CORDIS - EU research results

Photoredox catalysis as a tool for the site-specific labelling of proteins

Periodic Reporting for period 2 - PhotoChemBio (Photoredox catalysis as a tool for the site-specific labelling of proteins)

Reporting period: 2021-04-01 to 2022-03-31

The identification of disease-relevant protein targets that can be exploited for therapeutic benefit, broadly known as target ID, remains a fundamental goal of drug discovery. In recent years, the development of new biological and chemical techniques that accelerate target ID has become crucial amongst pharmaceutical research programmes. Despite significant advances, the identification of protein targets and their accompanying interaction networks remains a resource intensive, time-consuming, and often unsuccessful endeavour. While these methods remain state-of-the-art, the fundamental technology underpinning these approaches has evolved little in over 50 years. These traditional approaches to target ID rely on the activation of specific functional groups which are appended, via a chemical linker, to the drug of choice. Exposure to UV light allows these groups to decompose forming highly-reactive intermediates in close proximity to the protein of interest, which can rapidly insert into neighbouring chemical bonds, forming a formal 'covalent cross-link'. However, in the complex aqueous environment of a cell, near exclusive insertion into water occurs (>99%), and not the desired insertion into the target protein. Since the photoreactive group is directly appended to the small molecule ligand, unproductive quenching is far too frequently an insurmountable hurdle to successful target ID. As a result, the limitations in mechanistic understanding and subsequent off-target biological effects remains one of the leading causes of attrition for small-molecule drugs in the clinic. Therefore, the development of new methods to elucidate small molecule/protein interactions has the potential to increase significantly the success of medicinal target selections and ultimately reduce patient morbidity.
We have discovered a new platform for precision target identification (target ID) of proteins through the orchestration of small molecule-photocatalyst conjugates, blue light, and otherwise inert diazirines. Through a photophysical process known as Dexter energy transfer, we can, for the first time, effectively decouple the small molecule from the photoreactive group, which has been a mainstay of classical photoaffinity labelling for over half a century. Visible-light excitation of the small molecule-photocatalyst conjugate can sensitize diazirines within the immediate vicinity (0.1 nm), giving rise to a carbene with an extremely short solution half-life in water (~4 ns). Crucially, the catalytic generation of these reactive intermediates, at the precise location of the ligand binding site, gives rise to multiple labelling events. This unprecedented step, allows, for the first time, not only signal amplification at the protein level for target ID, but also at the peptide and single amino acid level, facilitating bind-site mapping. We've demonstrated the platform on a number of small molecules (including cyclic peptides) to label individual proteins and multi-protein complexes. Importantly, we have demonstrated that unlike traditional probe conjugates which require extensive SAR optimization, successful implementation of this platform requires no optimization of the catalyst or linker, saving critical time and expense. Moreover, the residue agnostic nature of carbene insertion, combined with the tight radius of activation, leads to precision mapping of the ligand binding site, which we have successfully highlighted on both single proteins and protein complexes.
The de novo identification of disease relevant GPCRs using traditional photoaffinity labelling is regarded as almost impossible within the pharmaceutical industry. Despite their therapeutic importance in areas ranging from oncology to neurodegenerative diseases, their druggability has been severely hampered by the challenges imposed by cell surface target validation, encumbered by low protein abundance, lack of exposed residues, and hydrophobic properties. Given the catalytic amplification afforded by photocatalytic target ID, we believe that this platform is ideally placed to overcome the critical challenges imposed by these limitations. Preliminary studies have demonstrated the first successful photocatalytic target ID of GPCRs in live cells and follow up studies are underway to not only establish the generality of this platform towards other membrane protein targets but also investigate its capability towards identifying drug off-targets and interactor proteins in a native biological context. Based upon the potential of this platform to elucidate new drug targets and contribute towards a greater understanding of biological pathways, we fully expect that this strategy for protein target ID will be rapidly adopted, especially for drug-discovery programmes across both academia and industry.