Topological pattern for the search of new active drugs against methicillin resistant Staphylococcus aureus.

Molecular topology was used to develop a mathematical model capable of classifying compounds according to antimicrobial activity against methicillin resistant Staphylococcus aureus (MRSA). Topological indices were used as structural descriptors and their relation to antimicrobial activity was determined by using linear discriminant analysis. This topological model establishes new structure activity relationships which show that the presence of cyclopropyl, chlorine and ramification pairs at a distance of two bonds favor this activity, while the presence of tertiary amines decreases it. This model was applied to a combinatorial library of a thousand and one 6-fluoroquinolones, from which 117 theoretical active molecules were obtained. The compound 10 and five new quinolones were tested against MRSA. They all showed some activity against MRSA, although compounds 6, 8 and 9 showed anti-MRSA activity similar to ciprofloxacin. This model was also applied to 263 theoretical antibacterial agents described by us in a previous work, from which 34 were predicted as theoretically active. Anti-MRSA activity was found bibliographically in 9 of them (ensuring at least 26% of success), and from the rest, 3 compounds were randomly chosen and tested, finding mitomycin C to be more active than ciprofloxacin. The results demonstrate the utility of the molecular topology approaches for identifying new drugs active against MRSA.

Population modelling to describe pharmacokinetics of amiodarone in rats: relevance of plasma protein and tissue depot binding.

The objective of this paper was to characterize the disposition phase of AM in rats, after different high doses and modalities of i.v. administration. Three fitting programs, WINNONLIN, ADAPT II and NONMEM were employed. The two-stage fitting methods led to different results, none of which can adequately explain amiodarone's behaviour, although a great amount of data per subject is available. The non-linear mixed effect modelling approach allows satisfactory estimation of population pharmacokinetic parameters, and their respective variability. The best model to define the AM pharmacokinetic profile is a two-compartment model, with saturable and dynamic plasma protein binding and linear tissular depot dynamic binding. These results indicate that peripheral tissues act as depots, causing an important fall in AM plasma levels in the first moment after dosing. Later, the return of the drug from these depots causes a slow increase in serum concentration whenever the dose is reduced.

Application of a mathematical topological pattern of antihistaminic activity for the selection of new drug candidates and pharmacology assays.

Molecular topology was used to achieve a mathematical model capable of classifying compounds according to their antihistaminic activity and low sedative effects. By application of this model of activity to databases containing chemical reagents and drugs exhibiting other pharmacological activity, we selected 30 compounds with possible antihistaminic activity. After those with possible sedative effects were discarded, activity tests were performed with five chemical reagents and three drugs searching for in vivo antihistaminic activity. The obtained results indicate that compounds such as 4-[(E)-2-(1,3-benzothiazol-2-yl)vinyl]-N,N-dimethylaniline (AH2), 2-ethyl-9,10-dimethoxyanthracene (AH4), and 2,4-bis(alpha,alpha-dimethylbenzyl)phenol (AH5) showed antihistaminic activity above terfenadine, the reference drug, whereas others, for instance, pergolide, miconazole, trihexyphenidyl, 2-(dibenzylamino-3-phenyl-1-propanol (AH1), and N-benzylquininium chloride (AH3), were less active than terfenadine.