A quantum chemical and chemometrical study of indolo[2,1-b]quinazoline and their analogues with cytotoxic activity against breast cancer cells

• Published in: SAR and QSAR in environmental research Vol. 20; no. 5-6; pp. 537 - 549
• Main Authors: Camargo, L.T.F.M., Sena, M.M., Camargo, A.J.
• Format: Journal Article
• Language: English
• Published: England Taylor & Francis Group 01.07.2009
• Subjects: anticancer activity; Antineoplastic Agents - chemistry; Antineoplastic Agents - pharmacology; Breast Neoplasms - drug therapy; Cell Line, Tumor; Computer Simulation; Female; Humans; Indoles - chemistry; Indoles - pharmacology; Inhibitory Concentration 50; PCA; PLS-DA; Principal Component Analysis; Quantitative Structure-Activity Relationship; Quinazolines - chemistry; Quinazolines - pharmacology; structure activity relationship; tryptanthrin
• ISSN: 1062-936X; 1029-046X
• DOI: 10.1080/10629360903278800

Some indolo[2,1-b]quinalozine (tryptanthrin) analogues present cytotoxic activity against human breast cancer cells. In this work, chemometric methods were applied in the search for building discriminant models between active and inactive analogues, based on the correlations among their in vitro cytotoxic activities and their electronic and geometric molecular descriptors. From 88 descriptors calculated with density functional theory with the exchange correlation functional B3LYP and the basis set 6-31G* (Gaussian 03), 29 were pre-selected based on their Fisher weights, and finally five descriptors (partial charge on atom 15, bond orders between atoms 12-13, 17-25 and 18-26, and energy difference between frontier molecular orbitals) were selected for principal component analysis. This analysis was able to discriminate 12 inactive from 22 active analogues by using only one principal component, accounting for 49% of the total variance and allowing us to better understand the influence of these electronic descriptors in the cytotoxic activity. In addition, a supervised partial least-squares discriminant model was build and successfully used to discriminate tryptanthrin analogues. The model was validated through an independent test set and considered robust to overfitting.