Functional Identification of Target by Expression Proteomics (FITExP) reveals protein targets and highlights mechanisms of action of small molecule drugs

Abstract

Phenomenological screening of small molecule libraries for anticancer activity yields potentially interesting candidate molecules, with a bottleneck in the determination of drug targets and the mechanism of anticancer action. We have found that, for the protein target of a small-molecule drug, the abundance change in late apoptosis is exceptional compared to the expectations based on the abundances of co-regulated proteins. Based on this finding, a novel method to drug target deconvolution is proposed. In a proof of principle experiment, the method yielded known targets of several common anticancer agents among a few (often, just one) likely candidates identified in an unbiased way from cellular proteome comprising more than 4,000 proteins. A validation experiment with a different set of cells and drugs confirmed the findings. As an additional benefit, mapping most specifically regulated proteins on known protein networks highlighted the mechanism of drug action. The new method, if proven to be general, can significantly shorten drug target identification, and thus facilitate the emergence of novel anticancer treatments.

Figure 1. Distributions of the prediction errors of protein regulations based on co-regulated proteins reveal unexpectedly large regulation of drug targets (3σ area is shaded)

.

Figure 2. General workflow of the FITExP method of drug target identification:

(a) a panel of cell lines is treated by a panel of drugs, in biological triplicates; (b) LC-MS/MS based proteomics identifies and quantifies ≥3,500 proteins, proteomic profiles are shown in a schematic heatmap with color-coded normalized abundances; the dendrogram shows hierarchical clustering of proteomic profiles with correlation-based distances; (c) for each protein, cell line and treatment, regulation Reg, specificity Spec and exceptionality Exc are calculated; (d) for each treatment, final protein ranks based on Reg and Exc are established and the p-values are calculated using Bonferroni correction; protein list is sorted in ascending order of p-values; (e) few proteins with p ≤ 0.05 (threshold p-value) represent the most likely drug targets; (f) top n proteins with p ≤ 0.05 according to Reg and Spec rankings are mapped on protein networks to identify the drug target mechanism.

Figure 3. Protein-protein interaction networks obtained with STRING, for proteins with rank-product p ≤ 0.05 according to Reg and Spec rankings for

(a) 5-FU, (b) TDX, (c) MTX, (d) PCTL and (e) DOXO treatments.