Search

Current in vitromodels for hepatotoxicity commonly suffer from low detection rates due to incomplete coverage of bioactivity space. Additionally, in vivoexposure measures such as Cmaxare used for hepatotoxicity screening and are unavailable early on. Here we propose a novel rule-based framework to extract interpretable and biologically meaningful multiconditional associations to prioritize in vitroend points for hepatotoxicity and understand the associated physicochemical conditions. The data used in this study were derived for 673 compounds from 361 ToxCast bioactivity measurements and 29 calculated physicochemical properties against two lowest effective levels (LEL) of rodent hepatotoxicity from ToxRefDB, namely 15 mg/kg/day and 500 mg/kg/day. To achieve 80% coverage of toxic compounds, 35 rules with accuracies ranging from 96% to 73% using 39 unique ToxCast assays are needed at a threshold level of 500 mg/kg/day, whereas to describe the same coverage at a threshold of 15 mg/kg/day, 20 rules with accuracies of between 98% and 81% were needed, comprising 24 unique assays. Despite the 33-fold difference in dose levels, we found relative consistency in the key mechanistic groups in rule clusters, namely (i) activities against Cytochrome P, (ii) immunological responses, and (iii) nuclear receptor activities. Less specific effects, such as oxidative stress and cell cycle arrest, were used more by rules to describe toxicity at the level of 500 mg/kg/day. Although the endocrine disruption through nuclear receptor activity formulated an essential cluster of rules, this bioactivity was not covered in four commercial assay setups for hepatotoxicity. Using an external set of 29 drugs with drug-induced liver injury (DILI) labels, we found that promiscuity over important assays discriminates between compounds with different levels of liver injury. In vitro–in vivoassociations were also improved by incorporating physicochemical properties especially for the potent, 15 mg/kg/day toxicity level as well for assays describing nuclear receptor activity and phenotypic changes. The most frequently used physicochemical properties, predictive for hepatotoxicity in combination with assay activities, are linked to bioavailability, which were the number of rotatable bonds (less than 7) at a of level of 15 mg/kg/day and the number of rings (of less than 3) at level of 500 mg/kg/day. In summary, hepatotoxicity cannot very well be captured by single assay end points, but better by a combination of bioactivities in relevant assays, with the likelihood of hepatotoxicity increasing with assay promiscuity. Together, these findings can be used to prioritize assay combinations that are appropriate to assess potential hepatotoxicity.

Adverse events resulting from drug therapy can be a cause of drug withdrawal, reduced and or restricted clinical use, as well as a major economic burden for society. To increase the safety of new drugs, there is a need to better understand the mechanisms causing the adverse events. One way to derive new mechanistic hypotheses is by linking data on drug adverse events with the drugs’ biological targets. In this study, we have used data mining techniques and mutual information statistical approaches to find associations between reported adverse events collected from the FDA Adverse Event Reporting System and assay outcomes from ToxCast, with the aim to generate mechanistic hypotheses related to structural cardiotoxicity (morphological damage to cardiomyocytes and/or loss of viability). Our workflow identified 22 adverse event-assay outcome associations. From these associations, 10 implicated targets could be substantiated with evidence from previous studies reported in the literature. For two of the identified targets, we also describe a more detailed mechanism, forming putative adverse outcome pathways associated with structural cardiotoxicity. Our study also highlights the difficulties deriving these type of associations from the very limited amount of data available.

Similarity between the ligand binding profiles of enzymes may aid functional characterization and be of greater relevance to inhibitor development than sequence similarity or structural homology. Fragment screening is an efficient approach for characterization of the ligand binding profile of an enzyme and has been applied here to study the family of cytochrome P450 enzymes (P450s) expressed by Mycobacterium tuberculosis(Mtb). The MtbP450s have important roles in bacterial virulence, survival, and pathogenicity. Comparing the fragment profiles of seven of these enzymes revealed that P450s which share a similar biological function have significantly similar fragment profiles, whereas functionally unrelated or orphan P450s exhibit distinct ligand binding properties, despite overall high structural homology. Chemical structures that exhibit promiscuous binding between enzymes have been identified, as have selective fragments that could provide leads for inhibitor development. The similarity between the fragment binding profiles of the orphan enzyme CYP144A1 and CYP121A1, a characterized enzyme that is important for Mtbviability, provides a case study illustrating the subsequent identification of novel CYP144A1 ligands. The different binding modes of these compounds to CYP144A1 provide insight into structural and dynamic aspects of the enzyme, possible biological function, and provide the opportunity to develop inhibitors. Expanding this fragment profiling approach to include a greater number of functionally characterized and orphan proteins may provide a valuable resource for understanding enzyme–ligand interactions.

Increasing knowledge of chromatin structure in various cell types raises the challenge of deciphering the contribution of epigenetic modifications to the regulation of nuclear functions in mammals. In a recent study, we have analysed the genome-wide distributions of thirteen epigenetic marks in the human cell line K562 at 100 kb resolution of Mean Replication Timing (MRT) data. Using classical clustering techniques, we have shown that the combinatorial complexity of these epigenetic data can be reduced to four predominant chromatin states that replicate at different periods of the S-phase. C1 is an early replicating transcriptionally active euchromatin state, C2 a mid-S repressive type of chromatin associated with Polycomb complexes, C3 a silent chromatin with lack of chromatin marks that replicates later than C2 but before C4, a HP1-associated heterochromatin state that replicates at the end of S-phase. These four chromatin states display remarkable similarities with those recently reported in fly, worm and plants at higher ∼ 1 kb resolution of gene expression data. Here, we extend our integrative analysis of epigenetic data in the K562 human cell line to this smaller scale by focusing on gene promoters (±3 kb around transcription start sites). We show that these promoters can similarly be classified into four main chromatin states: P1 regroups all the marks of transcriptionally active chromatin and corresponds to CpG rich promoters of highly expressed genes; P2 is notably associated with the histone modification H3K27me3 that is the mark of a polycomb repressed chromatin state; P3 corresponds to promoters that are not enriched for any available marks as the signature of a ‘null’ or ‘black’ silent heterochromatin state and P4 characterizes the few gene promoters that contain only the constitutive heterochromatin histone modification H3K9me3. When investigating the coherence between promoter activity (P1, P2, P3 or P4) and the large-scale chromatin environment (C1, C2, C3 or C4), we find that the higher the gene density in a considered 100 kb-window, the higher (resp. the lower) the probability of a P1 active promoter (resp. silent P2, P3 and P4 promoters) to be surrounded by an open euchromatin C1 (resp. facultative C2, black C3 or HP1-associated C4 heterochromatin) environment. From large to small scales, it is mainly C4 and to a lesser extent C3 heterochromatin environments both corresponding to gene poor regions, that strongly conditions promoters to belong to the inactive P3 and P4 classes. If C1 (resp. C2) environment surrounds a majority of corresponding active P1 (resp. P2) promoters, it also contains a non-negligible proportion of inactive P2 and P3 (resp. active P1 and inactive P3) promoters. When further investigating the large-scale organization of human genes with respect to ‘master’ replication origins that were shown to border megabase-sized U-shaped MRT domains, we reveal some significant enrichment of highly expressed P1 genes in a closed neighbourhood of these early initiation zones consistently with the gradient of chromatin states observed from C1 at U-domain borders followed by C2, C3 and C4 at U-domain centers. On the contrary to P2 promoters that are mainly found in the C2 environment at finite distance (∼200–300 kb) from U-domain borders, the inactive P3 and P4 promoters are distributed rather homogeneously inside U-domains. The generalization of our study to different cell types including ES, somatic and cancer cells is likely to provide new insight on the global reorganization of replication domains during differentiation (or disease) in relation to coordinated changes in chromatin environment and gene expression.