Discussion

Michelle Newman, Barbara Saxty and Ahmad Kamal

MRC Technology, Centre for Therapeutics Discovery,
1 – 3 Burtonhole Lane, Mill Hill, London NW7 1AD, UK

Drug efficacy and target engagement has traditionally been monitored by evaluation of downstream cellular responses. Alternative techniques, based on measuring changes in the stability of a target protein by direct binding of accessory proteins or small molecules, such as the Cellular Thermal Shift Assay (CETSA), can be used to investigate target protein stabilisation isolated from cells, animals or human samples. These approaches often rely on access to suitably selective antibodies for detection of the target protein following interaction with drug or compound, and melt curve analysis with a suitable immunoassay.

As an alternative approach, using the kinase ULK1 (Unc-51 like autophagy activating kinase 1) as an example, we evaluated the InCELL Hunter™ Target Engagement Assay (TEA; DiscoveRx® Corporation), to test a panel of ULK inhibitors. Data output were compared to other assays measuring indirect target engagement through assessment of a downstream ULK1 substrate, Atg13 using a phospho-specific antibody (Abnova). The TEA assay is based on Enzyme Fragment Complementation (EFC) between two inactive -galactosidase fragments, an enhanced ProLabel® (ePL) and an Enzyme Acceptor (EA). We fused the ePL fragment to the C-terminal of the kinase domain of human ULK1 and transiently transfected HEK293 cells to assess relative stabilisation of ULK1 with our inhibitors. The chemiluminescent signal generated, is directly proportional to the amount of transfected ePL-tagged ULK1, allowing comparative analysis with a gatekeeper mutant version of ULK1 (M92T).
Time course analysis, showed 3.5 hours to be optimal for both chemiluminescent signal and assay window using a test compound. Subsequent studies confirmed expression of ePL-tagged ULK1 in transfected cells and that the signal was not modulated transcriptionally, translationally or through proteosomal degradation, but partly through HSP90 interaction and stabilisation, suggesting ULK1 is a client protein for HSP90. Profiling of over 100 compounds in this optimised assay against both wild type and the M92T ULK1 mutant showed a range of activities which correlated well with phenotypic-based readouts of autophagy. Increased ULK1 stabilisation by compound treatment was lost or significantly attenuated in the gatekeeper M92T mutant suggesting an ATP-competitive interaction.

The TEA assay represents a simple, rapid, homogenous, high-throughput and quantitative approach for evaluation of intracellular protein stabilisation. Furthermore, it allows analysis in a native cellular environment, providing simultaneous assessment of compound potency, cell-permeability and target engagement.