Knowing the cellular targets of drugs is crucial if the process of drug discovery is to be made more efficient. Identifying the full spectrum of targets associated with a bioactive small molecule can lead to faster optimization, understanding of off-target side effects and the ability to minimize possible toxicities early on in the process.
Whether it is an FDA-approved drug with an unknown mechanism of action, a natural product or a small molecule derived from a cell based phenotypic or biochemical screen the need to identify mechanism of action becomes a priority. We have developed a robust and unbiased method of probing the proteins that bind to the small molecule of interest in a biologically relevant setting. It combines quantitative proteomics (SILAC) with biochemical affinity enrichment1.
Getting set up:
Generation of the affinity reagent (bait) is a critical first step, which involves tethering the small molecule of interest to a bead. This is generally done in a close collaboration with the Chemical Biology Platform team. If there are sufficient structure-activity relationship (SAR) data available, a tether is placed at an unimportant position. If, however, there is no SAR, a position that can be modified without compromising activity must be found. In either case, the modified compound must then be tested in the original system to ensure that the desired activity is retained. Once activity of the tethered compound has been confirmed it is attached to a solid support and the affinity reagent can be used in a pull-down studies.
While generation of the affinity reagent is ongoing the Target ID Proteomics group will test the cell line of choice for SILAC labeling efficiency and best media formulation.
The Quantitative Proteomic Experiment:
We use SILAC labeled lysates in pull down experiments where the bait is the small-molecule immobilized on an affinity matrix. Both lysates are incubated with small-molecule loaded beads, but excess free small molecule is added to one lysate to competitively displace target proteins. The beads from both lysates are then washed, combined and proteins that remain bound to the immobilized small molecule are eluted off. Eluted proteins are separated by molecular weight using gel electrophoresis. The entire gel lane is then cut into bands, trypsin digested and analyzed on a mass spectrometer (GeLCMS). We identify and compare relative enrichment of target proteins generating differential SILAC ratios between the two states. Routinely the data is analyzed using MaxQuant a quantitative proteomics software package the output of which contains the differential ratios.
The protein ratios obtained by MaxQuant are modeled using an empirical Bayes strategy2 developed within the Broad to give statistical significance to the ratios. The final result is a comprehensive rank ordered list of protein binders to the small molecule.
The “Target ID problem”:
Identifying the mechanism of action of a small molecule is not a task that can be easily achieved by a single method. Recognizing this, the Broad has undertaken an interdisciplinary approach in the context of genome-based drug discovery.
Early on in a project the RNAi platform and the Cancer and Chemical Biology programs are involved such that data acquired using these complementary approaches can be integrated using bioinformatics analyses and be leveraged to formulate a hypothesis on mechanism of action.
1. Ong, S. E., M. Schenone, et al. (2009) "Identifying the proteins to which small-molecule probes and drugs bind in cells." Proc Natl Acad Sci U S A 106:4617-22. Abstract
2. Margolin, A. A., S. E. Ong, et al. (2009) "Empirical Bayes analysis of quantitative proteomics experiments." PLoS One 4:e7454. Abstract