QSAR models for predicting enzymatic hydrolysis of new chemical entities in 'soft-drug' design.

The work described here is aimed at developing QSAR models capable of predicting in vitro human plasma lability/stability. They were built based on a dataset comprising about 200 known compounds. 3D structures of the molecules were drawn, optimized and submitted to the calculation of molecular descriptors that enabled selecting different TR/TS set pairs, subsequently exploited to develop QSAR models. Several 'machine learning' algorithms were explored in order to obtain suitable classification models, which were then validated on the relevant TS sets. Moreover the predictive ability of the best performing models was assessed on a Prediction set (PS) comprising about 40 molecules, not strictly related, from a structural point of view, to the initial dataset, but (obviously) comprised within the validity domain of the QSAR models obtained. The study allowed selecting predictive models enabling the classification of New Chemical Entities with regard to hydrolysis rate, that may be exploited for soft-drug design.

Cytotoxic activity of polyacetylenes and polyenes isolated from roots of Echinacea pallida.

BACKGROUND AND PURPOSE: The n-hexane extracts of the roots of three medicinally used Echinacea species exhibited cytotoxic activity on human cancer cell lines, with Echinacea pallida found to be the most cytotoxic. Acetylenes are present in E. pallida lipophilic extracts but essentially absent in extracts from the other two species. In the present study, the cytotoxic effects of five compounds, two polyacetylenes (namely, 8-hydroxy-pentadeca-(9E)-ene-11,13-diyn-2-one (1) and pentadeca-(9E)-ene-11,13-diyne-2,8-dione (3)) and three polyenes (namely, 8-hydroxy-pentadeca-(9E,13Z)-dien-11-yn-2-one (2), pentadeca-(9E,13Z)-dien-11-yne-2,8-dione (4) and pentadeca-(8Z,13Z)-dien-11-yn-2-one (5)), isolated from the n-hexane extract of E. pallida roots by bioassay-guided fractionation, were investigated and the potential bioavailability of these compounds in the extract was studied. EXPERIMENTAL APPROACH: Cytotoxic effects were assessed on human pancreatic MIA PaCa-2 and colonic COLO320 cancer cell lines. Cell viability was evaluated by the WST-1 assay and apoptotic cell death by the cytosolic internucleosomal DNA enrichment and the caspase 3/7 activity tests. Caco-2 cell monolayers were used to assess the potential bioavailability of the acetylenes. KEY RESULTS: The five compounds exhibited concentration-dependent cytotoxicity in both cell types, with a greater potency in the colonic cancer cells. Apoptotic cell death was found to be involved in the cytotoxic effect of the most active, compound 5. Compounds 2 and 5 were found to cross the Caco-2 monolayer with apparent permeabilities above 10 x 10(-6) cm s(-1). CONCLUSIONS AND IMPLICATIONS: Compounds isolated from n-hexane extracts of E. pallida roots have a direct cytotoxicity on cancer cells and good potential for absorption in humans when taken orally.

Torsadogenic cardiotoxicity of antipsychotic drugs: a structural feature, potentially involved in the interaction with cardiac HERG potassium channels.

Many non-cardiovascular drugs of common clinical use cause, as an unwanted accessory property, the prolongation of the cardiac repolarisation process, due to the block of the HERG (Human Ether-a-go-go Related Gene) potassium channel, responsible for the repolarising I(Kr) current. This delayed cardiac repolarisation process can be often unmasked by a prolongation of the QT interval of the ECG. In these conditions, premature action potentials can generate morphologically anomalous after-polarisations, and trigger a dangerous kind of polymorphic ventricular tachyarrhythmia, known as torsade de pointes, which can evolve in ventricular fibrillation and death. The risk associated with the torsadogenic cardiotoxicity of drugs, which prolong the QT interval has been the topic of documents produced by many health authorities, giving important issues about the preclinical and clinical evaluation of cardiac safety. Besides, public and private research laboratories developed several experimental in vitro or in vivo strategies, aimed to an early recognition of the influence of a drug (or of a drug-candidate) on the HERG channel and/or on the cardiac repolarisation process. Also the identification of a possible pharmacophore model, common in all or at least in numerous torsadogenic drugs, could represent a first step for the development of useful in silico approaches, allowing a preliminary indication about the potential torsadogenic property of a given molecule. In this work, we described the electrophysiological basis of torsade de pointes and listed several pharmacological classes of torsadogenic drugs. Among them, we focused our attention on antipsychotics, with an accurate overview on the experimental and clinical reports about their torsadogenic properties. Moreover, a common structural feature exhibited by these drugs, despite of their remarkable chemical differences, is evidenced by a computational approach and is indicated as a possible "facilitating" requirement for their torsadogenic properties. Together with other remarks, coming from different computational studies, the individuation of a satisfactory "toxicophore" model could be greatly useful, for the theoretical prediction of torsadogenic properties of a given chemical moiety and for the design of new drugs devoid of such an undesired and potentially lethal side-effect.

Analysis of the internal representations developed by neural networks for structures applied to quantitative structure--activity relationship studies of benzodiazepines.

An application of recursive cascade correlation (CC) neural networks to quantitative structure-activity relationship (QSAR) studies is presented, with emphasis on the study of the internal representations developed by the neural networks. Recursive CC is a neural network model recently proposed for the processing of structured data. It allows the direct handling of chemical compounds as labeled ordered directed graphs, and constitutes a novel approach to QSAR. The adopted representation of molecular structure captures, in a quite general and flexible way, significant topological aspects and chemical functionalities for each specific class of molecules showing a particular chemical reactivity or biological activity. A class of 1,4-benzodiazepin-2-ones is analyzed by the proposed approach. It compares favorably versus the traditional QSAR treatment based on equations. To show the ability of the model in capturing most of the structural features that account for the biological activity, the internal representations developed by the networks are analyzed by principal component analysis. This analysis shows that the networks are able to discover relevant structural features just on the basis of the association between the molecular morphology and the target property (affinity).

A 3D model for the human hepatic asialoglycoprotein receptor (ASGP-R).

The human hepatic Asialoglycoprotein Receptor (ASGP-R) consists of two different types of liver specific membrane glycoproteins that bind to terminal galactose and N-acetylgalactosamine residues of serum glycoproteins. The two different polypeptide chains are referred to as two receptor subunits, HH1 and HH2, which are both involved in the activity of the functional receptor. This receptor has served as a model for understanding receptor-mediated endocytosis and carbohydrate mediated recognition phenomena. Here models for the C-terminal extracellular region of both HH1 and HH2 subunit are presented. The standard homology building procedure was modified in order to make it suitable for the modeling problem at hand. The models for the extracellular regions of HH1 and HH2 were initially constructed by exploiting several fragments, belonging to proteins of known 3D structure, and showing high local sequence similarity with respect to the glycoproteins of interest. Putative binding sites were first hypothesized on the basis of the comparison with other complexes of lectins, the crystal structure of which was available in the Protein Data Bank. A model for the complex involving the HH2 subunit and the typical high affinity ligand N-acetylgalactosamine (NacGal) was refined as the first by a suitable combination of MD simulations and Energy Minimization calculations, since it seemed to quickly converge to a plausible structure. An intermediate model for HH1 was then rebuilt on the basis of the refined model for HH2. It was then submitted to a sequence of molecular dynamics simulations with templates which took into account the secondary structure prediction for a final refinement. The structures of small regions of the models, located around the binding sites, were compared with more recent crystallographic data regarding a complex involving the mutant of Mannose Binding Protein QPDWGH (1BCH entry in the Protein Data Bank) and NacGal. This mutant shows high local sequence similarity with HH1 and HH2 at the binding sites. On the basis of the above comparison, different locations of the binding sites were also considered. In addition to other expected interactions, two hydrophobic interactions were observed in the models with Trp residues (positions 243 in HH1 and 181 or 267 in HH2 respectively) and His residues (positions 256 in HHI and 184 in HH2.respectively). The quality of the models was evaluated by the Procheck program and they seemed plausible. This observation together with analogies found between binding sites of the models and IBCH supported the validity of the models. A further validation element arose by comparison between experimental binding data available in the literature about the homologous rat receptor subunits and theoretical interaction energies evaluated, by means of the DOCK 3.5 program, in models for the rat subunits obtained from the corresponding human ones. The new modeling procedure used here appears to be a well-suited method for structural analysis of small regions, located around the ligands, in proteins of unknown 3D structure.

Nuclear magnetic resonance structure of d(GCATATGATAG). d(CTATCATATGC): a consensus sequence for promoters recognized by sigma K RNA polymerase.

The three-dimensional structure of d(GCATATGATAG).d(CTATCATATGC), from the promoter region of a gene regulating sporulation in Bacillus subtilis mother cells, was determined utilizing two-dimensional nuclear Overhauser effect (2D NOE) and double-quantum-filtered COSY (2QF-COSY) spectra. To minimize the effect of methods used to obtain restraints and refine structure, several variables were studied. Interproton distance bounds were calculated very conservatively by running the complete relaxation matrix program MARDIGRAS hundreds of times using 2D NOE spectra for exchangeable and for nonexchangeable protons at different mixing times, assuming different overall correlation times and different starting structures. The 435 distance restraints were used with two different structural refinement methods: restrained molecular dynamics (rMD) and restrained Monte Carlo calculations (rMC). Refinement using different procedures and starting structures resulted in essentially the same structure (<0.8 A rmsd), indicating that the structure is defined by experimental restraints and not the refinement method or variables used. R factors indicate the structures fit the experimental NOE data very well. Some helical parameters, notably large negative X displacement, are characteristic of A-DNA, but others are characteristic of B-DNA. As with TG.CA steps in other duplex DNA sequences studied in our laboratory, the two TG.CA steps have a positive roll, with T6-G7 exhibiting the largest, and consequently a bent helix axis. The converged structure represents a time-averaged structure. However, multiple conformations, especially in deoxyriboses, were evident from vicinal coupling constants obtained from quantitative simulations of 2QF-COSY cross-peaks and from persistent inconsistencies in experimental distances due to nonlinear conformational averaging.

A 3D model of the human A1 adenosine receptor. An evaluation of the binding free-energy with ligands.

A model of A1 adenosine receptor was built on the basis of the prediction of transmembrane helices made by PHDtopology and forcing the rough initial model over the scaffold of the rhodopsin. Only helices were accurately modeled. Several complexes between the model of the A1 receptor and some ligands were built. The binding site was hypothesized on the basis of biochemical experiments (site directed mutagenesis). Ligands were selected so that their Kis range between millimolar to nanomolar. The validation of the model was carried out performing calculations of the binding free energy between ligands and the receptor model. The free energy calculations were accomplished by using the linear free energy approximation method (LIE). We could observe that the trend of the calculated delta delta Gs (differences in binding free energies between the antagonist 2, showing the lowest Ki, and the other antagonists analyzed) agreed with the one obtained from biological data.

Synthesis of 2-methylaminobenzimidazole derivatives tested for antiinflammatory activity.

A series of 1-alkyl substituted 2-methylaminobenzimidazole derivatives was prepared and tested. Some of them were assayed orally in the rat for antiinflammatory and analgesic properties. The compounds did not exhibit any significant activity compared with reference drug levels.

Synthesis and benzodiazepine receptor affinity of N-(indol-3-ylglyoxylyl)-dipeptide derivatives. Structural requirements for inverse agonist/antagonist receptor interactions.

Several N-(indol-3-ylglyoxylyl)dipeptide derivatives 1-12 were synthesized and tested for their affinity at the benzodiazepine receptor in bovine cortical membranes. They proved to bind with low or no affinity at the receptor site. It was hypothesized that this result was not due to the steric hindrance of the dipeptide side chain, but to the establishment of intramolecular hydrogen bonds involving the indole N-H and/or the glyoxylyl C = O(2). Conformational analysis indicated that coiled conformations, with intramolecular hydrogen bonds, were energetically more favoured than the staggered, completely unfolded ones. Therefore, the low or no affinity of these compounds should be attributed to the unavailability of the N-H and/or C = O(2) groups for the binding, again confirming that both these groups are necessary for interaction with the receptor.

Benzodiazepine receptor affinity and interaction of some N-(indol-3-ylglyoxylyl)amine derivatives.

Several derivatives, in which tryptamine, tyramine, and dopamine moieties are linked to the indole nucleus by an oxalyl bridge, were tested for their ability to displace the specific binding of [3H]flunitrazepam from bovine brain membranes. GABA ratio and in vivo tests for the most potent compounds showed they behave as inverse agonists at the benzodiazepine receptor (BzR). To better define the structure-activity relationship (SAR) of this kind of ligand, several phenylethylamine derivatives were synthesized to evaluate their affinity to BzR. Some of these derivatives (17, 21, 24, 26, and 30) were found to exhibit high affinity (Ki = 0.51-0.085 microM) for BzR and possessed a partial agonist activity, although their chemical structure is closely related to tryptamine 2-6, tyramine 7-11, and dopamine 12-16 derivatives. A different interaction of these ligands to the receptor site is hypothesized. Moreover, all the prepared 1-methyl derivatives exhibited very low binding affinity to BzR.

Solution structure of [d(GGTATACC)]2: wrinkled D structure of the TATA moiety.

Phase-sensitive two-dimensional nuclear Overhauser effect spectra of [d(GGTATACC)]2 in aqueous deuterium oxide solution at four mixing times were quantified to give all nonoverlapping cross-peak intensities. A structural model for [d(GGTATACC)]2 was built in which the GG- and -CC moieties were in the B-DNA form, while the middle -TATA- moiety was in the wrinkled-D form (BDB model). This model was subjected to energy refinement by molecular mechanics calculations with the program AMBER. Counterions (Na+) were added to neutralize the charges, and water molecules were placed bridging across the minor groove. A complete relaxation matrix analysis was used to calculate two-dimensional nuclear Overhauser effect spectra of [d(GGTATACC)]2 from the above models (before and after energy refinement) and from four other [d(GGTATACC)]2 structural models: regular A, crystalline A, regular B, and energy-minimized B. Among them, the energy-minimized BDB model yielded a set of theoretical spectra that gave the best fit to the experimental spectra. It was also the energetically most stable. Therefore, it is a good representation of the ensemble- and time-averaged structure of the octamer in solution. This model has backbone torsion angles similar to those of B-form DNA in the GG- and -CC moieties and torsion angles similar to those of wrinkled D form DNA in the -TATA- moiety. The base stacking and base pairing are not interrupted at the junctions between the two structural moieties. Its minor groove is narrower than that of B DNA, and the solvent-accessible surface of the minor groove forms a closed hydration tunnel in the middle -TATA- segment.

A conformational approach to histamine H2-receptor antagonists.

Four selective H2-antagonists were studied: burimamide, thiaburimamide, metiamide and cimetidine and their conformations were investigated using the PCILO method. The results show similar preferred conformations but also show that the most active drugs can more easily assume conformations different from the preferred ones. The selective H2-agonist dimaprit was also examined in order to find conformational features determining activity at the H2-histamine receptor.