Abstracts / Toxicology Letters 258S (2016) S62–S324
P05-020 Structure-based prediction of side effects of sortase A inhibitors G.M. Nitulescu ∗ , G. Nitulescu, O. Olaru, D. Margina “Carol Davila” University of Medicine and Pharmacy, Faculty of Pharmacy, 6 Traian Vuia, 020956 Bucharest, Romania Drugs side effects and adverse reactions are one of the major reason for market withdrawal, and their prediction during developmental phase could save drug companies both time and money. Sortases are a promising target for the development of antiinfective agents against Gram-positive bacteria, playing important roles in virulence and infection. Several small molecules that inhibit sortase A have been identified, but most of them are targeting the active cysteine thiol group. The selectivity of this mechanism raises the question of potential inhibition against other thiol proteins and consequently of possible toxicity. In order to develop new safer sortase A inhibitors we performed a virtual screening of a small library of known sortases inhibitors to predict potential side effects and the potential to interact with anti-targets, like hERG or CYPs. The analysis was performed using computer program PASS. The output data representing the probability of a particular side effect to manifest or not were analysed in correlation with the chemical structural patterns. A clear relationship between the chemical scaffold and the side effect spectrum was observed. Even if this method does not provide a good biological interpretations regarding the underlying molecular mechanisms of toxicity, it offers valuable information for the rational drug design of new safer sortases inhibitors as future anti-infective drugs. The authors acknowledge the financial support offered by Romanian National Authority for Scientific Research, UEFISCDI, through grant PN-II-RU-TE-2014-4-1670, number 342/2015. http://dx.doi.org/10.1016/j.toxlet.2016.06.1488 P05-021 A multi-compartment liver model for the prediction of toxicokinetics C. Fisher ∗ , O. Hatley, I. Gardner, M. Jamei Simcyp (a Certara Company), United Kingdom Hepatotoxicity represents a major concern in the drug development pipeline, accounting for a significant number of preclinical project closures and market withdrawals (Pandit et al., 2012; Cook et al., 2014). Drug metabolizing enzymes and transporters are differentially expressed across the liver resulting in regional differences in drug kinetics. This can impact the manifestation of drug induced liver injury, resulting in regionally specific lesions. For example, acetaminophen and methapyrilene cause centrilobular and periportal lesions, respectively, dictated by the differential expression of the cytochrome P450s that mediate the generation of their respective reactive metabolites. Well-stirred single compartment liver models have been widely used in predicting the pharmacokinetics of compounds in physiologically-based models. However, they do not represent the zonal nature of gene-expression. A number of approaches to modelling the liver have been published in the literature, proposing different structures and levels of complexity (Schwen et al., 2015; Lerapetritou et al., 2009; Watanabe et al., 2009; Anissimov and Roberts 2002). Here a multi-compartment liver model incorporating differential gene-expression, sinusoidal blood flow and biliary excretion was used to investigate the impact of regional differ-
S121
ences in drug metabolizing enzyme expression on the metabolism of acetaminophen. The model demonstrates spatial differences in hepatic concentration profiles across the acinus corresponding to both the concentration gradient along the sinusoid and heterogeneous CYP expression. Such profiles may be used to explain regional differences in observed toxicity data. Future work will allow for linking these PD effects to PK exposure. http://dx.doi.org/10.1016/j.toxlet.2016.06.1489 P05-022 Improving in silico methods of ecotoxicity hazard identification for active pharmaceutical ingredients (APIs) C.M. Ellison, J.C. Madden, C.L. Mellor, M.T. Cronin ∗ School of Pharmacy and Biomolecular Sciences, Liverpool John Moores University, Liverpool, England, United Kingdom In Europe a legislative requirement for an environmental risk assessment (ERA) of active pharmaceutical ingredients (APIs) to be completed prior to registration for marketing was established in 2006. Prior to this such assessments were not mandatory and hence legacy compounds have limited or no data available. Efficient methods of identifying ecotoxicity hazards are required to prioritise and/or assess these legacy compounds. The Verhaar scheme can help in the identification of ecotoxicity hazards as it enables users to classify compounds into one of four mechanistic categories; Class 1 – Baseline narcotics, Class 2 – Polar narcotics, Class 3 – Reactive compounds and Class 4 – Specifically acting compounds. These categories have been used to develop mechanistic QSARs for the prediction of acute fish lethality and other acute endpoints in surrogate species, for a large range of industrial compounds. However the scheme has rarely been implemented in the assessment of APIs. The majority of APIs are outside of the domain of the scheme with most compounds receiving an “unclassified” outcome. The aim of this study was to analyse a set of APIs that are outside of the domain of the Verhaar scheme to assess if it is possible to build additional rules for these structures and hence expand the structural domain of the scheme. Only a small minority (if any) APIs are likely to have a reactive mechanism of action and thus the category expansions focussed on rules for narcosis and mechanisms for specifically acting compounds. http://dx.doi.org/10.1016/j.toxlet.2016.06.1490 P05-023 The Novartis Translational Study Data Warehouse. A 21st century view on our animal data F. Hahne ∗ , P. Marc, T. Welker, J. Walker, D. Selinger, A. Mueller, K. Therrien, J. Demchak, K. Stango, R. Galicia, B. Huang, P. Bouchard Novartis Institutes for Biomedical Research, United States The CDISC SEND standard has been proposed to facilitate data exchange between pharmaceutical industry, CROS and the US Food and Drug Administration (FDA). Presented here is our SENDcompliant data warehouse SDW and its visualisation front-end TSP. The overarching goal of these systems is to provide access to internal and external study data to Novartis associates in a standardised fashion, and to simplify presentation of the high-dimensional data. The data warehouse utilises a non-SQL data base backend, and