ABSTRACT
222 Algorithms are proposed for classification and taxonomy, based on criteria as information entropy and its production. The 32 2-phenylindole-3-carb- aldehydes (PICs) present inhibition of breast cancer cells MDA-MB231 and MCF-7. On the basis of PIC structure-activity relationship (SAR) new derivatives are designed. The PICs are classified using five characteristic chemical properties of different moieties. Many classification algorithms are based on information entropy. When applying procedures to sets of moderate size, an excessive number of results appear compatible with data and suffer a combinatorial explosion. However, after equipartition conjecture one has a selection criterion among different variants, resulting from classification between hierarchical trees. Information entropy permits classifying compounds and agrees with principal component analysis. Features denote positions R1-2 on indole, R3-4 on phenyl and R5 on carbaldehyde. A periodic table of MB231/MCF-7 inhibitors is obtained. Inhibitors in the same group and period are suggested to present maximum similarity; the ones with the same group will present important resemblance. Indoles are attractive as inhibitors of tubulin polymerization. A number of PICs, with lipophilic substituents in both aromatic rings, were synthesized and evaluated for MB231/MCF-7 antitumor activity. Some 5-alkylindoles, with a 4-methoxy group in 2-phenyl ring, strongly inhibit MB231/MCF-7 growth with 50% inhibitory concentration (IC50) 5-20 nm. The action can be rationalized by the cell cycle arrest in phase G2/M, because of tubulin polymerization inhibition. Quantitative SAR (QSAR) was done on PICs to find out structural requirements, for more active antimitotic agents.
