KNeMAP: a network mapping approach for knowledge-driven comparison of transcriptomic profiles.

MOTIVATION: Transcriptomic data can be used to describe the mechanism of action (MOA) of a chemical compound. However, omics data tend to be complex and prone to noise, making the comparison of different datasets challenging. Often, transcriptomic profiles are compared at the level of individual gene expression values, or sets of differentially expressed genes. Such approaches can suffer from underlying technical and biological variance, such as the biological system exposed on or the machine/method used to measure gene expression data, technical errors and further neglect the relationships between the genes. We propose a network mapping approach for knowledge-driven comparison of transcriptomic profiles (KNeMAP), which combines genes into similarity groups based on multiple levels of prior information, hence adding a higher-level view onto the individual gene view. When comparing KNeMAP with fold change (expression) based and deregulated gene set-based methods, KNeMAP was able to group compounds with higher accuracy with respect to prior information as well as is less prone to noise corrupted data. RESULT: We applied KNeMAP to analyze the Connectivity Map dataset, where the gene expression changes of three cell lines were analyzed after treatment with 676 drugs as well as the Fortino et al. dataset where two cell lines with 31 nanomaterials were analyzed. Although the expression profiles across the biological systems are highly different, KNeMAP was able to identify sets of compounds that induce similar molecular responses when exposed on the same biological system. AVAILABILITY AND IMPLEMENTATION: Relevant data and the KNeMAP function is available at: https://github.com/fhaive/KNeMAP and 10.5281/zenodo.7334711.

AICAR and compound C negatively modulate HCC-induced primary human hepatic stellate cell activation in vitro.

Tumor stroma and microenvironment have been shown to affect hepatocellular carcinoma (HCC) growth, with activated hepatic stellate cells (HSC) as a major contributor in this process. Recent evidence suggests that the energy sensor adenosine monophosphate-activated kinase (AMPK) may mediate a series of essential processes during carcinogenesis and HCC progression. Here, we investigated the effect of different HCC cell lines with known TP53 or CTNBB1 mutations on primary human HSC activation, proliferation, and AMPK activation. We show that conditioned media obtained from multiple HCC cell lines differently modulate human hepatic stellate cell (hHSC) proliferation and hHSC AMPK activity in a paracrine manner. Pharmacological treatment of hHSC with AICAR and Compound C inhibited the HCC-induced proliferation/activation of hHSC through AMPK-dependent and AMPK-independent mechanisms, which was further confirmed using mouse embryonic fibroblasts (MEFs) deficient of both catalytic AMPKα isoforms (AMPKα1/α2-/-) and wild type (wt) MEF. Both compounds induced S-phase cell-cycle arrest and, in addition, AICAR inhibited the mTORC1 pathway by inhibiting phosphorylation of 4E-BP1 and S6 in hHSC and wt MEF. Data mining of the Cancer Genome Atlas (TCGA) and the Liver Cancer (LICA-FR) showed that AMPKα1 (PRKAA1) and AMPKα2 (PRKAA2) expression differed depending on the mutation (TP53 or CTNNB1), tumor grading, and G1-G6 classification, reflecting the heterogeneity in human HCC. Overall, we provide evidence that AMPK modulating pharmacological agents negatively modulate HCC-induced hHSC activation and may therefore provide a novel approach to target the mutual, tumor-promoting interactions between hHSC and HCC.NEW & NOTEWORTHY HCC is marked by genetic heterogeneity and activated hepatic stellate cells (HSC) are considered key players during HCC development. The paracrine effect of different HCC cell lines on the activation of primary hHSC was accompanied by differential AMPK activation depending on the HCC line used. Pharmacological treatment inhibited the HCC-induced hHSC activation through AMPK-dependent and AMPK-independent mechanisms. This heterogenic effect on HCC-induced AMPK activation was confirmed by data mining TCGA and LICA-FR databases.

Binding Site Interactions of Modulators of Breast Cancer Resistance Protein, Multidrug Resistance-Associated Protein 2, and P-Glycoprotein Activity.

ATP-binding cassette (ABC)-transporters protect tissues by pumping their substrates out of the cells in many physiological barriers, such as the blood-brain barrier, intestine, liver, and kidney. These substrates include various endogenous metabolites, but, in addition, ABC transporters recognize a wide range of compounds, therefore affecting the disposition and elimination of clinically used drugs and their metabolites. Although numerous ABC-transporter inhibitors are known, the underlying mechanism of inhibition is not well characterized. The aim of this study is to deepen our understanding of transporter inhibition by studying the molecular basis of ligand recognition. In the current work, we compared the effect of 44 compounds on the active transport mediated by three ABC transporters: breast cancer resistance protein (BCRP and ABCG2), multidrug-resistance associated protein (MRP2 and ABCC2), and P-glycoprotein (P-gp and ABCB1). Eight compounds were strong inhibitors of all three transporters, while the activity of 36 compounds was transporter-specific. Of the tested compounds, 39, 25, and 11 were considered as strong inhibitors, while 1, 4, and 11 compounds were inactive against BCRP, MRP2, and P-gp, respectively. In addition, six transport-enhancing stimulators were observed for P-gp. In order to understand the observed selectivity, we compared the surface properties of binding cavities in the transporters and performed structure-activity analysis and computational docking of the compounds to known binding sites in the transmembrane domains and nucleotide-binding domains. Based on the results, the studied compounds are more likely to interact with the transmembrane domain than the nucleotide-binding domain. Additionally, the surface properties of the substrate binding site in the transmembrane domains of the three transporters were in line with the observed selectivity. Because of the high activity toward BCRP, we lacked the dynamic range needed to draw conclusions on favorable interactions; however, we identified amino acids in both P-gp and MRP2 that appear to be important for ligand recognition.

Adenosine analogs bearing phosphate isosteres as human MDO1 ligands.

The human O-acetyl-ADP-ribose deacetylase MDO1 is a mono-ADP-ribosylhydrolase involved in the reversal of post-translational modifications. Until now MDO1 has been poorly characterized, partly since no ligand is known besides adenosine nucleotides. Here, we synthesized thirteen compounds retaining the adenosine moiety and bearing bioisosteric replacements of the phosphate at the ribose 5'-oxygen. These compounds are composed of either a squaryldiamide or an amide group as the bioisosteric replacement and/or as a linker. To these groups a variety of substituents were attached such as phenyl, benzyl, pyridyl, carboxyl, hydroxy and tetrazolyl. Biochemical evaluation showed that two compounds, one from both series, inhibited ADP-ribosyl hydrolysis mediated by MDO1 in high concentrations.

Structural Isosteres of Phosphate Groups in the Protein Data Bank.

We developed a computational workflow to mine the Protein Data Bank for isosteric replacements that exist in different binding site environments but have not necessarily been identified and exploited in compound design. Taking phosphate groups as examples, the workflow was used to construct 157 data sets, each composed of a reference protein complexed with AMP, ADP, ATP, or pyrophosphate as well other ligands. Phosphate binding sites appear to have a high hydration content and large size, resulting in U-shaped bioactive conformations recurrently found across unrelated protein families. A total of 16 413 replacements were extracted, filtered for a significant structural overlap on phosphate groups, and sorted according to their SMILES codes. In addition to the classical isosteres of phosphate, such as carboxylate, sulfone, or sulfonamide, unexpected replacements that do not conserve charge or polarity, such as aryl, aliphatic, or positively charged groups, were found.

Exploring the structure-activity relationships of ABCC2 modulators using a screening approach.

ABCC2 is a transporter with key influence on liver and kidney pharmacokinetics. In order to explore the structure-activity relationships of compounds that modulate ABCC2, and by doing so gain insights into drug-drug interactions, we screened a library of 432 compounds for modulators of radiolabeled ß-estradiol 17-(ß-d-glucuronide) (EG) and fluorescent 5(6)-carboxy-2',7'-dichlorofluorescein transport (CDCF) in membrane vesicles. Following the primary screen at 80µM, dose-response curves were used to investigate in detail 86 compounds, identifying 16 low µM inhibitors and providing data about the structure-activity relationships in four series containing 19, 24, 10, and eight analogues. Measurements with the CDCF probe were consistently more robust than for the EG probe. Only one compound was clearly probe-selective with a 50-fold difference in the IC50s obtained by the two assays. We built 24 classification models using the SVM and fused-XY Kohonen methods, revealing molecular descriptors related to number of rings, solubility and lipophilicity as important to distinguish inhibitors from inactive compounds. This study is to the best of our knowledge the first to provide details about structure-activity relationships in ABCC2 modulation.

SVM classification and CoMSIA modeling of UGT1A6 interacting molecules.

The human UDP-glucuronosyltransferase 1A6 (UGT1A6) plays important roles in elimination of many xenobiotics, including drugs. We have experimentally assessed inhibitory properties of 46 compounds toward UGT1A6 catalyzing the glucuronidation of 1-naphthol and built models for predicting compounds interactions with the enzyme. The tested compounds were divided into a training set (n = 31; evaluated by 10-fold cross-validation) and an external test set (n = 15), both of which yielded similar accuracies (80-81%) and Matthews correlation coefficients (0.61-0.63) when classified using support vector machines. Comparative molecular similarity index analysis (CoMSIA) modeling was conducted for nine methods of compound alignment. The most predictive CoMSIA model was analyzed in the light of a homology modeled UGT1A6 structure, with leave-one-out cross-validation, yielding a q² of 0.62 and r² of 0.91 on the training set and a r²(pred) of 0.82 on the test set. The CoMSIA contour plots highlighted the importance of H-bond donors and electrostatic field interactions, accounting for 28% and 25% contribution of the model, respectively.

Exploring cooperative molecular contacts using a PostgreSQL database system.

Cooperative molecular contacts play an important role in protein structure and ligand binding. Here, we constructed a PostgreSQL database that stores structural information in the form of atomic environments and allows flexible mining of molecular contacts. Taking the Ser-His-Asp/Glu catalytic triad as a first test case, we demonstrate that the presence of a carboxylate oxygen atom in the vicinity of a His is associated with shorter Ser-OH..N-His bond in the PDB30 subset. We prospectively mine catalytic triads in unannotated proteins, suggesting catalytic functions for unannotated proteins. As a second test case, we demonstrate that this database system can include ligand atoms, represented by Sybyl atom types, by evaluating the proportion of counter-ions for ligand carboxylate oxygens.

Analysis of Biologics Molecular Descriptors towards Predictive Modelling for Protein Drug Development Using Time-Gated Raman Spectroscopy.

Pharmaceutical proteins, compared to small molecular weight drugs, are relatively fragile molecules, thus necessitating monitoring protein unfolding and aggregation during production and post-marketing. Currently, many analytical techniques take offline measurements, which cannot directly assess protein folding during production and unfolding during processing and storage. In addition, several orthogonal techniques are needed during production and market surveillance. In this study, we introduce the use of time-gated Raman spectroscopy to identify molecular descriptors of protein unfolding. Raman spectroscopy can measure the unfolding of proteins in-line and in real-time without labels. Using K-means clustering and PCA analysis, we could correlate local unfolding events with traditional analytical methods. This is the first step toward predictive modeling of unfolding events of proteins during production and storage.

Installation of an aryl boronic acid function into the external section of N-aryl-oxazolidinones: Synthesis and antimicrobial evaluation.

N-aryl-oxazolidinones is a prominent family of antimicrobials used for treating infections caused by clinically prevalent Gram-positive bacteria. Recently, boron-containing compounds have displayed intriguing potential in the antibiotic discovery setting. Herein, we report the unprecedented introduction of a boron-containing moiety such as an aryl boronic acid in the external region of the oxazolidinone structure via a chemoselective acyl coupling reaction. As a result, we accessed a series of analogues with a distal aryl boronic pharmacophore on the oxazolidinone scaffold. We identified that a peripheric linear conformation coupled with freedom of rotation and no further substitution on the external aryl boronic ring, an amido linkage with hydrogen bonding character, in addition to a para-relative disposition between boronic group and linker, are the optimal combination of structural features in this series for antimicrobial activity. In comparison to linezolid, the analogue comprising all those features, compound 20b, displayed levels of antimicrobial activity augmented by an eight-fold to a thirty-two-fold against a panel of Gram-positive strains, and a near one hundred-fold against Escherichia coli JW5503, a Gram-negative mutant strain with a defective efflux capability.

Comparison of three preprocessing filters efficiency in virtual screening: identification of new putative LXRbeta regulators as a test case.

In silico screening methodologies are widely recognized as efficient approaches in early steps of drug discovery. However, in the virtual high-throughput screening (VHTS) context, where hit compounds are searched among millions of candidates, three-dimensional comparison techniques and knowledge discovery from databases should offer a better efficiency to finding novel drug leads than those of computationally expensive molecular dockings. Therefore, the present study aims at developing a filtering methodology to efficiently eliminate unsuitable compounds in VHTS process. Several filters are evaluated in this paper. The first two are structure-based and rely on either geometrical docking or pharmacophore depiction. The third filter is ligand-based and uses knowledge-based and fingerprint similarity techniques. These filtering methods were tested with the Liver X Receptor (LXR) as a target of therapeutic interest, as LXR is a key regulator in maintaining cholesterol homeostasis. The results show that the three considered filters are complementary so that their combination should generate consistent compound lists of potential hits.

Tamoxifen mechanically deactivates hepatic stellate cells via the G protein-coupled estrogen receptor.

Tamoxifen has been used for many years to target estrogen receptor signalling in breast cancer cells. Tamoxifen is also an agonist of the G protein-coupled estrogen receptor (GPER), a GPCR ubiquitously expressed in tissues that mediates the acute response to estrogens. Here we report that tamoxifen promotes mechanical quiescence in hepatic stellate cells (HSCs), stromal fibroblast-like cells whose activation triggers and perpetuates liver fibrosis in hepatocellular carcinomas. This mechanical deactivation is mediated by the GPER/RhoA/myosin axis and induces YAP deactivation. We report that tamoxifen decreases the levels of hypoxia-inducible factor-1 alpha (HIF-1α) and the synthesis of extracellular matrix proteins through a mechanical mechanism that involves actomyosin-dependent contractility and mechanosensing of tissue stiffness. Our results implicate GPER-mediated estrogen signalling in the mechanosensory-driven activation of HSCs and put forward estrogenic signalling as an option for mechanical reprogramming of myofibroblast-like cells in the tumour microenvironment. Tamoxifen, with half a century of safe clinical use, might lead this strategy of drug repositioning.

Predictive classification models and targets identification for betulin derivatives as Leishmania donovani inhibitors.

Betulin derivatives have been proven effective in vitro against Leishmania donovani amastigotes, which cause visceral leishmaniasis. Identifying the molecular targets and molecular mechanisms underlying their action is a currently an unmet challenge. In the present study, we tackle this problem using computational methods to establish properties essential for activity as well as to screen betulin derivatives against potential targets. Recursive partitioning classification methods were explored to develop predictive models for 58 diverse betulin derivatives inhibitors of L. donovani amastigotes. The established models were validated on a testing set, showing excellent performance. Molecular fingerprints FCFP_6 and ALogP were extracted as the physicochemical properties most extensively involved in separating inhibitors from non-inhibitors. The potential targets of betulin derivatives inhibitors were predicted by in silico target fishing using structure-based pharmacophore searching and compound-pharmacophore-target-pathway network analysis, first on PDB and then among L. donovani homologs using a PSI-BLAST search. The essential identified proteins are all related to protein kinase family. Previous research already suggested members of the cyclin-dependent kinase family and MAP kinases as Leishmania potential drug targets. The PSI-BLAST search suggests two L. donovani proteins to be especially attractive as putative betulin target, heat shock protein 83 and membrane transporter D1.

A structure-activity relationship study of ABCC2 inhibitors.

Multidrug resistance associated protein 2 (MRP2/ABCC2) is a membrane transport protein that can potentially affect the disposition of many substrate drugs and their metabolites. Recently, we studied the interaction of a library of 432 compounds with ABCC2, and the structure-activity relationship (SAR) of a subset of 64 compounds divided into four scaffolds (Wissel, G. et al., 2015. Bioorg Med Chem., 23(13), pp.3513-25). We have now expanded this test set by investigating 114 new compounds, of which 71 are representative of the previous four scaffolds and 43 compounds belong to a new scaffold. Interaction with ABCC2 was assessed by measuring the compounds effect on 5(6)-carboxy-2',7'-dichlorofluorescein transport in the vesicular transport assay. In line with our previous study, we observed that anionic charge is not essential for inhibition of ABCC2 transport, even though it often increases the inhibitory activity within the analogue series. Additionally, we found that halogen substitutions often increase the inhibitory activity. The results confirm the importance of structural features such as aromaticity and lipophilicity for ABCC2 inhibitory activity.

IDAAPM: integrated database of ADMET and adverse effects of predictive modeling based on FDA approved drug data.

BACKGROUND: The disposition of a pharmaceutical compound within an organism, i.e. its Absorption, Distribution, Metabolism, Excretion, Toxicity (ADMET) properties and adverse effects, critically affects late stage failure of drug candidates and has led to the withdrawal of approved drugs. Computational methods are effective approaches to reduce the number of safety issues by analyzing possible links between chemical structures and ADMET or adverse effects, but this is limited by the size, quality, and heterogeneity of the data available from individual sources. Thus, large, clean and integrated databases of approved drug data, associated with fast and efficient predictive tools are desirable early in the drug discovery process. DESCRIPTION: We have built a relational database (IDAAPM) to integrate available approved drug data such as drug approval information, ADMET and adverse effects, chemical structures and molecular descriptors, targets, bioactivity and related references. The database has been coupled with a searchable web interface and modern data analytics platform (KNIME) to allow data access, data transformation, initial analysis and further predictive modeling. Data were extracted from FDA resources and supplemented from other publicly available databases. Currently, the database contains information regarding about 19,226 FDA approval applications for 31,815 products (small molecules and biologics) with their approval history, 2505 active ingredients, together with as many ADMET properties, 1629 molecular structures, 2.5 million adverse effects and 36,963 experimental drug-target bioactivity data. CONCLUSION: IDAAPM is a unique resource that, in a single relational database, provides detailed information on FDA approved drugs including their ADMET properties and adverse effects, the corresponding targets with bioactivity data, coupled with a data analytics platform. It can be used to perform basic to complex drug-target ADMET or adverse effects analysis and predictive modeling. IDAAPM is freely accessible at http://idaapm.helsinki.fi and can be exploited through a KNIME workflow connected to the database.Graphical abstractFDA approved drug data integration for predictive modeling.

Predictive Modeling of Ocular Pharmacokinetics and Adverse Effects.

Treatment of ocular disorders is a challenge due to the difficulty of delivering drugs to the target tissues within the eye at sufficient concentrations to produce a therapeutic effect. The cornea and the blood-retinal barrier, comprising of the retinal pigment epithelium and the retinal capillaries, are the main barriers for delivering drugs to treat diseases in the anterior and posterior parts of the eye, respectively. The eye has a rich blood supply and relatively small mass, and drugs can distribute from the systemic blood circulation to the choroid through the fenestrated choroidal blood vessels, but further permeation into the eye is limited by the blood-retinal barrier. Computational prediction of the ocular pharmacokinetics of drugs can help improve drug delivery and predict ocular adverse effects resulting from ocular or systemic drugs. Computational models predicting ocular adverse effects of drugs are still scarce, even though prediction of eye irritation and corrosion of chemicals has been well studied as a consequence of recent European Union legislation. Predictive modeling of adverse effects suffers from the wide distribution of data resources, but databases that integrate data on adverse effects, drugs, targets and other related biological data from different sources offer improved prospects for predictive modeling.

Integrated in vitro-in silico screening strategy for the discovery of antibacterial compounds.

Multidrug-resistant bacterial infections are an increasing source of healthcare problems, and the research for new antibiotics is currently unable to respond to this challenge. In this work, we present a screening strategy that integrates cell-based high-throughput screening (HTS) with in silico analogue search for antimicrobial small-molecule drug discovery. We performed an HTS on a diverse chemical library by using an assay based on a bioluminescent Escherichia coli K-12 (pTetLux1) strain. The HTS yielded eight hit compounds with >50% inhibition. These hits were then used for structural similarity-based virtual screening, and of the 29 analogues selected for in vitro testing, four compounds displayed potential activity in the pTetLux1 assay. The 11 most active compounds from combined HTS and analogue search were further assessed for antimicrobial activity against clinically important strains of E. coli and Staphylococcus aureus and for in vitro cytotoxicity against human cells. Three of the compounds displayed antibacterial activity and low human cell cytotoxicity. Additionally, two compounds of the set fully inhibited S. aureus growth after 24 h, but also exhibited human cell cytotoxicity in vitro.

Screening and characterisation of antimicrobial properties of semisynthetic betulin derivatives.

Betulin (lup-20(29)-ene-3ß, 28-diol) is a naturally occurring triterpene, which is found in substantial amounts from the outer bark of birch trees. A library of 51 structurally diverse semisynthetic betulin derivatives was screened against five bacterial strains, Enterobacter aerogenes, Escherichia coli, Enterococcus faecalis, Pseudomonas aeruginosa, Staphylococcus aureus and a fungal strain Candida albicans, using broth microdilution assays. Primary antimicrobial screening at 50 µM concentration led to the identification of five compounds showing antimicrobial properties (inhibition of growth by >70% against one or more microbial strains). According to the dose-response results, 28-O-(N-acetylanthraniloyl)betulin (compound 5) was the most active, showing MIC90 of 6.25 µM against two Gram-positive bacteria, E. faecalis and S. aureus. However, the activity of this compound was affected by albumin binding, which was demonstrated by the loss of activity in a host-pathogen co-culture assay as well as in the antibacterial assay in the presence of increased concentration of albumin. Furthermore, the effects on mammalian cells were evaluated by cytotoxicity assessment on hepatocyte cell culture after 24 h exposure to the compounds. Betulinic aldehyde (18), betulin-28-oxime (31) and hetero cycloadduct with acetoxy groups at carbon atoms 3 and 28 and ethyl substituent at the triazolo ring (43) displayed cytotoxicity towards hepatocytes, with IC50 values of 47, 25 and 16 µM, respectively. The IC50 value for 28-O-(N-acetylanthraniloyl)betulin (5) was 56 µM. The current study presents an insight into using betulin scaffold for developing derivatives with antibacterial potential, and furthermore the necessity of in-depth analysis of found actives through selectivity profiling and follow-up studies including in silico ADMET predictions.

Retrospective molecular docking study of WY-25105 ligand to ß-secretase and bias of the three-dimensional structure flexibility.

ß-Secretase (BACE) is a very promising target in the search for a treatment for Alzheimer's disease using a protein-ligand inhibition approach. Given the many published X-ray structures of BACE protein, structure-based drug design has been used extensively to support new inhibitor discovery programs. Due to the high flexibility and large catalytic site of this protein, sampling of the huge conformational space of the binding site is the big challenge to overcome and is the main limitation of the most widely used docking programs. Incorrect treatment of these pitfalls can introduce bias into ligand docking and could affect the results. This is especially the case with the WY-25105 compound reported by the Wyeth Corporation as a BACE ligand that did not fit into any of the known crystal structures. In the present retrospective study, a set of available X-ray enzyme structures was selected and molecular dynamics simulations were conducted to generate more diverse representative BACE protein conformations. These conformations were then used for a docking study of the WY-25105 compound. The results confirmed the need to use an ensemble of structures in protein-ligand docking for identification of new binding modes in structure-based drug design of BACE inhibitors.

Applying linear and non-linear methods for parallel prediction of volume of distribution and fraction of unbound drug.

Volume of distribution and fraction unbound are two key parameters in pharmacokinetics. The fraction unbound describes the portion of free drug in plasma that may extravasate, while volume of distribution describes the tissue access and binding of a drug. Reliable in silico predictions of these pharmacokinetic parameters would benefit the early stages of drug discovery, as experimental measuring is not feasible for screening purposes. We have applied linear and nonlinear multivariate approaches to predict these parameters: linear partial least square regression and non-linear recursive partitioning classification. The volume of distribution and fraction of unbound drug in plasma are predicted in parallel within the model, since the two are expected to be affected by similar physicochemical drug properties. Predictive models for both parameters were built and the performance of the linear models compared to models included in the commercial software Volsurf+. Our models performed better in predicting the unbound fraction (Q(2) 0.54 for test set compared to 0.38 with Volsurf+ model), but prediction accuracy of the volume of distribution was comparable to the Volsurf+ model (Q(2) of 0.70 for test set compared to 0.71 with Volsurf+ model). The nonlinear classification models were able to identify compounds with a high or low volume of distribution (sensitivity 0.81 and 0.71, respectively, for test set), while classification of fraction unbound was less successful. The interrelationship between the volume of distribution and fraction unbound is investigated and described in terms of physicochemical descriptors. Lipophilicity and solubility descriptors were found to have a high influence on both volume of distribution and fraction unbound, but with an inverse relationship.