Senior Research Scientist Evotec Toulouse, Midi-Pyrenees, France
Abstract: Transcription factors (TF) regulate many key cellular processes and therefore represent major potential pharmacological targets. However, except nuclear receptor, TFs have been so far considered as undruggable and have suffered from limited efforts in drug development. As this paradigm is currently evolving, new pharmacological assays are needed including robust ones for High Throughput Screening application. Here we describe a Bioluminescence Resonance Energy Transfer (BRET) assay designed to identify lead-like compounds modulating the interaction of a transcription factor and its DNA binding site in raw cell lysates in a high throughput setup. K562 Chronic Myeloid Leukemia cells were engineered to stably express the TF of interest fused to NanoLuc® luciferase at the N- or C-terminus. 17-mer DNA duplexes corresponding to the TF consensus binding site were obtained by annealing complementary synthetic oligos, the forward strand conjugated to ATTO590 fluorochrome at the 5’ or 3’ ends. Scrambled 17-mer duplexes and unconjugated oligos were used as controls. Expression of N-ter- or C-ter- NanoLuc-TF fusions in K562 polyclonal pools was confirmed by measuring luminescence using NanoGlo® luciferase assay system (Promega). Similar luminescence levels at 450 nm were observed from both constructs. Binding of NanoLuc-TF to the ATTO590-target DNA should result in proximity bioluminescence energy transfer from the donor NanoLuc to the acceptor fluorochrome. Addition of ATTO590-target DNA prior to NanoLuc substrate addition resulted in measurable BRET-induced fluorescence. BRET was enhanced by freeze/thaw induced cell permeabilization, potentially facilitating TF:DNA interaction. Maximal BRET was observed before 10 min after substrate addition and decayed to background levels after 45-50 min. Nter-NanoLuc-TF and 5’-ATTO590-target DNA resulted in the largest BRET signals among all tested donor:acceptor couples. BRET ratios increased with concentration of ATTO590-target DNA with a typical binding hyperbolic curve saturating at high concentrations. Competition with unlabeled-target DNA duplex resulted in dose-dependent reduction of BRET signals, confirming signal specificity. Bystander BRET, between NanoLuc-TF and ATTO590-scrambled DNA, increased linearly with DNA concentration, but remained lower than specific BRET up to 1 µM DNA and was not competed by unlabeled DNA. A four-step assay was optimized in 384 well plates and automated for high throughput screening: 1. NanoLuc-TF expressing cells + ATTO590-target/scrambled DNA dispense in plates containing test-compounds; 2. Overnight freezing at -80°C; 3. Scheduled NanoGlo detection reagent dispense and 4. Reading 610/450 nm emission. The assay and robotic process were technically validated using plates containing DMSO or DAPI, a DNA binder identified to interfere with the assay. BRET signals were stable across plates, inter-well variability was lower than 5%, average target DNA/scrambled DNA was 2.6, RZ’ was 0.67 and DAPI IC50 ranged from 2.5 to 6.3 µM across plates. Altogether the assay performance was judged compatible with high throughput screening of large compound libraries.
