaction, due to binding affinity to so called “hot-spot” residues. In addition there is probability of high solubility and good membrane permeability of discovered initial hit that often could be a limiting factor for phenotypic screening or in vivo activity of following investigation.

Key features

• High novelty and core diversity
• Fast follow-up with stock available analogues & synthesis
• Cherry-picking is available, customer preferred delivery format

Examples of the molecules in the library

Z2010010696

Z1201621939

Z234897401

Z963458200

Z2418194212

Z785812168

Z1445591040

Z1410957296