A novel approach to predicting P450 mediated drug metabolism. CYP2D6 catalyzed N-dealkylation reactions and qualitative metabolite predictions using a combined protein and pharmacophore model for CYP2D6

COMPARISON OF METHODS FOR THE PREDICTION OF THE METABOLIC SITES FOR CYP3A4-MEDIATED METABOLIC REACTIONS

Predictions of the metabolic sites for new chemical entities, synthesized or only virtual, are important in the early phase of drug discovery to guide chemistry efforts in the synthesis of new compounds with reduced metabolic liability. This information can now be obtained from in silico predictions, and therefore, a thorough and unbiased evaluation of the computational techniques available is needed. Several computational methods to predict the metabolic hot spots are emerging. In this study, metabolite identification using MetaSite and a docking methodology, GLUE, were compared. Moreover, the published CYP3A4 crystal structure and computed CYP3A4 homology models were compared for their usefulness in predicting metabolic sites. A total of 227 known CYP3A4 substrates reported to have one or more metabolites adding up to 325 metabolic pathways were analyzed. Distance-based fingerprints and four-point pharmacophore derived from GRID molecular interaction fields were used to characterize the substrate and protein in MetaSite and the docking methodology, respectively. The CYP3A4 crystal structure and homology model with the reactivity factor enabled achieved a similar prediction success (78%) using the MetaSite method. The docking method had a relatively lower prediction success (approximately 57% for the homology model), although it still may provide useful insights for interactions between ligand and protein, especially for uncommon reactions. The MetaSite methodology is automated, rapid, and has relatively accurate predictions compared with the docking methodology used in this study.

Crystal Structure of Human Cytochrome P450 2D6

Cytochrome P450 2D6 is a heme-containing enzyme that is responsible for the metabolism of at least 20% of known drugs. Substrates of 2D6 typically contain a basic nitrogen and a planar aromatic ring. The crystal structure of human 2D6 has been solved and refined to 3.0A resolution. The structure shows the characteristic P450 fold as seen in other members of the family, with the lengths and orientations of the individual secondary structural elements being very similar to those seen in 2C9. There are, however, several important differences, the most notable involving the F helix, the F-G loop, the B'helix, beta sheet 4, and part of beta sheet 1, all of which are situated on the distal face of the protein. The 2D6 structure has a well defined active site cavity above the heme group, containing many important residues that have been implicated in substrate recognition and binding, including Asp-301, Glu-216, Phe-483, and Phe-120. The crystal structure helps to explain how Asp-301, Glu-216, and Phe-483 can act as substrate binding residues and suggests that the role of Phe-120 is to control the orientation of the aromatic ring found in most substrates with respect to the heme. The structure has been compared with published homology models and has been used to explain much of the reported site-directed mutagenesis data and help understand the metabolism of several compounds.

Utility of human/human-derived reagents in drug discovery and development: An industrial perspective

The shift to combinatorial chemistry and parallel synthesis in drug discovery has resulted in large numbers of compounds entering the lead seeking and lead development phases of the process. To support this, higher throughput computational (in silico) and in vitro approaches have become the forefront of the drug metabolism and pharmacokinetic (DMPK) input into drug discovery. This has been accompanied by a shift in focus from animal-derived data to human based studies, reflecting the realisation that extrapolation from animals to human has its limitations. In silico approaches may be regarded as human derived tools for DMPK, since models (template/pharmacophore and protein homology modelling), for example, for the human CYP enzymes, are widely used for identifying qualitatively enzyme/substrate interactions. Quantitative assessment of drug metabolism using human hepatocytes or sub-cellular fractions provide a valuable tool both for the screening out of high metabolic lability and in estimations of human intrinsic clearance. In terms of drug absorption, the human colon adenocarcinoma cell line, Caco-2, offers a versatile human derived system for measuring drug permeability, despite over expression of the efflux transporter P-glycoprotein (P-gp). The importance of P-gp can then be further assessed in recombinant systems expressing the human P-gp, where substrate affinity and inhibition potency can be measured, important factors when considering transporter mediated drug-drug interactions. The primary cause of pharmacokinetic-based drug-drug interactions (DDIs) is through enzyme inhibition or induction, with the CYP enzymes being of major importance. Human liver microsomes and hepatocytes are invaluable tools in assessment of DDI vulnerability of new chemical entities, having the capacity to identify enzymes responsible for specific routes of metabolism, and hence areas of vulnerability for a DDI. In addition, human-based screening tools can be used to identify the perpetrator of a DDI through enzyme inhibition/induction. Large differences in the nature of enzymes induced and the extent of induction when comparing animals to man are known. Thus, in vitro models allowing assessment of induction potential in human tissue, establishes some relevance to the clinical situation.

The role of phenylalanine 483 in cytochrome P450 2D6 is strongly substrate dependent

The polymorphic cytochrome P450 2D6 (CYP2D6) is involved in the metabolism of 30% of the drugs currently prescribed, and is thus clinically relevant. Typical CYP2D6 substrates generally contain a basic nitrogen atom and an aromatic moiety adjacent to the site of metabolism. Recently, we demonstrated the importance of active site residue F120 in substrate binding and catalysis in CYP2D6. On the basis of protein homology models, it is claimed that another active site phenylalanine, F483, may also play an important role in the interaction with the aromatic moiety of CYP2D6 substrates. Experimental data to support this hypothesis, however, is not yet available. In fact, in the only study performed, mutation of F483 to isoleucine or tryptophan did not affect the 1'-hydroxylation of bufuralol at all [Smith G, Modi S, Pillai I, Lian LY, Sutcliffe MJ, Pritchard MP, et al., Determinants of the substrate specificity of human cytochrome P-450 CYP2D6: design and construction of a mutant with testosterone hydroxylase activity. Biochem J 1998;331:783-92]. In the present study, the role of F483 in ligand binding and metabolism by CYP2D6 was examined experimentally using site-directed mutagenesis. Replacement of F483 by alanine resulted in a 30-fold lower V(max) for bufuralol 1'-hydroxylation, while the K(m) was hardly affected. The V(max) for 3,4-methylenedioxy-methylamphetamine O-demethylenation on the other hand decreased only two-fold, whereas the effect on the K(m) was much larger. For dextromethorphan, in addition to dextrorphan (O-demethylation) and 3-methoxymorphinan (N-demethylation), two other metabolites were formed that could not be detected for the wild-type. The substrate 7-methoxy-4-(aminomethyl)-coumarin was not metabolised at all by CYP2D6[F483A], a phenomenon that was reported also for CYP2D6[F120A]. The presented data show that next to F120, residue F483 plays a very important role in the metabolism of typical CYP2D6 substrates. The influence of F483 on metabolism was found to be strongly substrate-dependent.

A 3D-QSAR model for CYP2D6 inhibition in the aryloxypropanolamine series

A comparative molecular similarity index analysis (CoMSiA) has been performed for cytochrome P450 2D6 inhibition on a series of aryloxypropanolamines to determine the factors contributing to this activity. The model is in agreement with a CYP2D6 homology model constructed on the basis of the mammalian CYP2C5 crystal structure. The energy minimized conformations were generated using the systematic search methodology in Sybyl 6.7. The model not only elucidated the relationship between structure and biological activity but, more importantly, provided useful strategies to modulate CYP2D6 affinity in the aryloxypropanolamine series.

Computational prediction of human drug metabolism

There is an urgent requirement within the pharmaceutical and biotechnology industries, regulatory authorities and academia to improve the success of molecules that are selected for clinical trials. Although absorption, distribution, metabolism, excretion and toxicity (ADME/Tox) properties are some of the many components that contribute to successful drug discovery and development, they represent factors for which we currently have in vitro and in vivo data that can be modelled computationally. Understanding the possible toxicity and the metabolic fate of xenobiotics in the human body is particularly important in early drug discovery. There is, therefore, a need for computational methodologies for uncovering the relationships between the structure and the biological activity of novel molecules. The convergence of numerous technologies, including high-throughput techniques, databases, ADME/Tox modelling and systems biology modelling, is leading to the foundation of systems-ADME/Tox. Results from experiments can be integrated with predictions to globally simulate and understand the likely complete effects of a molecule in humans. The development and early application of major components of MetaDrug (GeneGo, Inc.) software will be described, which includes rule-based metabolite prediction, quantitative structure-activity relationship models for major drug metabolising enzymes, and an extensive database of human protein-xenobiotic interactions. This represents a combined approach to predicting drug metabolism. MetaDrug can be readily used for visualising Phase I and II metabolic pathways, as well as interpreting high-throughput data derived from microarrays as networks of interacting objects. This will ultimately aid in hypothesis generation and the early triaging of molecules likely to have undesirable predicted properties or measured effects on key proteins and cellular functions.

Utility of homology models in the drug discovery process

Advances in bioinformatics and protein modeling algorithms, in addition to the enormous increase in experimental protein structure information, have aided in the generation of databases that comprise homology models of a significant portion of known genomic protein sequences. Currently, 3D structure information can be generated for up to 56% of all known proteins. However, there is considerable controversy concerning the real value of homology models for drug design. This review provides an overview of the latest developments in this area and includes selected examples of successful applications of the homology modeling technique to pharmaceutically relevant questions. In addition, the strengths and limitations of the application of homology models during all phases of the drug discovery process are discussed.

Progress in cytochrome P450 active site modeling

Models capable of predicting the possible involvement of cytochromes P450 in the metabolism of drugs or drug candidates are important tools in drug discovery and development. Ideally, functional information would be obtained from crystal structures of all the cytochromes P450 of interest. Initially, only crystal structures of distantly related bacterial cytochromes P450 were available-comparative modeling techniques were used to bridge the gap and produce structural models of human cytochromes P450, and thereby obtain some useful functional information. A significant step forward in the reliability of these models came four years ago with the first crystal structure of a mammalian cytochrome P450, rabbit CYP2C5, followed by the structures of two human enzymes, CYP2C8 and CYP2C9, and a second rabbit enzyme, CYP2B4. The evolution of a CYP2D6 model, leading to the validation of the model as an in silico tool for predicting binding and metabolism, is presented as a case study.

Modelling in vitro hepatotoxicity using molecular interaction fields and SIMCA

There is currently a great deal of interest in creating computational tools for predicting the pharmacological properties of drug development candidates, ranging from physicochemical properties such as pK(a) and solubility to more complex biological properties such as oral bioavailability and toxicity. The limiting factor in many cases is a shortage of good data from which to construct training sets. In other cases, large amounts of data are available, but they use surrogate end-points or are comprised of compounds very different from those usually encountered in drug discovery and development. In such cases large training sets and global models are not necessarily better than local models based on smaller data sets. Such considerations make it as important to examine the available data carefully so as to avoid over-interpretation of the models obtained as it is to minimise errors in prediction per se. The kinds of complications likely to be encountered for in vitro hepatotoxicity modelling are discussed in general terms and illustrated in particular by SIMCA analysis of data obtained from assays of cultured hepatocytes for a large, structurally diverse data set and a smaller, much more focussed one.

Polymorphic cytochrome P450 2D6: humanized mouse model and endogenous substrates

Cytochrome P450 2D6 (CYP2D6) is the first well-characterized polymorphic phase I drug-metabolizing enzyme, and more than 80 allelic variants have been identified for the CYP2D6 gene, located on human chromosome 22q13.1. Human debrisoquine and sparteine metabolism is subdivided into two principal phenotypes--extensive metabolizer and poor metabolizer--that arise from variant CYP2D6 genotypes. It has been estimated that CYP2D6 is involved in the metabolism and disposition of more than 20% of prescribed drugs, and most of them act in the central nervous system or on the heart. These drug substrates are characterized as organic bases containing one nitrogen atom with a distance about 5, 7, or 10 A from the oxidation site. Aspartic acid 301 and glutamic acid 216 were determined as the key acidic residues for substrate-enzyme binding through electrostatic interactions. CYP2D6 transgenic mice, generated using a lambda phage clone containing the complete wild-type CYP2D6 gene, exhibits enhanced metabolism and disposition of debrisoquine. This transgenic mouse line and its wild-type control are models for human extensive metabolizers and poor metabolizers, respectively, and would have broad application in the study of CYP2D6 polymorphism in drug discovery and development, and in clinical practice toward individualized drug therapy. Endogenous 5-methoxyindole- thylamines derived from 5-hydroxytryptamine were identified as high-affinity substrates of CYP2D6 that catalyzes their O-demethylations with high enzymatic capacity and specificity. Thus, polymorphic CYP2D6 may play an important role in the interconversions of these psychoactive tryptamines, including a crucial step in a serotonin-melatonin cycle.

Docking and scoring in virtual screening for drug discovery: methods and applications

Computational approaches that 'dock' small molecules into the structures of macromolecular targets and 'score' their potential complementarity to binding sites are widely used in hit identification and lead optimization. Indeed, there are now a number of drugs whose development was heavily influenced by or based on structure-based design and screening strategies, such as HIV protease inhibitors. Nevertheless, there remain significant challenges in the application of these approaches, in particular in relation to current scoring schemes. Here, we review key concepts and specific features of small-molecule-protein docking methods, highlight selected applications and discuss recent advances that aim to address the acknowledged limitations of established approaches.

Influence of phenylalanine 120 on cytochrome P450 2D6 catalytic selectivity and regiospecificity: crucial role in 7-methoxy-4-(aminomethyl)-coumarin metabolism

The polymorphic human debrisoquine hydroxylase, cytochrome P450 2D6 (CYP2D6), is one of the most important phase I drug metabolising enzymes. It is responsible for metabolising a large number of compounds that mostly share similarity in having a basic N-atom and an aromatic moiety. In homology modelling studies, it has been suggested that in fixation of this aromatic moiety, there may be an important role for phenylalanine 120 (Phe(120)). In this study, the role of Phe(120) in ligand binding and catalysis was experimentally examined by mutating it into an alanine. Strikingly, this substitution led to a completely abolished 7-methoxy-4-(aminomethyl)-coumarin (MAMC) O-demethylating activity of CYP2D6. On the other hand, bufuralol metabolism was hardly affected (K(m) of 1-hydroxylation mutant: 1.2 microM, wild-type: 2.9 microM, 4-hydroxylation mutant: 1.5 microM, and wild-type: 3.2 microM) and neither was affected dextromethorphan O-demethylation (K(m) mutant: 1.2 microM, wild-type: 2 microM, k(cat) mutant: 4.5 min(-1), and wild-type: 3.3 min(-1)). However, the Phe(120)Ala mutant also formed 3-hydroxymorphinan, the double demethylated form of dextromethorphan, which was not detected using wild-type CYP2D6. 3,4-Methylenedioxymethamphetamine (MDMA) was demethylenated by both mutant and wild-type CYP2D6 to 3,4-dihydroxymethamphetamine (3,4-OH-MA K(m) of mutant: 55 microM and wild-type: 2 microM). In addition, the mutant formed two additional metabolites; 3,4-methylenedioxyamphetamine (MDA) and N-hydroxy-3,4-methylenedioxymethamphetamine (N-OH-MDMA). Inhibition experiments of dextromethorphan O-demethylation showed a decreased affinity of the Phe(120)Ala mutant for quinidine (IC(50) mutant: 240 nM and wild-type, 40 nM), while IC(50)s for quinine were equal (1 microM). These data indicate the importance of Phe(120) in the selectivity and regiospecificity in substrate binding and catalysis by CYP2D6.

A New Statistical Approach to Predicting Aromatic Hydroxylation Sites. Comparison with Model-Based Approaches

A new approach is described that is able to predict the most probable metabolic sites on the basis of a statistical analysis of various metabolic transformations reported in the literature. The approach is applied to the prediction of aromatic hydroxylation sites for diverse sets of substrates. Training is performed using the aromatic hydroxylation reactions from the Metabolism database (Accelrys). Validation is carried out on heterogeneous sets of aromatic compounds reported in the Metabolite database (MDL). The average accuracy of prediction of experimentally observed hydroxylation sites estimated for 1552 substrates from Metabolite is 84.5%. The proposed approach is compared with two electronic models for P450 mediated aromatic hydroxylation: the oxenoid model using the atomic oxygen and the model using the methoxy radical as a model for the heme active oxygen species. For benzene derivatives, the proposed method is inferior to the oxenoid model and as accurate as the methoxy-radical model. For hetero- and polycyclic compounds, the oxenoid model is not applicable, and the statistical method is the most accurate. Broad applicability and high speed of calculations provide the basis for using the proposed statistical approach for high-throughput metabolism prediction in the early stages of drug discovery.

QUANTITATIVE STRUCTURE-METABOLISM RELATIONSHIP MODELING OF METABOLICN-DEALKYLATION REACTION RATES

It is widely recognized that preclinical drug discovery can be improved via the parallel assessment of bioactivity, absorption, distribution, metabolism, excretion, and toxicity properties of molecules. High-throughput computational methods may enable such assessment at the earliest, least expensive discovery stages, such as during screening compound libraries and the hit-to-lead process. As an attempt to predict drug metabolism and toxicity, we have developed an approach for evaluation of the rate of N-dealkylation mediated by two of the most important human cytochrome P450s (P450), namely CYP3A4 and CYP2D6. We have taken a novel approach by using descriptors generated for the whole molecule, the reaction centroid, and the leaving group, and then applying neural network computations and sensitivity analysis to generate quantitative structure-metabolism relationship models. The quality of these models was assessed by using the cross-validated correlation coefficients of 0.82 for CYP3A4 and 0.79 for CYP2D6 as well as external test molecules for each enzyme. The relative performance of different neural networks was also compared, and modular neural networks with two hidden layers provided the best predictive ability. Functional dependencies between the neural network input and output variables, generalization ability, and limitations of the described approach are also discussed. These models represent an initial approach to predicting the rate of P450-mediated metabolism and may be applied and integrated with other models for P450 binding to produce a systems-based approach for predicting drug metabolism.

Towards integrated ADME prediction: past, present and future directions for modelling metabolism by UDP-glucuronosyltransferases

Undesirable absorption, distribution, metabolism, excretion (ADME) properties are the cause of many drug development failures and this has led to the need to identify such problems earlier in the development process. This review highlights computational (in silico) approaches that have been used to identify the characteristics of ligands influencing molecular recognition and/or metabolism by the drug-metabolising enzyme UDP-gucuronosyltransferase (UGT). Current studies applying pharmacophore elucidation, 2D-quantitative structure metabolism relationships (2D-QSMR), 3D-quantitative structure metabolism relationships (3D-QSMR), and non-linear pattern recognition techniques such as artificial neural networks and support vector machines for modelling metabolism by UGT are reported. An assessment of the utility of in silico approaches for the qualitative and quantitative prediction of drug glucuronidation parameters highlights the benefit of using multiple pharmacophores and also non-linear techniques for classification. Some of the challenges facing the development of generalisable models for predicting metabolism by UGT, including the need for screening of more diverse structures, are also outlined.

Metabolic stability for drug discovery and development: pharmacokinetic and biochemical challenges

Metabolic stability refers to the susceptibility of compounds to biotransformation in the context of selecting and/or designing drugs with favourable pharmacokinetic properties. Metabolic stability results are usually reported as measures of intrinsic clearance, from which secondary pharmacokinetic parameters such as bioavailability and half-life can be calculated when other data on volume of distribution and fraction absorbed are available. Since these parameters are very important in defining the pharmacological and toxicological profile of drugs as well as patient compliance, the pharmaceutical industry has a particular interest in optimising for metabolic stability during the drug discovery and development process. In the early phases of drug discovery, new chemical entities cannot be administered to humans; hence, predictions of these properties have to be made from in vivo animal, in vitro cellular/subcellular and computational systems. The utility of these systems to define the metabolic stability of compounds that is predictive of the human situation will be reviewed here. The timing of performing the studies in the discovery process and the impact of recent advances in research on drug absorption, distribution, metabolism and excretion (ADME) will be evaluated with respect to the scope and depth of metabolic stability issues. Quantitative prediction of in vivo clearance from in vitro metabolism data has, for many compounds, been shown to be poor in retrospective studies. One explanation for this may be that there are components used in the equations for scaling that are missing or uncertain and should be an area of more research. For example, as a result of increased biochemical understanding of drug metabolism, old assumptions (e.g. that the liver is the principal site of first-pass metabolism) need revision and new knowledge (e.g. the relationship between transporters and drug metabolising enzymes) needs to be incorporated into in vitro-in vivo correlation models. With ADME parameters increasingly being determined on automated platforms, instead of using results from high throughput screening (HTS) campaigns as simple go/no-go filters, the time saved and the many compounds analysed using the robots should be invested in careful processing of the data. A logical step would be to investigate the potential to construct computational models to understand the factors governing metabolic stability. A rational approach to the use of HTS assays should aim to screen for many properties (e.g. physicochemical parameters, absorption, metabolism, protein binding, pharmacokinetics in animals and pharmacology) in an integrated manner rather than screen against one property on many compounds, since it is likely that the final drug will represent a global average of these properties.

Predicting undesirable drug interactions with promiscuous proteins in silico

Although computational tools have been used to predict toxic responses resulting from molecules binding either as substrates or inhibitors to proteins, there are complications to be resolved. Some proteins appear promiscuous in their ability to bind a diverse array of hydrophobic molecules. This promiscuity arises from the binding site simultaneously accommodating more than one molecule, multiple separate binding sites, protein flexibility, or a combination of all these properties. With the availability of more crystal structures for these non-target proteins, we should be able to predict binding in silico with a greater accuracy, thus avoiding or managing toxic side effects, therefore ultimately improving the success of drug discovery.

PREDICTINGHUMANDRUGGLUCURONIDATIONPARAMETERS: Application of In Vitro and In Silico Modeling Approaches

Cytochrome P450 (CYP) and UDP-glucuronosyltransferase (UGT), which both exist as enzyme "superfamilies," are together responsible for the metabolism of most hepatically cleared drugs. There is currently intense interest in the development of techniques that permit identification of the CYP and UGT isoform(s) involved in the metabolism of a newly discovered drug, and hence prediction of factors likely to alter elimination in vivo. In addition, the quantitative scaling of kinetic parameters for a metabolic pathway assumes importance for identifying newly discovered drugs with undesirable in vivo pharmacokinetic properties. Although qualitative and quantitative in vitro-in vivo correlation based on data generated using human liver tissue or recombinant enzymes have been applied successfully to many drugs eliminated by CYP, these strategies have proved less definitive for glucuronidated compounds. Computational (in silico) modeling techniques that potentially provide a facile and economic alternative to the in vitro methods are now emerging. This review assesses the utility of in vitro and in silico approaches for the qualitative and quantitative prediction of drug glucuronidation parameters and the challenges facing the development of generalizable models.

Comparative molecular field analysis and QSAR on substrates binding to cytochrome P450 2D6

In this study, we utilized comparative molecular field analysis (CoMFA) to gain a better understanding of the steric and electrostatic features of the cytochrome p450 2D6 (CYP2D6) active site. The training set consists of 24 substrates with reported K(M) values from liver microsomal CYP2D6 spanning an activity range of almost three log units. The low energy conformers were fit by root mean square (RMS) to minaprine at the site of metabolism and to the protonated nitrogen. In this manner, we constructed two CoMFA models, one model with a distance constraint and another without. The model with the distance parameter (non-cross-validated R(2)=0.99) was approximately equal to the CoMFA without a distance parameter (non-cross-validated R(2)=0.98). Validation of our CoMFA was accomplished by predicting the K(M) values of 15 diverse CYP2D6 substrates not in the original training set resulting in a predictive R(2)=0.62. Finally, we also pursued correlations of pK(a) and log P with CYP2D6 substrate K(M) in an effort to investigate other physicochemical properties.

Influence of lipophilicity on the interactions of N-alkyl-4-phenyl-1,2,3,6-tetrahydropyridines and their positively charged N-alkyl-4-phenylpyridinium metabolites with cytochrome P450 2D6

The relationship between lipophilicity and CYP2D6 affinity of cyclic tertiary (N-alkyl-4-phenyl-1,2,3,6-tetrahydropyridines) and quaternary (N-alkyl-4-phenylpyridinium) amines was examined. The 1,2,3,6-tetrahydropyridine scaffold was chosen due to its common occurrence in the structures of CYP2D6 ligands such as the Parkinsonian neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) and the dehydrated haloperidol metabolite N-[4-(4-fluorophenyl)-4-oxobutyl]-4-(4-chlorophenyl)-1,2,3,6-tetrahydropyridine (HPTP). Likewise, the pyridinium framework is found in and 4-(4-chlorophenyl)-1-[4-(4-fluorophenyl)-4-oxobutyl]pyridinium and N-methyl-4-phenylpyridinium (MPP(+)), the positively charged metabolites of MPTP and haloperidol. The lack of CYP2D6 inhibition by MPTP and its pyridinium metabolite MPP(+) was due to their hydrophilic nature since higher N-alkyl homologs revealed substantial increases in inhibitory potency against recombinant CYP2D6-mediated bufuralol-1'-hydroxylation. The reasonable correlation between lipophilicity and CYP2D6 inhibition by pyridiniums and 1,2,3,6-tetrahydropyridines was only limited to straight chain N-alkyl analogs, since certain N-alkylaryl analogs of lower lipophilicity were better CYP2D6 inhibitors. CYP2D6 substrate properties of straight chain N-alkyltetrahydropyridines were also governed by lipophilicity, and N-heptyl-4-phenyl-1,2,3,6-tetrahydropyridine was the optimal substrate (K(mapp) = 0.63 microM). Metabolism studies indicated that the N-heptyl analog underwent monohydroxylation on the aromatic ring and on the N-heptyl group suggesting that 1,2,3,6-tetrahydropyridines can bind in more than one conformation in the CYP2D6 active site. Increased lipophilicity of haloperidol metabolites did not correlate with inhibitory potency since the more lipophilic HPTP metabolite was less potent as an inhibitor than reduced-haloperidol and reduced-HPTP. Furthermore, HPTP and reduced-HPTP, of comparable lipophilicity to the N-heptyltetrahydropyridine analog were inactive as CYP2D6 substrates. This observation suggests that steric constraints rather than lipophilicity are responsible for the lack of CYP2D6 substrate properties of cyclic tertiary amines tethered to bulky N-substituents. This phenomenon appears to be a common theme among several cyclic tertiary amine-containing anti-depressants and should be taken into consideration when designing central nervous system agents devoid of CYP2D6 substrate properties.

Do drug metabolism and pharmacokinetic departments make any contribution to drug discovery?

The alignment of drug metabolism and pharmacokinetic departments with drug discovery has not produced a radical improvement in the pharmacokinetic properties of new chemical entities. The reason for this is complex, reflecting in part the difficulty of combining potency, selectivity, water solubility, metabolic stability and membrane permeability into a single molecule. This combination becomes increasingly problematic as the drug targets become more distant from aminergic seven-transmembrane-spanning receptors (7-TMs). The leads available for aminergic 7-TMs, like the natural agonists, are invariably small molecular weight, water soluble and potent. Even moving to 7-TMs for which the agonist is a peptide invariably produces lead matter that is less drug-like (higher molecular weight and lipophilic). The role of drug metabolism departments, therefore, has been to guide chemistry to obtaining adequate, rather than optimal, pharmacokinetic properties for these 'difficult' drug targets. A consistent belief of many researchers is that a high value is placed on optimal, rather than adequate, pharmacokinetic properties. One measure of value is market sales, and when these are examined no clear pattern emerges. Part of the success of amlodipine in the calcium channel antagonist sector must be due to its excellent pharmacokinetic profile, but the best-selling drugs among the angiotensin antagonists and beta-blockers have a much greater market share than other agents with better pharmacokinetic properties. Clearly, many other factors are important in the successful launch of a medicine, some reflected in the manner the compound is developed and the subsequent structure of the labelling. Overall, therefore the presence of drug metabolism in drug discovery has probably contributed most by allowing 'difficult' drug targets to be prosecuted, rather than by guiding medicinal chemists to optimal pharmacokinetics. These 'difficult' target candidates become successful drugs when skilfully developed. There is no doubt that skilful development relies heavily on drug metabolism and pharmacokinetic departments, in this case those with a clinical rather than a preclinical orientation.

Residues glutamate 216 and aspartate 301 are key determinants of substrate specificity and product regioselectivity in cytochrome P450 2D6

Cytochrome P450 2D6 (CYP2D6) metabolizes a wide range of therapeutic drugs. CYP2D6 substrates typically contain a basic nitrogen atom, and the active-site residue Asp-301 has been implicated in substrate recognition through electrostatic interactions. Our recent computational models point to a predominantly structural role for Asp-301 in loop positioning (Kirton, S. B., Kemp, C. A., Tomkinson, N. P., St.-Gallay, S., and Sutcliffe, M. J. (2002) Proteins 49, 216-231) and suggest a second acidic residue, Glu-216, as a key determinant in the binding of basic substrates. We have evaluated the role of Glu-216 in substrate recognition, along with Asp-301, by site-directed mutagenesis. Reversal of the Glu-216 charge to Lys or substitution with neutral residues (Gln, Phe, or Leu) greatly decreased the affinity (K(m) values increased 10-100-fold) for the classical basic nitrogen-containing substrates bufuralol and dextromethorphan. Altered binding was also manifested in significant differences in regiospecificity with respect to dextromethorphan, producing enzymes with no preference for N-demethylation versus O-demethylation (E216K and E216F). Neutralization of Asp-301 to Gln and Asn had similarly profound effects on substrate binding and regioselectivity. Intriguingly, removal of the negative charge from either 216 or 301 produced enzymes (E216A, E216K, and D301Q) with elevated levels (50-75-fold) of catalytic activity toward diclofenac, a carboxylate-containing CYP2C9 substrate that lacks a basic nitrogen atom. Activity was increased still further (>1000-fold) upon neutralization of both residues (E216Q/D301Q). The kinetic parameters for diclofenac (K(m) 108 microm, k(cat) 5 min(-1)) along with nifedipine (K(m) 28 microm, k(cat) 2 min(-1)) and tolbutamide (K(m) 315 microm, k(cat) 1 min(-1)), which are not normally substrates for CYP2D6, were within an order of magnitude of those observed with CYP3A4 or CYP2C9. Neutralizing both Glu-216 and Asp-301 thus effectively alters substrate recognition illustrating the central role of the negative charges provided by both residues in defining the specificity of CYP2D6 toward substrates containing a basic nitrogen.

Oxidation of methoxyphenethylamines by cytochrome P450 2D6. Analysis of rate-limiting steps

Cytochrome P450 (P450) 2D6 is involved in the oxidation of a large fraction ( approximately 30%) of drugs used by humans and also catalyzes the O-demethylation of the model substrates 3- and 4-methoxyphenethylamine followed by subsequent ring hydroxylation to dopamine. Burst kinetics were not observed; rate-limiting step(s) must occur prior to product formation. Rates of reduction of ferric P450 2D6 were stimulated by 3- or 4-methoxyphenethylamine or the inhibitor quinidine; reduction is not the most rate-limiting step. The non-competitive intramolecular deuterium isotope effect, an estimate of the intrinsic isotope effect, for 4-methoxyphenethylamine O-demethylation was 9.6. Intermolecular non-competitive deuterium isotope effects of 3.1-3.8 were measured for k(cat) and k(cat)/K(m) for both O-demethylation reactions, implicating at least partially rate-limiting C-H bond breaking. Simulation of steady-state kinetic data yielded a catalytic mechanism dominated by the rates of (i) Fe(2+)O(2)(-) protonation (plus O-O bond scission) and (ii) C-H bond breaking, consistent with the appearance of the spectral intermediates in the steady state, attributed to iron-oxygen complexes. However, all the rates of individual steps (or rates of combined steps) are considerably higher than k(cat), and the contributions of several steps must be considered in understanding rates of the P450 2D6 reactions.

Designer drugs that are potent inhibitors of CYP2D6

Some abused drugs are substrates of CYP2D6 (e.g., paramethoxyamphetamine, methylenedioxy-methamphetamine). CYP2D6 inhibition by concurrently used drugs of abuse could potentiate such drugs and increase acute toxicity. Ten designer drugs were tested as inhibitors of CYP2D6. Only 1-methyl-4-phenyl-4-propionoxypiperidine (MPPP) and 1-[2-phenylethyl]-4-phenyl-4-acetoxypiperidine (PEPAP) interacted significantly (Ki 1.6 microM and 0.3 microM, respectively). Both are synthetic analogues of meperidine sold as "synthetic heroin." No CYP2D6-mediated metabolites were detected for either compound. Concurrent oral use of CYP2D6 substrates with MPPP and PEPAP may represent a kinetic drug interaction risk, but this risk must be confirmed clinically.

Pharmacophore modeling of cytochromes P450

Understanding the binding of ligands in the active site of a membrane-bound protein is difficult in the absence of a crystal structure. When these proteins are the enzymes involved in drug metabolism, it leaves little option but to use site-directed mutagenesis and in vitro studies to provide critical information relating to determinants of binding affinity. Pharmacophore models and three-dimensional quantitative structure-activity relationships have been used either alone or in combination with protein homology models to provide this information for cytochrome P450s. At present, their application has been directed to the major enzymes but this may escalate in future as more in vitro data are generated for other P450s. The following review outlines the methodologies and models as well as future prospects for applying these technologies to P450s in the hope that future drugs will be selected with increased metabolic stability and fewer incidences of undesirable drug-drug interactions.

Computational models for cytochrome P450: a predictive electronic model for aromatic oxidation and hydrogen atom abstraction

Experimental observations suggest that electronic characteristics play a role in the rates of substrate oxidation for cytochrome P450 enzymes. For example, the tendency for oxidation of a certain functional group generally follows the relative stability of the radicals that are formed (e.g., N-dealkylation > O-dealkylation > 2 degrees carbon oxidation > 1 degree carbon oxidation). In addition, results show that useful correlations between the rates of product formation can be developed using electronic models. In this article, we attempt to determine whether a combined computational model for aromatic and aliphatic hydroxylation can be developed. Toward this goal, we used a combination of experimental data and semiempirical molecular orbital calculations to predicted activation energies for aromatic and aliphatic hydroxylation. The resulting model extends the predictive capacity of our previous aliphatic hydroxylation model to include the second most important group of oxidations, aromatic hydroxylation. The combined model can account for about 83% of the variance in the data for the 20 compounds in the training set and has an error of about 0.7 kcal/mol.

Combined three-dimensional quantitative structure-activity relationship analysis of cytochrome P450 2B6 substrates and protein homology modeling

Understanding the basis of the substrate specificity of cytochrome P450 2B6 (CYP2B6) is important for determining the role of this enzyme in drug metabolism and for predicting new substrates. Pharmacophores were generated for 16 structurally diverse CYP2B6 substrates with Catalyst after overlapping the reaction sites. Two pharmacophores were determined for the CYP2B6 binding site. Both include two hydrophobes and one hydrogen bond acceptor. The three-dimensional structure of CYP2B6 was then modeled based on the crystal structure of CYP2C5. Benzyloxyresorufin and 7-ethoxy-4-trifluoromethylcoumarin, the two lowest K(m) substrates in the training set, were then docked in the active site of CYP2B6. The pharmacophores were combined with the CYP2B6 model by comparing the docking results and the mapping of the two substrates with the pharmacophores. The results indicated that the active site of CYP2B6 complements the pharmacophores. The pharmacophores and the CYP2B6 model were used in conjunction to predict the K(m) values of substrates in a test set of five compounds and yielded satisfactory predictions for benzphetamine, cinnarizine, bupropion, and verapamil but not lidocaine. The CYP2B6 model, the pharmacophores, and the combination of the model with these pharmacophores provide insight into the interactions of CYP2B6 with substrates. The pharmacophores may be used as queries to search a database to predict new substrates for CYP2B6 when the reaction site is known (N- or O-dealkylation). For C-hydroxylation, the CYP2B6 model is helpful in evaluating the possible reaction sites in order for the pharmacophores to predict corresponding K(m) values.

The impact of informatics and computational chemistry on synthesis and screening

High-throughput synthesis and screening technologies have enhanced the impact of computational chemistry on the drug discovery process. From the design of targeted, drug-like libraries to 'virtual' optimization of potency, selectivity and ADME/Tox properties, computational chemists are able to efficiently manage costly resources and dramatically shorten drug discovery cycle times. This review will describe some of the successful strategies and applications of state-of-the-art algorithms to enhance drug discovery, as well as key points in the drug discovery process where computational methods can have, and have had, greatest impact.

Functional evaluation of cytochrome P450 2D6 with Gly42Arg substitution expressed in Saccharomyces cerevisiae

A single amino acid-substituted mutant protein, CYP2D6 (G42R) was expressed in Saccharomyces cerevisiae and its enzymatic properties were compared with those of other single (P34S, R296C and S486T) and double amino acid-substituted mutant proteins (P34S/S486T and R296C/S486T) expressed in yeast cells, all of which were known to occur in the CYP2D6 gene as single nucleotide polymorphisms. The protein levels of G42R, P34S and P34S/S486T in microsomal fractions and their oxidation capacities towards debrisoquine as a prototypic substrate and bunitrolol as a chiral substrate were different from those of wild-type CYP2D6, while the R296C, S486T and R296C/S486T behaved similarly to the wild-type in these indices. The CYP contents both in yeast microsomal and in whole cell fractions indicated that some part of G42R protein was localized in the endoplasmic reticulum membrane fraction, whereas most of G42R protein was in some subcellular fractions other than endoplasmic reticulum. In kinetic analysis, the G42R substitution increased apparent Km and decreased Vmax for debrisoquine 4-hydroxylation, while it increased both Km and Vmax for bunitrolol 4-hydroxylation. The P34S substitution did not drastically change Km but decreased Vmax for debrisoquine 4-hydroxylation, whereas Km was increased and Vmax unchanged or decreased for bunitrolol 4-hydroxylation by P34S substitution. These results suggest that the G42R substitution causes a change in the CYP2D6 conformation, which may be different from the change produced by the P34S substitution.

Heterologous expression of cytochrome P450 2D6 mutants, electron transfer, and catalysis of bufuralol hydroxylation: the role of aspartate 301 in structural integrity

Cytochrome P450 (P450) 2D6 is a polymorphic human enzyme involved in the oxidation of >50 drugs, most of which contain a basic nitrogen. In confirmation of previous work by others, substitutions at Asp301 decreased rates of substrate oxidation by P450 2D6. An anionic residue (Asp, Glu) at this position was found to be important in proper protein folding and heme incorporation, and positively charged residues were particularly disruptive in bacterial and also in baculovirus expression systems. Truncation of 20 N-terminal amino acids had no significant effect on catalytic activity except to attenuate P450 2D6 interaction with membranes and NADPH-P450 reductase. The truncation of the N-terminus increased the level of bacterial expression of wild-type P450 2D6 (Asp301) but markedly reduced expression of all codon 301 mutants, including Glu301. Reduction of ferric P450 2D6 by NADPH-P450 reductase was enhanced in the presence of the prototypic substrate bufuralol. Bacterial flavodoxin, an NADPH-P450 reductase homolog, binds tightly to P450 2D6 but is inefficient in electron transfer to the heme. These results collectively indicate that the acidic residue at position 301 in P450 2D6 has a structural role in addition to any in substrate binding and that the N-terminus of P450 2D6 is relatively unimportant to catalytic activity beyond a role in facilitating binding to NADPH-P450 reductase.

Optimization of metabolic stability as a goal of modern drug design

Metabolism and other pharmacokinetic (PK) studies have always played a critical role in helping to optimize the bioavailability and duration of action of new drugs thereby increasing their success rate. With the advent of automated combinatorial synthesis, high-throughput pharmacological testing, and the ability to create extensive databases in the past decade, drug discovery has undergone an amazing evolution. With the increased throughput of drug discovery, metabolism and other PK studies have evolved to keep pace. Often called "early ADME" studies, these studies are characterized by parallel processing and higher throughput than before. This article focuses on a particular class of early ADME (absorption, distribution mechanism, and excretion) studies known as "metabolic stability" studies. The theoretical basis for metabolic stability and its relationship to the concept of metabolic intrinsic clearance is briefly presented. Some key relationships between structure and metabolism are summarized. Several case studies from recent medicinal chemistry literature are reviewed to exemplify how metabolic stability studies influenced drug design and led to improvements in bioavailability and half-life. Finally, future trends in drug metabolism and analytical chemistry and how they may influence metabolic stability studies are reviewed.

Screening of drug candidates for their drug--drug interaction potential

Within the past year, additional papers have been published that focus on higher-throughput drug-interaction screening. Some papers have described enzyme assays that can be used to evaluate inhibition or induction of the human cytochrome P450s. At the same time, numerous investigators have developed computational (in silico) methods to predict interactions and have validated the approach using in vitro (assay-derived) data. These so called 'in silico--in vitro' correlations have great potential and may complement existing 'in vitro--in vivo' correlations.

A virtual high throughput screen for high affinity cytochrome P450cam substrates. Implications for in silico prediction of drug metabolism

Structure-based virtual screening techniques require reliable scoring functions to discriminate potential substrates effectively. In this study we compared the performance of GOLD, PMF, DOCK and FlexX scoring functions in FlexX flexible docking to cytochrome P450cam binding site. Crystal structures of protein-substrate complexes were most effectively reproduced by the FlexX/PMF method. On the other hand, the FlexX/GOLD approach provided the best correlation between experimental binding constants and predicted scores. Binding modes selected by the FlexX/PMF approach were rescored by GOLD to obtain a reliable measure of binding energetics. The effectiveness of the FlexX/PMF/GOLD method was demonstrated by the correct classification of 32 out of the 33 experimentally studied compounds and also in a virtual HTS test on a library of 10,000 compounds. Although almost all the available functions were developed to be general, our study on cytochrome P450cam substrates suggests that careful selection or even tailoring the scoring function might increase the prediction power of virtual screens significantly. The FlexX/PMF/GOLD methodology was tested on cytochrome P450 3A4 substrates and inhibitors. This preliminary study revealed that the combined function was able to recognise 334 out of the 345 compounds bound to 3A4.

Evaluation of designed ligands by a multiple screening method: application to glycogen phosphorylase inhibitors constructed with a variety of approaches

Glycogen phosphorylase (GP) is an important enzyme that regulates blood glucose level and a key therapeutic target for the treatment of type II diabetes. In this study, a number of potential GP inhibitors are designed with a variety of computational approaches. They include the applications of MCSS, LUDI and CoMFA to identify additional fragments that can be attached to existing lead molecules; the use of 2D and 3D similarity-based QSAR models (HQSAR and SMGNN) and of the LUDI program to identify novel molecules that may bind to the glucose binding site. The designed ligands are evaluated by a multiple screening method, which is a combination of commercial and in-house ligand-receptor binding affinity prediction programs used in a previous study (So and Karplus, J. Comp.-Aid. Mol. Des., 13 (1999), 243-258). Each method is used at an appropriate point in the screening, as determined by both the accuracy of the calculations and the computational cost. A comparison of the strengths and weaknesses of the ligand design approaches is made.

Influence of phenylalanine-481 substitutions on the catalytic activity of cytochrome P450 2D6

Homology models of the active site of cytochrome P450 2D6 (CYP2D6) have identified phenylalanine 481 (Phe(481)) as a putative ligand-binding residue, its aromatic side chain being potentially capable of participating in pi-pi interactions with the benzene ring of ligands. We have tested this hypothesis by replacing Phe(481) with tyrosine (Phe(481)-->Tyr), a conservative substitution, and with leucine (Phe(481)-->Leu) or glycine (Phe(481)-->Gly), two non-aromatic residues, and have compared the properties of the wild-type and mutant enzymes in microsomes prepared from yeast cells expressing the appropriate cDNA-derived protein. The Phe(481)-->Tyr substitution did not alter the kinetics [K(m) (microM) and V(max) (pmol/min per pmol) respectively] of oxidation of S-metoprolol (27; 4.60), debrisoquine (46; 2.46) or dextromethorphan (2; 8.43) relative to the respective wild-type values [S-metoprolol (26; 3.48), debrisoquine (51; 3.20) and dextromethorphan (2; 8.16)]. The binding capacities [K(s) (microM)] of a range of CYP2D6 ligands to the Phe(481)-->Tyr enzyme (S-metoprolol, 22.8; debrisoquine, 12.5; dextromethorphan, 2.3; quinidine, 0.13) were also similar to those for the wild-type enzyme (S-metoprolol, 10.9; debrisoquine, 8.9; dextromethorphan, 3.1; quinidine, 0.10). In contrast, the Phe(481)-->Leu and Phe(481)-->Gly substitutions increased significantly (3-16-fold) the K(m) values of oxidation of the three substrates [S-metoprolol (120-124 microM), debrisoquine (152-184 microM) and dextromethorphan (20-31 microM)]. Similarly, the K(s) values of the ligands to Phe(481)-->Leu and Phe(481)-->Gly mutants were also increased 3 to 10-fold (S-metoprolol, 33.2-41.9 microM; debrisoquine, 85-90 microM; dextromethorphan, 15.7-18.8 microM; quinidine 0.35-0.53 microM). However, contrary to a recent proposal that Phe(481) has the dominant role in the binding of substrates that undergo CYP2D6-mediated N-dealkylation routes of metabolism, the Phe(481)-->Gly substitution did not substantially decrease the capacity of the enzyme to N-deisopropylate metoprolol (wild-type, 1.12 pmol/min per pmol of P450; Phe(481)-->Gly, 0.71), whereas an Asp(301)-->Gly substitution decreased the N-dealkylation reaction by 95% of the wild-type rate. Overall, our results are consistent with the proposal that Phe(481) is a ligand-binding residue in the active site of CYP2D6 and that the residue interacts with ligands via a pi-pi interaction between its phenyl ring and the aromatic moiety of the ligand. However, the relative importance of Phe(481) in binding is ligand-dependent; furthermore, its importance is secondary to that of Asp(301). Finally, contrary to predictions of a recent homology model, Phe(481) does not seem to have a primary role in CYP2D6-mediated N-dealkylation.

Competitive CYP2C9 inhibitors: enzyme inhibition studies, protein homology modeling, and three-dimensional quantitative structure-activity relationship analysis

This study describes the generation of a three-dimensional quantitative structure activity relationship (3D-QSAR) model for 29 structurally diverse, competitive CYP2C9 inhibitors defined experimentally from an initial data set of 73 compounds. In parallel, a homology model for CYP2C9 using the rabbit CYP2C5 coordinates was built. For molecules with a known interaction mode with CYP2C9, this homology model, in combination with the docking program GOLD, was used to select conformers to use in the 3D-QSAR analysis. The remaining molecules were docked, and the GRID interaction energies for all conformers proposed by GOLD were calculated. This was followed by a principal component analysis (PCA) of the GRID energies for all conformers of all compounds. Based on the similarity in the PCA plot to the inhibitors with a known interaction mode, the conformer to be used in the 3D-QSAR analysis was selected. The compounds were randomly divided into two groups, the training data set (n = 21) to build the model and the external validation set (n = 8). The PLS (partial least-squares) analysis of the interaction energies against the K(i) values generated a model with r(2) = 0.947 and a cross-validation of q(2) = 0.730. The model was able to predict the entire external data set within 0.5 log units of the experimental K(i) values. The amino acids in the active site showed complementary features to the grid interaction energies in the 3D-QSAR model and were also in agreement with mutagenesis studies.

Computational toxicology and the generation of mechanistic hypotheses: gamma-butyrolactone

In this study, we use SAR approaches in an attempt to elucidate the action of gamma-butyrolactone (GBL), an illicit drug and a dietary supplement, that can cause coma and deaths in humans while exhibiting low systemic toxicity towards rodents. The lack of systemic toxicity of GBL and of its metabolite(s) was also predicted by validated SAR models. In fact using diverse SAR models, the only significant SAR prediction was that GBL had the potential for inhibiting human cytochrome P4502D6 (CYP2D6). However, inhibition of that isozyme is not necessarily associated with toxicity. It is suggested that GBL users also abuse other substances. When GBL inhibits CYP2D6 this may prevent the CYP2D6-mediated detoxification of other toxicants simultaneously consumed by the GBL user.

Application of in silico approaches to predicting drug--drug interactions

In an environment driven to find the next blockbuster drug, failure years into a project should not be an option. Recent studies have shown that poor absorption, distribution, metabolism, and excretion (ADME), and the related properties of toxicity and pharmacokinetics are responsible for a large proportion of failures. One way to understand and potentially predict molecules likely to be successful in humans as drugs from an ADME point of view is to use simulations. Such simulations may include simple rule-based approaches, structure--activity relationships, three-dimensional quantitative structure--activity relationships (3D-QSAR), and pharmacophores. All of these represent useful tools in understanding metabolism by the cytochromes P450, predicting drug--drug interactions (DDIs), and other pharmacokinetic parameters. The present paper briefly reviews the application of some computational tools applied to predicting DDIs and will provide the reader with an idea of their utility.