Local motifs involved in the canonical structure of the ligand-binding domain in the nuclear receptor superfamily

A unified solution for different scenarios of predicting drug-target interactions via triple matrix factorization

Background

During the identification of potential candidates, computational prediction of drug-target interactions (DTIs) is important to subsequent expensive validation in wet-lab. DTI screening considers four scenarios, depending on whether the drug is an existing or a new drug and whether the target is an existing or a new target. However, existing approaches have the following limitations. First, only a few of them can address the most difficult scenario (i.e., predicting interactions between new drugs and new targets). More importantly, none of the existing approaches could provide the explicit information for understanding the mechanism of forming interactions, such as the drug-target feature pairs contributing to the interactions.

Results

In this paper, we propose a Triple Matrix Factorization-based model (TMF) to tackle these problems. Compared with former state-of-the-art predictive methods, TMF demonstrates its significant superiority by assessing the predictions on four benchmark datasets over four kinds of screening scenarios. Also, it exhibits its outperformance by validating predicted novel interactions. More importantly, by using PubChem fingerprints of chemical structures as drug features and occurring frequencies of amino acid trimer as protein features, TMF shows its ability to find out the features determining interactions, including dominant feature pairs, frequently occurring substructures, and conserved triplet of amino acids.

Conclusions

Our TMF provides a unified framework of DTI prediction for all the screening scenarios. It also presents a new insight for the underlying mechanism of DTIs by indicating dominant features, which play important roles in the forming of DTI.

Electronic supplementary material

The online version of this article (10.1186/s12918-018-0663-x) contains supplementary material, which is available to authorized users.

The xenobiotic sensor PXR in a marine flatfish species (Solea senegalensis): Gene expression patterns and its regulation under different physiological conditions

The pregnane X receptor (PXR) is a nuclear receptor belonging to the NR1I sub-family and a known master regulator of xenobiotic metabolism. New roles have been recently proposed in mammals through its activation by vitamin K (VK) such as regulation of glucose metabolism, bone homeostasis, reproduction, neuronal development and cognitive capacities. In marine fish species little is known about PXR and its potential roles. Here, expression patterns of pxr transcripts and conservation of protein domains were determined in the Senegalese sole (Solea senegalensis), a marine flatfish model species in aquatic ecotoxicology. In addition to a full coding sequence transcript (sspxr1), two variants lacking DNA and/or ligand binding domains (sspxr2 and sspxr3) were also identified. The expression of sspxr1 during early development and in adult tissues was ubiquitous, but highest levels were observed in liver, intestine and skin. Expression was also detected by in situ hybridization in chondrocytes and cells from the granular and inner nuclear layers in three month old fish. Finally, sspxr1 expression was shown to be differentially regulated under physiological conditions related with fasting, VK and warfarin metabolism. The present work provides new and basic knowledge regarding pxr sequence and expression patterns in a marine flatfish species to unveil the potential impact of xenobiotics on marine fish physiology, and will allow a better and more ecosystemic environmental risk assessment of different pollutants over the marine environments with the development of reporter assays using PXR sequences from evolutionary distantly marine species (such as vertebrate and invertebrate marine species).

Antagonist-perturbation mechanism for activation function-2 fixed motifs: active conformation and docking mode of retinoid X receptor antagonists

HX531, which contains a dibenzodiazepine skeleton, is one of the first retinoid X receptor (RXR) antagonists. Functioning via RXR-PPARγ heterodimer, this compound is receiving a lot of attention as a therapeutic drug candidate for diabetic disease controlling differentiation of adipose tissue. However, the active conformation of HX531 for RXRs is not well established. In the present study, quantum mechanics calculations and molecular mechanical docking simulations were carried out to precisely study the docking mode of HX531 with the human RXRα ligand-binding domain, as well as to provide a new approach to drug design using a structure-based perspective. It was suggested that HX531, which has the R configuration for the bent dibenzodiazepine plane together with the equatorial configuration for the N-methyl group attached to the nitrogen atom in the seven-membered diazepine ring, is a typical activation function-2 (AF-2) fixed motif perturbation type antagonist, which destabilizes the formation of AF-2 fixed motifs. On the other hand, the docking simulations supported the experimental result that LG100754 is an RXR homodimer antagonist and an RXR heterodimer agonist.

Identifying the receptor subtype selectivity of retinoid X and retinoic acid receptors via quantum mechanics

Understanding and identifying the receptor subtype selectivity of a ligand is an important issue in the field of drug discovery. Using a combination of classical molecular mechanics and quantum mechanical calculations, this report assesses the receptor subtype selectivity for the human retinoid X receptor (hRXR) and retinoic acid receptor (hRAR) ligand‐binding domains (LBDs) complexed with retinoid ligands. The calculated energies show good correlation with the experimentally reported binding affinities. The technique proposed here is a promising method as it reveals the origin of the receptor subtype selectivity of selective ligands.

A ligand-entry surface of the nuclear receptor superfamily consists of the helix H3 of the ligand-binding domain

We successfully simulated receptor-ligand complex holo-form formation using the human retinoid X receptor-α ligand-binding domain (LBD) and its natural ligand, 9-cis retinoic acid. The success of this simulation was strongly dependent on the findings for an initial structure between the apo-LBD and the ligand as well as the discovery of the driving forces underlying the ligand-trapping and subsequent ligand-induction processes. Here, we would like to propose the "helix H3 three-point initial-binding hypothesis," which was instrumental in simulating the nuclear receptor (NR) superfamily. Using this hypothesis, we also succeeded in simulating holo-form formation of the human retinoic acid receptor-γ LBD and its natural ligand, all-trans retinoic acid. It is hoped that this hypothesis will facilitate novel understanding of both the ligand-trapping mechanism and the simultaneous C-terminal folding process in NR LBDs, as well as provide a new approach to drug design using a structure-based perspective.

Predicting drug-target interaction for new drugs using enhanced similarity measures and super-target clustering

Predicting drug-target interaction using computational approaches is an important step in drug discovery and repositioning. To predict whether there will be an interaction between a drug and a target, most existing methods identify similar drugs and targets in the database. The prediction is then made based on the known interactions of these drugs and targets. This idea is promising. However, there are two shortcomings that have not yet been addressed appropriately. Firstly, most of the methods only use 2D chemical structures and protein sequences to measure the similarity of drugs and targets respectively. However, this information may not fully capture the characteristics determining whether a drug will interact with a target. Secondly, there are very few known interactions, i.e. many interactions are "missing" in the database. Existing approaches are biased towards known interactions and have no good solutions to handle possibly missing interactions which affect the accuracy of the prediction. In this paper, we enhance the similarity measures to include non-structural (and non-sequence-based) information and introduce the concept of a "super-target" to handle the problem of possibly missing interactions. Based on evaluations on real data, we show that our similarity measure is better than the existing measures and our approach is able to achieve higher accuracy than the two best existing algorithms, WNN-GIP and KBMF2K. Our approach is available at http://web.hku.hk/∼liym1018/projects/drug/drug.html or http://www.bmlnwpu.org/us/tools/PredictingDTI_S2/METHODS.html.