Computational modeling of pathway dynamics for detecting drug effects: paradoxical effects of LYS303511 on TRAIL-induced apoptosis

Yuan Shi, Simi M. Varghese, Sinong Huang, Jacob White, Shazib Pervaiz, Lisa Tucker-Kellogg*

Singapore-MIT Alliance, E4-04-10, 4 Engineering Drive 3, Singapore 117576, Singapore. tucker@comp.nus.edu.sg

Proc LSS Comput Syst Bioinform Conf. August, 2009. Vol. 8, p. 213-224. Full-Text PDF

*To whom correspondence should be addressed.

Based on previous work showing that the drug LY303511 sensitizes cancer cell lines to TRAIL-induced apoptosis, we used computational modeling and ordinary differential equations to represent how each agent individually affects the system. We then simulated the combined effects of both agents together, and found that the level of death in silico correlated well with experimental observations, including the synergistic (greater than additive) level of death from the combination treatment. However, simulation and experimental data exhibited numerous mismatches for some molecules upstream in the pathway, at time points shortly after treatment, such as differences in the slope and onset of the synergistic activation of caspase-8. Additional measurements of caspase-8 revealed a paradoxical caspase-inhibitory effect of LY303511 at early time points after treatment. This recent finding is seemingly at odds with the long-term effects of LY303511 towards promoting cell death. Moreover, LY303511 treatment was found to cause up-regulation of the protein cFLIP-s within one hour, followed by a decrease relative to untreated. Because cFLIP-s has often been found to antagonize caspase-8 activation, our cFLIP-s results suggest a possible explanation for the early caspase-inhibitory effect of LY303511. Further work remains, but the mathematical modeling approach used here may be of general use for identifying unknown effects of any drug (or perturbation), provided the drug is acting on a well-understood molecular pathway.