Amine promiscuity and toxicology analysis

Drug discovery programs often face challenges to obtain sufficient duration of action of the drug (i.e. seek longer half-lives). If the pharmacodynamic response is driven by free plasma concentration of the drug then extending the plasma drug concentration is a valid approach. Half-life is dependent on the volume of distribution, which in turn can be dependent upon the ionization state of the molecule. Basic compounds tend to have a higher volume of distribution leading to longer half-lives. However, it has been shown that bases may also have higher promiscuity. In this work, we describe an analysis of in vitro pharmacological profiling and toxicology data investigating the role of primary, secondary, and tertiary amines in imparting promiscuity and thus off-target toxicity. Primary amines are found to be less promiscuous in in vitro assays and have improved profiles in in vivo toxicology studies compared to secondary and tertiary amines.     

Use of the Zebrafish Larvae as a Model to Study Cigarette Smoke Condensate Toxicity

The smoking of tobacco continues to be the leading cause of premature death worldwide and is linked to the development of a number of serious illnesses including heart disease, respiratory diseases, stroke and cancer. Currently, cell line based toxicity assays are typically used to gain information on the general toxicity of cigarettes and other tobacco products. However, they provide little information regarding the complex disease-related changes that have been linked to smoking. The ethical concerns and high cost associated with mammalian studies have limited their widespread use for in vivo toxicological studies of tobacco. The zebrafish has emerged as a low-cost, high-throughput, in vivo model in the study of toxicology. In this study, smoke condensates from 2 reference cigarettes and 6 Canadian brands of cigarettes with different design features were assessed for acute, developmental, cardiac, and behavioural toxicity (neurotoxicity) in zebrafish larvae. By making use of this multifaceted approach we have developed an in vivo model with which to compare the toxicity profiles of smoke condensates from cigarettes with different design features. This model system may provide insights into the development of smoking related disease and could provide a cost-effective, high-throughput platform for the future evaluation of tobacco products.     

Predictive metabolite profiling applying hierarchical multivariate curve resolution to GC-MS data--a potential tool for multi-parametric diagnosis

A method for predictive metabolite profiling based on resolution of GC-MS data followed by multivariate data analysis is presented and applied to three different biofluid data sets (rat urine, aspen leaf extracts, and human blood plasma). Hierarchical multivariate curve resolution (H-MCR) was used to simultaneously resolve the GC-MS data into pure profiles, describing the relative metabolite concentrations between samples, for multivariate analysis. Here, we present an extension of the H-MCR method allowing treatment of independent samples according to processing parameters estimated from a set of training samples. Predictions or inclusion of the new samples, based on their metabolite profiles, into an existing model could then be carried out, which is a requirement for a working application within, e.g., clinical diagnosis. Apart from allowing treatment and prediction of independent samples the proposed method also reduces the time for the curve resolution process since only a subset of representative samples have to be processed while the remaining samples can be treated according to the obtained processing parameters. The time required for resolving the 30 training samples in the rat urine example was approximately 13 h, while the treatment of the 30 test samples according to the training parameters required only approximately 30 s per sample (approximately 15 min in total). In addition, the presented results show that the suggested approach works for describing metabolic changes in different biofluids, indicating that this is a general approach for high-throughput predictive metabolite profiling, which could have important applications in areas such as plant functional genomics, drug toxicity, treatment efficacy and early disease diagnosis.     

Methods for Identifying a Default Cross-Species Scaling Factor

The need to identify “toxicologically equivalent” doses across different species is a major issue in toxicology and risk assessment. In this article, we describe an approach for establishing default cross-species extrapolation factors used to scale oral doses across species for non-carcinogenic endpoints. This work represents part of an on-going effort to harmonize the way animal data are evaluated for carcinogenic and non-carcinogenic endpoints. In addition to considering default scaling factors, we also discuss how chemical-specific data (e.g., metabolic or mechanistic data) can be incorporated into the dose extrapolation process. After first examining the required properties of a default scaling methodology, we consider scaling approaches based on empirical relationships observed for particular classes of compounds and also more theoretical approaches based on general physiological principles (i.e, allometry). The available data suggest that the empirical and allometric approaches each provide support for the idea that toxicological risks are approximately equal when daily oral doses are proportional to body weight raised to the 3/4-power. We also discuss specific challenges for dose scaling related to different routes of exposure, acute versus chronic toxicity, and extrapolations related to particular life stages (e.g., childhood).     

Systematic analysis of native membrane protein complexes in Escherichia coli

The cell envelope of Escherichia coli is an essential structure that modulates exchanges between the cell and the extra-cellular milieu. Previous proteomic analyses have suggested that it contains a significant number of proteins with no annotated function. To gain insight into these proteins and the general organization of the cell envelope proteome, we have carried out a systematic analysis of native membrane protein complexes. We have identified 30 membrane protein complexes (6 of which are novel) and present reference maps that can be used for cell envelope profiling. In one instance, we identified a protein with no annotated function (YfgM) in a complex with a well-characterized periplasmic chaperone (PpiD). Using the guilt by association principle, we suggest that YfgM is also part of the periplasmic chaperone network. The approach we present circumvents the need for engineering of tags and protein overexpression. It is applicable for the analysis of membrane protein complexes in any organism and will be particularly useful for less-characterized organisms where conventional strategies that require protein engineering (i.e., 2-hybrid based approaches and TAP-tagging) are not feasible.     

Metabolic response to low-level toxicant exposure in a novel renal tubule epithelial cell system

Toxicity testing is vital to protect human health from exposure to toxic chemicals in the environment. Furthermore, combining novel cellular models with molecular profiling technologies, such as metabolomics can add new insight into the molecular basis of toxicity and provide a rich source of biomarkers that are urgently required in a 21st Century approach to toxicology. We have used an NMR-based metabolic profiling approach to characterise for the first time the metabolome of the RPTEC/TERT1 cell line, an immortalised non-tumour human renal epithelial cell line that recapitulates phenotypic characteristics that are absent in other in vitro renal cell models. RPTEC/TERT1 cells were cultured with either the dosing vehicle (DMSO) or with exposure to one of six compounds (nifedipine, potassium bromate, monuron, D-mannitol, ochratoxin A and sodium diclofenac), several of which are known to cause renal effects. Aqueous intracellular and culture media metabolites were profiled by (1)H NMR spectroscopy at 6, 24 and 72 hours of exposure to a low effect dose (IC(10)). We defined the metabolome of the RPTEC/TERT1 cell line and used a principal component analysis approach to derive a panel of key metabolites, which were altered by chemical exposure. By considering only major changes (±1.5 fold change from control) across this metabolite panel we were able to show specific alterations to cellular processes associated with chemical treatment. Our findings suggest that metabolic profiling of RPTEC/TERT1 cells can report on the effect of chemical exposure on multiple cellular pathways at low-level exposure, producing different response profiles for the different compounds tested with a greater number of major metabolic effects observed in the toxin treated cells. Importantly, compounds with established links to chronic renal toxicity produced more diverse and severe perturbations to the cellular metabolome than non-toxic compounds in this model. As these changes can be rationalised with the different pharmacological and toxicity profiles of the chemicals it is suggested that metabolic profiling in the RPTEC/TERT1 model would be useful in investigating the mechanism of action of toxins at a low dose.     

Machine learning methods for predictive proteomics

The search for predictive biomarkers of disease from high-throughput mass spectrometry (MS) data requires a complex analysis path. Preprocessing and machine-learning modules are pipelined, starting from raw spectra, to set up a predictive classifier based on a shortlist of candidate features. As a machine-learning problem, proteomic profiling on MS data needs caution like the microarray case. The risk of overfitting and of selection bias effects is pervasive: not only potential features easily outnumber samples by 10(3) times, but it is easy to neglect information-leakage effects during preprocessing from spectra to peaks. The aim of this review is to explain how to build a general purpose design analysis protocol (DAP) for predictive proteomic profiling: we show how to limit leakage due to parameter tuning and how to organize classification and ranking on large numbers of replicate versions of the original data to avoid selection bias. The DAP can be used with alternative components, i.e. with different preprocessing methods (peak clustering or wavelet based), classifiers e.g. Support Vector Machine (SVM) or feature ranking methods (recursive feature elimination or I-Relief). A procedure for assessing stability and predictive value of the resulting biomarkers' list is also provided. The approach is exemplified with experiments on synthetic datasets (from the Cromwell MS simulator) and with publicly available datasets from cancer studies.     

Assessment of the functional integrity of the humoral immune response: the plaque-forming cell assay and the enzyme-linked immunosorbent assay.

The plaque-forming cell (PFC) assay and enzyme-linked immunosorbent assay (ELISA) appear to have comparable sensitivity and reproducibility for measuring IgM antibody production in mice and rats immunized with sheep red blood cells (sRBCs). Both assays can be manipulated, with respect to the immunizing antigen (e.g., T-dependent vs T-independent antigen), to provide evidence as to which cell type(s) may be adversely affected by a given compound. However, the PFC assay has more utility in dissecting out the target cell(s) involved. Since both the PFC assay and the ELISA may be readily conducted in the rat, it is feasible to incorporate either of these assays into standard acute and repeat dose toxicology studies. This may be accomplished by inclusion of satellite groups in the study. However, it has been suggested that the primary antibody response to sRBCs, as measured by an ELISA, may also be evaluated in the main group of animals in a toxicology study without compromise to the integrity of other toxicological endpoints (e.g., hematology, clinical chemistry, histopathology). Both approaches will provide a more extensive delineation of the safety profile of a drug or chemical. The latter approach will also reduce the number of animals needed and the cost of the study.     

Cross-Platform Toxicogenomics for the Prediction of Non-Genotoxic Hepatocarcinogenesis in Rat

In the area of omics profiling in toxicology, i.e. toxicogenomics, characteristic molecular profiles have previously been incorporated into prediction models for early assessment of a carcinogenic potential and mechanism-based classification of compounds. Traditionally, the biomarker signatures used for model construction were derived from individual high-throughput techniques, such as microarrays designed for monitoring global mRNA expression. In this study, we built predictive models by integrating omics data across complementary microarray platforms and introduced new concepts for modeling of pathway alterations and molecular interactions between multiple biological layers. We trained and evaluated diverse machine learning-based models, differing in the incorporated features and learning algorithms on a cross-omics dataset encompassing mRNA, miRNA, and protein expression profiles obtained from rat liver samples treated with a heterogeneous set of substances. Most of these compounds could be unambiguously classified as genotoxic carcinogens, non-genotoxic carcinogens, or non-hepatocarcinogens based on evidence from published studies. Since mixed characteristics were reported for the compounds Cyproterone acetate, Thioacetamide, and Wy-14643, we reclassified these compounds as either genotoxic or non-genotoxic carcinogens based on their molecular profiles. Evaluating our toxicogenomics models in a repeated external cross-validation procedure, we demonstrated that the prediction accuracy of our models could be increased by joining the biomarker signatures across multiple biological layers and by adding complex features derived from cross-platform integration of the omics data. Furthermore, we found that adding these features resulted in a better separation of the compound classes and a more confident reclassification of the three undefined compounds as non-genotoxic carcinogens.     

A different approach to validating screening assays for developmental toxicity

BACKGROUND: There continue to be many efforts around the world to develop assays that are shorter than the traditional embryofetal developmental toxicity assay, or use fewer or no mammals, or use less compound, or have all three attributes. Each assay developer needs to test the putative assay against a set of performance standards, which traditionally has involved testing the assays against a list of compounds that are generally recognized as “positive” or “negative” in vivo. However, developmental toxicity is highly conditional, being particularly dependent on magnitude (i.e. dose) and timing of exposure, which makes it difficult to develop lists of compounds neatly assigned as developmental toxicants or not. APPROACH: Here we offer an alternative approach for the evaluation of developmental toxicity assays based on exposures. Exposures are classified as “positive” or “negative” in a system, depending on the compound and the internal concentration. Although this linkage to “internal dose” departs from the recent approaches to validation, it fits well with widely accepted principles of developmental toxicology. CONCLUSIONS: This paper introduces this concept, discusses some of the benefits and drawbacks of such an approach, and lays out the steps we propose to implement it for the evaluation of developmental toxicity assays. Birth Defects Res (Part B) 89:526–530, 2010. © 2010 Wiley‐Liss, Inc.     

Quantifying Burial,the Transport of Matter from the Lake Biosphere to the Geosphere

Burial is one of the most fundamental processes in contexts of massbalance calculations for substances (such as nutrients, organics, metals and radionuclides) in lakes. Substances can leave a lake by two processes, outflow, i.e., the transport to a downstream system, and burial, i.e., the transport by sedimentation from the lake biosphere to the geosphere. This work gives for the first time, to there best of the author's knowledge, a review on the factors and processes regulating burial and presents a general model for burial. This approach accounts for bottom dynamic conditions (i.e., where areas of fine sediment erosion, transport and accumulation prevail), sedimentation, bioturbation, mineralisation, and the depth and age of the bioactive sediment layer. This approach has been critically tested with very good results for radiocesium, radiostrontium, many metals, calcium from liming and phosphorus, but it has not been presented before in a comprehensive way. This model for burial is meant to be used in massbalance models based on ordinary differential equations (i.e., box models) in contexts where burial is not a target y‐variable but a necessary model variable (an x‐variable). This means that there are also specific demands on this approach, e.g., it must be based on readily accessible driving variables so that it is not too difficult to use the model in practice within the context of an overall lake model. The factors influencing burial, e.g., the deposition of materials and the depth of the bioactive sediment layer, are also needed in calculations of sediment concentrations and to determine amounts of substances or pollutants in sediments. To carry out such calculations, one also needs information on sediment bulk density, water content and organic content. This paper also presents new empirical models for such calculations to be used in the new model for burial.     

Computational prediction of inter-species relationships through omics data analysis and machine learning

Background

Antibiotic resistance and its rapid dissemination around the world threaten the efficacy of currently-used medical treatments and call for novel, innovative approaches to manage multi-drug resistant infections. Phage therapy, i.e., the use of viruses (phages) to specifically infect and kill bacteria during their life cycle, is one of the most promising alternatives to antibiotics. It is based on the correct matching between a target pathogenic bacteria and the therapeutic phage. Nevertheless, correctly matching them is a major challenge. Currently, there is no systematic method to efficiently predict whether phage-bacterium interactions exist and these pairs must be empirically tested in laboratory. Herein, we present our approach for developing a computational model able to predict whether a given phage-bacterium pair can interact based on their genome.

Results

Based on public data from GenBank and phagesDB.org, we collected more than a thousand positive phage-bacterium interactions with their complete genomes. In addition, we generated putative negative (i.e., non-interacting) pairs. We extracted, from the collected genomes, a set of informative features based on the distribution of predictive protein-protein interactions and on their primary structure (e.g. amino-acid frequency, molecular weight and chemical composition of each protein). With these features, we generated multiple candidate datasets to train our algorithms. On this base, we built predictive models exhibiting predictive performance of around 90% in terms of F1-score, sensitivity, specificity, and accuracy, obtained on the test set with 10-fold cross-validation.

Conclusion

These promising results reinforce the hypothesis that machine learning techniques may produce highly-predictive models accelerating the search of interacting phage-bacteria pairs.

     

Do predictive signatures really predict response to cancer chemotherapy?

Predictive signatures are gene expression profiles that should predict the response of tumors to chemotherapy in patients. Such signatures have been derived from the response of tumor cell lines to chemotherapy, but their usefulness in patients remains controversial, as the most spectacular published signatures are based on unreliable data. We discuss why it is difficult to derive meaningful predictive signatures from cell line panels and we argue that it is implausible that fully predictive signatures can be obtained for classical chemotherapy from oligo-based gene expression arrays. One reason is that resistance to chemotherapy can be caused by alterations in (the expression of) a single gene. We do not expect that such subtle alterations will be reliably picked up by standard gene expression profiling. We delineate alternative approaches that should be able to yield predictive markers that can be used for optimizing patient treatment.     

Immortalisation of Primary Cells

Knowledge of the target cells is fundamental to maximise efficiency in attempts at immortalisation of specific cell types. It is also important to optimise the primary cell culture system to promote the survival of the target cell population. Other important factors that may influence the success in obtaining immortalised cells include the toxicity and efficiency of the immortalisation procedure. These can be assessed experimentally and if necessary appropriate techniques can be employed to purify the target cells. When cell lines have been established it is vital to assess them at an early stage for desired scientific and practical features as well as determining their stability and life-span. Furthermore, early characterisation of cell line authenticity (e.g. genetic characters, species of origin) and quality control testing will avoid wasted time and resources should contamination with micro-organisms or another cell line occur. Establishing a programme of immortalisation is a serious undertaking that should only be considered when there are no candidate continuous cell lines available. However, new approaches to modify the biology of cells to give extended life-span, whilst retaining the characteristics of differentiated cells in vivo, will hopefully provide valuable new substrates for in vitro toxicology.     

Linguistic feature analysis for protein interaction extraction

Background  

The rapid growth of the amount of publicly available reports on biomedical experimental results has recently caused a boost of text mining approaches for protein interaction extraction. Most approaches rely implicitly or explicitly on linguistic, i.e., lexical and syntactic, data extracted from text. However, only few attempts have been made to evaluate the contribution of the different feature types. In this work, we contribute to this evaluation by studying the relative importance of deep syntactic features, i.e., grammatical relations, shallow syntactic features (part-of-speech information) and lexical features. For this purpose, we use a recently proposed approach that uses support vector machines with structured kernels.