Current Drug Metabolism - Volume 11, Issue 4, 2010

Many authors have been used Graph and Complex Network theory to approach very large metabolic networks with low computational cost. These large networks are graphical representations of real metabolic systems with essentially two components nodes and links. Nodes (represented by dots) are usually metabolites, enzymes, substrates, intermediary substances, metabolic reactions or transition states. Node-to-node links (edges or arcs) express metabolic relationships between two nodes as for instance: substrate-enzyme pairs or metabolite-reaction pair. We can use these networks to describe and study all the set of metabolic processes (Metabolome) related to one organism, tissue, or diseases. Including nodes representing body compartments we can represent and study also the Absorption, Distribution, Metabolism, Excretion and Toxicity (ADMET) processes of drugs or hazardous compounds. On the other hand, other authors with a more chemical background have used graphs to represent the structure of drugs, xenobiotic substances, hazardous compounds, and metabolites. These graphs are essentially, in mathematical terms, the same objects than the metabolic networks referred in the previous paragraph. The main difference is that in molecular graphs nodes represent atoms and edges represent chemical bonds. Consequently, molecular graphs express the structure of organic compounds in terms of atom connectivity and metabolic networks represent the structure of the metabolic system in terms of metabolome connectedness. In addition, we can associate both types of graphs with different classes of numeric matrices to carry out computational studies of at the two levels of matter structural organization. The Boolean or Adjacency matrices are perhaps the more simple to explain. These matrices are square tables (number of rows = number of columns) of nxn elements, where n is the number of nodes of the system. The element matrix cell bij = 1 if the element ith link to jth in the graph. I meant, the atom ith is chemically bonded to atom jth in the structure of the drug or, for instance, the substrate ith is metabolized by the the enzyme jth. Yet we can mention a third type of complex network that lie in-between molecular graphs of drugs or metabolites and large graph of complex metabolic networks. We refer to the complex networks used to represent the 3D structures of proteins (enzymes, molecular targets, channels, receptors) involved in natural or disease metabolism or in drug ADMET processes as well. The construction of this type of graphs and matrices is straightforward to realize in an intuitive form taking into consideration the analogy between the previous situations. In these networks, aminoacids often play the role of nodes and links express spatial contact between two aminoacids (see also, contact maps or residue networks). In the same group with proteins we can find the graphs used to represent another type or biomolecules that may play also the role of enzymes, drug targets, and also participate in biological processes regulation. We refer to the graph used to represent the secondary structure (or less common the 3D structure) or RNAs. In this last class of networks, nucleotides often play the role of nodes and links express that this pair of bases are sequence neighbors or are involved in a hydrogen bond. In all these cases, we can easily calculate different invariant parameters of the matrices associated to the graphs that may be used to describe the structure of these objects (drugs, proteins, or metabolome). As this numbers are based only on connectivity information they are often named as Connectivity measures or Topological Indices (TIs). In fact, in our days, there is an explosion on the use of Topological Indices (TIs) of Graphs and Complex Networks on a broad spectrum of topics related to Drug metabolism and distribution research. One reason for the success of TIs, is the high flexibility of this theory to solve in a fast but rigorous way many apparently unrelated problems in all these disciplines. This determined the recent development of several interesting software and theoretical methods to handle with structurefunction information and data mining in this field. However, another important advantage is that: The theoretical basis is straightforward to realize for experimental scientists non-expert on computational techniques....

The integration of early ADMET (absorption, distribution, metabolism, excretion and toxicity) profiling, or simply prediction, of 'lead' molecules to speed-up the 'lead' selection further for phase-I trial without losing large amount of revenue. The ADMET profiling and prediction is mostly dependent of a number of molecular descriptors, for example, Lipinski's 'Rule of 5' (Ro5). Recently a large number of articles have been reporting that it possible to do some prediction of the ADMET properties using the structural features of the molecules, utilizing several and multiple approaches. One of the most important approaches is the QSAR/QSPR modelling of the data derived from their activity profiles and their different structural features (i.e., quantitative molecular descriptors).

PF-02341066 is a selective c-Met/Alk tyrosine kinase inhibitor currently in clinical development as an anticancer agent. Non-clinical toxicokinetic evaluation in rats revealed gender-related differences in pharmacokinetics with at least 2-fold higher PF-02341066 plasma concentrations in males than females when administered the same dose. In general, lower systemic exposure of drugs that undergoes oxidative metabolism in male than female rats has been well known to be attributed to gender-specific expression of CYP genes in rats. It is of interest to understand why the gender-related pharmacokinetics in rats for PF-02341066 was opposite to the general observations and if the gender-related pharmacokinetics would be seen in humans that may impact the drug efficacy and toxicity profiles. The potential gender-related differences in PF-02341066 metabolism were investigated both in vitro and in vivo using [3H]PF-02341066. Oxidation was found to be the major metabolic pathway in male rat liver S9 incubations whereas sulfoconjugation was the predominant metabolic pathway in females. There was no qualitative difference in metabolite profiles of PF-02341066 between man and woman liver S9 incubations. Following a single oral administration of [3H]PF-02341066 to rats at 150 mg/kg, the primary route of excretion of the radioactivity was via feces, in which, the most abundant radio-component in male rat was the parent drug (29% of dose) and in female rat was the parent sulfate (44% of dose). The more extensive formation of the parent solfoconjugate in female rats most likely explains why the female rat had lower drug exposure compared to male rat, as gender-related changes of sulfotransferase expression were widely reported in rats. The human liver S9 study suggests that gender-related pharmacokinetics of PF-02341066 are unlikely to occur in humans.

Efficient drugs such as statins or mevinic acids are inhibitors of the rate-limiting enzyme of cholesterol biosynthesis, 3- hydroxy-3-methyl-glutaryl coenzyme A reductase (HMGR), an enzyme responsible for the double reduction of 3-hydroxy-3-methyl-glutaryl coenzyme A. These compounds promoted the synthesis and evaluation of new inhibitors for HMGR, named HMGRIs. The high number of possible candidates creates the necessity of Quantitative Structure-Activity Relationship models in order to guide the HMGRI (3-hydroxy-3-methyl-glutaryl coenzyme A inhibitor) synthesis. In this work, we revised different computational studies for a very large and heterogeneous series of HMGRIs. First, we revised QSAR studies with conceptual parameters such as flexibility of rotation, probability of availability, etc; we then used the method of regression analysis; and QSAR studies in order to understand the essential structural requirement for binding with receptor. Next, we reviewed 3D QSAR, CoMFA and CoMSIA with different compounds to find out the structural requirements for 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR) inhibitory activity

Complexity of metabolic systems can be undertaken at different scales (metabolites, metabolic pathways, metabolic network map, biological population) and under different aspects (structural, functional, evolutive). To analyse such a complexity, metabolic systems need to be decomposed into different components according to different concepts. Four concepts are presented here consisting in considering metabolic systems as sets of metabolites, chemical reactions, metabolic pathways or successive processes. From a metabolomic dataset, such decompositions are performed using different mathematical methods including correlation, stoichiometric, ordination, classification, combinatorial and kinetic analyses. Correlation analysis detects and quantifies affinities/oppositions between metabolites. Stoichiometric analysis aims to identify the organisation of a metabolic network into different metabolic pathways on the hand, and to quantify/optimize metabolic flux distributions through the different chemical reactions of the system. Ordination and classification analyses help to identify different metabolic trends and their associated metabolites leading to highlight chemical polymorphism representing different variability poles of the metabolic system. Then, metabolic processes/correlations responsible for such a polymorphism can be extracted in silico by combining metabolic profiles representative of different metabolic trends according to a weighting bootstrap approach. Finally, evolution of metabolic processes in time can be analysed by different kinetic/dynamic modelling approaches.

The cytochrome P450 family is a large and diverse group of hemoproteins that are located in virtually all types of organism, such as bacteria, eukaryotes and even Archaea. These proteins are found throughout the body, however the highest concentrations are associated with liver. As the Human Genome Project completed, there are 57 genes and more than 59 pseudogenes divided among 18 families of CYP genes and 43 subfamilies have been detected. In humans, CYPs are the major enzymes involved in drug metabolism and bioactivation, accounting for almost 75% of the total drug metabolism. The variability in drug metabolisms that are mainly induced by the CYP polymorphisms is reflected on the differences of the maximal plasma concentrations, half lives of some drugs and their clearance. Besides, it can also lead to adverse drug reactions that are considered as a major factor in drug toxicity. So, the genotype-activity relationships of the CYP proteins have become a hot topic in recent years. It is important to further understand why a certain genotype influences enzyme activity and how to predict more structure-activity relationships.

Colorectal cancer is one of the most frequent types of cancer in the world and generates important social impact. The understanding of the specific metabolism of this disease and the transformations of the specific drugs will allow finding effective prevention, diagnosis and treatment of the colorectal cancer. All the terms that describe the drug metabolism contribute to the construction of ontology in order to help scientists to link the correlated information and to find the most useful data about this topic. The molecular components involved in this metabolism are included in complex network such as metabolic pathways in order to describe all the molecular interactions in the colorectal cancer. The graphical method of processing biological information such as graphs and complex networks leads to the numerical characterization of the colorectal cancer drug metabolic network by using invariant values named topological indices. Thus, this method can help scientists to study the most important elements in the metabolic pathways and the dynamics of the networks during mutations, denaturation or evolution for any type of disease. This review presents the last studies regarding ontology and complex networks of the colorectal cancer drug metabolism and a basic topology characterization of the drug metabolic process subontology from the Gene Ontology.

Using graphic rules to deal with kinetic systems is an elegant approach by combining the graph representation (schematic representation) and rigorous mathematical derivation. It bears the following advantages: (1) providing an intuitive picture or illuminative insights; (2) helping grasp the key points from complicated details; (3) greatly simplifying many tedious, laborious, and error-prone calculations; and (4) able to double-check the final results. In this mini review, the non-steady state graphic rule in enzyme-catalyzed kinetics and protein-folding kinetics was extended to cover drugmetabolic systems. As a demonstration, a step-by-step illustration is presented showing how to use the graphic rule to derive the concentrations of the parent drug and its metabolites vs. time for the seliciclib, vildagliptin, and cyclin-dependent kinase inhibitor (AG-024322) metabolic systems, respectively. It can be seen from these paradigms that the graphic rule is particularly useful to analyze complicated drug metabolic systems and ensure the correctness of the derived results. Meanwhile, the intuitive feature of graphic representation may facilitate analyzing and classifying drug metabolic systems; e.g., according to their directed graphs, the metabolism of seliciclib and the metabolism of vildagliptin can be categorized as 0→5 mechanism while that of AG-024322 as 0→4→3mechanism.

In this communication we carry out an in-depth review of a very versatile QSPR-like method. The method name is MARCH-INSIDE (MARkov CHains Ivariants for Network Selection and DEsign) and is a simple but efficient computational approach to the study of QSPR-like problems in biomedical sciences. The method uses the theory of Markov Chains to generate parameters that numerically describe the structure of a system. This approach generates two principal types of parameters Stochastic Topological Indices (sto-TIs). The use of these parameters allows the rapid collection, annotation, retrieval, comparison and mining structures of molecular, macromolecular, supramolecular, and non-molecular systems within large databases. Here, we review and comment by the first time on the several applications of MARCH-INSIDE to predict drugs ADMET, Activity, Metabolizing Enzymes, and Toxico-Proteomics biomarkers discovery. The MARCH-INSIDE models reviewed are: a) drug-tissue distribution profiles, b) assembling drug-tissue complex networks, c) multi-target models for anti-parasite/anti-microbial activity, c) assembling drug-target networks, d) drug toxicity and side effects, e) web-server for drug metabolizing enzymes, f) models in drugs toxico-proteomics. We close the review with some legal remarks related to the use of this class of QSPR-like models.