A field based 3D QSAR model of novel anti-microtubule agent noscapine and its derivatives

Abstract

BACKGROUD & OBJECTIVE: Noscapinoids are a new class of microtubule binding compounds which show great
promise as chemotherapeutic agents for the treatment of human cancers. To investigate the structural determinants of
noscapinoids responsible for anti-cancer activity in order to design more potent derivatives, attempts were made to
developed a 3D QSAR model based on “field point” descriptors using Forge V10 software(Cresset group).
METHODOLOGY: We have used 53 structurally diverse noscapinoids in a single panel and experimentally determined
their IC50 value using human acute lymphoblastic leukemia cells (IC50 values vary from 1.2 to 56.0 μM). Molecular
models of these compounds were built,energy minimized and geometry optimized. The data set was randomly divided
into 43 training and 10 test set molecules. Amino noscapine was considered as template molecule for the calculation of
hydrophobic, steric, electrostatic and volume field points. These field based descriptors were used to align the training set
molecules with the template molecule.A Partial Least Square (PLS) model was built based on field points using sphere
exclusion algorithm. RESULTS:A statistically significant model (Rtrain2 = 0.884; R2LOO = 0.875) was obtained with the
field point descriptors. The robustness of the QSAR model was characterized by the values of the internal leave-one-out
cross-validated regression coefficient (R2LOO) for the training set and Rtest2 for the test set. The overall root mean square
error (RMSE) between the experimental and predicted IC50 value was 1.75 and Rtest2 = 0.713, revealing good
predictability of the QSAR model. The 3D QSAR model developed in this study shall aid further design of novel potent
noscapine derivatives.