A cheminformatic study on chemical space characterization and diversity analysis of 5-LOX inhibitors
Graphical abstract
Introduction
Leukotrienes (LTs) are a class of potent inflammatory mediators produced by 5-LOX from arachidonic acid. The overproduction of LT causes severe allergic and inflammatory conditions. Inhibition of LT formation, therefore, has valuable therapeutic advantages in excessive and chronic inflammatory and allergic responses like asthma, rhinitis, rheumatoid arthritis, gastroesophageal reflux disease, and atherosclerosis [1]. Moreover, the role of LTs in carcinogenesis has also been documented recently [[2], [3], [4], [5]]. Because traditional anti-inflammatory treatments, like treatment with NSAID, are still far from effective in many of these diseases, new and improved approaches are being actively sought to counter these conditions. Methods that inhibit the biosynthesis of LTs are, therefore, of interest as potential therapies for such diseases. The non-heme iron-containing enzyme 5-LOX catalyzes the first step in the biosynthesis of leukotriene. 5-LOX selective inhibition offers a definite means of reducing the effects of all leukotrienes, and such an inhibitor could form a new class of therapeutic agents [6].
Scientists have been discovered several types of 5-LOX inhibitors depending upon on the mechanism of action such as (i) iron-chelating inhibitors, (ii) competitive, reversible (non-redox) inhibitors (iii) inhibitors of the 5-lipoxygenase activating protein (FLAP) and (IV) redox inhibitors [7]. These large amounts of structure-activity data of 5-LOX inhibitors obtained as a result of general screening, high-throughput screening (HTS) and combinatorial chemistry is collected and deposited not only in research papers but also in public domain databases such as ChEMBL [8], PubChem [9], BindingDB [10], etc. These compound collections contribute to the development of ’Biologically Relevant Chemical Space (BRCS) of 5-LOX. Cheminformatic characterization of these chemical spaces is a significant step towards virtual or experimental testing to recognize novel biologically active molecules [[11], [12], [13]]. Besides, data mining of these chemical spaces could help the development of models that can be beneficial to predict the activity of a new compound [14]. Computational approaches have been playing an essential role in identifying 5-LOX inhibitors without any limitations of known 5-LOX inhibitors found. With the increase in the quantity and complexity of 5-LOX inhibitor’s structural data, chemical space analysis, and cheminformatic modeling are becoming increasingly important to understand and predict the interactions between inhibitors and 5-LOX protein.
So, visualization and characterization of biologically relevant chemical space of 5-LOX and their compound distributions are critical to medicinal chemists as it can assist in understanding molecular features that are pharmaceutically important [11,15]. Different computational data mining and visualization methods and machine learning algorithms that originally developed for computer science is now used mainly for the understanding the chemical space of biological interest [13,[16], [17], [18], [19]]. Standard statistical and classification techniques can also be used to organize datasets and evaluate the chemical neighborhood of potent hit [[20], [21], [22], [23]]. With the help of these methods, cheminformatic characterization of chemical space of inhibitors of various targets or disease has been reported for the last few decades by many eminent scientists, especially the team lead by Jose L. Medina-Franco [12,[24], [25], [26], [27], [28], [29]].
In recent years, several Structure-Activity Relationship (SAR), Quantitative Structure-Activity Relationship (QSAR), and other predictive models of 5-LOX and its activating protein ’FLAP’ inhibitors have been developed [[30], [31], [32], [33], [34]]. However, there are no systematic cheminformatic studies on the structural diversity and chemical space distribution analysis of 5-LOX and FLAP inhibitor’s chemical space. Therefore, exploring and navigating the biologically relevant 5-LOX and FLAP inhibitor’s chemical space is of the utmost importance. It can provide an opportunity for the analysis and enumeration of the compound in the chemical space of 5-LOX and FLAP.
So, in this work, we have tried to locate the chemical space of 5-LOX and FLAP currently stored in a major public databases ’ChEMBL’ and characterized these spaces using multiple criteria including physicochemical properties (PCP), structural fingerprints, and molecular scaffolds. In all these studies, we have compared the chemical space of 5-LOX, and FLAP inhibitors to the dataset of FDA approved drugs from ’DrugBank’ because this FDA approved drug database is a commonly used reference compound database in drug discovery campaigns. We have also tried to understand and validate the Target focused virtual LOX library created and offered by the Enamine database by comparing it to the 5-LOX dataset.
Section snippets
5-LOX and FLAP inhibitor dataset
The study of the complexity and diversity of large molecular databases allows for the creation of high-quality leads. It thus increases the performance of real and virtual drug design compound libraries. Here we too tried to understand the complexity and diversity of 5-LOX and FLAP chemical space. For this purpose, molecules in the ChEMBL database, which are experimentally tested for 5-LOX and FLAP protein inhibitory potency, were chosen. The chemical structures and activity data of 5-LOX and
Physicochemical properties of 5-LOX chemical space
The property distribution was evaluated with SPSS box plots. Fig. 2 shows the box and whisker plots spread out of the basic PCP of all four datasets. The primary portion of the chart is a box, which indicates the interquartile range, i.e., it shows where the central portion of the data is. The first quartile, Q1 (25% mark), and the third quartile, Q3 (75% mark), are situated at lower and upper ends of the box, respectively. The ’median’ represents a horizontal bar in the middle of the box. The
Conclusion
This study discusses a comprehensive cheminformatic characterization of the chemical space of 5-LOX and FLAP inhibitors obtained from the CHEMBL database by comparing it with the Approved Drug space. Also, we have validated the LOX virtual library created by the Enamine database by comparing it to the 5-LOX dataset. Analysis of the distributions of PCPs like HBA, HDB, and TPSA indicated that the compounds screened as inhibitors of 5-LOX and FLAP are, in general, less or comparable polar than an
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
The author, T.K. Shameera Ahamed, express her sincere gratitude to Human Resource Development Group’s Council of Scientific & Industrial Research (CSIR), India, for the financial support.
References (53)
- et al.
Arachidonic acid stimulates prostate cancer cell growth: critical role of 5-lipoxygenase
Biochem. Biophys. Res. Commun.
(1997) Inhibitors of 5-lipoxygenase: a therapeutic potential yet to be fully realized?
Eur. J. Med. Chem.
(1999)- et al.
Computational mapping tools for drug discovery
Drug Discov. Today
(2009) - et al.
Data structures and computational tools for the extraction of SAR information from large compound sets
Drug Discov. Today
(2010) - et al.
Chapter one - molecular modeling and chemoinformatics to advance the development of modulators of epigenetic targets: a focus on DNA methyltransferases
- et al.
5-Lipoxygenase as a drug target: a review on trends in inhibitors structural design, SAR and mechanism based approach
Bioorg. Med. Chem.
(2019) - et al.
Drug discovery approaches targeting 5-lipoxygenase-activating protein (FLAP) for inhibition of cellular leukotriene biosynthesis
Eur. J. Med. Chem.
(2018) - et al.
Insights from pharmacological similarity of epigenetic targets in epipolypharmacology
Drug Discov. Today
(2018) - et al.
Computational approaches for combinatorial library design and molecular diversity analysis
Curr. Opin. Chem. Biol.
(1997) - et al.
Lipoxygenase and leukotriene pathways: biochemistry, biology, and roles in disease
Chem. Rev.
(2011)
Overexpression of 5-lipoxygenase in colon polyps and cancer and the effect of 5-LOX Inhibitors in vitro and in a murine model
Clin. Canc. Res.
5-Lipoxygenase, a marker for early pancreatic intraepithelial neoplastic lesions
Canc. Res.
Loss of the Alox5 gene impairs leukemia stem cells and prevents chronic myeloid leukemia
Nat. Genet.
Brash, lipoxygenases: occurrence, functions , catalysis , and acquisition of substrate
J. Biol. Chem.
ChEMBL: a large-scale bioactivity database for drug discovery
Nucleic Acids Res.
PubChem substance and compound databases
Nucleic Acids Res.
BindingDB: a web-accessible database of experimentally determined protein-ligand binding affinities
Nucleic Acids Res.
The chemical space project
Acc. Chem. Res.
Epigenetic relevant chemical space: a chemoinformatic characterization of inhibitors of DNA methyltransferases
RSC Adv.
Chemoinformatic analysis of combinatorial libraries, drugs, natural products, and molecular libraries small molecule repository
J. Chem. Inf. Model.
Data mining and machine learning models for predicting drug likeness and their disease or organ category
Front. Chem.
Interrogating novel areas of chemical space for drug discovery using chemoinformatics
Drug Dev. Res.
Computational methods in drug discovery
Pharmacol. Rev.
Navigating chemical space for biology and medicine
Nature
Recent applications of deep learning and machine intelligence on in silico drug discovery: methods, tools and databases
Briefings Bioinf.
Cited by (1)
Concepts and applications of chemical fingerprint for hit and lead screening
2022, Drug Discovery TodayCitation Excerpt :For example, CSNAP3D, proposed by Lo et al., combines shape and pharmacophore characteristics to measure the 3D similarities among ligands and showed a significant improvement in target prediction.71 Recently, Ahamed et al. proposed a novel methodology to analyze the cheminformatic characteristics and diversity of chemical space of 5-LOX through combining fingerprints of ECFP, MACC, PC, molecular scaffolds, and basic pharmaceutical relevance property descriptors.72 Capecchi et al. designed a universal fingerprint (suitable for drugs, biomolecules, and the metabolome) to describe and search chemical space called MAP4 (MinHashed atom-pair fingerprint up to a diameter of four bonds) by combining advantages of AP (i.e., ability to describe large molecules and with excellent perception of molecular shape) and circular fingerprints (i.e., ability to encode molecular substructures in detail, performing well in small-molecule benchmarking studies).73