Target deconvolution—no longer a needle in the haystack
In recent years there has been a resurgence in the use of phenotypic-based screening for drug discovery, largely driven by the increasingly sophisticated cell models and assay readouts available to drug discovery scientists. There are many advantages of this approach which include the potential to identify novel druggable proteins and the fact that active compounds are already proven to work in a cellular setting. However, a major challenge of this drug development strategy is the subsequent need to identify the target of the drug molecule and elucidate a mechanism of action.
Whilst there is no regulatory requirement to know the molecular target of a new molecular entity (NME), having this knowledge can prove invaluable in the clinical progression of a compound, or the advancement of subsequent chemical series. Defining your on- and off- targets allows monitoring of efficacy and safety during the lead optimization phase often making this a key step in the drug development pipeline. Discovering the mechanism by which your compound works, or target deconvolution, has typically been seen as a time- and resource-intensive process, almost a needle in the haystack search. In recent years though the number of available strategies for target deconvolution has increased significantly, and now with a multi-pronged approach the chances of success are greatly improved.
These target deconvolution strategies can be based on functional genetic approaches which are concerned with identifying the genes responsible for the observed phenotypic response, whilst cellular profiling is used to interrogate the effects of treatment on a variety of factors including signalling, gene expression and/or metabolism. Affinity-based approaches are used to identify cellular interactions of the drug, for example via chemo-proteomics methodologies, and in this field in particular there are many powerful tools now available to us. Finally, knowledge-based strategies utilise existing information and often computational methods to make inferences about targets and MOAs. These various techniques are based on fundamentally different principles and thus can provide complementary information, which when considered as a whole may significantly speed up the path to target identification.
Advances in the field of chemo-proteomics have provided us with various strategies for target deconvolution. In particular, photo-affinity labelling (PAL) is proving a popular approach. The premise of PAL is to take an active ligand and functionalise it in such a way as to give a compound that can covalently bind it’s interacting protein. From there the protein can be isolated from the biological system, by binding the drug-protein complex to a solid support in a technique known as a pull-down, before being identified through proteomics. Such methods allow enrichment of target proteins above usual cell proportions, and are a more robust approach compared to traditional affinity-based pull-downs, in which the drug in not covalently bound to target proteins. Advances in bio-orthogonal reactions allow in-cell labelling with such compounds, expanding their scope to encompass imaging techniques and support target ID with cellular localisation studies.
Alternative proteomics-based strategies for studying drug-protein interactions exploit the fact that the presence of a drug molecule in a biological system can cause changes to the thermal stability of proteins within that system. Techniques such as thermal proteome profiling (TPP) use proteomics to measure these protein stability changes, and the resulting information can reveal relevant drug-protein interactions. This technique can be used to measure drug-protein interactions in living cells, with no additional chemical modification required to the drug molecule of interest.
With all these tools at our disposal target deconvolution efforts have a much higher chance of success. Phenotypic screening in a disease relevant cellular assay followed by robust target deconvolution using a combination of increasingly sophisticated strategies is a powerful approach to modern drug discovery.