Emerging Hit Finding Technology: High-throughput Spectral Shift (HT-SpS) for Binder and Molecular Glue Identification

Figure 1. Spectral Shift principle. A. Schematic overview of the Spectral Shift assay with recombinant protein freshly labelled with a dye. Ligand binding may lead to a decrease of the red dyes emission at 650 nm and an increase of the 670 nm fluorescence or vice versa1.B. SDS-PAGE analysis of the purified PIM3. C. Single dose analysis for hit identification. Neutral control (DMSO) was used as reference and appropriate threshold was defined. Binders inducing red-shift (positive) and blueshift (negative) are depicted as purple triangles and orange circles, respectively. D. Plotting the 670 nm/650 nm ratio against the logarithmic concentration of a non-fluorescent ligand led to a sigmoidal binding curve that was used to extract the KD value of the interaction. Dose response curves of reference compound (AZD1208) at different incubation times allow for a time course analysis and the identification of the best assay window. E. High-Throu

Figure 2. PIM3 primary screen, statistics and hit confirmation. A. 48,252 compounds from the diversity library were tested at 28 µM. 1866 binder hits were identified by HT-Spectral Shift. B. Quality of the assay was evaluated using SSMD values per plate. The overall value is > 6, indicating a high-quality data set. C. Dose-response curves of hit selected for the hit confirmation stage. Data were fitted using the tight binding fit equation from Genedata Screener. D. Examples of X-scan graphs used as quality controls for well homogeneity and aggregation. The blue curve depicts a compound inducing PIM3 aggregation (scan anomaly) whereas the purple curve is typical of a normal pattern. E. Chemical structures of selected PI

Figure 3. Screening cascade and Medchem comparative study. A. Affinity-based screening strategy using HT-SpS. Various conditions can be established in parallel to allow for the early qualification of the hits. B. Hits identified in the two assays (Spectral Shift and ADP-GloTM) were analyzed and showed a minimal overlap. C. Principal Component Analysis (PCA) performed on the Morgan Fingerprints (atom-based similarity, 1024 bits) revealed that the hits identified in the HT-SpS screen target additional chemical space not explored in the biochemical screen (represented by purple circles).

• Sensitive biophysical technique with direct binding measurement
• Simplified set-up with one labeled partner reducing the number of false positives (vs biochemical assays)
• Miniaturization in 1536-well plate format for throughput and reduced sample consumption
• Solution-based assay, proteins are not constrained by immobilization, allowing them to undergo structural changes upon binding (vs DEL or SPR)
• Discriminate affinity binding efficiency from each partner (cooperativity factor α)

WEE1 is a member of the serine/threonine kinase family and plays a crucial role in regulating cell cycle progression. Inhibiting WEE1 presents a promising strategy for cancer therapy. However, a major challenge is selectively targeting WEE1 in cancer cells.4 Targeted protein degradation provides an innovative therapeutic approach by leveraging small molecules to engage the ubiquitin-proteasome system, leading to the degradation of selective target proteins. This can be achieved using bifunctional PROTACs or molecular glues (MGs), both of which recruit the target protein to an E3 ligase. We focused on identifying PROTACs and MGs that bind to WEE1, with Cereblon (CRBN) acting as the E3 ligase. For screening, we produced a truncated CRBN construct (CRBNmidi) with similar function to the full-length wild-type CRBN.5 A focused-library of 106 molecular glues and PROTAC compounds were tested against WEE1 and CRBNmidi.

Figure 4: Primary screen for binder identification using HT-SpS. A library of 106 compounds was tested at 10 µM final concentration in parallel using 3 different conditions. A. Direct binding of the compounds to labeled WEE1. Adavosertib was used as positive control for WEE1 at 30 µM and depicted as yellow circles. Horizontal purple dashed lines represent ± 4 RSD of DMSO controls depicted as purple circles. B. Direct binding of the compounds to labeled CRBNmidi. Lenalidomide was used as positive control at 30 µM and depicted as yellow circles. Horizontal purple dashed lines represent ± 3 RSD of DMSO controls depicted as purple circles. C. Direct binding of the compounds to labeled WEE1 co-incubated with CRBNmidi to identify ternary complex formation. Adavosertib was used as positive control at 30 µM and depicted as yellow circles. Horizontal purple dashed lines represent ± 4 RSD of DMSO controls depicted as purple circles.

Figure 5: Parallel hit confirmation by HT-SpS. A. Example of binding curves obtained with known WEE1 binders (Staurosporine and WEE1 PROTAC: Pomalidomide-C3-adavosertib). The WEE1 glue EDMG001037 does not bind WEE1 in absence of E3 ligase. B. Example of binding curves obtained with known CRBN binders (Lenalidomide and WEE1 PROTAC: Pomalidomide-C3-adavosertib). C. Example of binding curves obtained with known WEE1 binders (WEE1 PROTAC: Pomalidomide-C3-adavosertib). The WEE1 glue EDMG00103 binds to WEE1 in presence of CRBN E3 ligase.

Figure 6. PROTAC and MG hit validation. A. Venn diagram summarizing the hit list in each screen condition (CRBN binders, WEE1 binders and WEE1/CRBN binders) and highlighting the shared hit list. Pomalidomide-C3-adavosertib EDMG0001006 showed binding to both CRBN and WEE1 and induced the ternary complex formation. B. Ternary binding affinities against WEE1/CRBN complex. Cooperativity (α) values were calculated based on binary and ternary KD with WEE1 C. Example of hit confirmation for the molecular glue hit EDMG000103. The experiment was performed in Dianthus equipment using 384w plate forma