engleski

QSAR in prediction of target and antitarget drug properties

Drug discovery has entered a new century with a wealth of sophisticated technologies and information generation platforms that theoretically will allow the more rapid development of medicines with improved selectivity and safety profiles. The most used, the quantitative structure activity relationship (QSAR) provides a great deal of information regarding the nature of investigated NCEs and so far many approaches were developed. In an extension of our earlier studies in attempting the role of the use of topological indices (TIs) for QSAR studies in lead compound search among 1, 4- (I) and 3, 1-benzoxazinones (II) ((n=62) we have undertaken the present investigation in that we have used Wiener (W) and molecular connectivity ( 1,  MOD, CID) indices, molecular descriptors (MDs), the lipophilicity parameter, log P (AlogP and miLogP) and van der Waals volume (Sv) in correlations with antimicrobial activity (MICx10-3moldm-3) obtained in tests with various microorganisms, including bacteria and fungi. The Lipinski's rule-of-five study was also undertaken with the aim to explore, the potential oral bioavailability and druglikeness properties of these compounds, in order to lead identification with the lowest potential toxophoric features. For this purpose the 'n toxophoric scores' were calculated and used in correlations with data computed using Molinspiration software for druglikeness (GPCR ligand, ion channel modulator, kinase inhibitor, nuclear receptor ligand), lipophylicity parameter, miLogP and molecular polar surface area (TPSA). Computed data for efavirenz and its 1, 4- isomer were also included in this study. All compounds I and II were previously synthesized and biologically tested. The TIs (W, χ 1, χ MOD and CID) and MDs (Sv, Mv and AlogP) were computed using Dragon for Windows, 2007. The Lipinski's rule-of-five analysis and parameters (miLogP, MW, TPSA, n atoms, nON, nOH NH, nrotab, volume) as well as druglikeness scores were computed using Molinspiration molecular processing engine, 2007.10. The 'n toxophoric score' for each compound was calculated on the basis of presence one or more main toxophoric groups in molecule. All analyses were performed and data were graphed using OriginPro 7.5 software (Origin Laboratories, Northampton, MA). RESULTS AND DISCUSSION QSAR study In present QSAR study W, χ 1, χ MOD and CID were correlated with AlogP, Sv and MICs (x10-3 moldm-3) for Staphilococcus aureus ATCCC 56511, Klebsiella pneumoniae ATCC 10031 and Candida monosa. The 2D correlations with similar significancy between TIs (W, χ 1, χ MOD) and Sv versus MICx10-3moldm-3 against S. aureus, were obtained and they are as follows: i. MIC x 10-3 (mol dm-3) = 4.6115 – 0.00125 W (n = 20, r = - 0.8627) ; ii. MIC x 10-3 (mol dm-3) = 7.35675 – 0.39864 χ 1 (n = 20, r = - 0.8952) ; iii. MIC x 10-3 (mol dm-3) = 7.35268 – 0.06072 χ MOD (n = 20, r = - 0.88495) ; and iv. MIC x 10-3 (mol dm-3) = 6.87874 – 0.16685 Sv (n = 20, r = - 0.86517). The lipophilicity parameter, AlogP were used in 2D and 3D correlations. The 3D multiparameter correlations offered an additional insight in relationships among TIs, lipophilicity and MIC. QSAR results revealed that with an increasing of W, χ 1 and AlogP, the MIC is decreasing for this set of compounds. The 2D correlations with the same TIs (W, χ 1, χ MOD and CID) and Sv versus MIC against K. pneumoniae and C. monosa, showed comparatively less colinearities between investigated parameters. Lipinski's rule-of-five analysis Lipinski's rule-of-five predict poor absorption or permeation of orally administered compound if it meets the following criteria: molecular mass greater than 500 Da, high lipophilicity (expressed as cLogP greater than 5), more than 5 hydrogen bond donors and more than 10 hydrogen bond acceptors. Lipinski's rule-of-five analysis was applied on a set (n = 64) of I and II derivatives, including efavirenz and its 1, 4- isomer. Eight compounds out of 64 did not meet Lipinski's rule-of-five requirements (18, 19, 23, 27, 28, 30, 35 and 56), mostly because of TPSA and number of H-bond acceptors in molecule. In addition to druglikeness data estimated using various molecular properties, the 'n toxophoric score' were calculated on the basis of presence one or more main toxophoric groups in molecule. In correlations calculated n toxophoric scores with estimated druglikeness scores, the miLogP and TPSA, the 6 subgroups with relative and potential toxicity from 2 to 7 were obtained. The 'n toxophoric score' have to be considered as relative value of indication for possible toxicity. By this correlations only relative position of each compound in particular subgroup and in relation to efavirenz for the particular druglikeness score, estimated miLogP and TPSA could be evaluated. Fig. 1. 2D and 3D correlations of W versus AlogP and MIC (Staphylococcus aureus ATCC 56511). The results of QSAR study could be used in predicting more active analogues of this series of compounds. The Lipinski's rules-of-five analysis revealed valuable information relevant to potential drug candidate properties. The 'n toxophoric score' have to be more sofisticated for the prediction of potential toxicity of NCEs. References 1.Höltje HD, Sippl W, Rognan D, Folkers G, Moelcular modeling: Basic principles and applications, Wiley-VCH Weinheim, DE, 2008. 2.Vaz RJ, Klabunde T, Antitargets: Prediction and prevention of drug side effects, Vol. 38, Wiley-VCH, Weinheim, DE, 2008. 3.Lipinski CA, Lombardo F, Dominy BW, Feeney PJ, Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings, Adv.Drug.Delivery Rev. 1997 ; 23: 4-25. 4. Jadrijević-Mladar Takač M., Kos I., Takač V., QSAR studies on antimicrobial activity of benzoxazindione analogues, 68th International Congress of FIP. Abstracts Book. Basel, CH, September 2008.

QSAR; 3; 1- and 1; 4-benzoxazinones; efavirenz; DIBOA; 2D and 3D correlations; molecular descriptors; topological descriptors; Wiener index; Randić connectivity index; constitutional descriptors; number of H-bond acceptors; number of H-bond donors; molecular volume; lipophyilicity; number of toxophoric scores; Lipinski's rule-of-five analysis; molecular polar surface area; druglikeness properties; GPCR ligand; nuclear receptor ligand; kinase inhibitor; ion channel modulator

Sažetak usmenog izlaganja objavljen je u European Journal of Pharmaceutical Sciences 38 (2009) (S1) (ISSN 0928-0987).