Fragment Screening Using the ForteBio Octet RED System

Sriram Kumaraswamy, Ph. D., Product Manager, skumaraswamy@fortebio.com
Shirley Li, Principal Scientist, Roche Discovery Technologies

The Octet RED system allows rapid screening against an immobilized target with minimal protein degradation, and has the sensitivity to detect the binding of small molecules in the 200 MW range. Compounds were screened at a single concentration, followed by three point dose response measurements against the primary and a secondary protein target. Parallel biochemical functional assays were also run to gather supporting data that hits were targeting the correct binding sites on the proteins.

Super Streptavidin biosensors were loaded with biotinylated protein that was generated in vivo from E. coli by co-expression of a biotin-tag (GLNDIFEAQKIEWR) appended target with E. coli biotin ligase (BirA), resulting in protein modified specifically at the Lysine residue in the tag (underlined). Super Streptavidin biosensors loaded with biocytin were used as reference.

For the biochemical assay, three point dose responses against two protein targets were run using a standard TR-FRET format to monitor the binding of biotinylated peptide substrates to His6 tagged proteins. Energy transfer between the Europium streptavidin donor and APC-antiHis6 acceptor was quantified using an Envision reader (615/665nm).

A reference compound was first monitored on the Octet RED for binding affinity (KD = 1.8 μM) and confirmed on a Biacore S51 instrument (KD = 0.8 μM). The response window from the reference compound was used to set up the parameters for the subsequent screen. A response maximum of 200 pm (picometers) corresponding to the saturating signal for the reference compound and a response minimum of 15 pm corresponding to 3X of standard deviation for running buffer over the response of the baseline were set.

The simple and easy dip and read method employed by the Octet RED provides an advantage for screening compound fragments. Binding of compounds can be detected, even when they are partially precipitated from solution. The results of binding experiments in such cases are reproducible and correlate well with solubility measurements. The absence of microfluidics eliminates clogging issues.

The overall fragment screening strategy employed is illustrated in Figure 1. The fragment screen was conducted on 96-well plates with 68 unknown and two reference compounds per plate, each at 100 μM concentration. The corresponding plate map is shown in Figure 2. The average throughput is roughly 1.75 to 2 hours per plate, and a combined total of 4 to 5 plates per day, which translates to a maximum of 420 compounds screened in one working day (eight hours).

An example of data from the screening is shown in Figure 3. The screen was followed by confirmatory assays done for 21 compounds per plate at three different concentrations of 33, 100 and 300 μM. The rules were set as:

• Binding signal to be between 15 and 200 pm
• Dissociation signal returns to a value less than 50 pm
• Binding responses display the following logic: Conc 1 < Conc 2 and/or Conc 2 < Conc 3.

Biochemical assays performed in a follow-up experiment verified that the fragments were binding to the correct sites of the target protein. True stoichiometric binders were then elucidated in characterization experiments involving full concentration curves on the Biacore S51.

In summary, the Octet RED system’s dip and read method allows quick assay development, has enough sensitivity to detect 200 Da fragments, and sufficient throughput for rapid screening of thousands of compounds. The system does not consume target protein, and its disposable biosensors maintain protein integrity throughout the screen. Another major advantage of the Octet RED is its ability to detect binding of partially precipitated samples without the complications found in using microfluidics-based biosensors.

^ TOP