Cancer Cell's Achilles Heels: Considerations for Design of Anti-Cancer Drug Combinations

Abstract

Loss of function screens using shRNA (short hairpin RNA) and CRISPR (clustered regularly interspaced short palindromic repeats) are routinely used to identify genes that modulate responses of tumor cells to anti-cancer drugs. Here, by integrating GSEA (Gene Set Enrichment Analysis) and CMAP (Connectivity Map) analyses of multiple published shRNA screens, we identified a core set of pathways that affect responses to multiple drugs with diverse mechanisms of action. This suggests that these pathways represent "weak points" or "Achilles heels", whose mild disturbance should make cancer cells vulnerable to a variety of treatments. These "weak points" include proteasome, protein synthesis, RNA splicing, RNA synthesis, cell cycle, Akt-mTOR, and tight junction-related pathways. Therefore, inhibitors of these pathways are expected to sensitize cancer cells to a variety of drugs. This hypothesis was tested by analyzing the diversity of drugs that synergize with FDA-approved inhibitors of the proteasome, RNA synthesis, and Akt-mTOR pathways. Indeed, the quantitative evaluation indicates that inhibitors of any of these signaling pathways can synergize with a more diverse set of pharmaceuticals, compared to compounds inhibiting targets distinct from the "weak points" pathways. Our findings described here imply that inhibitors of the "weak points" pathways should be considered as primary candidates in a search for synergistic drug combinations.

Keywords: CMAP; GSEA; drug combination; shRNA screening; signaling pathway; synergy.

Figure 1

(A) Radar plot representing GSEA analysis of sensitizing hits from the shRNA screens. Each blue dotted line represents a screen. The red dots mark the number of screens that contain indicated pathways. FDR values were below 0.25, which is a standard for generating a hypothesis [23]. (B) A complementary analysis of shRNA screen hits. Genes common for at least two screens were identified followed by the GSEA analysis. Radar plot representing GSEA analysis.

Figure 2

Finding synergistic drugs. (A) The graphic represents an overall summary of the analysis of synergistic drugs against “weak points” and regular pathways. (B) The MDS plot demonstrates the diversity of drugs that synergize with a “weak point” pathway inhibitor bortezomib and a “control therapeutics” Top1 inhibitor irinotecan. The area covered by drug effects-representing triangles reflects the diversity of the drugs’ actions. (C) Bar plots represent the quantitative diversity of drugs that synergize with inhibitors of “weak points” pathways and control therapeutics. The left plot shows an average of drug diversity calculated based on Euclidian distances between the drugs on the MDS plots. The right plot represents a diversity of drugs synergizing with individual therapeutics. The table shows diversity scores for drugs synergizing with each of the therapeutics. ***—p-value < 0.001. Statistical analysis was conducted using Student’s t-test by considering 4 drugs per group (n = 4).