Study of the Effects of Basic Di‐ and Tri‐phenyl Derivatives on Malignant Cell Proliferation: An Example of the Application of Correspondence Factor Analysis to Structure‐Activity Relationships (SAR)

Abstract

AbstractThe descriptive multivariate method known as Correspondence Factor Analysis (CFA) was used to establish correlations between the structures of three chemical classes of compounds (triphenyl‐acrylonitriles (TPEs), diphenylethylenes (DPEs), and diphenylalkyls) substituted in the para position by either hydroxy or basic groups and their responses in a battery of three biochemical tests, namely the induction of the proliferation of the MCF7 human breast cancer cell‐line, the estrogen‐irreversible inhibition of MCF7 cell proliferation (herein denoted cytotoxicity), and binding to the estrogen receptor (ER). The power of CFA was illustrated by performing several analyses: (a) Construction of factorial maps that described only the specificity of the response of the TPE population in the tests or both the specificity and amplitude of the response; (b) Use of the factorial maps as mathematical models for the introduction of new variables. These variables were either further biochemical tests (cytotoxicity under different conditions, inhibition of the activation of protein kinase C) on which the TPE population had been screened or further compounds (DPEs and diphenylalkyls). Relationships among the different tests were thus assessed as well as affiliations of the new compounds with TPEs. The analyses revealed the importance of the presence and configuration of hydroxy groups in ER binding and cell proliferation, but also the ability of non‐hydroxylated compounds to induce cell growth independently of their relative affinity for ER. Cytotoxicity could be related to the presence of basic groups but also to resonance of conjugated bis‐para‐hydroxy diphenyl derivatives. Overall, the analyses stressed the complexity of the relationships between growth‐promoting and growth‐inhibitory potential of the test‐compound populations and suggested the involvement of multiple mechanisms of action.