A QSAR study of macrocyclic diterpenes with P-gp inhibitory activity isolated from Euphorbia species

A set of 50 compounds, constituted mainly by macrocyclic diterpenes of the jatrophane and lathyrane-type, isolated from Euphorbia species, was used in the present study. These compounds were found to be P-glycoprotein-mediated multidrug-resistance (MDR) reversing agents in cancer cells. The reversal of MDR was investigated by flow cytometry, at two different concentrations, 4 and 40µg/mL, by measuring rhodamine-123 accumulation, a fluorescent substrate analogue of doxorubicin [1, 2]. The quantitative structure-activity relationship (QSAR) methodology was applied to the whole set of diterpenes and to the jatrophanes subset at both concentrations studied. We used multiple linear regression (MLR) with forward features (from five classes of molecular descriptors: constitutional, topological, geometrical, electrostatic and quantum-chemical) selection to establish QSAR models. The best model obtained, at 4µg/mL, for the diterpenes set was constructed using 4 descriptors with R² of 0.80 for a training set of 40 compounds and R² _pred of 0.71 for a test set of 9 compounds. The best model obtained, at 4µg/mL, for jatrophanes subset was constructed using 5 descriptors with R² of 0.82 for a training set of 28 compounds and R² _pred of 0.54 for a test set of 6 compounds. All models were statistically valid with high predictability, and in both models the descriptors used (logP and quantum descriptor related to resonance energy) can easily be translated into the design of novel diterpene derivatives with a forecasted improved P-gp inhibitory activity.

Keywords: QSAR, P-gp inhibitory activity, macrocyclic diterpenes

Acknowledgement: This study was supported by FCT (PTDC/QUI-QUI/099815/2008), Portugal.

References: 1. Molnár J et al. (2006) Curr Pharm Des 12: 287–311.

2. Duarte N et al. (2008) Bioorg Med Chem 16: 9323–30.