Ligand-gated ion channels (LGICs) represent an important class of drug targets that play a critical role in the function of the central nervous system. Compounds acting on ligand-gated channels are used to treat a wide range of conditions.
Effective screening campaigns against these targets require both the ability to conduct true electrophysiology measurements at high throughput and the ability to measure real-time changes in current during compound addition. The real-time recording requirement is driven by the fact that many LGICs desensitize quickly after the application of an activator molecule. Ion-channel screening strategies can be rendered ineffective if the current is recorded following a long delay after the activating ligand is applied.
Of the few commercially available systems for automated electrophysiology, there is no solution that offers fast compound addition and meets the rigorous cost and throughput requirements of HTS. Therefore, strategies for large screens against ligand-gated targets have typically required alternate approaches from voltage-gated target screens.
LGIC screens often start with indirect fluorescence or radioligand binding assays. While useful because of their speed and relatively low cost, indirect assays have often produced false negatives and positives. Furthermore, these indirect assays have shown poor correlation with downstream electrophysiology assays.
The IonFlux HT platform from Fluxion Biosciences was developed to provide high-throughput patch-clamp measurements for a diverse range of ion-channel targets. The system enables fast compound additions and continuous electrophysiological recording, both of which are necessary to screen against fast-acting ligand-gated targets.
The IonFlux platform has a unique design based on standard well-plate formats and has a plate-reader style form factor. These attributes enable integration with standard HTS equipment including fluid-handling robots. In this article, we will describe assay characteristics and compound pharmacology for the GABAA receptor, a well-studied LGIC target.
The IonFlux system (Figure 1) integrates all of the ion-channel recording hardware (amplifiers, electrode assembly) and the compound perfusion control interface into a plate-reader style enclosure. Two versions are available, the IonFlux 16 and IonFlux HT. The IonFlux 16 is suitable for assay development and medium-throughput applications. This system records from 16 cell ensembles in parallel on 96-well plate devices. The IonFlux HT is geared toward high-throughput screening and has 64 cell ensembles in parallel on 384-well plate devices.
The entire recording protocol, including cell trapping, whole-cell voltage clamp, compound perfusion, and ion-channel recording, is managed automatically via software. Plates are preloaded with cells, compounds, and reagents using either a fluid liquid handler (IonFlux HT) or multihead pipettors (IonFlux 16). Once filled, plates are fed automatically into the IonFlux system.
All fluidic control necessary to capture and voltage-clamp cells and deliver compounds is handled within the instrument. The system also houses integrated electrodes for continuous recording of the current response from all cell ensembles across the plate.
A key feature of the system is the ability to apply 32 compounds simultaneously, reducing the total assay time for a 384-well plate (as little as 10 minutes). Another unique feature is the ability to control the temperature of the plate and fluids during the experiment. This becomes critical for temperature-sensitive channels as well as maximizing ion-channel current density.
The IonFlux system uses well-plate microfluidic devices called IonFlux Plates. These are SBS-standard well plates with a network of micron-scale channels that run underneath and in between wells of the plate. The network consists of a repeated pattern (Figure 2) interfacing to 12 wells on the plate. This pattern is repeated eight times on 96-well plates (IonFlux 16) and 32 times on 384-well plates (IonFlux HT).
Each pattern is an independent experimental unit that obtains recordings from two cell ensembles in parallel. Both ensembles in each pattern are exposed to eight independent compounds or a concentration series of a single compound.
Cells expressing GABAA (α1/β3/γ2-HEK, Millipore PrecisION) were cultured in 225 cm2 filter-top flasks containing DMEM/F12 glutamax, G-418, Hygromycin B, and Puromycin at 37°C and 5% CO2. The cells were kept below 90% confluency. For cell isolation, flasks were first washed with 2 mL of Ca- and Mg-free PBS, followed by Detachin solution after which cells were given two to five minutes of further Detachin treatment.
After release, the cell suspension was spun for 90 seconds at 1,000 rpm prior to being resuspended in extracellular solution (mM): 137 NaCl, 4 KCl, 1 MgCl21, 1.8 CaCl2, 10 HEPES, 9 glucose, pH 7.4 with NaOH 10. The intracellular solution for the whole-cell voltage clamp contained (mM): 130 K Aspartate, 5 MgCl2, 5 EGTA, 4 Tris-ATP, 10 HEPES, pH 7.2 with KOH. To evoke inward Cl- currents, cells were trapped in ensembles of 30 cells each. Whole-cell voltage clamp was established by applying suction, and GABA was applied while holding the membrane at -80 mV.
Assay validation was performed by first running a detailed analysis of the response to increasing doses of GABA, in order to determine the EC50 and EC20 concentrations, which were 3.4 µm (literature EC50 value 6 µM) and 1 µM, respectively. Next, we performed concentration-response measurements for known GABA antagonists: Picrotoxin showed an IC50 of 4.6 µM (literature IC50=2.4 µM), bicuculine IC50=15 µM (literature IC50=2 µM) and gabazine IC50=7.2 µM (literature IC50=6.6 µM).
A number of positive modulators were also tested by starting at the GABA EC20 concentration, preincubating with the modulator for one minute, followed by a GABA+modulator co-application. Using this protocol, we obtained an EC50 of 425 nM for diazepam (literature EC50=160 nM) (Figure 3), EC50 of 12 nM for triazolam (literature EC50=22 nM), and an EC50 of 10 nM for zolpidem (literature EC50=30 nM).
The IonFlux system has been validated to screen compounds quickly for their modulation of ligand-gated ion-channel response. The well-plate microfluidic interface allows the user to apply compounds simultaneously to a large number of continuously recorded cell ensembles. In this format, a 384-well plate generates data for seven compounds across 32 patterns yielding 224 unique modulators per plate.
A typical experiment including priming, voltage clamping, and voltage/compound protocol can be performed in 10 minutes. An instrument with a fluid-handling robot can perform six well-plate experiments (1,332 unique compounds) per hour. Because the IonFlux instrument is in a plate-reader format, up to four instruments can be integrated with a single liquid-handling robot. This further maximizes throughput and efficiency by enabling loading of plates on the deck while an assay is being run.
Currently, it is impossible to conduct early screens for fast-acting LGICs. At a throughput of 1,332 unique compounds per hour (tested in duplicate) and a cost per compound of approximately $0.50, the IonFlux system enables true electrophysiology screening that will lead to more relevant hit generation than indirect screening methods.