Profiling withanolide A for therapeutic targets in neurodegenerative diseases.

To identify new potential therapeutic targets for neurodegenerative diseases, we initiated activity-based protein profiling studies with withanolide A (WitA), a known neuritogenic constituent of Withania somnifera root with unknown mechanism of action. Molecular probes were designed and synthesized, and led to the discovery of the glucocorticoid receptor (GR) as potential target. Molecular modeling calculations using the VirtualToxLab predicted a weak binding affinity of WitA for GR. Neurite outgrowth experiments in human neuroblastoma SH-SY5Y cells further supported a glucocorticoid-dependent mechanism, finding that WitA was able to reverse the outgrowth inhibition mediated by dexamethasone (Dex). However, further GR binding and transactivation assays found no direct interference of WitA. Further molecular modeling analysis suggested that WitA, although forming several contacts with residues in the GR binding pocket, is lacking key stabilizing interactions as observed for Dex. Taken together, the data suggest that WitA-dependent induction of neurite outgrowth is not through a direct effect on GR, but might be mediated through a closely related pathway. Further experiments should evaluate a possible role of GR modulators and/or related signaling pathways such as ERK, Akt, NF-κB, TRα, or Hsp90 as potential targets in the WitA-mediated neuromodulatory effects.

In Silico Prediction of the Toxic Potential of Lupeol.

Lupeol is a natural triterpenoid found in many plant species such as mango. This compound is the principal active component of many traditional herbal medicines. In the past decade, a considerable number of publications dealt with lupeol and its analogues due to the interest in their pharmacological activities against cancer, inflammation, arthritis, diabetes, and heart disease. To identify further potential applications of lupeol and its analogues, it is necessary to investigate their mechanisms of action, particularly their interaction with off-target proteins that may trigger adverse effects or toxicity. In this study, we simulated and quantified the interaction of lupeol and 11 of its analogues toward a series of 16 proteins known or suspected to trigger adverse effects employing the VirtualToxLab. This software provides a thermodynamic estimate of the binding affinity, and the results were challenged by molecular-dynamics simulations, which allow probing the kinetic stability of the underlying protein-ligand complexes. Our results indicate that there is a moderate toxic potential for lupeol and some of its analogues, by targeting and binding to nuclear receptors involved in fertility, which could trigger undesired adverse effects.

Molecular mechanisms of endocrine and metabolic disruption: An in silico study on antitrypanosomal natural products and some derivatives.

The VirtualToxLab is an in silico technology for estimating the toxic potential - endocrine and metabolic disruption, as well as aspects of carcinogenicity and cardiotoxicity - of drugs, chemicals and natural products. The technology is based on an automated protocol that simulates and quantifies the binding of small molecules towards a series of currently 16 proteins, known or suspected to trigger adverse effects. The simulations are conducted at the atomic level and explicitly allow for a mechanistic interpretation of the results (in real-time 3D/4D), thereby complying with the Setubal principles put forward in 2002 for computational approaches to toxicology. Moreover, the underlying "ab-initio" protocol is independent from any training data and makes the approach universal with respect to the applicability domain. The VirtualToxLab runs in client-server mode and is freely available to academic and non-profit organizations. As the underlying technology yields a thermodynamic estimate of the binding affinity, the associated ligand-protein complexes have been challenged by molecular-dynamics simulations to probe their kinetic stability. Human African trypanosomiasis is a neglected tropical disease caused by two subspecies of Trypanosoma brucei. The control of this parasitic infection relies on a few chemotherapeutic agents, most of which were discovered decades ago and pose many challenges including adverse side effects, poor efficacy, and the occurrence of drug resistances. Natural products, on the other hand, offer a high potential for the discovery of new drug leads due to their chemical diversity. In this in silico study, we analyze a series of 89 natural products and derivatives displaying anti-trypanosomal activity for their potential to trigger adverse effects. Our results indicate a moderate potential for a number of those compounds to bind to nuclear receptors and thereby ease the development of endocrine disregulation. A few others would seem to inhibit enzymes of the cytochrome P450 family and, hence, sustain drug-drug interactions.

OpenVirtualToxLab--a platform for generating and exchanging in silico toxicity data.

The VirtualToxLab is an in silico technology for estimating the toxic potential--endocrine and metabolic disruption, some aspects of carcinogenicity and cardiotoxicity--of drugs, chemicals and natural products. The technology is based on an automated protocol that simulates and quantifies the binding of small molecules towards a series of currently 16 proteins, known or suspected to trigger adverse effects: 10 nuclear receptors (androgen, estrogen α, estrogen ß, glucocorticoid, liver X, mineralocorticoid, peroxisome proliferator-activated receptor γ, progesterone, thyroid α, thyroid ß), four members of the cytochrome P450 enzyme family (1A2, 2C9, 2D6, 3A4), a cytosolic transcription factor (aryl hydrocarbon receptor) and a potassium ion channel (hERG). The toxic potential of a compound--its ability to trigger adverse effects--is derived from its computed binding affinities toward these very proteins: the computationally demanding simulations are executed in client-server model on a Linux cluster of the University of Basel. The graphical-user interface supports all computer platforms, allows building and uploading molecular structures, inspecting and downloading the results and, most important, rationalizing any prediction at the atomic level by interactively analyzing the binding mode of a compound with its target protein(s) in real-time 3D. Access to the VirtualToxLab is available free of charge for universities, governmental agencies, regulatory bodies and non-profit organizations.

Migration of cyclo-diBA from coatings into canned food: method of analysis, concentration determined in a survey and in silico hazard profiling.

Cyclo-diBA, the cyclic product formed from bisphenol A and bisphenol A diglycidyl ether during production of epoxy resins, was measured in canned food using reversed phase HPLC with fluorescence detection. Half (9 of 17) of the samples of canned fish in oil collected in April 2010 contained cyclo-diBA with an average concentration of 1025 µg/kg and a maximum of 1980 µg/kg. In September 2012, cyclo-diBA was detectable (>25 µg/kg) in merely 13 from 44 such products; the average concentration in these was 807 µg/kg and the maximum now reached 2640 µg/kg. Fish in brine contained far less cyclo-diBA. The majority of the canned meat products contained cyclo-diBA at a mean concentration of 477 µg/kg and a maximum of 1050 µg/kg. All prepared meals, such as ravioli or soups, contained cyclo-diBA, with a mean at 287 µg/kg. In canned tomatoes, peas and other vegetables in water or fruits in syrup, no cyclo-diBA was detected (<25 µg/kg). Since no experimental toxicity data are available except for its cytotoxicity, an in silico hazard profiling was performed. Cyclo-diBA seems to be stable and of low reactivity. There is indication for considerable oral bioavailability and for the potential to accumulate in the human body. Cyclo-diBA can be metabolized into cyclic and acyclic compounds. Based on SAR assessment for cyclo-diBA and read-across from BADGE to linear cyclo-diBA metabolites, genotoxic effects are improbable. Specific binding of cyclo-diBA to nuclear receptors, such as ERß, can be predicted, indicating a potential endocrine-disrupting potency. The limit by the EFSA guidelines of 50 µg/person/d for compounds shown not to be genotoxic as well as the TTC-based Cramer structural class III value of 90 µg/person/d could be exceeded several fold by high consumers of canned fish in oil with high brand loyalty. As a consequence, risk reduction measures were taken.

VirtualToxLab - a platform for estimating the toxic potential of drugs, chemicals and natural products.

The VirtualToxLab is an in silico technology for estimating the toxic potential (endocrine and metabolic disruption, some aspects of carcinogenicity and cardiotoxicity) of drugs, chemicals and natural products. The technology is based on an automated protocol that simulates and quantifies the binding of small molecules towards a series of proteins, known or suspected to trigger adverse effects. The toxic potential, a non-linear function ranging from 0.0 (none) to 1.0 (extreme), is derived from the individual binding affinities of a compound towards currently 16 target proteins: 10 nuclear receptors (androgen, estrogen α, estrogen ß, glucocorticoid, liver X, mineralocorticoid, peroxisome proliferator-activated receptor γ, progesterone, thyroid α, and thyroid ß), four members of the cytochrome P450 enzyme family (1A2, 2C9, 2D6, and 3A4), a cytosolic transcription factor (aryl hydrocarbon receptor) and a potassium ion channel (hERG). The interface to the technology allows building and uploading molecular structures, viewing and downloading results and, most importantly, rationalizing any prediction at the atomic level by interactively analyzing the binding mode of a compound with its target protein(s) in real-time 3D. The VirtualToxLab has been used to predict the toxic potential for over 2500 compounds: the results are posted on http://www.virtualtoxlab.org. The free platform - the OpenVirtualToxLab - is accessible (in client-server mode) over the Internet. It is free of charge for universities, governmental agencies, regulatory bodies and non-profit organizations.

VirtualToxLab - in silico prediction of the toxic (endocrine-disrupting) potential of drugs, chemicals and natural products. Two years and 2,000 compounds of experience: a progress report.

The VirtualToxLab is an in silico tool for predicting the toxic (endocrine-disrupting) potential of drugs, chemicals and natural products. It is based on a fully automated protocol and calculates the binding affinity of any molecule of interest towards a series of 12 proteins, known or suspected to trigger adverse effects and estimates the resulting toxic potential. In contrast to other approaches in the field, the technology allows to rationalize a prediction at the molecular level by interactively analyzing the binding mode of the tested compound with any target protein in 3D. The technology is accessible over the Internet (via a secure SSH protocol) and available for any science-oriented organization. The toxic potential - a complex value derived from the individual binding affinities, their standard deviation and the quality of the underlying model (number and ratio of training and test compounds, activity range covered) - of existing and hypothetical compounds is estimated by simulating and quantifying their interactions towards a series of macromolecular targets at the molecular level using automated flexible docking combined with multidimensional QSAR (mQSAR). Currently, those targets comprise 12 proteins: the androgen, aryl hydrocarbon, estrogen alpha/beta, glucocorticoid, mineralocorticoid, thyroid alpha/beta liver X and the peroxisome proliferator-activated receptor gamma as well as the enzymes cytochrome P450 3A4 (CYP 3A4) and 2A13 (CYP 2A13). Up to date, the technology has been used to predict the toxic potential for more than 2,000 drugs, chemicals and natural compounds. All results are posted in the Internet - in this account, a few will be discussed in detail with reference to the molecular mechanisms triggering the adverse effect.

VirtualToxLab - in silico prediction of the toxic potential of drugs and environmental chemicals: evaluation status and internet access protocol.

We present a receptor-modeling concept based on multidimensional QSAR (mQSAR) developed at our laboratory for the in silico prediction of the toxic potential of drugs and environmental chemicals. Presently, the VirtualToxLab includes nine so-called virtual test kits for the estrogen (alpha/beta), androgen, thyroid (alpha/beta), glucocorticoid, aryl hydrocarbon, and peroxisome proliferator-activated receptor gamma, as well as for the enzyme cytochrome P450 3A4. The surrogates have been tested against a total of 798 compounds and are able to predict the binding affinity close to the experimental uncertainty, with only six of the 188 test compounds being calculated more than a factor of 10 off the experimental binding affinity and the maximal individual deviation not exceeding a factor of 15. These results suggest that our approach is suited for the in silico identification of adverse effects triggered by drugs and environmental chemicals. In this account, we summarise the current evaluation status of the models and introduce an Internet access portal, immediately available to selected laboratories, and aimed at a peer evaluation of our concept.

Predicting the toxic potential of drugs and chemicals in silico: a model for the peroxisome proliferator-activated receptor gamma (PPAR gamma).

Poor pharmacokinetics, side effects and compound toxicity are frequent causes of late-stage failures in drug development. A safe in silico identification of adverse effects triggered by drugs and chemicals would therefore be highly desirable as it not only bears economical potential but also spawns a variety of ecological benefits: sustainable resource management, reduction of animal models and possibly less risky clinical trials as in silico studies are typically based on human proteins. In the recent past, our laboratory has developed a 6D-QSAR concept and validated a series of "virtual test kits" based on the aryl hydrocarbon, estrogen, androgen, thyroid, and glucocorticoid receptor as well as on the enzyme cytochrome P450 3A4. The test kits were trained using a representative selection of 610 substances and validated with 188 compounds different therefrom. These models were subsequently compiled into a database for the virtual screening of drugs and environmental chemicals. In this account, we report the validation of a model for the peroxisome proliferator-activated receptor gamma (PPAR gamma). Its receptor surrogate is based on the experimental structure of the protein and 95 tyrosine-based compounds. The simulation reached a cross-validated r(2)=0.832 (75 training ligands) and yielded a predictive r(2)=0.723 (20 test compounds). The model was challenged by a series of scramble tests as well as with the prediction of a few structurally different compounds.

Simulating alpha/beta selectivity at the human thyroid hormone receptor: consensus scoring using multidimensional QSAR.

We present a consensus-scoring study on the human thyroid hormone receptor alpha and beta using two receptor-modeling concepts (software Quasar and Raptor) that are based on multidimensional QSAR and allow for the explicit simulation of induced fit. The binding mode of 82 agonists and indirect antagonists, spanning an activity range of seven orders of magnitude in K(i), was identified through flexible docking to the respective X-ray crystal structures (Yeti software) and represented by a 4D data set with up to four conformations per compound. The receptor surrogates for the thyroid alpha receptor converged at a cross-validated r(2) of 0.846/0.919 (64 training compounds; for Quasar and Raptor, respectively) and yielded a predictive r(2) of 0.812/0.814 (18 test compounds); the models for the thyroid beta receptor resulted in a cross-validated r(2) of 0.823/0.909 and a predictive r(2) of 0.665/0.796, respectively. Consensus was achieved as, on average, the calculated activities of the training set differ only by a factor of 2.2 in K(i) and those of the test set by a factor of 2.8 when predicted by Quasar and Raptor, respectively.

Predicting the toxic potential of drugs and chemicals in silico.

Based on the 3D structure of the target protein (ERalphabeta, AR, PPARgamma, TRalphabeta, GR; CYP3A4) or a surrogate thereof (AhR), the Biographics Laboratory 3R has generated a series of virtual test kits and validated them against 693 compounds. In a pilot project (ToxDataBase), both existing and new drugs or environmental chemicals can be screened for their endocrine-disrupting potential or the probability to trigger drug-drug interactions in silico. After peer testing (2007-8), it is planned to make the database available on the Internet.

Virtual test kits for predicting harmful effects triggered by drugs and chemicals mediated by specific proteins.

Poor pharmacokinetics and toxicity are not only frequent causes of late-stage failures in drug development but also a source for unnecessary animal tests. In drug discovery and for the assessment of the toxic potential of chemicals, in silico techniques are nowadays considered as valuable alternatives to in vivo approaches. Based on a receptor-modelling concept developed at our laboratory (multidimensional QSAR), we have developed and validated virtual test kits for the estrogen, androgen and aryl hydrocarbon receptor (endocrine disruption), for cytochrome P450 3A4 (metabolic transformations) and most recently for the thyroid receptor. These surrogates have been tested against a total of 430 compounds and are able to predict the binding affinity close to the experimental uncertainty. These results suggest that our approach is suited for the in silico identification of adverse effects triggered by drugs and chemicals. Consequently, we are prepared to offer a free testing to selected academic institutions and non-profit oriented organisations.

Impact of induced fit on ligand binding to the androgen receptor: a multidimensional QSAR study to predict endocrine-disrupting effects of environmental chemicals.

We investigated the influence of induced fit of the androgen receptor binding pocket on free energies of ligand binding. On the basis of a novel alignment procedure using flexible docking, molecular dynamics simulations, and linear-interaction energy analysis, we simulated the binding of 119 molecules representing six compound classes. The superposition of the ligand molecules emerging from the combined protocol served as input for Raptor, a receptor-modeling tool based on multidimensional QSAR allowing for ligand-dependent induced fit. Throughout our study, protein flexibility was explicitly accounted for. The model converged at a cross-validated r(2) = 0.858 (88 training compounds) and yielded a predictive r(2) = 0.792 (26 test compounds), thereby predicting the binding affinity of all compounds close to their experimental value. We then challenged the model by testing five molecules not belonging to compound classes used to train the model: the IC(50) values were predicted within a factor of 4.5 compared to the experimental data. The demonstrated predictivity of the model suggests that our approach may well be beneficial for both drug discovery and the screening of environmental chemicals for endocrine-disrupting effects, a problem that has recently become a cause for concern among scientists, environmental advocates, and politicians alike.

In silico prediction of harmful effects triggered by drugs and chemicals.

While the computer-assisted discovery and optimization of drug candidates based on the known three-dimensional structure of the macromolecular target (structure-based design) or a binding-site surrogate (receptor modeling) is doubtless one of the more potent approaches in rational drug design, the simulation and quantification of side effects triggered by drugs and chemicals are still in their infancy. Major obstacles include the often not available 3D structure of the molecular target, the low specificity of the involved bioregulators and the identification of the controlling metabolic pathways. In the recent past, our laboratory has explored concepts allowing to simulate receptor-mediated toxic phenomena by developing algorithms, allowing to construct realistic 3D binding-site surrogates of receptors known or assumed triggering adverse effects and validating them against large batches of molecular data. The underlying technology (software Quasar and Raptor, respectively) specifically allows for induced fit, solvation phenomena and entropic effects. It has been applied to various systems both of pharmacological and toxicological interest including the neurokinin-1, chemokine-3, bradykinin B(2), steroid, 5 HT(2A), aryl hydrocarbon, estrogen and androgen receptor, respectively. In this account, we describe the design of a virtual laboratory allowing for a reliable estimation of harmful effects triggered by drugs, chemicals and their metabolites in silico. In the recent past, the Biographics Laboratory 3R has compiled a 3D database including the surrogates of three major receptor systems known to mediate adverse effects (the aryl hydrocarbon, the estrogen and the androgen receptor, respectively) and validated them against a total of 345 compounds (drugs, chemicals, toxins) using multidimensional QSAR technologies. Within this pilot project, we could demonstrate that our virtual laboratory is able to both recognize toxic compounds substantially different from those used in the training set as well as to classify harmless compounds as being nontoxic. This suggests that our approach may be used for the prediction of adverse effects of drug molecules and chemicals. It is the aim to provide cost-covering access to this technology--particularly to universities, hospitals and regulatory bodies--as it bears a significant potential to recognize hazardous compounds early in the development process and hence improve resource and waste management as well as reduce animal testing. The Biographics Laboratory 3R is a non-profit-oriented organization aimed at reducing animal experimentation in the biomedical sciences by computational approaches (cf. http://www.biograf.ch).

Combining protein modeling and 6D-QSAR. Simulating the binding of structurally diverse ligands to the estrogen receptor.

We present a concept for the in silico simulation of adverse effects triggered by drugs and chemicals. The underlying philosophy combines flexible docking (software Yeti) for the identification of the binding mode(s) and 6D-QSAR (software Quasar) for their quantification. The results obtained for 106 diverse molecules binding to the estrogen receptor (q2 = 0.903; p2 = 0.885) suggest that our approach is suitable for the identification of an endocrine-disrupting potential associated with drugs and chemicals.

Raptor: combining dual-shell representation, induced-fit simulation, and hydrophobicity scoring in receptor modeling: application toward the simulation of structurally diverse ligand sets.

We present a novel receptor-modeling approach (software Raptor) based on multidimensional quantitative structure-activity relationships (QSARs). To accurately predict relative free energies of ligand binding, it is of utmost importance to simulate induced fit. In Raptor, we explicitly and anisotropically allow for this phenomenon by a dual-shell representation of the receptor surrogate. In our concept, induced fit is not limited to steric aspects but includes the variation of the physicochemical fields along with it. The underlying scoring function for evaluating ligand-receptor interactions includes directional terms for hydrogen bonding and hydrophobicity and thereby treats solvation effects implicitly. This makes the approach independent from a partial-charge model and, as a consequence, allows one to smoothly model ligand molecules binding to the receptor with different net charges. We have applied the new concept toward the estimation of ligand-binding energies associated with the chemokine receptor-3 (50 ligands: r(2) = 0.965; p(2) = 0.932), the bradykinin B(2) receptor (52 ligands: r(2) = 0.949; p(2) = 0.859), and the estrogen receptor (116 ligands: r(2) = 0.908; p(2) = 0.907), respectively.

Internet laboratory for predicting harmful effects triggered by drugs and chemicals--a progress report.

The main objective of our institution is to establish a virtual laboratory on the Internet to allow for a reliable in silico estimation of harmful effects triggered by drugs, chemicals and their metabolites. In the past two years, we have compiled a pilot system including the 3D models of five receptors known to mediate adverse effects (the Ah, 5HT(2A), cannabinoid, GABA(A), and estrogen receptor, respectively) and tested them against 280 compounds (drugs, chemicals, toxins). Within this set-up we could demonstrate that our concept is able to both recognise toxic compounds substantially different from those used in the training set as well as to classify harmless compounds clearly as being non-toxic at low-level doses. This suggests that our approach can be used for the prediction of adverse effects of drug molecules and chemicals. It is the aim to provide free access to this 3D data base, particularly to universities, hospitals and regulatory bodies as it bears a significant potential to recognise hazardous compounds early in the development process and withdraw them from the evaluation pipeline. Hence, for substances recognised as hazardous in silico, subsequent toxicity tests involving animal models become obsolete.

[Ochratoxins: Molecular strategies for developing an antidote]

Ochratoxin A (OcA) is a prominent member of a group of mycotoxins which display nephrotoxic, genotoxic, teratogenic, carcinogenic and immunosuppressive effects and which have also been linked to Balkan Endemic Nephropathy. The toxicity of OcA is thought to be primarily due to its inhibition of phenylalanine-t-RNA synthetase, a phenylalanine-metabolizing enzyme. Based on the three-dimensional structure of phenylalanine-t-RNA synthetase, we have analyzed its interactions with OcA by means of molecular-dynamical simulations and identified three quite different binding modes, all of which suggest an affinity only in the millimolar range. This would seem to be in conflict with toxicological findings frequently cited in textbooks but is in agreement with recent in vitro studies on purified phenylalanine-t-RNA synthetase, which also exclude this enzyme as the main target for OcA action. In vivo, OcA binds preferentially to serum albumin, a plasma protein, with a corresponding effect on its toxicokinetics (retention). Antagonizing this effect would lead to an enhanced elimination rate, thereby reducing all adverse effects of OcA, as has been demonstrated using albumin-deficient mice. Based on the three-dimensional structure of serum albumin, we have simulated its interaction with OcA. The long-term goal is the animal-free identification of a synthetic antagonist with an affinity between that of the endogenous ligands (e.g. billirubin) and OcA. Such a substance could - by reducing the retention time of the toxin in the body - potentially eliminate all toxic effects of OcA.

[Target proteins and mechanisms of ochratoxin toxicity. A contribution to the identification of potential ochratoxin antagonists]

Ochratoxins are mycotoxins released by moulds on grain, peanuts and vegetables. Toxicological investigations have shown that ochratoxin A displays nephrotoxic, genotoxic, teratogenic, cancerogenic and immunosuppressive effects. Increased blood levels observed in humans would seem to suggest a link to a kidney desease (Balcan Endemic Nephropaty) frequently observed in the Balkan countries. The adverse effects of ochratoxin A are mainly associated with its impact on phenylalanine-metabolizing enzymes. Based on the three-dimensional structure of phenylalanine-t-RNA-synthetase, its interactions with ochratoxins are analyzed as well as with Aspartam. In animal models, Aspartam has been shown to almost fully prevent toxic effects of ochratoxin A. The topology of the binding site of phenylalanine-t-RNA-synthetase would seem to be favorable towards a few affinity-enhancing modifications of the Aspartame molecule. A known molecular mechanism is a prerequisite for a systematic search of antagonizing substances for toxins. Based on a receptor structure, binding properties of such drugs can be identified and optimized using computer-aided drug design. Susequently, only the most potent candidate structures must be subjected to a determination of their biological activity, which can lead to a significant reduction of substances to be tested in vivo. Such experiments are particularly stressful as the animals must be intoxicated beforehand. The extent of an antagonistic impact on humans suffering from a chronical ochratoxin A intoxication must be subject of clinical studies.