(A) systems biological study of cell-fate decision processes based on the dynamics of cancer cell line-specific networks

Abstract

Cancer is a complex disease associated with various genomic mutations that influence the dynamics of molecular interaction network, leading to heterogenous properties of cancer cells. The inherent heterogeneous properties of cancer often result in variable anti-cancer drug resistance. One of the essential tumor suppressors, p53 has been an attractive cancer therapeutic target and also considered to be a major determinant of such drug resistance Despite the pivotal role of p53 and the recent progress of targeted cancer therapy, its regulatory mechanism and the underlying mechanism of variable drug resistance remains unclear and need to be investigated. To tackle this problem in view of p53 regulation, in our first study we used Boolean network model-ing and attractor landscape analysis to analyze the state transition dynamics of a simplified p53 network for which particular combinations of activation states of the molecules corresponded to specific cellular out-comes. Our results identified five critical interactions in the network that determined the cellular response to DNA damage, and simulations lacking any of these interactions produced states associated with sustained p53 activity, which corresponded to a cell death response. Attractor landscape analysis of the cellular response to DNA damage of the breast cancer cell line MCF7 and the effect of the Mdm2 inhibitor nutlin-3 indicated that nutlin-3 would exhibit limited efficacy in triggering cell death. It also suggested that combining nutlin-3 with inhibition of Wip1 would synergize to stimulate a sustained increase in p53 activity and promote p53-mediated cell death. We validated this synergistic effect in stimulating p53 activity and triggering cell death with single-cell imaging of a fluorescent p53 reporter in MCF7 cells. Thus, attractor landscape analysis of p53 network dynamics can identify potential therapeutic strategies for treating cancer. In our second study, we have reconstructed differential p53 networks by mapping p53 network from cancer genomics data and analyzed their state transition dynamics for various anti-cancer drug treatments. We could categorize heterogeneous cancer cell types into three groups: sensitive, moderate, or resistant to drugs. For each group, we found distinct p53 network dynamics on the basis of attractor landscape analysis. From combinatorial perturbation analysis, we identified an optimal drug combination that can maximize p53-mediated cell death regardless of heterogeneous cancer cell types. Our study shows that the attractor landscape analysis of p53 network dynamics can unravel the hidden mechanism underlying individual variation in drug responses and provide a new therapeutic strategy that can overcome variable drug resistance.