Pushing the boundaries in pharmaceutical hit discovery with the novel waveRAPID kinetic assay

Summary

Drug discovery high-throughput screening campaigns of thousands of compounds can produce a large number of hits showing potential interaction with the drug target. Binding affinities can be derived from steady-state analysis from many high-throughput assays. For the selection of those hits with a potential for further development, a full characterization of binding kinetics is required. This identifies compounds specifically bringing to the desired target with favorable kinetic binding rates. In practice, this means that extensive kinetic studies must be performed on large numbers of screening hit compounds. To date, this is an extremely time- and material-consuming undertaking. Suitable dilution series from hundreds of hit compounds have to be prepared and each concentration has to be injected in replicates. We have developed a novel technology that can make such projects much more efficient, enabling a robust kinetic characterization from a sample at a single concentration. waveRAPID on the WAVEsystem, a Creoptix technology, represents a paradigm change for kinetic measurements in a screening setting.

In the current study, conducted by an HTS drug discovery group of our customer Idorsia Pharmaceuticals Ltd, we demonstrate the power of waveRAPID in streamlining the kinetic characterization of a large number of drug hit compounds. In several experiments, binding kinetics of up to 146 hit compounds were determined by waveRAPID, using pulsed injections of samples at a single concentration onto the drug target previously immobilized on the sensor matrix. All analyzed compounds had been studied before on state-of-the-art SPR instrumentation, using traditional multi-cycle kinetic measurements. 

Comparing the data obtained from the two technologies revealed an excellent correlation for the calculated dissociation constants (K_D) with only a few outliers. This finding validates the waveRAPID technology on the WAVEsystem for application in early drug discovery, enabling the generation of in-depth kinetic data in a screening setup with high throughput. As the time for characterizing each compound is reduced to only a few minutes, significant time and materials savings can be achieved. Large campaigns to determine the binding kinetics of hundreds of compounds can be carried out in days instead of weeks as compared to current state-of-the-art SPR instruments. 

Furthermore, the waveRAPID technology generated robust data even for difficult compounds (such as those showing large bulk refractive index contribution or slow dissociation) throughout the assay series. Such compounds often necessitate surface regeneration and re-immobilization of the ligand protein. This interrupts the workflow during larger measurement campaigns using traditional multicycle assays. Finally, the WAVEsystem, as well as the novel waveRAPID method, enable an extremely straightforward assay setup, offering full binding kinetic analysis from a single concentration per sample. Additionally, direct kinetics, our unique and novel evaluation engine based on statistical parameter estimation, offers an automated and robust analysis workflow including real statistical error calculation. The waveRAPID technology and the direct kinetics automated evaluation engine represent a powerful package that has the potential to transform the way kinetic binding assays are performed in early drug discovery today.

Materials and Methods

WAVEsystem Protocol

Histidine-tagged target protein 1 and Histidine-tagged target protein 2 were captured to a density of 4000 to 5000 pg/mm² on a 4PCP-NTA WAVEchip previously loaded with Ni²⁺ ions. Pulsed waveRAPID injections of compounds were done from a 10 μM sample solution for 25 s, followed by buffer injection for 250 s with a flow rae of 32.5 μl/min per flow cell. All measurements were performed in 50 mM Phosphate Buffer with 0.05% Tween-20 and 5% DMSO. Data were double-referenced by subtracting the signal of the reference channel and blank injections. The apparent concentration of the compounds in the sensing area over the 6 injection pulses was calibrated with a pulsed injection of running buffer supplemented with an additional 0.5% DMSO, using the same injection parameters as for the compound samples. Kinetic analysis was performed by the Direct Kinetics estimation engine using a 1:1 binding model.

Biacore Protocol

Histidine-tagged target protein 1 and Histidine-tagged target protein 2 were captured to a density of 4000 to 5000 pg/mm² on a Ni-NTA chip previously activated with Ni²⁺ ions. The analytes were adjusted to the desired concentration by a serial dilution in a running buffer (50 mM Phosphate Buffer with 0.05% Tween-20 and 5% DMSO). The sensorgrams were obtained with an association time of 100 s, a dissociation time of 250 s, and a flow rate of 40 μl/min. To determine the binding kinetics (k_on, k_off and K_d), the sensorgrams obtained after double referencing were analyzed by the Biacore T200 Evaluation software using a 1:1 binding model.

Results

The novel waveRAPID technology for fast acquisition of kinetic binding data in a screening-like setting was validated against state-of-the-art instrumentation and methodology in this study. As shown in Figure 1, kinetic binding data can be generated rapidly for a large number of screening hit compounds, by pulsed injection of each sample at a single concentration. Including analyte dissociation, each measurement took around 5 minutes. The typical response profiles generated by waveRAPID are shown in Figure 1, panel A and B (data in red, with the kinetic 1:1 model in black). The evaluated parameters are shown in a rate map in panel C (control compound injected over the entire assay in the black circle), demonstrating excellent reproducibility. Typical assay times and sample throughput comparing an exemplary assay with 384 samples on two targets are listed in Table 1. 

Figure 1: Application of GCI on the WAVEdelta for early drug hit discovery and characterization using the novel waveRAPID technology. Full kinetic characterization can be derived from compounds injected at a single concentration in a screening setting with waveRAPID, enabling large time and cost savings in early drug hit discovery.
(a) A representative selection of waveRAPID response profiles from a screening of 80 compounds is shown.
(b) Using tailored injection settings from selectable presets in WAVEcontrol, the entire range of affinities can be covered. Here, settings for intermediate binders were used (20 seconds pulsed injection, 250 seconds dissociation), enabling the resolution of affinities from low nanomolar up to low micromolar.
(c) Kinetic rates are shown in a rate map with identical affinities on the diagonal lines. Datapoints in the circle show the excellent reproducibility of the control compound over the entire assay.

	WAVE	Biacore T200	Biacore 8K
Number of Targets	2	2	2
Number of Samples	384	384	384
Type of Assay	waveRAPID	Kinetics	Kinetics
Affinity Range	nM	nM	nM
Captured Steps per Sample	0	0	0
Reagent Steps per Sample	0	0	0
Runtime (h)	58	490	122
Samples / 24h	159	19	75

Table 1: Exemplary assay run times and sample throughput comparing waveRAPID on the WAVEdelta to state-of-the-art technologies (Biacore T200, 8k) for full kinetic characterization of 384 samples on 2 targets.

The kinetic data obtained was filtered for errors on the binding rates (below 100% of the absolute k_a and below 50% of the absolute k_d) and compared to data from standard SPR multi-cycle measurements. As shown in Figure 2, the comparison to traditional SPR multi-cycle measurements reveals a very good correlation between the two datasets. 

Figure 2: Drug screening hit binding kinetics measured with waveRAPID on the WAVEdelta are plotted against values obtained by traditional multi-cycle kinetic measurements performed on a Biacore T200 instrument. The waveRAPID data was filtered for statistical errors on the association rate (k_a) below 100% and dissociation rate (k_d) below 50% of the calculated rate parameter. The correlation validates the ten-fold faster waveRAPID kinetics method as compared to the traditional method.

To further challenge the system, another measurement series was performed, where more difficult compounds causing workflow interruptions and surface regeneration in the traditional multi-cycle measurements were run throughout the assay. Presumably due to the significantly shorter contact time of the analyte with the target protein, these difficult molecules (showing high bulk response or slow dissociation) do not disrupt the overall workflow on the WAVE and make any intermittent surface regeneration unnecessary. As shown in Figure 3, the kinetic binding parameters of almost all compounds of interest could still be robustly evaluated with low errors and the positive control still demonstrated a high reproducibility.

Figure 3: Thirty previously characterized compounds were measured again in an assay containing also 30 more challenging compounds showing high bulk refractive index (RI) contribution or slow dissociation, as well as the challenging compound injected throughout the assay. Whereas the challenging compounds caused major workflow limitations in the traditional multi-cycle kinetics, making regular surface regeneration necessary, waveRAPID enabled a seamless workflow with continuous unattended measuring time. The statistical errors in percentage of the absolute calculated association and dissociation rate constants, as determined by the Direct Kinetics estimation engine, are plotted on the horizontal and vertical axis, respectively. Compounds that showed no difficult behavior in previous assays (green marks), as well as replicates of the control compound (purple marks), are clearly identified by low errors. Also, compounds showing difficulties in previous assays due to large bulk refractive index contribution can be robustly analyzed on the WAVE with low errors (blue marks). The remaining difficult-to-analyze compounds are clearly characterized by larger errors (red marks).

Conclusions

With high sensitivity, the ability to resolve extremely fast interactions and waveRAPID assay technology, the WAVEsystem enables the kinetic characterization of large numbers of drug hit compounds in a significantly shorter time as compared to traditional multi-cycle kinetic assays. By obtaining kinetics from a single well in one injection, reagent and sample consumption are reduced to benefit early drug discovery significantly. In addition, waveRAPID provides:

• Results comparable to traditional SPR in a much shorter time
• Reduced contact time, preserving target protein and enabling longer uninterrupted assays
• High reproducibility and confidence in results
• Binding kinetics from a single injection to select hits based on relevant parameters
• More interactions characterized in days rather than weeks
• Profound statistical error calculation to discover relevant hits and filter out problematic compounds

Key takeaways

• Reduced sample contact time preserves target integrity for longer runs
• Binding kinetics from one single injection
• Reproducible and comparable to traditional SP technologies

Great for:

• Screening of crude mixtures: small molecules, fragments and peptides
• Fast hit-to-lead progression
• Lead optimization
• Unstable targets
• Difficult compounds
  

Acknowledgments

We thank Geoffroy Bourquin, Laksmei Goglia, Solange Meyer and Oliver Peter for carrying out all experiments and contributing to this document.