The binding affinities (IC) reported for diverse structural and chemical classes of human β-secretase 1 (BACE-1) inhibitors in literature were modeled using multiple in silico ligand based modeling approaches and statistical techniques. The descriptor space encompasses simple binary molecular fingerprint, one- and two-dimensional constitutional, physicochemical, and topological descriptors, and sophisticated three-dimensional molecular fields that require appropriate structural alignments of varied chemical scaffolds in one universal chemical space. The affinities were modeled using qualitative classification or quantitative regression schemes involving linear, nonlinear, and deep neural network (DNN) machine-learning methods used in the scientific literature for quantitative-structure activity relationships (QSAR). In a departure from tradition, ∼20% of the chemically diverse data set (205 compounds) was used to train the model with the remaining ∼80% of the structural and chemical analogs used as part of an external validation (1273 compounds) and prospective test (69 compounds) sets respectively to ascertain the model performance. The machine-learning methods investigated herein performed well in both the qualitative classification (∼70% accuracy) and quantitative IC predictions (RMSE ∼ 1 log). The success of the 2D descriptor based machine learning approach when compared against the 3D field based technique pursued for hBACE-1 inhibitors provides a strong impetus for systematically applying such methods during the lead identification and optimization efforts for other protein families as well.

The bromodomain containing proteins TRIM24 (Tripartite motif containing protein 24) and BRPF1 (bromodomain and PHD finger containing protein 1) are involved in the epigenetic regulation of gene expression and have been implicated in human cancer. Overexpression of TRIM24 correlates with poor patient prognosis and BRPF1 is a scaffolding protein required for the assembly of histone acetyltransferase complexes, where the gene of MOZ (monocytic leukemia zinc finger protein) was first identified as a recurrent fusion partner in leukemia patients (8p11 chromosomal rearrangements). Here, we present the structure guided development of a series of N,N-dimethyl benzimidazolone bromodomain inhibitors through the iterative use of X-ray cocrystal structures. A unique binding mode enabled the design of a potent and selective inhibitor, 8i (IACS-9571) with low nanomolar affinities for TRIM24 and BRPF1 (ITC Kd = 31 nM and 14 nM, respectively). With its excellent cellular potency (EC50 = 50 nM) and favorable pharmacokinetic properties (F = 29%), 8i is a high-quality chemical probe for the evaluation of TRIM24 and/or BRPF1 bromodomain function in vitro and in vivo.

No abstract

The indole moiety in the delta-opioid antagonist, naltrindole (2, NTI), was employed as a scaffold to hold an "address" for interaction with the kappa-opioid receptor. The attachment of the address to the 5'-position of the indole moiety was based on superposition of NTI upon the kappa antagonist, norbinaltorphimine (1, norBNI). A variety of cationic groups were employed as a kappa address in an effort to investigate its interaction with the anionic address subsite, Glu297, on the kappa receptor. Some of the groups that were employed for this purpose were amines, amidines, guanidines, and quaternary ammonium. Members of the series were found to have a varying degree of kappa antagonist potency and kappa selectivity when tested in smooth muscle preparations. The 5'-guanidine derivative 12a (GNTI) was the most potent member of the series and had the highest kappa selectivity ratio. GNTI was 2 times more potent and 6-10-fold more selective than norBNI (1). In general, the order of potency in the series was: guanidines > amidines approximately quaternary ammonium > amines. The kappa antagonist potency appeared to be a function of a combination of the pK(a) and distance constraint of the cationic substituent of the ligand. Receptor binding studies were qualitatively in agreement with the pharmacological data. Molecular modeling studies on 12a suggested that the protonated N-17 and guanidinium groups of GNTI are associated with Asp138 (TM3) and Glu297 (TM6), respectively, while the phenolic hydroxyl may be involved in donor-acceptor interactions with the imidazole ring of His291. It was concluded that the basis for the high kappa selectivity of GNTI is related both to association with the nonconserved Glu297 residue and to unfavorable interactions with an equivalent position in mu- and delta-opioid receptors.

The nature and strength of the cation-pi interaction in protein-ligand binding are modeled by considering a series of nonbonded complexes involving N-substituted piperidines and substituted monocylic aromatics that mimic the delta-opioid receptor-ligand binding. High-level ab initio quantum mechanical calculations confirm the importance of such cation-pi interactions, whose intermolecular interaction energy ranges from -6 to -12 kcal/mol. A better understanding of the electrostatics, polarization, and other intermolecular interactions is obtained by appropriately decomposing the total interaction energy into their individual components. The energy decomposition analysis is also useful for parametrizing existing molecular mechanics force fields that could then account for energetic contributions arising out of cation-pi interactions in biomolecules. The present results further provide a framework for interpreting experimental results from point mutation reported for the delta-opioid receptor.