There is a general consensus that supports the need for standardized reporting of metadata or information describing large-scale metabolomics and other functional genomics data sets. Reporting of ...standard metadata provides a biological and empirical context for the data, facilitates experimental replication, and enables the re-interrogation and comparison of data by others. Accordingly, the Metabolomics Standards Initiative is building a general consensus concerning the minimum reporting standards for metabolomics experiments of which the Chemical Analysis Working Group (CAWG) is a member of this community effort. This article proposes the minimum reporting standards related to the chemical analysis aspects of metabolomics experiments including: sample preparation, experimental analysis, quality control, metabolite identification, and data pre-processing. These minimum standards currently focus mostly upon mass spectrometry and nuclear magnetic resonance spectroscopy due to the popularity of these techniques in metabolomics. However, additional input concerning other techniques is welcomed and can be provided via the CAWG on-line discussion forum at http://msi-workgroups.sourceforge.net/ or http://Msi-workgroups-feedback@lists.sourceforge.net. Further, community input related to this document can also be provided via this electronic forum.
Although early detection of toxicant induced kidney injury during drug development and chemical safety testing is still limited by the lack of sensitive and reliable biomarkers of nephrotoxicity, ...omics technologies have brought enormous opportunities for improved detection of toxicity and biomarker discovery. Thus, transcription profiling has led to the identification of several candidate kidney biomarkers such as kidney injury molecule (Kim-1), clusterin, lipocalin-2, and tissue inhibitor of metalloproteinase 1 (Timp-1), and metabonomic analysis of urine is increasingly used to indicate biochemical perturbations due to renal toxicity. This study was designed to assess the value of a combined 1H-NMR and gas chromatography–mass spectrometry (GC-MS) metabonomics approach and a set of novel urinary protein markers for early detection of nephrotoxicity following treatment of male Wistar rats with gentamicin (60 and 120 mg/kg bw, sc) for 7 days. Time- and dose-dependent separation of gentamicin-treated animals from controls was observed by principal component analysis of 1H-NMR and GC-MS data. The major metabolic alterations responsible for group separation were linked to the gut microflora, thus related to the pharmacology of the drug, and increased glucose in urine of gentamicin-treated animals, consistent with damage to the S1 and S2 proximal tubules, the primary sites for glucose reabsorption. Altered excretion of urinary protein biomarkers Kim-1 and lipocalin-2, but not Timp-1 and clusterin, was detected before marked changes in clinical chemistry parameters were evident. The early increase in urine, which correlated with enhanced gene and protein expression at the site of injury, provides further support for lipocalin-2 and Kim-1 as sensitive, noninvasive biomarkers of nephrotoxicity.
At the confluence of predictive and regulatory toxicologies, negative predictions may be the thin green line that prevents populations from being exposed to harm. Here, two novel approaches to making ...confident and robust negative in silico predictions for mutagenicity (as defined by the Ames test) have been evaluated. Analyses of 12 data sets containing >13,000 compounds, showed that negative predictivity is high (∼90%) for the best approach and features that either reduce the accuracy or certainty of negative predictions are identified as misclassified or unclassified respectively. However, negative predictivity remains high (and in excess of the prevalence of non-mutagens) even in the presence of these features, indicating that they are not flags for mutagenicity.
Display omitted
•Two methods for making negative predictions have been developed.•Negative predictivity for (Ames test) mutagenicity of ∼90% for >13,000 proprietary compounds.•Features that may increase uncertainty or reduce accuracy highlighted for expert review.
The relative wealth of bacterial mutagenicity data available in the public literature means that in silico quantitative/qualitative structure activity relationship (QSAR) systems can readily be built ...for this endpoint. A good means of evaluating the performance of such systems is to use private unpublished data sets, which generally represent a more distinct chemical space than publicly available test sets and, as a result, provide a greater challenge to the model. However, raw performance metrics should not be the only factor considered when judging this type of software since expert interpretation of the results obtained may allow for further improvements in predictivity. Enough information should be provided by a QSAR to allow the user to make general, scientifically-based arguments in order to assess and overrule predictions when necessary. With all this in mind, we sought to validate the performance of the statistics-based in vitro bacterial mutagenicity prediction system Sarah Nexus (version 1.1) against private test data sets supplied by nine different pharmaceutical companies. The results of these evaluations were then analysed in order to identify findings presented by the model which would be useful for the user to take into consideration when interpreting the results and making their final decision about the mutagenic potential of a given compound.
Display omitted
•The statistics-based model, Sarah Nexus, was evaluated using 14 proprietary data sets from 9 pharmaceutical companies.•High-level performance metrics were explored along with potential for expert analysis of different prediction scenarios.•Performance was generally good. Analysis of example predictions indicated they may be improved following expert analysis.
For early detection of toxicity and improved mechanistic understanding, GC/MS-, 1H NMR-, and LC/MS-based metabonomics were applied to urine samples from a rodent toxicity study on the mycotoxin and ...renal carcinogen ochratoxin A (OTA). OTA was administered at doses of 0, 21, 70, and 210 μg/kg body wt for up to 90 days. Urine samples were collected at 24 h intervals 14, 28, and 90 days after the start of treatment and analyzed with GC/MS, 1H NMR, and LC/MS. Principal component analysis and orthogonal projection to latent structures discriminate analysis (OPLS-DA) based on GC/MS and 1H NMR data discriminated controls from animals dosed with 210 μg/kg body wt OTA as early as 14 days and animals dosed with 70 μg/kg body wt 28 days after the start of treatment, correlating with mild histopathological changes in the kidney. Integration of histopathology scores as discriminators in OPLS-DA models resulted in better multivariate model predictivity and facilitated marker identification. Decreased 2-oxoglutarate and citrate excretion and increased glucose, creatinine, pseudouridine, 5-oxoproline, and myo-inositol excretion were detected with GC/MS. Decreased 2-oxoglutarate and citrate excretion and increased amino acid excretion were found with 1H NMR. Increased urinary glucose is a well-established indicator of kidney damage, and altered excretion of TCA cycle intermediates (citrate and 2-oxoglutarate) is found as a general response to toxic insult in many metabonomics studies. Other markers are associated with cell proliferation (pseudouridine), changes in renal osmolyte handling (myo-inositol), and oxidative stress (5-oxoproline), established mechanisms of OTA toxicity. LC/MS was also able to discriminate controls and treated animals but contained more noise, and marker annotation was only speculative due to lack of reference databases. Use of multiple analytical platforms for metabonomics analysis may result in a more comprehensive metabolite coverage and may be applied to obtain mechanistic information from conventional rodent toxicity studies.
In this publication, we report the outcome of the integrated EU Framework 6 Project: Predictive Toxicology (PredTox), including methodological aspects and overall conclusions. Specific details ...including data analysis and interpretation are reported in separate articles in this issue. The project, partly funded by the EU, was carried out by a consortium of 15 pharmaceutical companies, 2 SMEs, and 3 universities.
The effects of 16 test compounds were characterized using conventional toxicological parameters and “omics” technologies. The three major observed toxicities, liver hypertrophy, bile duct necrosis and/or cholestasis, and kidney proximal tubular damage were analyzed in detail.
The combined approach of “omics” and conventional toxicology proved a useful tool for mechanistic investigations and the identification of putative biomarkers. In our hands and in combination with histopathological assessment, target organ transcriptomics was the most prolific approach for the generation of mechanistic hypotheses. Proteomics approaches were relatively time-consuming and required careful standardization. NMR-based metabolomics detected metabolite changes accompanying histopathological findings, providing limited additional mechanistic information. Conversely, targeted metabolite profiling with LC/GC-MS was very useful for the investigation of bile duct necrosis/cholestasis. In general, both proteomics and metabolomics were supportive of other findings.
Thus, the outcome of this program indicates that “omics” technologies can help toxicologists to make better informed decisions during exploratory toxicological studies. The data support that hypothesis on mode of action and discovery of putative biomarkers are tangible outcomes of integrated “omics” analysis. Qualification of biomarkers remains challenging, in particular in terms of identification, mechanistic anchoring, appropriate specificity, and sensitivity.
In silico
toxicology protocols are meant to support computationally-based assessments using principles that ensure that results can be generated, recorded, communicated, archived, and then evaluated ...in a uniform, consistent, and reproducible manner. We investigated the availability of
in silico
models to predict the carcinogenic potential of pregabalin using the ten key characteristics of carcinogens as a framework for organizing mechanistic studies. Pregabalin is a single-species carcinogen producing only one type of tumor, hemangiosarcomas in mice via a nongenotoxic mechanism. The overall goal of this exercise is to test the ability of
in silico
models to predict nongenotoxic carcinogenicity with pregabalin as a case study. The established mode of action (MOA) of pregabalin is triggered by tissue hypoxia, leading to oxidative stress (KC5), chronic inflammation (KC6), and increased cell proliferation (KC10) of endothelial cells. Of these KCs,
in silico
models are available only for selected endpoints in KC5, limiting the usefulness of computational tools in prediction of pregabalin carcinogenicity. KC1 (electrophilicity), KC2 (genotoxicity), and KC8 (receptor-mediated effects), for which predictive
in silico
models exist, do not play a role in this mode of action. Confidence in the overall assessments is considered to be medium to high for KCs 1, 2, 5, 6, 7 (immune system effects), 8, and 10 (cell proliferation), largely due to the high-quality experimental data. In order to move away from dependence on animal data, development of reliable
in silico
models for prediction of oxidative stress, chronic inflammation, immunosuppression, and cell proliferation will be critical for the ability to predict nongenotoxic compound carcinogenicity.
Liquid chromatography coupled to mass spectrometry (LC–MS) is a major platform in metabolic profiling but has not yet been comprehensively assessed as to its repeatability and reproducibility across ...multiple spectrometers and laboratories. Here we report results of a large interlaboratory reproducibility study of ultra performance (UP) LC–MS of human urine. A total of 14 stable isotope labeled standard compounds were spiked into a pooled human urine sample, which was subject to a 2- to 16-fold dilution series and run by UPLC coupled to time-of-flight MS at three different laboratories all using the same platform. In each lab, identical samples were run in two phases, separated by at least 1 week, to assess between-day reproducibility. Overall, platform reproducibility was good with median mass accuracies below 12 ppm, median retention time drifts of less than 0.73 s and coefficients of variation of intensity of less than 18% across laboratories and ionization modes. We found that the intensity response was highly linear within each run, with a median R 2 of 0.95 and 0.93 in positive and negative ionization modes. Between-day reproducibility was also high with a mean R 2 of 0.93 for a linear relationship between the intensities of ions recorded in the two phases across the laboratories and modes. Most importantly, between-lab reproducibility was excellent with median R 2 values of 0.96 and 0.98 for positive and negative ionization modes, respectively, across all pairs of laboratories. Interestingly, the three laboratories observed different amounts of adduct formation, but this did not appear to be related to reproducibility observed in each laboratory. These studies show that UPLC–MS is fit for the purpose of targeted urinary metabolite analysis but that care must be taken to optimize laboratory systems for quantitative detection due to variable adduct formation over many compound classes.
N-Nitrosamines (NAs) are a class of reactive organic chemicals that humans may be exposed to from environmental sources, food but also impurities in pharmaceutical preparations. Some NAs were ...identified as DNA-reactive mutagens and many of those have been classified as probable human carcinogens. Beyond high-potency mutagenic carcinogens that need to be strictly controlled, NAs of low potency need to be considered for risk assessment as well. NA impurities and nitrosylated products of active pharmaceutical ingredients (APIs) often arise from production processes or degradation. Most NAs require metabolic activation to ultimately become carcinogens, and their activation can be appropriately described by first-principles computational chemistry approaches. To this end, we treat NA-induced DNA alkylation as a series of subsequent association and dissociation reaction steps that can be calculated stringently by density functional theory (DFT), including α-hydroxylation, proton transfer, hydroxyl elimination, direct SN2/SNAr DNA alkylation, competing hydrolysis and SN1 reactions. Both toxification and detoxification reactions are considered. The activation reactions are modeled by DFT at a high level of theory with an appropriate solvent model to compute Gibbs free energies of the reactions (thermodynamical effects) and activation barriers (kinetic effects). We study congeneric series of aliphatic and cyclic NAs to identify trends. Overall, this work reveals detailed insight into mechanisms of activation for NAs, suggesting that individual steric and electronic factors have directing and rate-determining influence on the formation of carbenium ions as the ultimate pro-mutagens and thus carcinogens. Therefore, an individual risk assessment of NAs is suggested, as exemplified for the complex API-like 4-(N-nitroso-N-methyl)aminoantipyrine which is considered as low-potency NA by in silico prediction.
The European InnoMed–PredTox project was a collaborative effort between 15 pharmaceutical companies, 2 small and mid-sized enterprises, and 3 universities with the goal of delivering deeper insights ...into the molecular mechanisms of kidney and liver toxicity and to identify mechanism-linked diagnostic or prognostic safety biomarker candidates by combining conventional toxicological parameters with “omics” data. Mechanistic toxicity studies with 16 different compounds, 2 dose levels, and 3 time points were performed in male Crl: WI(Han) rats.
Three of the 16 investigated compounds, BI-3 (FP007SE), Gentamicin (FP009SF), and IMM125 (FP013NO), induced kidney proximal tubule damage (PTD). In addition to histopathology and clinical chemistry, transcriptomics microarray and proteomics 2D-DIGE analysis were performed. Data from the three PTD studies were combined for a cross-study and cross-omics meta-analysis of the target organ. The mechanistic interpretation of kidney PTD-associated deregulated transcripts revealed, in addition to previously described kidney damage transcript biomarkers such as KIM-1, CLU and TIMP-1, a number of additional deregulated pathways congruent with histopathology observations on a single animal basis, including a specific effect on the complement system. The identification of new, more specific biomarker candidates for PTD was most successful when transcriptomics data were used. Combining transcriptomics data with proteomics data added extra value.