Automated patch clamp for PRIMARY SCREENING

While fluorescent and luminescent assays have been the default choice for primary screening for decades, the introduction of 384 well, automated patch clamp (APC) platforms and the ability to run extended overnight runs of up to 16 plates makes patch clamping an appealing alternative.

Qube VS. FLIPR

When it comes to ion channel physiology and pharmacology, the technology of choice for many years was manual patch clamp. An art with high precision but with very low throughput. The low throughput issue has been changed today. Since APC was invented in the early 2000’s the technology has now developed to the stage where it is applicable as the primary screening technology.

Drug discovery remains to be a trial and error exercise in the quest for the next drug. Once a target has been identified the next step is often to screen compound libraries to generate leads for later drug candidates. This process is influenced basically by two parameters:

1. Number of experiments per time unit

2. Degree of the usefulness of results from the screening

While 1) is easy to measure 2) takes more insight to evaluate. In most discovery processes the flow can be described with the funnel depicted in Figure 1 where the high throughput has been followed by secondary, or confirmatory, screen, then lead optimization, safety studies and profiling. When targets are ion channels, everything from secondary screening has been patch-clamp, but due to the low throughput, the high throughput screen has been synonymous with fluorescence-based technologies. Fluorescence-based technologies have high performance on parameter 1

Since 384-format became available in APC it has been possible to conduct electrophysiology as primary screening. Therefore, it is possible to omit the secondary screen, as depicted in Figure 1 by the graded out portion, simply because e-phys confirmation is already done with the primary screening.

“But isn’t that expensive compared to FLIPR?” you may ask.

 

Using an APC screen, the price per data point is slightly higher since FLIPR is around 5-10 cents per well and APC around 20-30 cents/well, so looking at it from an isolated point it is. However, consider the entire in vitro part of the drug discovery process as depicted in Figure 2, where “Traditional….” means using fluorescence-based technology as primary screening.

The target discovery and validation are the same for the two processes. The assay development takes some time to find e.g. Ca²⁺-dyes, membrane potential sensing dyes etc. for the FLIPR-assay, whereas the Qube APC is a patch-clamp instrument and hence does the same as during the target discovery and validation. The screening part is faster on the FLIPR but APC will more than catch up since:

A. No need for hit validation (already patch clamp)

B. Better hits/leads since the data are genuine ion channel recording in contrast to indirect measures of downstream/ secondary messenger systems when using fluorescence-based technology.

In particular parameter b) means that the medicinal chemists get better data to work with and they get the same type of data when the refined molecules are synthesized and tested since that is also done with patch-clamp. This means that both the more sophisticated on-target effects, such as mode-of-action and subtype selectivity and the off-target effects like cardiac liability, is available very early in the drug discovery and will ensure a faster and more cost-efficient process. Looking at Figure 2 and imagining that the time saved in using APC as primary screening is converted into prolonged patent life once the molecule is developed to the final drug, will enhance the value of using APC upfront tremendously.

FAQ

Concentration-response with much fewer wells

Do you run the usual 22-point concentration-response experiments on your FLIPR to determine IC₅₀ values? That, of course, fits with the compound plate layout that you likely have but it also requires one assay plate with seeding of cells for every compound plate and most importantly; you don’t have the control cell for every concentration. Consider running such an assay with a cumulative application on an automated electrophysiology platform:

• All 22 concentrations will have the same cell response as control
• All 22 concentrations will only use one well instead of 22
• The electrophysiological readout is direct on the ion channel and hence more predictive

Reference: High throughput screening for mode-of-action on Na_V1.4

Doesn’t it take long to screen with electrophysiology?

Admitted, automated electrophysiology is not as fast as fluorescence-based screening. Notably, e-phys so far only comes in 384-format and not 1536-format, but automated e-phys can run unattended so that makes up for some of it:

• ≥8 hours run without any need for personnel around the instrument
• >9,000 wells assayed unattended
• >25,000 wells assays per 24 hours
• Z’-values that support large screening windows
• The electrophysiological readout is direct on the ion channel and hence more predictive

Reference: Na_V1.1 currents on Qube

High-Throughput Screening of Na_V1.7 Modulators Using a Giga-Seal Automated Patch Clamp Instrument

How do you help your chemist the best way – faster hit to lead?

When you have done your initial screening on a fluorescence-based instrument you are often obliged to confirm this by electrophysiology. Why not do that, to begin with? Utilize your expertise in running a large screening campaign with all the logistics required for compound handling and get the output values of pharmacological effects with high fidelity electrophysiology, which can go directly to your chemists modelling program. In other words, faster from hit to lead.

• More precise pharmacology with adaptive protocols
• More relevant questions asked to your molecules
• More subtle effects of your molecules can be pursued

Reference: Adaptive voltage protocols increase precision of voltage-gated ion channel measurements on high-throughput automated patch clamp platforms

How do you help your chemist the best way – better leads?

In the ideal world, modelling could generate the molecule you want to develop for any disease. In the real world, we still need data and the more predictive data the better. When interrogating ion channels, the best data come from electrophysiology. And you can get that without jeopardizing the need for speed. With 384-format high fidelity electrophysiology, you can run unattended, your chemist will always have the data needed in due time, and more importantly, with the data content for creating a better drug:

• State-dependent mode of action
• on/off rates
• IC₅₀ values
• All of the above at any temperature between 10-40^◦C

Reference: Multi-parameter ion channel screening: mechanism-of-action data directly from HTS

Adaptive voltage protocols increase precision of voltage-gated ion channel measurements on high-throughput automated patch clamp platforms

Early de-risking of cardiac liability, hERG and Na_V1.5

When do you take care of investigating the cardiac liability of your compounds? Wouldn’t it be great to have that early in the drug development and for multiple targets? You can get that in one operation with electrophysiology on e.g. hERG and Na_V1.5 and early on find out if you need to direct your medicinal chemistry program around a problem.

• Take advantage of constant voltage clamp
• Take advantage of temporal resolution
• Take advantage of multihole recording mode
• E.g. hERG and Na_V1.5 pharmacology tested in the same assay

Reference: Simultaneous Measurement of Cardiac hERG and Na_V1.5 Currents Using an Automated Qube Patch Clamp Platform

Ligand-gated ion channels that desensitize – no problem

Some ligand-gated targets only allow short exposure to their ligand before they desensitize and don’t lend themselves to re-stimulation. That is a problem when your screening technology does not provide wash-out. But it is not a problem with electrophysiology with microfluidic flow channels and very short exposure times where the target can be stimulated multiple times and hence serve as their control read-out.

• Possible with repeated stimulation of fast-desensitizing ligand-gated channels like AchRα₇
• Possible with cumulative dose-response where each cell is its control
• Possible to follow the kinetics of the target in control as well as the compound situation
• Possible to do all of the above in 384-format unattended

Reference: Characterization of the rapidly-desensitizing α7 nicotinic acetylcholine receptor using the Qube