Prediction of eye irritation from organic chemicals using membrane-interaction QSAR analysis

In silico prediction of the full United Nations Globally Harmonized System eye irritation categories of liquid chemicals by IATA-like bottom-up approach of random forest method

As an alternative to in vivo Draize rabbit eye irritation test, this study aimed to construct an in silico model to predict the complete United Nations (UN) Globally Harmonized System (GHS) for classification and labeling of chemicals for eye irritation category [eye damage (Category 1), irritating to eye (Category 2) and nonirritating (No category)] of liquid chemicals with Integrated approaches to testing and assessment (IATA)-like two-stage random forest approach. Liquid chemicals (n = 219) with 34 physicochemical descriptors and quality in vivo data were collected with no missing values. Seven machine learning algorithms (Naive Bayes, Logistic Regression, First Large Margin, Neural Net, Random Forest (RF), Gradient Boosted Tree, and Support Vector Machine) were examined for the ternary categorization of eye irritation potential at a single run through 10-fold cross-validation. RF, which performed best, was further improved by applying the 'Bottom-up approach' concept of IATA, namely, separating No category first, and discriminating Category 1 from 2, thereafter. The best performing training dataset achieved an overall accuracy of 73% and the correct prediction for Category 1, 2, and No category was 80%, 50%, and 77%, respectively for the test dataset. This prediction model was further validated with an external dataset of 28 chemicals, for which an overall accuracy of 71% was achieved.

An overview of current alternative models in the context of ocular surface toxicity

The 21st century has seen a steadily increasing social awareness of animal suffering, with increased attention to ethical considerations. Developing new integrated approaches to testing and assessment (IATA) strategies is an Organisation for Economic Co-operation and Development (OECD) goal to reduce animal testing. Currently, there is a lack of alternative models to test for ocular surface toxicity (aside from irritation) in lieu of the Draize eye irritation test (OECD guideline No. 405) performed in rabbits. Five alternative in vitro or ex vivo methods have been validated to replace this reference test, but only in combination. However, pathologies like Toxicity-Induced Dry Eye (TIDE), cataract, glaucoma, and neuropathic pain can occur after exposure to a pharmaceutical product or chemical and therefore need to be anticipated. To do so, new models of lacrimal glands, lens, and neurons innervating epithelia are required. These models must take into account real-life exposure (dose, time, and tear film clearance). The scientific community is working hard to develop new, robust, alternative, in silico, and in vitro models, while attempting to balance ethics and availability of biological materials. This review provides a broad overview of the validated methods for analyzing ocular irritation and those still used by some industries, as well as promising models that need to be optimized according to the OECD. Finally, we give an overview of recently developed innovative models, which could become new tools in the evaluation of ocular surface toxicity within the scope of IATAs.

AOP and IATA applied to ocular surface toxicity

Until now, the Draize test on rabbits has been the only test performed to anticipate ocular toxicity of pharmaceutical compounds, mainly irritation. The OECD is urging the scientific community to develop and validate alternative methods to reduce the need for animal testing. Since the models and tests used cannot reflect the entire biologic response, it is necessary to combine them into integrated approaches to testing and assessment (IATA) to obtain robust data. IATAs, along with adverse outcome pathways (AOP) that encompass molecular cascades and key events, require the best combinations of tests. This commentary manuscript describes these OECD tools and proposes original approaches for ocular surface AOP and an IATA for toxicity-induced dry eye (TIDE).

Functional and Material Properties in Nanocatalyst Design: A Data Handling and Sharing Problem

(1) Background: Properties and descriptors are two forms of molecular in silico representations. Properties can be further divided into functional, e.g., catalyst or drug activity, and material, e.g., X-ray crystal data. Millions of real measured functional property records are available for drugs or drug candidates in online databases. In contrast, there is not a single database that registers a real conversion, TON or TOF data for catalysts. All of the data are molecular descriptors or material properties, which are mainly of a calculation origin. (2) Results: Here, we explain the reason for this. We reviewed the data handling and sharing problems in the design and discovery of catalyst candidates particularly, material informatics and catalyst design, structural coding, data collection and validation, infrastructure for catalyst design and the online databases for catalyst design. (3) Conclusions: Material design requires a property prediction step. This can only be achieved based on the registered real property measurement. In reality, in catalyst design and discovery, we can observe either a severe functional property deficit or even property famine.

An alternative predictor of eye irritation that utilizes potential parameters of the human corneal epithelium model calculated based on Hansen solubility parameters

Eye irritation predictions are very important in the development of cosmetics and pharmaceuticals. For animal protection, alternative tests are being developed to replace the Draize test, which involves the use of rabbits to test eye irritation. The Vitrigel-eye irritancy test (Vitrigel-EIT), is one such alternative. As a preliminary study, we evaluated if Hansen solubility parameter (HSP) values can be used to predict Vitrigel-EIT results. An Hansen sphere was created based on the HSP values and Vitrigel-EIT results from 61 substances. Substances inside and outside of the sphere were designated as dangerous and safe substances, respectively. The safety of each test substance was predicted by comparing the center point (R_o) of the sphere with the relative energy difference, i.e., the ratio of each test substance (R_a). The accuracy, false negativity, and false positivity of the "irritant" and "nonirritant" designations, as determined by the Vitrigel-EIT results and Hansen sphere, were 91.8% (56/61), 2.3% (1/43), and 22.2% (4/18), respectively. These results indicated that HSP values can be used to predict Vitrigel-EIT results with high reproducibility, and thus are useful for evaluating the safety of substances.

Ocular surface and tear film changes in workers exposed to organic solvents used in the dry-cleaning industry

Workers in the dry-cleaning industry are exposed to organic solvents that may cause eye irritation and tear film changes. Objective To quantify changes in the ocular surface and tear film in dry cleaners exposed to organic solvents and associate these changes with ocular irritation as reported in a symptom questionnaire for dry eye diagnosis. Methods This was a case and control study in which the characteristics and eye-irritation symptoms were compared between two groups of 62 participants that were either exposed or not exposed to organic solvents. A general optometric examination and the following test were performed: lipid interferometry, Lissamine Green Stain, tear breakup time, Schirmer I, conjunctival impression cytology and the Donate dry eye symptoms questionnaire. Results Sixty-five percent of exposed workers obtained a higher score than 13 on the Donate dry eye symptoms questionnaire which indicated the presence of more irritation symptoms than those in the non- exposed group. A Chi-square analysis indicated the exposed group reported significantly higher incidences (P <0.005) for eye irritation symptoms of sandy sensation; tearing eyes sensation; foreign body sensation; tearing; dry eye; dryness; eyestrain and heavy eyelids. A Mann Whitney-U indicated greater severity only for symptoms relating to dry eye; sandy sensation; foreign body sensation, tearing; tearing eyes and dryness. There was a statistically significant difference (P <0.05) for Schirmer I; tear break up time; and the ocular surface assessed with Lissamine green staining and conjunctival impression cytology between groups. A reduction in the thickness of the lipid layer in the exposed group compared to the non-exposed group was observed. Surprisingly, clinical test outcomes were not significantly correlated with dry eye symptoms nor years of exposure. Conclusion Workers in the dry-cleaning industry exposed to organic solvents are associated with changes in ocular surface and tear film generating irritation symptoms commonly present in evaporative dry eye.

Assessment of the Eye Irritating Properties of Chemicals by Applying Alternatives to the Draize Rabbit Eye Test: The Use of QSARs and In Vitro Tests for the Classification of Eye Irritation

Huggins has reported on the current situation relating to the development of alternatives to the Draize eye irritation test with rabbits, and an ECVAM Working Group have reviewed the efforts needed in order to replace this animal test within the next 10 years by using the results of non-animal assessment methods. Our report reviews regulatory experience gained over the last 20 years with the EU chemicals notification procedure with respect to the assessment of eye lesions observed in Draize tests. The nature of eye lesions and their importance for classification and labelling of possible hazards to human eyes are evaluated and discussed, with a view to promoting the development of specific in vitro assays which are able to discriminate between eye damage, moderate eye irritation, and minor irritation effects which are completely reversible within a few days. Structural alerts for the prediction of eye irritation/corrosion hazards to be classified and labelled according to international classification criteria, are presented, which should be validated in accordance with internationally agreed (OECD) principles for (Q)SAR system validation. Physicochemical limit values for prediction of the absence of any eye irritation potential relevant for human health can make available a definition of the applicability domains of alternative methods developed for the replacement of the Draize eye irritation test.

Predicting Skin Permeability by Means of Computational Approaches: Reliability and Caveats in Pharmaceutical Studies

The skin is the main barrier between the internal body environment and the external one. The characteristics of this barrier and its properties are able to modify and affect drug delivery and chemical toxicity parameters. Therefore, it is not surprising that permeability of many different compounds has been measured through several in vitro and in vivo techniques. Moreover, many different in silico approaches have been used to identify the correlation between the structure of the permeants and their permeability, to reproduce the skin behavior, and to predict the ability of specific chemicals to permeate this barrier. A significant number of issues, like interlaboratory variability, experimental conditions, data set building rationales, and skin site of origin and hydration, still prevent us from obtaining a definitive predictive skin permeability model. This review wants to show the main advances and the principal approaches in computational methods used to predict this property, to enlighten the main issues that have arisen, and to address the challenges to develop in future research.

In silico prediction of serious eye irritation or corrosion potential of chemicals

Chemical fingerprints combined with machine learning methods were used to build binary classification models for predicting the potential EC/EI of compounds.

Estimation of the chemical-induced eye injury using a weight-of-evidence (WoE) battery of 21 artificial neural network (ANN) c-QSAR models (QSAR-21): part I: irritation potential

Evaluation of potential chemical-induced eye injury through irritation and corrosion is required to ensure occupational and consumer safety for industrial, household and cosmetic ingredient chemicals. The historical method for evaluating eye irritant and corrosion potential of chemicals is the rabbit Draize test. However, the Draize test is controversial and its use is diminishing - the EU 7th Amendment to the Cosmetic Directive (76/768/EEC) and recast Regulation now bans marketing of new cosmetics having animal testing of their ingredients and requires non-animal alternative tests for safety assessments. Thus, in silico and/or in vitro tests are advocated. QSAR models for eye irritation have been reported for several small (congeneric) data sets; however, large global models have not been described. This report describes FDA/CFSAN's development of 21 ANN c-QSAR models (QSAR-21) to predict eye irritation using the ADMET Predictor program and a diverse training data set of 2928 chemicals. The 21 models had external (20% test set) and internal validation and average training/verification/test set statistics were: 88/88/85(%) sensitivity and 82/82/82(%) specificity, respectively. The new method utilized multiple artificial neural network (ANN) molecular descriptor selection functionalities to maximize the applicability domain of the battery. The eye irritation models will be used to provide information to fill the critical data gaps for the safety assessment of cosmetic ingredient chemicals.

Predicting chemical ocular toxicity using a combinatorial QSAR approach

Regulatory agencies require testing of chemicals and products to protect workers and consumers from potential eye injury hazards. Animal screening, such as the rabbit Draize test, for potential environmental toxicants is time-consuming and costly. Therefore, virtual screening using computational models to tag potential ocular toxicants is attractive to toxicologists and policy makers. We have developed quantitative structure-activity relationship (QSAR) models for a set of small molecules with animal ocular toxicity data compiled by the National Toxicology Program Interagency Center for the Evaluation of Alternative Toxicological Methods. The data set was initially curated by removing duplicates, mixtures, and inorganics. The remaining 75 compounds were used to develop QSAR models. We applied both k nearest neighbor and random forest statistical approaches in combination with Dragon and Molecular Operating Environment descriptors. Developed models were validated on an external set of 34 compounds collected from additional sources. The external correct classification rates (CCR) of all individual models were between 72 and 87%. Furthermore, the consensus model, based on the prediction average of individual models, showed additional improvement (CCR = 0.93). The validated models could be used to screen external chemical libraries and prioritize chemicals for in vivo screening as potential ocular toxicants.

An integrated QSAR-PBPK modelling approach for predicting the inhalation toxicokinetics of mixtures of volatile organic chemicals in the rat

The objective of this study was to predict the inhalation toxicokinetics of chemicals in mixtures using an integrated QSAR-PBPK modelling approach. The approach involved: (1) the determination of partition coefficients as well as V(max) and K(m) based solely on chemical structure for 53 volatile organic compounds, according to the group contribution approach; and (2) using the QSAR-driven coefficients as input in interaction-based PBPK models in the rat to predict the pharmacokinetics of chemicals in mixtures of up to 10 components (benzene, toluene, m-xylene, o-xylene, p-xylene, ethylbenzene, dichloromethane, trichloroethylene, tetrachloroethylene, and styrene). QSAR-estimated values of V(max) varied compared with experimental results by a factor of three for 43 out of 53 studied volatile organic compounds (VOCs). K(m) values were within a factor of three compared with experimental values for 43 out of 53 VOCs. Cross-validation performed as a ratio of predicted residual sum of squares and sum of squares of the response value indicates a value of 0.108 for V(max) and 0.208 for K(m). The integration of QSARs for partition coefficients, V(max) and K(m), as well as setting the K(m) equal to K(i) (metabolic inhibition constant) within the mixture PBPK model allowed to generate simulations of the inhalation pharmacokinetics of benzene, toluene, m-xylene, o-xylene, p-xylene, ethylbenzene, dichloromethane, trichloroethylene, tetrachloroethylene and styrene in various mixtures. Overall, the present study indicates the potential usefulness of the QSAR-PBPK modelling approach to provide first-cut evaluations of the kinetics of chemicals in mixtures of increasing complexity, on the basis of chemical structure.

Semi-synthetic ecdysteroids as gene-switch actuators: synthesis, structure-activity relationships, and prospective ADME properties

The ligand-inducible, ecdysteroid receptor (EcR) gene-expression system can add critical control features to protein expression in cell and gene therapy. However, potent natural ecdysteroids possess absorption, distribution, metabolism and excretion (ADME) properties that have not been optimised for use as gene-switch actuators in vivo. Herein we report the first systematic synthetic exploration of ecdysteroids toward modulation of gene-switch potency. Twenty-three semi-synthetic O-alkyl ecdysteroids were assayed in both a natural insect system (Drosophila B(II) cells) and engineered gene-switch systems in mammalian cells using Drosophila melanogaster, Choristoneura fumiferana, and Aedes aegypti EcRs. Gene-switch potency is maintained, or even enhanced, for ecdysteroids methylated at the 22-position in favourable cases. Furthermore, trends toward lower solubility, higher permeability, and higher blood-brain barrier penetration are supported by predicted ADME properties, calculated using the membrane-interaction (MI)-QSAR methodology. The structure-activity relationship (SAR) of alkylated ecdysteroids indicates that 22-OH is an H-bond acceptor, 25-OH is most likely an H-bond donor, and 2-OH and 3-OH are donors and/or acceptors in network with each other, and with the EcR. The strategy of alkylation points the way to improved ecdysteroidal actuators for switch-activated gene therapy.

An Integrated Decision-tree Testing Strategy for Eye Irritation with Respect to the Requirements of the EU REACH Legislation

This paper presents some results of a joint research project, sponsored by Defra and conducted by FRAME and Liverpool John Moores University, on the status of alternatives to animal testing with regard to the European Union REACH (Registration, Evaluation and Authorisation of Chemicals) system for the safety testing and risk assessment of chemicals. The project covered all the main toxicity endpoints associated with REACH. This paper focuses on the use of alternative (non-animal) methods (both in vitro and in silico) for eye irritation testing. The manuscript reviews numerous in vitro tests and their possible collation into test batteries, in silico models and a refined in vivo method (the low volume eye test), before combining the use of all these methods into an integrated testing strategy. The aim of this strategy is a reduction in the number of animal tests which would need to be performed in the process of fulfilling the REACH system criteria; this would also lead to a lowering of the number of animals required in compliance with the REACH system requirements.

An integrated decision-tree testing strategy for eye irritation with respect to the requirements of the EU REACH legislation

This paper presents some results of a joint research project, sponsored by Defra and conducted by FRAME and Liverpool John Moores University, on the status of alternatives to animal testing with regard to the European Union REACH (Registration, Evaluation and Authorisation of Chemicals) system for the safety testing and risk assessment of chemicals. The project covered all the main toxicity endpoints associated with REACH. This paper focuses on the use of alternative (non-animal) methods (both in vitro and in silico) for eye irritation testing. The manuscript reviews numerous in vitro tests and their possible collation into test batteries, in silico models and a refined in vivo method (the low volume eye test), before combining the use of all these methods into an integrated testing strategy. The aim of this strategy is a reduction in the number animal tests which would need to be performed in the process of fulfilling the REACH system criteria; this would also lead to a lowering of the number of animals required in compliance with the REACH system requirements.

Combined 4D‐Fingerprint and Clustering Based Membrane‐Interaction QSAR Analyses for Constructing Consensus Caco‐2 Cell Permeation Virtual Screens

A set of 30 structurally diverse molecules, for which Caco-2 cell permeation coefficients were determined, formed the training set for construction of Caco-2 cell permeation models based upon membrane-interaction (MI) QSAR analysis and a new QSAR method called 4D-fingerprint QSAR analysis. The descriptor terms of the 4D-fingerprints equation are molecular similarity eigenvalues, and this set of descriptors is being evaluated as a potential "universal" QSAR descriptor set. The 4D-fingerprint model suggests that Caco-2 cell permeation is governed by the spatial distribution of hydrogen bonding and nonpolar groups over the molecular shape of a molecule. Moreover, a complementary resampling of the original Caco-2 cell permeation training set, followed by the construction of several "clustered" MI-QSAR models, led to a consensus model consistent in interpretation with the 4D-fingerprint model.

Categorical QSAR Models for Skin Sensitization based upon Local Lymph Node Assay Classification Measures Part 2: 4D-Fingerprint Three-State and Two-2-State Logistic Regression Models

Three and four state categorical quantitative structure-activity relationship (QSAR) models for skin sensitization have been constructed using data from the murine Local Lymph Node Assay studies. These are the same data we previously used to build two-state (sensitizer, nonsensitizer) QSAR models (Li et al., 2007, Chem. Res. Toxicol. 20, 114-128). 4D-fingerprint descriptors derived from the 4D-molecular similarity paradigm are used to generate these models. A training set of 196 and a test set of 22 structurally diverse compounds were used in this study. Logistic regression, and partial least square coupled logistic regression were used to build the models. The three-state QSAR model gives a classification accuracy of 73.4% for the training set and 63.6% for the test set, while the random average value of classification accuracy for any three-state data set is 33.3%. The two-2-state [four categories in total] QSAR model gives a classification accuracy of 83.2% for the training set and 54.6% for the test set, while the random average value of classification accuracy for any two-2-state data set is 25%. An analysis of the skin-sensitization models developed in this study, as well as the two-state QSAR models developed in our previous analysis, suggests that the "moderate" sensitizers may be the main source of limited model accuracy.

Prediction and Mechanistic Interpretation of Human Oral Drug Absorption Using MI-QSAR Analysis

Membrane-interaction [MI]-QSAR analysis, which includes descriptors explicitly derived from simulations of solutes [drugs] interacting with phospholipid membrane models, was used to construct QSAR models for human oral intestinal drug absorption. A data set of 188 compounds, which are mainly drugs, was divided into a parent training set of 164 compounds and a test set of 24 compounds. Stable, but not highly fit [R2 = 0.68] MI-QSAR models could be built for all 188 compounds. However, the relatively large number [47] of drugs having 100% absorption, as well as all zwitterionic compounds [11], had to be eliminated from the training set in order to construct a linear five-term oral absorption diffusion model for 106 compounds which was both stable [R2 = 0.82, Q2 = 0.79] and predictive given the test set compounds were predicted with nearly the same average accuracy as the compounds of the training set. Intermolecular membrane-solute descriptors are essential to building good oral absorption models, and these intermolecular descriptors are displaced in model optimizations and intramolecular solute descriptors found in published oral absorption QSAR models. A general form for all of the oral intestinal absorption MI-QSAR models has three classes of descriptors indicative of three thermodynamic processes: (1) solubility and partitioning, (2) membrane-solute interactions, and (3) flexibility of the solute and/or membrane. The intestinal oral absorption MI-QSAR models were compared to MI-QSAR models previously developed for Caco-2 cell permeation and for blood-brain barrier penetration. The MI-QSAR models for all three of these ADME endpoints share several common descriptors, and suggest a common mechanism of transport across all three barriers. A further analysis of these three types of MI-QSAR models has been done to identify descriptor-term differences across these three models, and the corresponding differences in thermodynamic transport behavior of the three barriers.

Effect of cholesterol on DMPC phospholipid membranes and QSAR model construction in membrane-interaction QSAR study through molecular dynamics simulation

In this study, both pure DMPC and DMPC/cholesterol mixed membrane monolayer were built to compare the physical-chemical properties and dynamics properties through molecular dynamics simulation and normal-mode analysis. The results show that the addition of cholesterol decreases the area of per molecule of membrane, increases the lipid amplitude motion, and changes the solute diffusion coefficient. It is also found that the addition of cholesterol greatly changes the solute-membrane 1,4-nonbonded interaction energy (deltaE14). MI-QSAR models were constructed based on solute-membrane interaction energy descriptors and other intramolecular descriptors. The results show that deltaE14 substitutes deltaE(HB) as the second important descriptor compared with the previous study. Final results suggest that short range solute-membrane interaction energy changes due to the uptake of the solute may play an important decision on permeability in DMPC/cholesterol membrane. A test set was applied to evaluate the predictivity of MI-QSAR models. The result suggests that the combination of F(H2O) and deltaE14 not only improves r2 and q2, but also greatly improves the model predictivity. Based on the combination of q2 and r(pre)2 values, a two-term model is better used to predict the solute permeability in this study (r2 = 0.859, q2 = 0.803, and r(pre)2 = 0.540). Due to the small sample both in training set and test set, more datasets are necessary to make a final decision about the model construction and prediction.

Predicting MDCK cell permeation coefficients of organic molecules using membrane-interaction QSAR analysis

AIM

To use membrane-interaction quantitative structure-activity relationship analysis (MI-QSAR) to develop predictive models of partitioning of organic compounds in gastrointestinal cells.

METHODS

A training set of 22 structurally diverse compounds, whose apparent permeability across cellular membranes of Madin-Darby canine kidney (MDCK) cells were measured, were used to construct MI-QSAR models. Molecular dynamic simulations were used to determine the explicit interaction of each test compound (solute) with a dimyristoyl-phosphatidyl-choline monolayer membrane model. An additional set of intramolecular solute descriptors were computed and considered in the trial pool of descriptors for building MI-QSAR models. The QSAR models were optimized using multidimensional linear regression fitting and the stepwise method. A test set of 8 compounds were evaluated using the MI-QSAR models as part of a validation process.

RESULTS

MI-QSAR models of the gastrointestinal absorption process were constructed. The descriptors found in the best MI-QSAR models are as follows: 1) ClogP (the logarithm of the 1-octanol/water partition coefficient); 2) E(HOMO) (the highest occupied molecular orbital energy); 3) E(s) (stretch energy); 4) PM(Y) (the principal moment of inertia Y, the inertia along the y axis in the rectangular coordinates; 5) C(t) (total connectivity); and 6) E(nb) (the energy of interactions between all of the non-bonded atoms). The most important descriptor in the models is ClogP.

CONCLUSION

Permeability is not only determined by the properties of drug molecules, but is also very much influenced by the molecule-membrane interaction process.

Predicting permeability coefficient in ADMET evaluation by using different membranes-interaction QSAR

Membrane-interaction quantitative structure activity relationship (MI-QSAR) analysis was applied to a data set with 18 compounds in 18 different membranes. MI-QSAR was used to estimate the ADMET properties including the transport of organic solutes through biological membranes. The most important descriptors are the aqueous solvation free energy, FH2O, and diffusion coefficient for all membranes. The correlation coefficient, r2, and cross-validation correlation coefficient, q2, for DMPG membrane is 0.850 and 0.770, respectively. The relationship between FH2O and permeability is nonlinear. But the detail effect of aqueous solvation free energy and diffusion coefficient to the permeability depends on the type of membrane. The final models also support the solution-diffusion mechanism of transport is important in membrane.

A study of the relationship between cornea permeability and eye irritation using membrane-interaction QSAR analysis

A methodology termed membrane-interaction QSAR (MI-QSAR) analysis has been used to develop QSAR models to predict drug permeability coefficients across cornea and its component layers (epithelium, stroma, and endothelium). From a training set of 25 structurally diverse drugs, significant QSAR models are constructed and compared for the permeability of the cornea, epithelium, and stroma plus endothelium. Cornea permeability is found to depend on the measured distribution coefficient of the drug, the cohesive energy of the drug, the total potential energy of the drug-membrane "complex," and three other energy refinement descriptor terms. The endothelium may be a more important barrier in cornea permeation than the stroma. Moreover, an investigation of the correlation between cornea permeation and eye irritation is presented as an example of a cross study on different ADMET properties using MI-QSAR analysis. Thirteen structurally diverse drugs, whose molar-adjusted eye irritation scores (MES) have been measured using the Draize rabbit-eye test, were chosen as an eye irritation comparison set. A poor correlation (R(2) = 0.0232) between the MES measures and the predicted cornea permeability coefficients for the drugs in the eye irritation set suggests there is no significant relationship between eye irritation potency and the cornea permeability.

Predictive model of blood-brain barrier penetration of organic compounds

AIM

To build up a theoretical model of organic compounds for the prediction of the activity of small molecules through the blood-brain barrier (BBB) in drug design.

METHODS

A training set of 37 structurally diverse compounds was used to construct quantitative structure-activity relationship (QSAR) models. Intermolecular and intramolecular solute descriptors were calculated using molecular mechanics, molecular dynamics simulations, quantum chemistry and so on. The QSAR models were optimized using multidimensional linear regression fitting and stepwise method. A test set of 8 compounds was evaluated using the models as part of a validation process.

RESULTS

Significant QSAR models (R=0.955, s=0.232) of the BBB penetration of organic compounds were constructed. BBB penetration was found to depend upon the polar surface area, the octanol/water partition coefficient, Balaban Index, the strength of a small molecule to combine with the membrane-water complex, and the changeability of the structure of a solute-membrane-water complex.

CONCLUSION

The QSAR models indicate that the distribution of organic molecules through BBB is not only influenced by organic solutes themselves, but also relates to the properties of the solute-membrane-water complex, that is, interactions of the molecule with the phospholipid-rich regions of cellular membranes.

Characterization of Skin Penetration Processes of Organic Molecules Using Molecular Similarity and QSAR Analysis

Molecular similarity and QSAR analyses have been used to develop compact, robust, and definitive models for skin penetration by organic compounds. The QSAR models have been sought to provide an interpretation and characterization of plausible molecular mechanisms of skin penetration. A training set of 40 structurally diverse compounds were selected to be representative of a parent set of 152 compounds in terms of both structural diversity and range in measured skin penetration. The subset of 40 compounds was used in a series of QSAR analyses in the search for the most significant, compact, and straightforward skin penetration QSAR models. Molecular dynamics simulations were employed to determine a set of MI (membrane-interaction) descriptors for each test compound (solute) interacting with a model DMPC monolayer membrane model. The MI-QSAR models may capture features of cellular membrane lateral transverse transport involved in the overall skin penetration process by organic compounds. An additional set of intramolecular solute descriptors, the non-MI-QSAR descriptors, were computed and added to the trial pool of descriptors for building QSAR models. All QSAR models were constructed using multidimensional linear regression fitting and a genetic algorithm optimization function. QSAR models were constructed using only non-MI-QSAR descriptors and using a combination of both these descriptor sets. It was found that a combination of non-MI-QSAR and MI-QSAR descriptors yielded the optimum models, not only with respect to the statistical measures of fit but also regarding model predictivity.

Validation tools for variable subset regression

Variable selection is applied frequently in QSAR research. Since the selection process influences the characteristics of the finally chosen model, thorough validation of the selection technique is very important. Here, a validation protocol is presented briefly and two of the tools which are part of this protocol are introduced in more detail. The first tool, which is based on permutation testing, allows to assess the inflation of internal figures of merit (such as the cross-validated prediction error). The other tool, based on noise addition, can be used to determine the complexity and with it the stability of models generated by variable selection. The obtained statistical information is important in deciding whether or not to trust the predictive abilities of a specific model. The graphical output of the validation tools is easily accessible and provides a reliable impression of model performance. Among others, the tools were employed to study the influence of leave-one-out and leave-multiple-out cross-validation on model characteristics. Here, it was confirmed that leave-multiple-out cross-validation yields more stable models. To study the performance of the entire validation protocol, it was applied to eight different QSAR data sets with default settings. In all cases internal and external model performance was good, indicating that the protocol serves its purpose quite well.

Tiered testing strategies--acute local toxicity

More work has been done to develop alternatives to animal use in the areas of eye and skin irritation than in any area other than carcinogenicity. There has long been a belief both in the scientific community and among the public that the development of nonanimal tests in these areas should be simple and straightforward. After more than 20 yr of research, we can identify materials corrosive to the skin without using animals, but the assessment of irritation using in vitro methods alone is still an illusive goal. This review of current recommendations and industry practices that reduce the number of animals needed for these two tests concludes that animal use for skin irritation testing is not necessary today, with currently available and accepted methodology, except for regulatory reasons. Scientifically sound improvements in current eye irritation methods are also available. Advances in the understanding of the mechanisms of eye irritation that have been made in the last 5 yr should lead to improved in vitro methods for this endpoint. In the meantime, changes should be made in the current animal protocol to reduce pain and distress. This paper provides an overview of the progress that has been made toward discontinuing the use of animals in tests to determine the potential of materials to cause skin or eye damage after a single acute exposure. It also discusses some additional changes that could be made now to reduce animal use further or to reduce pain and distress in the testing that must still be done until such time as we can meet the ultimate goal--validated and accepted nonanimal methods for these endpoints.

Methods for reliability and uncertainty assessment and for applicability evaluations of classification- and regression-based QSARs

This article provides an overview of methods for reliability assessment of quantitative structure-activity relationship (QSAR) models in the context of regulatory acceptance of human health and environmental QSARs. Useful diagnostic tools and data analytical approaches are highlighted and exemplified. Particular emphasis is given to the question of how to define the applicability borders of a QSAR and how to estimate parameter and prediction uncertainty. The article ends with a discussion regarding QSAR acceptability criteria. This discussion contains a list of recommended acceptability criteria, and we give reference values for important QSAR performance statistics. Finally, we emphasize that rigorous and independent validation of QSARs is an essential step toward their regulatory acceptance and implementation.

Mapping property distributions of molecular surfaces: algorithm and evaluation of a novel 3D quantitative structure-activity relationship technique

A novel molecular descriptor called MaP (mapping property distributions of molecular surfaces) is presented. It combines facile computation, translational and rotational invariance, and straightforward interpretability of the computed models. A three-step procedure is used to compute the MaP descriptor. First, an approximation to the molecular surface with equally distributed surface points is computed. Next, molecular properties are projected onto this surface. Finally, the distribution of surface properties is encoded into a translationally and rotationally invariant molecular descriptor that is based on radial distribution functions (distance-dependent count statistics). The calculated descriptor is correlated with biological data through chemometric regression techniques in combination with a variable selection. The latter is used to identify variables that are highly relevant for the model and hence for its interpretation. Three applications of the new descriptor are presented, each representing a different area of 3D-QSAR. For reasons of comparability, the new descriptor was tested on the steroid "benchmark" data set. Furthermore, a highly diverse data set with potentially eye-irritating compounds was studied, and third, a set of flexible structures with a modulating effect on the muscarinic M(2) receptor were studied. Not only were all models highly predictive but interpretation of the back-projected variables into the original molecular space led to biologically and chemically relevant conclusions.

Looking for an unambiguous geometrical definition of organic series from 3-d molecular similarity indices

A mathematically consistent definition of series was derived by the application of Principal Component Analysis to 3-D similarity indices (ASP Software). For two data sets of aromatic molecules, a mono-dimensional numerical ordering was derived, corresponding to the consensus distance from the series lead. The series is unambiguously defined in terms of location along the mono-dimensional axis, along which shape and electrostatic features of the molecules vary in a coordinated way. Molecules along the axis are characterized by the same pattern of electrostatic potential and shape shared by the lead compound, of which they represent linearly "shifted" replicas. This approach can contribute to the exploration of the chemical spaces from a local "fine grain" perspective, thus complementing the "coarse grain" descriptions normally used in the analysis of large data sets.

Predicting blood-brain barrier partitioning of organic molecules using membrane-interaction QSAR analysis

PURPOSE

Membrane-interaction quantitative structure-activity relationship (OSAR) analysis (MI-QSAR) has been used to develop predictive models of blood-brain barrier partitioning of organic compounds by, in part, simulating the interaction of an organic compound with the phospholipid-rich regions of cellular membranes.

METHOD

A training set of 56 structurally diverse compounds whose blood-brain barrier partition coefficients were measured was used to construct MI-QSAR models. Molecular dynamics simulations were used to determine the explicit interaction of each test compound (solute) with a model DMPC monolayer membrane model. An additional set of intramolecular solute descriptors were computed and considered in the trial pool of descriptors for building MI-QSAR models. The QSAR models were optimized using multidimensional linear regression fitting and a genetic algorithm. A test set of seven compounds was evaluated using the MI-QSAR models as part of a validation process.

RESULTS

Significant MI-QSAR models (R2 = 0.845, Q2 = 0.795) of the blood-brain partitioning process were constructed. Blood-brain barrier partitioning is found to depend upon the polar surface area. the octanol/water partition coefficient, and the conformational flexibility of the compounds as well as the strength of their binding" to the model biologic membrane. The blood-brain barrier partitioning measures of the test set compounds were predicted with the same accuracy as the compounds of the training set.

CONCLUSION

The MI-QSAR models indicate that the blood-brain barrier partitioning process can be reliably described for structurally diverse molecules provided interactions of the molecule with the phospholipids-rich regions of cellular membranes are explicitly considered.

Predicting Caco-2 cell permeation coefficients of organic molecules using membrane-interaction QSAR analysis

A methodology termed membrane-interaction QSAR (MI-QSAR) analysis has been developed in order to predict the behavior of organic compounds interacting with the phospholipid-rich regions of biological membranes. One important application of MI-QSAR analysis is to estimate ADME properties including the transport of organic solutes through biological membranes as a computational approach to forecasting drug intestinal absorption. A training set of 30 structurally diverse drugs, whose permeability coefficients across the cellular membranes of Caco-2 cells were measured, was used to construct significant MI-QSAR models of Caco-2 cell permeation. Cellular permeation is found to depend primarily upon aqueous solvation free energy (solubility) of the drug, the extent of drug interaction with a model phospholipid (DMPC) monolayer, and the conformational flexibility of the solute within the model membrane. A test set of eight drugs was used to evaluate the predictivity of the MI-QSAR models. The permeation coefficients of the test set compounds were predicted with the same accuracy as the compounds of the training set.

The computational prediction of toxicity

Recent developments in the prediction of toxicity from chemical structure have been reviewed. Attention has been drawn to some of the problems that can be encountered in the area of predictive toxicology, including the need for a multi-disciplinary approach and the need to address mechanisms of action. Progress has been hampered by the sparseness of good quality toxicological data. Perhaps too much effort has been devoted to exploring new statistical methods rather than to the creation of data sets for hitherto uninvestigated toxicological endpoints and/or classes of chemicals.