Jadomycin B, an Aurora-B kinase inhibitor discovered through virtual screening

Abstract

Aurora kinases have emerged as promising targets for cancer therapy because of their critical role in mitosis. These kinases are well-conserved in all eukaryotes, and IPL1 gene encodes the single Aurora kinase in budding yeast. In a virtual screening attempt, 22 compounds were identified from nearly 15,000 microbial natural products as potential small-molecular inhibitors of human Aurora-B kinase. One compound, Jadomycin B, inhibits the growth of ipl1-321 temperature-sensitive mutant more dramatically than wild-type yeast cells, raising the possibility that this compound is an Aurora kinase inhibitor. Further in vitro biochemical assay using purified recombinant human Aurora-B kinase shows that Jadomycin B inhibits Aurora-B activity in a dose-dependent manner. Our results also indicate that Jadomycin B competes with ATP for the kinase domain, which is consistent with our docking prediction. Like other Aurora kinase inhibitors, Jadomycin B blocks the phosphorylation of histone H3 on Ser10 in vivo. We also present evidence suggesting that Jadomycin B induces apoptosis in tumor cells without obvious effects on cell cycle. All the results indicate that Jadomycin B is a new Aurora-B kinase inhibitor worthy of further investigation. [Mol Cancer Ther 2008;7(8):2386-93]

Structural studies of B-type Aurora kinase inhibitors using computational methods

“…A series of 48 quinazoline derivatives with their inhibitory activities was retrieved from the literature [24,32] . We have considered only the IC 50 value for the cell-based assay (not an enzyme-based assay), which is the inhibition of histone H3 in SW620 cells. Histone H3 is a cellular substrate of Aurora B and a marker of Aurora B kinase inhibition in vivo.…”

“…Histone H3 is a cellular substrate of Aurora B and a marker of Aurora B kinase inhibition in vivo. The IC 50 values in the micro-molar (µmol/L) range were converted into the molar (M) range and then into pIC 50 values using the relationship pIC 50 =-logIC 50 values. The pIC 50 values were used as the dependent variables in the statistical modeling.…”

“…The IC 50 values in the micro-molar (µmol/L) range were converted into the molar (M) range and then into pIC 50 values using the relationship pIC 50 =-logIC 50 values. The pIC 50 values were used as the dependent variables in the statistical modeling. The dataset of Aurora B kinase inhibitors was divided into training (37 compounds) and test (11 compounds) sets considering the fact that the test set molecules must present a range of biological activity and the typical chemical structures similar to those of the training set as reported in Table 1.…”

“…Partial Least Square (PLS) analysis and validation of the QSAR models To derive the 3D-QSAR models, the CoMFA and CoMSIA descriptors were used as independent variables and the pIC 50 values as the dependent variable. The PLS method [37,38] was used to linearly correlate the CoMFA and CoMSIA descriptors to the activity.…”

“…where, y pred , y actual , and y mean are the predicted, actual and mean values of the target property (pIC 50 ), respectively, and PRESS is the sum of the predictive sum of squares. The non-crossvalidated PLS analyses were performed with a column filter value of 2.0 to reduce the analysis time with a small effect on the q 2 values.…”