Open PHACTS computational protocols for in silico target validation of cellular phenotypic screens: knowing the knowns.

Phenotypic screening is in a renaissance phase and is expected by many academic and industry leaders to accelerate the discovery of new drugs for new biology. Given that phenotypic screening is per definition target agnostic, the emphasis of in silico and in vitro follow-up work is on the exploration of possible molecular mechanisms and efficacy targets underlying the biological processes interrogated by the phenotypic screening experiments. Herein, we present six exemplar computational protocols for the interpretation of cellular phenotypic screens based on the integration of compound, target, pathway, and disease data established by the IMI Open PHACTS project. The protocols annotate phenotypic hit lists and allow follow-up experiments and mechanistic conclusions. The annotations included are from ChEMBL, ChEBI, GO, WikiPathways and DisGeNET. Also provided are protocols which select from the IUPHAR/BPS Guide to PHARMACOLOGY interaction file selective compounds to probe potential targets and a correlation robot which systematically aims to identify an overlap of active compounds in both the phenotypic as well as any kinase assay. The protocols are applied to a phenotypic pre-lamin A/C splicing assay selected from the ChEMBL database to illustrate the process. The computational protocols make use of the Open PHACTS API and data and are built within the Pipeline Pilot and KNIME workflow tools.

Inhibition of human glutathione S-transferase P1-1 by tocopherols and alpha-tocopherol derivatives.

alpha-Tocopherol inhibits glutathione S-transferase P1-1 (GST P1-1) (R.I.M. van Haaften, C.T.A. Evelo, G.R.M.M. Haenen, A. Bast, Biochem. Biophys. Res. Commun. 280 (2001)). In various cosmetic and dietary products alpha-tocopherol is added as a tocopherol ester. Therefore we have studied the effect of various tocopherol derivatives on GST P1-1 activity. It was found that GST P1-1 is inhibited, in a concentration dependent manner, by these compounds. Of the compounds tested, the tocopherols were the most potent inhibitors of GST P1-1; the concentration giving 50% inhibition (IC(50)) is <1 microM. The esterified tocopherols and alpha-tocopherol quinone also inhibit the GST P1-1 activity at a very low concentration: for most compounds the IC(50) was below 10 microM. RRR-alpha-Tocopherol acetate lowered the V(max) values, but did not affect the K(m) for either 1-chloro-2,4-dinitrobenzene or GSH. This indicates that the GST P1-1 enzyme is non-competitively inhibited by RRR-alpha-tocopherol acetate. The potential implications of GST P1-1 inhibition by tocopherol and alpha-tocopherol derivatives are discussed.

No reduction of alpha-tocopherol quinone by glutathione in rat liver microsomes.

The cell membrane is protected against lipid peroxidation by endogenous antioxidants such as vitamin E (alpha-tocopherol). The oxidised form of alpha-tocopherol (alpha-tocopherol quinone) does not have this antioxidant function. However, the literature indicates that alpha-tocopherol quinone can be reduced to alpha-tocopherol in vivo and thereby will add to the total antioxidant potential (Moore AN, Ingold KU. Free Radic Biol Med 1997;22:931-4). We found that GSH (reduced glutathione) did not mediate the reduction of alpha-tocopherol quinone, either directly in solution or in rat liver microsomes fortified with alpha-tocopherol quinone. This renders GSH a less likely candidate for alpha-tocopherol quinone reduction in vivo. In addition, alpha-tocopherol quinone did not enhance GSH-dependent protection against lipid peroxidation, either in control microsomes, or in vitamin E-extracted microsomes. Indeed, alpha-tocopherol quinone blocked GSH-dependent protection against lipid peroxidation in vitamin E-extracted microsomes. This indicates that alpha-tocopherol quinone can act as a pro-oxidant.

Physiologically based toxicokinetic modeling of 1,3-butadiene lung metabolism in mice becomes more important at low doses.

This paper describes a physiologically based toxicokinetic model for 1,3-butadiene uptake, distribution, and metabolic clearance in mice. Model parameters for metabolic activity were estimated from the correspondence between computer simulation studies and experimental results as published in the literature. The parameterized model was validated with independent literature data. With the resulting model, the relative importance of lung metabolism as compared to metabolism in the liver increased with decreasing ambient air concentrations. This was due to saturation of metabolism in the alveolar area of the lung, which occurred in the simulations at ambient air concentrations well below current threshold limit values. At higher air concentration, liver metabolism became relatively more important. The tendency toward increased importance of lung metabolism at low doses indicates the necessity of careful extrapolation of in vivo results to low doses. Moreover, this trend may also contribute to species difference in susceptibility to the carcinogenic activity of butadiene.

Glutathione depletion in human erythrocytes as an indicator for microsomal activation of cyclophosphamide and 3-hydroxyacetanilide.

A model system for the detection of reactive metabolites, using glutathione depletion after microsomal activation, has been described previously. We developed a battery of complementary test systems using rat liver microsomes for metabolism and aqueous glutathione solutions, human erythrocytes or hemolysate derived therefrom, as target. Reactive metabolite formation and the ability of metabolites to pass the erythrocyte membrane were tested using 3-hydroxyacetanilide (3-HAA) and cyclophosphamide (CP) as substrates. Neither unchanged 3-HAA nor CP depleted glutathione in erythrocytes or in aqueous reduced glutathione solutions (GSH solutions). Addition of fortified normal or liver microsomes from rats pretreated with phenobarbital (PB microsomes) induced a 3-HAA/CP concentration-dependent glutathione depletion in both systems. With PB microsomes, higher depletions were found. While unchanged 3-HAA did not deplete aqueous GSH solutions or glutathione in erythrocytes, a significant depletion in hemolysate was found. The results indicate that both CP and 3-HAA metabolites are able to pass through the erythrocyte membrane. While both substances can metabolically be activated by rat liver microsomes, only 3-HAA can be activated by soluble factors in erythrocytes. However, unchanged 3-HAA has no effect on GSH in erythrocytes. This might be caused by an inability of unchanged 3-HAA to enter the erythrocyte. More generally, an adequate combination of the test systems described can be used to detect (a) the reactivity of unchanged substances and their metabolites, and (b) the ability of unchanged substances and their reactive metabolites to pass through the erythrocyte membrane.

Influence of glutathione on the formation of cysteine alkylation products in human hemoglobin.

Human blood samples were treated in vitro with iodoacetamide. At low concentrations - less than 1 mM - only a low fraction of the beta 93 cysteine in hemoglobin was alkylated, whereas the alkylating reaction with glutathione was extensive. At higher iodoacetamide concentrations the glutathione pool became exhausted leading to more than proportional increases in the alkylation of the sulfhydryl group in hemoglobin. When diethyl maleate was used as a glutathione depletor prior to incubation with iodoacetamide, low concentrations of iodoacetamide were sufficient to obtain high degrees of hemoglobin sulfhydryl alkylation. N-Ethylmaleimide could not be used as glutathione depletor because the reaction with glutathione appeared to be reversible. The lower reactivity of the thiol group in hemoglobin in comparison with that of glutathione was also found for the isolated biomolecules. The protection of hemoglobin by glutathione present in the human erythrocyte renders the measurement of hemoglobin alkylation less attractive for biological effect monitoring. The sensitivity of such methods is lowered, while the important relation between the alkylation of hemoglobin and that of DNA in the target tissues is affected by interindividual differences in the ratios of the effectiveness of glutathione protection between erythrocytes and target cells.

Hypochlorous acid is a potent inhibitor of GST P1-1.

Glutathione S-transferase is a phase II detoxification enzyme that can be inactivated by H(2)O(2). During oxidative stress various other reactive oxygen species are generated that are more reactive than the relatively stable H(2)O(2). Hypochlorous acid (HOCl) is a powerful oxidant which is highly reactive towards a range of biological substrates. We studied the influence of HOCl on the activity of GST P1-1. HOCl inhibits purified glutathione S-transferase P1-1 in a concentration dependent manner with an IC(50)-value of 0.6 microM, which is more than 1000 times as low as IC(50) reported for H(2)O(2). HOCl lowered the V(max) value, but did not affect the K(m) for CDNB. Our results show that HOCl is a potent, non-competitive inhibitor of GST P1-1. The relevance of this effect is discussed.

Erythrocyte glutathione S transferase as a marker of oxidative stress at birth.

AIMS: To determine the level of oxidative stress and cell damage as a result of exposure to O(2) at birth. METHODS: Using glutathione S transferase (GST) as an indicator of oxidative stress, GST activity in cord blood was compared with that in samples taken three hours after birth. Twenty four prematurely born infants and eight full term infants were studied. To test whether stronger effects occur under less favourable conditions, the neonates were divided in three groups: healthy premature; sick premature; and healthy full term infants. RESULTS: GST activity three hours after birth was significantly decreased compared with that at birth in all three groups tested. There were no significant differences in the magnitude of this effect among the three groups. CONCLUSIONS: These results indicate that a sudden increase in oxygenation exposes the neonate to oxidative stress. Measurement of GST activity might be useful for the evaluation of protective treatment in trials considering antioxidant strategies.

Glutathione depletion in human erythrocytes and rat liver: a study on the interplay between bioactivation and inactivation functions of liver and blood.

The interplay between bioactivation and inactivation functions of human erythrocytes and rat liver was studied. Glutathione depletion was used as a measure of the amount of reduced glutathione (GSH)-reactive compound. Iodoacetamide (IAcA), N-ethylmaleimide (NEM) and diethyl maleate (DEM), which are electrophiles that need no metabolic activation, were able to deplete GSH in incubations with either aqueous GSH solution or erythrocytes. These results indicate that these compounds can pass the erythrocyte membrane. Cyclophosphamide (CP), 3-hydroxyacetanilide (3-HAA) and 2-methylfurane (2-MF) needed metabolic activation by rat liver microsomes to deplete glutathione in incubations with aqueous GSH solution or erythrocytes. By measuring the sum of both reduced and oxidized glutathione [ = total glutathione (GT)] it became clear that GSH-reactive metabolites are generated out of CP, 3-HAA and 2-MF by the action of microsomes and that these metabolites can pass through the erythrocyte membrane. As GT depletion was higher when microsomes of phenobarbital-pretreated rats were used, the metabolites were (are expected to be) generated by phenobarbital-inducible enzymes. GT was also depleted in incubations with haemolysate and 3-HAA or 2-MF but not in incubations with aqueous GSH solution, which indicates that erythrocyte cytosol can metabolize 3-HAA and 2-MF into GSH-reactive compounds. The pesticides monuron and monulinuron did not affect GT concentrations when aqueous GSH solution, haemolysate or erythrocytes with or without microsomal activating system were tested. When hepatocytes were incubated with 3-HAA or CP (2 mm), about 2 mm of internal GT concentration was depleted. The hepatocytes excreted GSH-reactive metabolites generated from 3-HAA and CP (about 20% of the metabolites formed for 3-HAA). Erythrocyte GT was not depleted in co-incubations of hepatocytes and erythrocytes with 3-HAA. This can be explained by the amounts of GSH-reactive metabolites excreted by the hepatocytes, which would require very effective uptake by the erythrocytes in order to be detectable.

Influence of oxygen supply on liver condition and elimination of dimethylacetamide in the isolated perfused rat liver.

The effects were studied of improved oxygen supply on the integrity and metabolic activity towards dimethylacetamide of the isolated perfused rat liver. Improvement of oxygen supply by increased medium oxygenation or addition of chemical oxygen carriers (perfluortributylamine) or erythrocytes led to increased bile secretion. Leakage of lactate dehydrogenase and aspartate aminotransferase could be prevented during a 1-hr perfusion when either chemical oxygen carriers or erythrocytes were added. Improved medium oxygenation alone was not sufficient to prevent high enzyme leakage during the second half of the perfusion period. Histological evaluation confirmed the conclusion that less damage occurred when erythrocytes or perfluortributylamine were added to the perfusion medium. The metabolic clearance of dimethylacetamide by the perfused rat liver was not significantly improved when erythrocytes were added to the medium. The results show that addition of perfluortributylamine, or erythrocytes at a level of 4 g haemoglobin/litre, is necessary to maintain liver integrity for at least 1 hr in the liver perfusion system used in this study.

Only the glutathione dependent antioxidant enzymes are inhibited by haematotoxic hydroxylamines.

Hydroxylamine and some of its derivatives are known to cause oxidative effects both in vitro and in vivo. In the current study we investigated the effects of hydroxylamines on the enzymatic antioxidant defense system in human erythrocytes. The activity of catalase and superoxide dismutase was not significantly influenced by any of the hydroxylamines tested. However, the activity of glutathione peroxidase (GPX) and glutathione S-transferase (GST) was strongly inhibited by hydroxylamine and its O-derivatives (O-methyl and O-ethyl hydroxylamine). GPX was also inhibited by two N-derivatives of hydroxylamine (i.e. N-dimethyl and N,O-dimethyl hydroxylamine). This indicates that exposure to hydroxylamines not only changes the cellular oxidation-reduction status but also leads to inhibition of the glutathione dependent antioxidant enzymes. GST as well as GPX have cysteine residues at the active site of the enzymes. Such an accessible thiol group is generally susceptible to formation of protein-mixed disulphides or intramolecular disulphides. If these thiol groups are essential for activity this would be accompanied by an increase or decrease in the enzyme activity. In principle this is also true for glutathione reductase (GR), which in this study was only inhibited by N,O-dimethyl and N-methyl hydroxylamines. However, GR is capable to reduce these disulphides by taking up two electrons, either from its substrate NAPDH or from another reductant. Oxidation of these thiol groups in GR would thus not lead to impairment of GR activity. The fact that NODMH and NMH do decrease the GR activity can therefore only be explained by other modifications. The activity loss of GST and GPX on the other hand, is likely to involve oxidation of critical cysteine residues. The practical consequence of these findings is that the cellular prooxidant state that may arise in erythrocytes exposed to hydroxylamines can be further increased by activity loss of protective enzymes, which may decrease the average life span of the red blood cell.

Toxicokinetics of dimethylacetamide (DMAc) in rat isolated perfused liver.

Dimethylacetamide (DMAc) is a skin-penetrating solvent able to induce hepatic damage after chronic exposure. Previous research has indicated that metabolism may be saturated at its present TLV/TWA (10 ppm). Biological monitoring of monomethylacetamide (MMAc), the primary metabolite of DMAc, might therefore underestimate exposure to DMAc and related health hazards. We used the recirculating perfusion technique in isolated rat liver to evaluate DMAc metabolism. Medium concentrations starting at about 30, 50, 100 and 275 microM, respectively, were tested. Perfusate samples were taken regularly and analysed for DMAc; pharmacokinetic parameters (extraction ratio and clearance) were calculated for each perfusion. Inlet DMAc concentrations were calculated and concentration groups divided in 16, 36, 70, 160, 225 microM. The extraction ratio of the 16 microM group differed significantly from the other concentration groups tested. DMAc metabolism was saturated at a DMAc concentration of 36 microM. Extraction ratios were unaffected when cimetidine, an inhibitor of cytochrome P450 activity, was added to the perfusion medium or when cimetidine-pretreated animals were used. DMAc clearance was 2.20 ml min-1 at a medium concentration of about 36 microM. Extrapolation of the observed (rat) liver clearance to man showed that airborne concentrations of 18 ppm would, under the presumptions used, lead to saturated metabolism of DMAc; however, saturation at even lower concentrations could not be excluded.

Development of a software package for computer simulations: the use of a sorted event list for reduction of calculation times.

A new program for continuous simulation is described. The program is block oriented, simulating the function of an analogue computer. C structures containing pointers to the calculation of their input functions were used for the individual blocks. The actual calculation order was determined from a sorted event list. After calculation events were inserted at the appropriate place in this list again. This procedure allowed for a variable and independent stepsize for the main blocks and for the introduction of discrete events. With typical applications a speed increase of about 500% was attained.

Transcriptome analysis of peripheral blood mononuclear cells in human subjects following a 36 h fast provides evidence of effects on genes regulating inflammation, apoptosis and energy metabolism.

There is growing interest in the potential health benefits of diets that involve regular periods of fasting. While animal studies have provided compelling evidence that feeding patterns such as alternate-day fasting can increase longevity and reduce incidence of many chronic diseases, the evidence from human studies is much more limited and equivocal. Additionally, although several candidate processes have been proposed to contribute to the health benefits observed in animals, the precise molecular mechanisms responsible remain to be elucidated. The study described here examined the effects of an extended fast on gene transcript profiles in peripheral blood mononuclear cells from ten apparently healthy subjects, comparing transcript profiles after an overnight fast, sampled on four occasions at weekly intervals, with those observed on a single occasion after a further 24 h of fasting. Analysis of the overnight fasted data revealed marked inter-individual differences, some of which were associated with parameters such as gender and subject body mass. For example, a striking positive association between body mass index and the expression of genes regulated by type 1 interferon was observed. Relatively subtle changes were observed following the extended fast. Nonetheless, the pattern of changes was consistent with stimulation of fatty acid oxidation, alterations in cell cycling and apoptosis and decreased expression of key pro-inflammatory genes. Stimulation of fatty acid oxidation is an expected response, most likely in all tissues, to fasting. The other processes highlighted provide indications of potential mechanisms that could contribute to the putative beneficial effects of intermittent fasting in humans.

From nanotechnology to nanomedicine: applications to cancer research.

Scientific advances have significantly improved the practice of medicine by providing objective and quantitative means for exploring the human body and disease states. These innovative technologies have already profoundly improved disease detection, imaging, treatment and patient follow-up. Today's analytical limits are at the nanoscale level (one-billionth of a meter) enabling a detailed exploration at the level of DNA, RNA, proteins and metabolites which are in fact nano-objects. This translational review aims at integrating some recent advances from micro- and nano-technologies with high potential for improving daily oncology practice.

The role of bioinformatics in pathway curation.

Diagrams and models of biological pathways are useful tools in biology. Pathway diagrams are mainly used for illustrative purposes for instance in textbooks and in presentations. Pathway models are used in the analysis of genomic data. Bridging the gap between diagrams and models allows not only the analysis of genomics data and interactions but also the visualisation of the results in a variety of different ways. The knowledge needed for pathway creation and curation is available from three distinct sources: databases, literature and experts. We describe the role of bioinformatics in facilitating the creation and curation of pathway.

Validating nutrient-related gene expression changes from microarrays using RT(2) PCR-arrays.

Microarray technology allows us to perform high-throughput screening of changes in gene expression. The outcome of microarray experiments largely depends on the applied analysis methods and cut-off values chosen. Results are often required to be verified using a more sensitive detection technique, such as quantitative real-time PCR (qPCR or RT-PCR). Throughout the years, this technique has become a de facto golden standard. Individual qPCRs are time-consuming, but the technology to perform high-throughput qPCR reactions has become available through PCR-arrays that allow up to 384 PCR reactions simultaneously. Our current aim was to investigate the usability of a RT(2) Profilertrade mark PCR-array as validation in a nutritional intervention study, where the measured changes in gene expression were low. For some differentially expressed genes, the PCR-array confirmed the microarray prediction, though not for all. Furthermore, the PCR-array allowed picking up the expression of genes that were not measurable on the microarray platform but also vice versa. We conclude that both techniques have their own (dis)advantages and specificities, and for less pronounced changes using both technologies may be useful as complementation rather than validation.

Bioinformatic interrogation of expression array data to identify nutritionally regulated genes potentially modulated by DNA methylation.

DNA methylation occurs at CpG dinucleotide sites within the genome and is recognised as one of the mechanisms involved in regulation of gene expression. CpG sites are relatively underrepresented in the mammalian genome, but occur densely in regions called CpG islands (CGIs). CGIs located in the promoters of genes inhibit transcription when methylated by impeding transcription factor binding. Due to the malleable nature of DNA methylation, environmental factors are able to influence promoter CGI methylation patterns and thus influence gene expression. Recent studies have provided evidence that nutrition (and other environmental exposures) can cause altered CGI methylation but, with a few exceptions, the genes influenced by these exposures remain largely unknown. Here we describe a novel bioinformatics approach for the analysis of gene expression microarray data designed to identify regulatory sites within promoters of differentially expressed genes that may be influenced by changes in DNA methylation.

Genetic variation in thioredoxin interacting protein (TXNIP) is associated with hypertriglyceridaemia and blood pressure in diabetes mellitus.

AIMS: Thioredoxin interacting protein (TXNIP) is an attractive candidate gene for diabetes or diabetic dyslipidaemia, since TXNIP is the strongest glucose-responsive gene in pancreatic B-cells, TXNIP deficiency in a mouse model is associated with hyperlipidaemia and TXNIP is located in the 1q21-1q23 chromosomal Type 2 diabetes mellitus (DM) locus. We set out to investigate whether metabolic effects of TXNIP that were previously reported in a murine model are also relevant in human Type 2 DM. METHODS: The frequency distribution of a 3' UTR single nucleotide polymorphism (SNP) in TXNIP was investigated in subjects with normal glucose tolerance (NGT; n = 379), impaired glucose tolerance (IGT; n = 228) and Type 2 DM (n = 230). Metabolic data were used to determine the effect of this SNP on parameters associated with lipid and glucose metabolism. RESULTS: The frequency of the TXNIP variation did not differ between groups, but within the group of diabetic subjects, carriers of the TXNIP-T variant had 1.6-fold higher triglyceride concentrations (P = 0.015; n = 136) and a 5.5-mmHg higher diastolic blood pressure (P = 0.02; n = 212) than homozygous carriers of the common C-allele, whereas in non-diabetic subjects fasting glucose was 0.26 mmol/l lower (P = 0.002; n = 478) in carriers of the T-allele. Moreover, a significant interaction between plasma glucose concentrations and TXNIP polymorphism on plasma triglycerides was observed (P = 0.012; n = 544). CONCLUSION: This is the first report to implicate TXNIP in a human disorder of energy metabolism, Type 2 diabetes. The effect of TXNIP on triglycerides is influenced by plasma glucose concentrations, suggesting that the biological relevance of TXNIP variations may be particularly relevant in recurrent episodes of hyperglycaemia.

Gene profiling of cathepsin K deficiency in atherogenesis: profibrotic but lipogenic.

Recently, we showed that cathepsin K deficiency reduces atherosclerotic plaque progression, induces plaque fibrosis, but aggravates macrophage foam cell formation in the ApoE -/- mouse. To obtain more insight into the molecular mechanisms by which cathepsin K disruption evokes the observed phenotypic changes, we used microarray analysis for gene expression profiling of aortic arches of CatK -/-/ApoE -/- and ApoE -/- mice on a mouse oligo microarray. Out of 20 280 reporters, 444 were significantly differentially expressed (p-value of < 0.05, fold change of > or = 1.4 or < or = - 1.4, and intensity value of > 2.5 times background in at least one channel). Ingenuity Pathway Analysis and GenMAPP revealed upregulation of genes involved in lipid uptake, trafficking, and intracellular storage, including caveolin - 1, - 2, - 3 and CD36, and profibrotic genes involved in transforming growth factor beta (TGFbeta) signalling, including TGFbeta2, latent TGFbeta binding protein-1 (LTBP1), and secreted protein, acidic and rich in cysteine (SPARC), in CatK -/-/ApoE -/- mice. Differential gene expression was confirmed at the mRNA and protein levels. In vitro modified low density lipoprotein (LDL) uptake assays, using bone marrow derived macrophages preincubated with caveolae and scavenger receptor inhibitors, confirmed the importance of caveolins and CD36 in increasing modified LDL uptake in the absence of cathepsin K. In conclusion, we suggest that cathepsin K deficiency alters plaque phenotype not only by decreasing proteolytic activity, but also by stimulating TGFbeta signalling. Besides this profibrotic effect, cathepsin K deficiency has a lipogenic effect owing to increased lipid uptake mediated by CD36 and caveolins.