Label-Free 7,000 Fragment Screening
Label-Free 7,000 Fragment Screening
Authored by: Nicolas Andre1, Oliver Esser 1, Christian Bergsdorf 2, Johannes Ottl 2, Volker Eckelt1, 1Corning SAS, France, 2Novartis NIBR, Basel, CH
Originally published as a Corning Technical Note, reprinted with permission.
The Epic System is a high-throughput label-free screening platform that allows for the observation of direct biomolecular interactions in a biochemical binding assay and pharmacologically more relevant cellular response to a drug in a cell-based assay. It consists of a disposable SBS standard 384-well microplate with resonant waveguide optical biosensors integrated into each well and a high throughput screening (HTS)-compatible optical reader. It eliminates detection issues, such as background fluorescent interference associated with the use of labels, while providing access to new biochemical and pharmacological information for drug discovery.
Fragment-based drug discovery (FBDD) is a recent approach that uses a structurally defined fragment library in conjunction with chemical structure information to screen new potential drugs . FBDD has the advantage of covering a large chemical diversity with a fewer number of compounds. “Fragments” represent low molecular weight compounds suitable for screening at high concentration. Traditionally, fragment screening is more challenging, due partly to the difficulty of developing labels for conventional detection techniques. Conventional biophysical techniques used for fragment screening, including NMR, X-ray crystallography, SPR or mass spectroscopy, may face limitations, such as protein consumption, the need for amino acid labeling, availability of crystals and solved 3D structures, and throughput (see  for review). Most of these issues are not encountered when using the Epic system. In the present study, we performed a successful fragmentbased screen on the well-characterized enzyme bovine carbonic anhydrase II (bCAII) (~30 kDa) with the Epic system. We screened seven thousand compounds with an average molecular weight of 230 Da (Novartis, Basel, CH) at a final concentration of 520 μM.
Materials and Methods
Bovine carbonic anhydrase II was obtained from Sigma- Aldrich (Cat. No. C2522) and the 7,000 fragment library was provided by Novartis. Epic User Activated Biochemical Assay Microplates (Cat. No. 5046XX1) were used for all experiments.
Epic microplates were activated with 10 μL of 200 mM 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC )/50 mM sulfo-N-Hydroxysolfosuccinimide (sNHS) (Pierce, Cat. No. 22981/24510) and incubated for 30 minutes at room temperature. The microplates were subsequently washed 3 times with 30 μL H2O/0.005% Tween® 20 on a Biotek Platewasher ELX 405 HT.
Protein immobilization was accomplished by applying 10 μL of 50 μg/mL bovine carbonic anhydrase II in 20 mM sodium acetate pH 5.5 using a 384-channel pipettor (CyBi®-Well) and incubating overnight at 4°C. The microplates were subsequently blocked with 500 mM ethanolamine/phosphate buffered saline (PBS) pH 8 for 10 minutes and washed three times with assay buffer (PBS, pH 7.4/0.005% Tween 20/3% DMSO) and, finally, 15 μL assay buffer was added.
|Figure 1. Furosemide binding to immobilized carbonic anhydrase II was evaluated using the Epic® instrument and Epic Microplate (Cat. No. 5046XX1).|
Compounds were diluted with assay buffer to a concentration of 833 μM and 3% DMSO for the working solution.
The binding assay was run on an Epic instrument integrated into a CyBio Workstation consisting of an incubator (Cytomat 6001, Thermo Scientific), a 384-channel pipettor (CyBi®-Well, CyBio) and a rotating arm for plate transportation. After 25 minutes thermal equilibration, a baseline reading was taken on the Epic microplates with protein immobilized in assay buffer. In the next step, 25 μL of reconstituted compounds were added to the plates to a final compound concentration of 520 μM and then mixed 13 times. The final reading was taken after 25 minutes thermal equilibration. The Epic response is measured as a shift in reflected wavelength and is expressed in picometers (pm).
Furosemide at 10 μM final concentration was used as positive control. Buffer only was used as negative control.
The Positive Control
The binding of furosemide to the bCAII is specific and saturable as seen with the dose response curve (Figure 1). The calculated affinity (KD) is in agreement with published data . Furthermore, the small error bars indicated that the data is high quality and that the assay conditions are suitable for obtaining relevant information.
Over the 20 Epic microplates used for the 7,000 fragment screen, the positive control (furosemide) gave an average signal of 21 pm with only limited variation between each microplate (<7%) (Figure 2). This lower signal compared to the 26 pm obtained in the saturation experiment reflects the step from assay optimization to scale-up and higher throughput. The robustness of the assay over the entire screen was high; specifically, the average Z' for all 20 microplates was 0.7, and each individual microplate had a Z' >0.5.
|Figure 2. Assay robustness (Z’, red cross) and signals of positive control (orange bars) over the 20 microplates of the 7,000 fragment screen is shown. Every microplate had a Z' factor above 0.5 (green dashed line).|
The corrected signals for the 7,000 compounds screened are plotted on a graph (Figure 3). A majority of the compounds give a signal very close to zero. The hit threshold has been set arbitrarily at 5 times the median average deviation (5MAD), and its absolute value is close to 8 pm in each of the 20 microplates used during this fragment screen. The threshold is represented by the green dashed line. It is clearly seen that the hits are well distributed along the 20 microplates of the screen, therefore indicating that the hit distribution is not biased.
Based on this selection criterion, the overall hit rate for the 7,000 fragments screen is ~6.4%, which is in the expected range for a fragment screen run at a high final compound concentration of >500 μM . Altering the hit threshold to 4, 5 or 6 times MAD does not dramatically change the hit rate (5.5, 6.4, and 7.5%, respectively). The present fragment library contains 69 known bCAII inhibitors, out of which 97% were selected with this screen (Novartis). Therefore, the screen generated new hits that would require further confirmation and validation.
- The Epic® system enables high throughput label-free direct biochemical assay for small molecule binding to an immobilized target without the need for labels. The results of this 7,000 fragment screen demonstrate successfully high assay robustness (average Z' of 0.7).
- The hit rate determined under the present conditions is in agreement with the expected hit rate for this particular library (6%hit rate).
- ↑ Hubbart, R.E. fragment approaches in structure-based drug discovery. J. Synchrotron Rad. 15,227-230 (2008).
- ↑ Carr, R.A., Congreve, M., Murray, C.W., Rees, D.C. Fragment-based lead discovery: leads by design. Drug Discov. Today 14, 987-92 (2005).
- ↑ Myszka, D.G. Analysis of small-molecule interactions using Biacore S51 technology. Anal. Biochem. 329, 316-323 (2004).
- ↑ Erlanson, D.A., Jahnke, W. The concept of fragment-based drug discovery. In Fragment-based Approaches in Drug Discovery, Jahnke, W. and Erlanson, D.A. (ed.), WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim (2006).
|Click [+] for other articles on||The Market Place for Lab Automation & Screening||Analytical Equipment, Measuring, Testing Detection Devices|