Elucidating the role of retroactive signaling and kinase inhibitors on off-target drug effects

MICHELLE WYNN; ALEJANDRA C VENTURA; HECTOR GARCIA; JACQUES-A. SEPULCHRE; SOFIA D MERAJVER

Conferencia; 2011 AACR Annual Meeting; 2011

The primary objective of targeted cancer therapies is to modulate cancer progression by perturbing specific molecules involved inaberrant proliferation and invasion. Kinase inhibitors are targeted therapies which are designed to interfere with a specific kinasemolecule in a dysregulated oncogenic signaling cascade. While extremely promising as anti-cancer agents, such inhibitors may haveundesirable off-target effects, whether by non-specific interactions or by effects from pathway cross-talk. We have shown inpublished experimental and theoretical work that covalently modified signaling cascades naturally exhibit bidirectional signalpropagation. This phenomenon is termed retroactivity and challenges the widespread notion that information in cascades only flowsfrom the cell surface to the nucleus. Previous work has demonstrated that increasing the concentration of a phosphatase in theterminal cycle of a covalently modified cascade may result in a measurable decrease in the concentration of the previous cycle’sactivated kinase. Thus, a downstream perturbation in a signaling cascade can produce a reverse (or retroactive) response withoutthe need for direct negative feedback connections. This led us to hypothesize that the use of an inhibitory drug in a signalingnetwork may cause an upstream off-target effect simply by inhibiting the activation or deactivation of a downstream kinase. To testthe hypothesis that retroactivity contributes to off-target effects, we extended our previous work to a computational model thattested a series of signaling networks. The objective of our approach was two-fold: (1) to probe the effect of retroactivity on a kinaseinhibitor in a signaling network and (2) to test whether retroactivity is likely to produce a measurable off-target effect underphysiologically realistic conditions. Specifically, our model simulates the targeted inhibition of an activated kinase in a series ofmulti-cycle networks. The results of our work indicate that at physiologically and therapeutically relevant concentrations, a targetedinhibitor may induce a measurable off-target effect via retroactivity. We also performed local sensitivity analyses to predict thekinetic parameters that most affect the off-target response. Surprisingly, the drug disassociation constant is predicted to have verylittle effect while parameters such as the enzyme saturation and maximum velocity of some cycles are predicted to be veryimportant. A proper characterization of a pathway’s structure is important for identifying which protein in the pathway representsthe optimal drug target as well as what concentration of the targeted therapy is likely to modulate the pathway in the mannerdesired. We believe our results support the position that such characterizations should consider the role of retroactivity as a sourceof a potential off-target effects by kinase inhibitors.