Sage Journals: Discover world-class research

Abstract

Japanese

Korean

Chinese

We have developed an automated patch clamp module for high-throughput ion channel screening, recording from 384 cells simultaneously. The module is incorporated into a laboratory pipetting robot and uses a 384-channel pipettor head for application of cells and compounds. The module contains 384 amplifier channels for fully parallel recordings using a digital amplifier. Success rates for completed experiments (1- to 4-point concentration–response curves for cells satisfying defined quality control parameters) of greater than 85% have been routinely achieved with, for example, HEK, CHO, and RBL cell lines expressing hNa_V1.7, hERG, Kir2.1, GABA, or glutamate receptors. Pharmacology experiments are recorded and analyzed using specialized software, and the pharmacology of hNa_V1.7 and hERG is described. Fast external solution exchange rates of <50 ms are demonstrated using Kir2.1. Short exposure times are achieved by stacking the external solutions inside the pipette of the robot to minimize exposure of the ligand on the receptor. This ensures that ligand-gated ion channels, for example, GABA and glutamate described in this report, can be reproducibly recorded. Stem cell–derived cardiomyocytes have also been used with success rates of 52% for cells that have a seal resistance of >200 MΩ, and recordings of voltage-gated Na⁺ and Ca²⁺ are shown.

Keywords

automated patch clamp ion channels high-throughput screening voltage-gated ion channels ligand-gated ion channels

Introduction

Patch clamp remains the gold standard for obtaining information about ion channels since it was first described in the 1970s by Neher and Sakmann.¹ However, conventional patch clamp is technically demanding and is unsuitable for high-throughput screening (HTS) experiments. Attempts to automate the technique began in the late 1990s with two approaches being adopted, either a pipette-based automation approach, for example, the Apatchi-1² (Sophion Biolin Scientific, Copenhagen, Denmark) and Flyscreen³ (Flyion GmbH, Tübingen, Germany), or a planar chip approach, for example, the Port-a-Patch^4,5 (Nanion Technologies GmbH, Munich, Germany), Patchliner⁶ (Nanion Technologies), QPatch^2,7 (Biolin Scientific/Sophion A/S, Ballerup, Denmark), PatchXpress^8,9 (Molecular Devices, LLC, Sunnyvale, CA), SyncroPatch 96¹⁰ (Nanion Technologies), IonWorks^11,12 (Molecular Devices), IonWorks Barracuda^13,14 (Molecular Devices), Qube (Biolin Scientific/Sophion A/S), CytoPatch¹⁵ (Cytocentrics AG, Rostock, Germany), and IonFlux¹⁶ (Fluxion Biosciences, South San Francisco, CA). The planar chip approach has been by far the most successful of the two methods and has done much to improve ease of use and increase throughput.^17,18 However, the demands of HTS laboratories require the constant improvement of existing devices to include features such as temperature control, which are available on the Patchliner,¹⁰ IonFlux, and QPatch, and the introduction of new devices to increase throughput, data quality, and flexibility. Devices such as the Patchliner, SyncroPatch 96, QPatch, CytoPatch, and IonFlux promise gigaseal recordings, a desirable attribute for accurate voltage control and reliable pharmacology, but these devices lack the throughput required for true HTS experiments. The IonWorks Quattro and, particularly, the IonWorks Barracuda, with the provider’s claim of up to 6000 data points per hour, address the throughput needs, but with a seal resistance of <100 MΩ, data quality may be compromised. Only the Qube, which was first sold in 2014 with its estimated throughput of 30,000 compounds in 24 h coupled with gigaseal recordings, begins to address both the throughput and quality needs of an HTS laboratory. However, all currently available instruments are stand-alone devices, resulting in potential complications when integrating into existing HTS environments. The SyncroPatch 384PE (PE for Patch Engine) differs from the currently available systems because of its revolutionary modular design. The SyncroPatch 384PE is a patch clamp module ( Fig. 1A ) capable of recording from 384 cells simultaneously with success rates routinely greater than 80% achieved for completed experiments (cells satisfying quality control criteria such as a seal resistance of >500 MΩ and a defined peak current amplitude at the start of the experiment). The module is connected to a digital amplifier (Tecella, Foothill Ranch, CA) with 384 channels. The module can be integrated into existing state-of-the-art pipetting robots with a 384 pipettor arm so that all 384 cells are truly recorded and serviced with solutions in parallel. The pipetting robots that have been used routinely so far are the Biomek FX (Beckman Coulter GmbH, Krefeld, Germany; Fig. 1C–E ) and the CyBi FeliX (CyBio, Jena, Germany; Fig. 1B ). The CyBi FeliX has a smaller footprint than the Biomek FX and therefore is a good choice for laboratories where space is limited. The Biomek FX has a larger footprint, but has the advantage that two PE modules can be incorporated into the system, thus vastly increasing throughput. The quality of the data and the success rate for completed experiments (based on quality control criteria such as seal resistance and current amplitude) do not differ between the systems, and the user is free to choose which pipetting robot best suits his or her laboratory needs based on aspects such as size, prior experience, and throughput requirements. The SyncroPatch 384PE uses a borosilicate glass chip in a 384-well plate format with one or more patch clamp apertures per well.

Figure 1.

(A) Recording module of the SyncroPatch 384PE showing the patch clamp chip loaded onto the chip wagon. For recordings, the chip wagon moves inside the module and the electrodes are inserted for patch clamp measurements. (B) CyBi FeliX (CyBio) showing the SyncroPatch 384PE module positioned for patch clamp measurements. The 384 pipetting head is shown for simultaneous pipetting of all 384 wells. Multiple compound positions separated onto two movable levels reduce space requirements, making this robot compact while maintaining functionality. (C) Biomek FX pipetting robot (Beckmann Coulter) showing the SyncroPatch 384PE module situated for patch clamp measurements. The 384 pipetting head is shown above the module for simultaneous pipetting to all wells. A 384-channel amplifier (Tecella; not shown) ensures that all 384 cells are measured simultaneously. (D) Photograph of the Biomek FX robot during an experiment. SyncroPatch 384PE module, compound plates, and wash solutions can be seen. (E) The 384 pipettor head of the Biomek FX positioned inside the SyncroPatch 384PE module during addition of solutions.

A powerful and easy-to-use data acquisition and analysis software package is paramount to the efficient performance of such a high-throughput device. The data acquisition software, PatchControl 384 (Nanion Technologies), and data analysis software, DataControl 384 (Nanion Technologies), have been developed in collaboration with academic and pharmaceutical HTS laboratories. PatchControl 384 controls the running of the experiment, from positioning of compounds through definition of voltage protocols and basic online analysis functions, to acquisition and display of raw data. A color-coded display of all 384 wells based on, for example, seal resistance, and an expanded view of 1 and 16 wells provides an at-a-glance overview of the quality of the experiment for the user (see Fig. 3 for an annotated screenshot of the PatchControl 384 software during an actual experiment). The user can then toggle between raw data and basic online analysis such as current–voltage plots or current–time plots. The PatchControl 384 software and the software controlling the movement of the pipetting robot, for example, Biomek FX (Beckman Coulter) or CyBi FeliX (CyBio), work together for efficient running of the experiment.

The data analysis software, DataControl 384, is used for subsequent offline analysis of the data. In this software package, data are loaded and displayed in the same way as PatchControl 384 and then concentration–response curves for individual wells are generated. The data from wells can be included or excluded based on quality (e.g., seal resistance > 500 MΩ, RSeries < 25 MΩ, current amplitude >100 pA at the start and/or end of the experiment). Average concentration–response curves for each compound are then generated. The data can be exported in different formats and easily integrated into existing HTS databases. The data analysis is permitted directly after the experiment on the computer where the experiment is being performed, or later offline on a remote computer.

In this report, we describe a new automated patch clamp module that meets the needs of HTS laboratories studying ion channels. We present data recorded with the SyncroPatch 384PE using the Biomek FX pipetting robot from Beckman Coulter. Cells used so far are HEK293, CHO, Ltk, Jurkat, 1321N1, and RBL, with success rates for completed experiments of routinely >85% (cells satisfying certain quality criteria for a 1- to 4-point concentration–response curve lasting approximately 20–30 min). Stem cell–derived cardiomyocytes (iCell cardiomyocytes², Cellular Dynamics International, Madison, WI; Cor.4U, Axiogenesis AG, Cologne, Germany) have also been recorded with success rates of >50% for cells with a seal resistance of >200 MΩ and data are shown.

Materials and Methods

System Components

The SyncroPatch 384PE (Nanion Technologies) described here is an automated patch clamp module in combination with a 384-channel digital amplifier (Tecella), a Biomek FX pipetting robot (Beckman Coulter) with 384 pipettor, software for control of the robot (Biomek Software, Beckman Coulter), and proprietary software for data acquisition (PatchControl 384, Nanion Technologies) and data analysis (DataControl 384, Nanion Technologies). Data output and compound information is compatible with most database formats. For the patch clamp recordings, planar borosilicate glass patch clamp chips were used¹⁸ in a 384-microtiter-plate format. For Patchliner experiments, a Patchliner Octo (Nanion Technologies) was used, including HEKA EPC10 Quadro amplifiers (HEKA Elektronik, Lambrecht, Germany). PatchControlHT (Nanion Technologies), the software for controlling the experiment, was integrated with PatchMaster (HEKA Elektronik) for acquisition of data and online analysis of the recorded data. Data were exported and analyzed offline using IGOR Pro (WaveMetrics, Lake Oswego, OR).

Cell Culture and Harvesting

Standard cell lines

The cell lines used here were cultured as previously described.^5,19-21 In brief, cells were cultured in T75 culture flasks in the media recommended by the supplier and passaged every 2–3 days when they were 50%–80% confluent. The cells were passaged regularly to ensure that the cells were single when passaged and harvested. The cells were prohibited from reaching 100% confluency so that the cells remained healthy, single, and expressed the ion channel of interest when they were harvested into a cell suspension for recordings. For experiments on the SyncroPatch 384PE and Patchliner, cells were harvested into suspension and suction was used to attract a cell to the patch clamp aperture of each well. Since it was a blind method for capture, cells had to be single with not too many clusters and the cell suspension was free from cell debris, as the presence of cell clusters and debris could have decreased the success rate. Cells were typically harvested as described previously^5,19-21 using trypsin, other suitable enzymes, or even enzyme-free detachment protocols. Cells were then resuspended in a mixture of 50% culture media/50% external recording solution at a density of 50,000–500,000 cells/mL.

Stem cell–derived cardiomyocytes

The human-induced pluripotent stem cell–derived cardiomyocytes (iPSCs) were cultured as per the manufacturer’s instructions (iCell cardiomyocytes², Cellular Dynamics International) or were provided by the manufacturer as beating monolayers in T25 culture flasks (Cor.4U, Axiogenesis). Cells were maintained in culture for 2–3 weeks before being used for patch clamp measurements. Cells were harvested as previously described.^10,21

Patch clamp solutions

The internal solution for hERG, Kir2.1, GABA, glutamate experiments, and iPSCs was as follows: 50 mM KCl, 10 mM NaCl, 60 mM KF, 20 mM EGTA, 10 mM HEPES/KOH, pH 7.2. The internal solution for hNa_V1.7 experiments (SyncroPatch 384PE) differed slightly and was composed of 10 mM CsCl, 110 mM CsF, 20 mM EGTA, 10 mM HEPES/CsOH, pH 7.2. The internal solution for hNa_V1.7 experiments on the Patchliner used the following solution: 50 mM CsCl, 10 mM NaCl, 60 mM CsF, 20 mM EGTA, 10 mM HEPES/KOH, pH 7.2. The external recording solution for Kir2.1, hERG IV, GABA, and glutamate experiments used the following composition: 140 mM NaCl, 4 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 5 mM d-glucose monohydrate, 10 mM HEPES/NaOH, pH 7.4. The external recording solution for hERG pharmacology, hNa_V1.7, and iPSCs was as follows: 80 mM NaCl, 60 mM NMDG, 4 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 5 mM d-glucose monohydrate, 10 mM HEPES/NaOH, pH 7.4. A high K⁺ solution for activation of Kir2.1 was made up of 126 mM KCl, 20 mM NaCl, 1 mM MgCl₂, 2 mM CaCl₂, 5 mM d-glucose monohydrate, 10 mM HEPES/NaOH, pH 7.4. GABA (Sigma-Aldrich, St. Louis, MO) was made as a 10 mM stock solution in external recording solution. l-Glutamic acid (monosodium salt monohydrate, Sigma-Aldrich) was made as a 100 mM stock solution in external recording solution. Astemizole (Sigma-Aldrich), cisapride (Sigma-Aldrich), terfenadine (Sigma-Aldrich), and pimozide were made as 1 mM stock solutions in DMSO. Tetracaine (Sigma-Aldrich) was made as a 10 mM stock solution in external recording solution. Pentobarbital was made as a 100 mM stock solution in DMSO. Stock solutions were then diluted to the appropriate concentration in external recording solution prior to experiments.

Electrophysiology

All cells were recorded in the whole-cell mode of the patch clamp technique using the SyncroPatch 384PE incorporated into a Biomek FX pipetting robot (Beckman Coulter) or the Patchliner Octo. On the SyncroPatch 384PE, voltage protocols were constructed and data acquired using PatchControl 384 (Nanion Technologies). For experiments on the Patchliner, voltage protocols were constructed and data acquired using PatchMaster (HEKA Elektronik). Data were analyzed and concentration–response curves were generated using DataControl 384 (Nanion Technologies) or IgorPro (Wavemetrics) for Patchliner experiments. For hERG concentration–response curves on the SyncroPatch 384PE, one concentration of compound was applied to each cell and the concentration–response curve calculated across the whole plate. In this case, IC₅₀ values were calculated from the averaged fits. For hERG pharmacology on the Patchliner, tetracaine block of hNa_V1.7 on the SyncroPatch 384PE, and GABA on the SyncroPatch 384PE, cumulative concentration–response curves were generated and the average IC₅₀ calculated as the mean of the individual plots. In these experiments, the IC₅₀ is given as mean ± SEM. For current–voltage plots, points were plotted as the mean peak amplitude or mean normalized amplitude ± SEM. For hNa_V1.7, the points were fitted using a Boltzmann equation.

Results and Discussion

The SyncroPatch 384PE automated patch clamp module has been developed for high-throughput screening of ion channels important in drug discovery and safety pharmacology. It uses a 384-channel digital amplifier (Tecella), and each cell is clamped individually with a common ground for 32 wells. The SyncroPatch 384PE uses a borosilicate glass substrate for the patch clamp recordings that has already proven successful with the Port-a-Patch,^4,5,18,19 Patchliner,^6,18,19 and SyncroPatch 96.^10,19 The borosilicate glass patch clamp chip has a 384-well plate format, typically with a single hole in each well, and so far the success rate for a seal resistance of >500 MΩ at the start of the experiment for CHO or HEK cells expressing hNa_V1.7 or hERG is >90% (see Figs. 3 and 6 ). Although the wells typically have one hole per well, patch clamp chips containing multiple holes per well are also available. This is particularly useful when current size is very small or not all cells express the ion channel of interest, for example, when transiently transfected cells are used. In this case, the sum of the currents recorded in each well is measured. A seal resistance of 50–200 MΩ can be expected with these kinds of recordings (see Fig. 10 ).

The SyncroPatch 384PE has been incorporated into a Biomek FX (Beckman Coulter) or CyBi FeliX (CyBio) robot with a 384 pipetting arm. With a 384-channel amplifier, 384 pipetting arm, and success rate of >85% for a seal resistance of >500 MΩ (at the start of the experiment) coupled with an experiment time of around 20–30 min for a 4-point concentration–response curve, the SyncroPatch 384PE provides a theoretical throughput for ion channel screening using a patch clamp of 20,000 data points per 8 h day. In addition, two SyncroPatch 384PE modules can be incorporated into one pipetting robot (data not shown), thus almost doubling throughput.

We have performed pharmacology experiments on several different voltage-gated ion channels using the SyncroPatch 384PE routinely with success rates of >80% for completed experiments (using strict quality control parameters described below), and two examples are shown here in detail: hNa_V1.7 and hERG.

The voltage-gated sodium channel, Na_V1.7, found primarily in the peripheral nervous system and thought to play a role in nociception and pain sensing, has been recorded on the SyncroPatch 384PE. The current–voltage plot for an average of 275 CHO cells expressing hNa_V1.7 (Anaxon AG, Berne, Switzerland) recorded in one experiment on the SyncroPatch 384PE is shown in Figure 2A . An example of the raw traces is shown in the inset. The current–voltage plot for an average of eight CHO cells expressing hNa_V1.7 (Anaxon) recorded on the Patchliner is shown in Figure 2B for comparison. The Vhalf of activation calculated from the average current–voltage plot using a Boltzmann equation was −24 mV for the SyncroPatch 384PE. This is in good agreement with the value obtained for the Patchliner (−19 mV) and with values reported in the literature.^22-24 For pharmacology experiments using tetracaine, CHO cells expressing hNa_V1.7 (Anaxon) were used on the SyncroPatch 384PE with a success rate of 95% for cells that have a seal resistance of >500 MΩ at the start of the experiment ( Fig. 3 ). In Figure 3 , a screenshot of the PatchControl 384 software with current traces in response to a voltage step protocol (panel A) and the corresponding online analysis (panel B) are shown. The PatchControl 384 software was developed to be user-friendly and intuitive. There is a color-coded overview of all 384 wells; in the case shown in Figure 3 , the color coding is based on seal resistance, but cell capacitance or series resistance could also be used. The settings for the color coding are set in the bottom left corner; in this case, green wells are cells with a seal resistance of >500 MΩ, blue wells are cells with a seal resistance of >100 MΩ, light blue wells are cells with a resistance of <100 MΩ, and gray wells are disabled. Sixteen wells can be highlighted and are shown on the right of the screen for a slightly larger view of the recordings, and one well is expanded and shown at the bottom of the screen. In this way, the user gains a quick, at-a-glance overview of the quality of the experiment. In panel B, a background of different shades of blue indicates three successive additions of tetracaine at increasing concentrations (150 nM, 1.58 µM, 15.8 µM). The voltage protocol (panel C) consists of two consecutive depolarization steps from −120 to 0 mV. It is designed to show the current response and compound potency on the resting state (cursor C1) as well as the inactivated state (cursor C2) of the channel. Inactivation is achieved by holding the membrane potential for 2 s at 0 mV between C1 and C2 (not displayed, represented by a broken vertical line) and followed by partial recovery at −120 mV for a defined period of 30 ms. It has been previously established that many compounds that act on voltage-gated Na⁺ channels display a state-dependent change in potency.^25,26 For this reason, being able to reliably measure the potency of compounds on different activation states of Na_V1.7 is critical. Importantly, seal resistance and current amplitude remained stable over the course of the experiment, which took 18 min ( Fig. 4 ). Three applications of control solution (including vehicle; 0.1% DMSO) were made to ensure that the current amplitude was stable before application of tetracaine.

Figure 2.

Average current–voltage plot for hNa_V1.7 recorded on the SyncroPatch 384PE and the Patchliner. (A) Average current–voltage plot for 275 CHO cells expressing hNa_V1.7 recorded on the SyncroPatch 384PE is shown with the raw current traces from a single cell. Current was normalized to the maximum peak current, and for each point, mean ± SEM (n = 275) is plotted. As there is not much well-to-well variation of the data, the error bars on some points are not visible. hNa_V1.7 currents were elicited using a voltage step protocol from the holding potential of −120 mV to the test potential for 20 ms. The test potential was −40 mV at the start and increased in 5 mV increments up to 55 mV. The Vhalf of activation calculated using a Boltzmann fit of the averaged IV curve was −24 mV (n = 275). (B) Average current–voltage plot for eight CHO cells expressing hNa_V1.7 recorded on the Patchliner is shown with the raw current traces from a single cell in the inset. The current was normalized to the peak current amplitude, and for each point, mean ± SEM (n = 8 cells) is plotted. hNa_V1.7 currents were elicited using a voltage step protocol from the holding potential of −120 mV to the test potential for 20 ms. The test potential was −60 mV at the start and increased in 10 mV increments up to 50 mV. The Vhalf of activation calculated using a Boltzmann fit of the averaged IV curve was −19 mV (n = 8).

Figure 3.

Screenshot of the PatchControl 384PE software showing an experiment using CHO cells expressing hNa_V1.7 during a voltage step protocol. (A) Raw current traces to a double voltage step protocol from −120 mV to 0 mV are shown in a screenshot of PatchControl 384, the data acquisition software for the SyncroPatch 384PE. An overview of all 384 wells showing the active trace is shown in the top left part of the screen. The wells are color-coded according to, in this case, seal resistance. All cells with a seal resistance of >500 MΩ are shown in green, cells with a seal resistance between 100 and 500 MΩ are shown in blue, cells with a seal resistance of <100 MΩ are shown in light blue, and wells that are disabled are shown in gray. In this experiment, 95% of cells had a seal resistance of >500 MΩ at the start of the experiment. Sixteen wells are highlighted and all recorded sweeps are shown on the right. One cell is highlighted and shown at the bottom middle of the screen, and the statistical analysis parameters are shown at the bottom left of the screen. In this way, the user gains a useful overview of the quality of the experiment at a glance in real time. (B) Online analysis window of PatchControl 384 is shown for the hNa_V1.7 recording shown in A. In this case, the peak current amplitude of the first peak (light blue) and the second peak (dark blue) is plotted against time. The vertical lines and color-coded regions show addition of compound and length of exposure in each compound concentration, respectively. First, control solution without tetracaine was added, shown in white, and then increasing concentrations of tetracaine in different shades of blue: lightest blue 150 nM, followed by 1.58 µM, and finally, darkest blue 15.8 µM (cumulative concentration response). (C) Raw current traces from an exemplar cell showing the voltage step protocol overlaid and the positions of the cursors for online analysis. (D) Bar graph showing seal resistance at the start (dark blue) and end (light blue) of the experiment. In this experiment, 95% of cells had a seal resistance of >500 MΩ at the start of the experiment and 91% of cells had a seal resistance of >500 MΩ at the end of the experiment (after 20 min).

Figure 4.

Stability of recordings on the SyncroPatch 384PE. (A) Seal stability for a representative 16 cells during a recording of CHO cells expressing hNa_V1.7. Shown is seal resistance versus time for the course of a cumulative (3-point) concentration–response curve. The seal stability was stable over the entire 18 min recording. (B) Current stability for a representative 16 cells during a recording of CHO cells expressing hNa_V1.7. Shown is peak amplitude (of the second peak of a two-pulse protocol; see Figure 3 ) versus time. Vehicle was added (0.1% DMSO) to represent a cumulative (3-point) concentration–response curve. The peak amplitude was stable over the entire 18 min recording.

In the subsequent experiment, three different concentrations of tetracaine were added to each cell sequentially and the corresponding concentration response was calculated. Figure 5 shows the analysis of Na_V1.7 inhibition by tetracaine in DataControl 384. After defining strict quality criteria for a successful experiment (seal resistance >500 MΩ, first peak amplitude >550 pA, second peak amplitude >200 pA at the start of the experiment), only valid wells were selected for calculation and analysis of the averaged concentration–response curve. Depending on the chosen cursor, the fitting results are displayed for the block of the resting state (C1, panel B), the inactivated state (C2, panel C), or both (panel D). The averaged fitting results from all valid wells (panel A), as well as the raw traces, time plots of peak amplitude, and concentration–response curves from each individual well (panel D), can be retrieved and exported from the software. As expected,^25,26 tetracaine exhibits state dependence of inhibition of Na_V1.7. For activation from the closed state (holding potential −120 mV), the IC₅₀ was 14.9 ± 0.6 µM (n = 290; C1), whereas the IC₅₀ on the inactivated state (C2) was 266 ± 6 nM (n = 296).

Figure 5.

Analysis of tetracaine block of hNa_V1.7 expressed in CHO cells using DataControl 384. (A) Results file is loaded into DataControl 384 and sorted by compound name. All successful cells are displayed and wells can be included or excluded from the analysis using the check box. In this example, a cumulative concentration–response curve to tetracaine was performed using a double pulse protocol. (B) Average concentration–response curve for tetracaine block of the first peak is shown. The IC₅₀ stated in the figure (13.0 µM) is the IC₅₀ calculated from the average concentration–response curve. The IC₅₀ calculated from the average of the individual plots was 14.9 ± 0.6 µM (n = 290). (C) Average concentration–response curve for tetracaine block of the second peak. The IC₅₀ stated in the figure (249 nM) is the IC₅₀ calculated from the average concentration–response curve. The IC₅₀ calculated from the average of the individual plots was 266 ± 6 nM (n = 296). (D) DataControl 384 software shows the raw data traces, online analysis, and concentration–response curves for each individual cell. The concentration–response curves for the first and second peaks are displayed overlaid. Once the user is satisfied with the data set, the average concentration curves can be generated (shown in panels B and C). Only cells that satisfied the quality control criteria (seal resistance >500 MΩ, first peak current amplitude >550 pA, second peak current >200 pA at the start of the experiment) were used for the concentration–response curve analysis. Using these quality control criteria, the success rate of this experiment was 77% for completed experiments (experiment time 20 min).

The human ether-a-go-go-related gene (hERG) encodes a potassium channel that mediates the delayed rectifying current (IKr) in cardiac myocytes.²⁷ The hERG channel has become an important target for drug safety testing, as compound interaction with this channel has been shown to prolong the QT interval, resulting in torsades de pointes, a potentially fatal ventricular tachycardia.^27,28 Figure 6 shows the current–voltage response of CHO cells expressing hERG (Anaxon) recorded on the SyncroPatch 384PE. Panel A shows the online analysis (peak current vs. voltage) for all 384 cells recorded. Ninety-three percent of cells had a seal resistance of >500 MΩ at the end of the voltage protocol, with a further 5% of cells that had a seal resistance of >100 MΩ. Panel B shows the voltage protocol used to elicit the hERG current and an example of the raw traces. Panel C shows the current–voltage relationship for an average of 322 cells. The current–voltage relationship is consistent with IKr reported in the literature²⁷ and with hERG expressed in HEK cells (Merck Millipore, Darmstadt, Germany) recorded on the Patchliner (see Fig. 8A ). HEK cells expressing hERG were used for pharmacology experiments on the SyncroPatch 384PE. Four known blockers of the hERG channel were applied via the external solution, and the average concentration–response curves are shown in Figure 7 . A summary of the IC₅₀ values is given in Table 1 . The IC₅₀ values agreed well with those reported in the literature^29-37 and with those obtained using the Patchliner ( Fig. 8 , Table 1 ).

Figure 6.

Current–voltage plot for hERG expressed in CHO cells recorded on the SyncroPatch 384PE. (A) Screenshot of the data acquisition software (PatchControl 384) showing online analysis (peak current vs. voltage). More than 93% of cells had a seal resistance of >500 MΩ. (B) hERG was activated using the voltage step protocol shown with a prepulse to 40 mV for 500 ms, followed by the test potential for 500 ms (holding potential −80 mV). The raw current responses from an exemplar cell are also shown. (C) Average current–voltage plot for 322 cells.

Figure 7.

Concentration–response curves for four known hERG blockers. (A) Average concentration–response curve for block of hERG by astemizole. hERG expressed in HEK cells was elicited using a voltage step to −40 mV for 500 ms after a depolarizing step to 60 mV for 500 ms (holding potential −110 mV). Peak amplitude at −40 mV was used for analysis. Each cell was exposed to a single concentration of astemizole and the average concentration–response curve calculated across the whole plate. The block of the hERG current was normalized to the maximum block achieved with the highest concentration of astemizole and fitted with a Hill equation. IC₅₀ = 19.8 nM (n = 315), in good agreement with the literature^29-31 and with the value obtained from the Patchliner (see Fig. 8 ). (B) Average concentration–response curve for block of hERG by pimozide. As described above, each cell was exposed to a single concentration of pimozide and the average concentration–response curve calculated across the whole plate. IC₅₀ = 5.9 nM (n = 301), in good agreement with the literature.^32,33 (C) Average concentration–response curve for block of hERG by cisapride. IC₅₀ = 33.1 nM (n = 297), in good agreement with the literature^30,32-34 and with the value obtained from the Patchliner (see Fig. 8 ). (D) Average concentration–response curve for block of hERG by terfenadine. IC₅₀ = 35.9 nM (n = 321), in good agreement with the literature^30,32,35-37 and with the value obtained from the Patchliner (see Fig. 8 ). Only cells that satisfied the quality criteria (seal resistance >100 MΩ and peak current amplitude >100 pA at the start of the experiment) were used for the concentration–response curve analysis. Using these quality control criteria, the success rate of these experiments was 77%–84% for completed experiments (experiment time 15 min).

Figure 8.

hERG current–voltage plot and pharmacology recorded on the Patchliner. (A) hERG was activated using the voltage step protocol described in Figure 6 ; the raw current traces of an example cell and the current–voltage plot for an average of four cells are shown. (B) Average concentration–response curve for block of hERG by astemizole. hERG expressed in HEK cells was elicited using a voltage step to −40 mV for 500 ms after a depolarizing step to 40 mV for 500 ms (holding potential −80 mV). Peak amplitude at −40 mV was used for analysis. Each cell was exposed to four concentrations of astemizole and the concentration–response curve constructed for each individual cell. The block of the hERG current was normalized to the maximum block achieved with the highest concentration of astemizole and fitted with a Hill equation. The average IC₅₀ value was calculated from the average of the individual IC₅₀ plots, IC₅₀ = 2.4 ± 0.4 (n = 34), in good agreement with the literature.^29-31 (C) Average concentration–response curve for block of hERG by cisapride with the corresponding raw traces from an example cell. The average IC₅₀ value was calculated from the average of the individual IC₅₀ plots, IC₅₀ = 42.8 ± 6.8 (n = 23), in good agreement with the literature.^30,32-34 (D) Average concentration–response curve for block of hERG by terfenadine with the corresponding raw traces from an example cell. IC₅₀ = 9.7 ± 1.0 (n = 25), in good agreement with the literature.^30,32,35-37

Table 1.

IC₅₀ Values for Astemizole, Pimozide, Cisapride, and Terfenadine on hERG-Mediated Currents Recorded on the SyncroPatch 384PE.

Compound	IC₅₀ (nM)	Success Rate (%)	Literature Range	IC₅₀ (nM) Patchliner
Astemizole	19.8 (315)	82	0.9–26^29 -31	2.4 ± 0.4 (34)
Pimozide	5.9 (391)	78	1–18^32,33	n.d.
Cisapride	33.1 (297)	77	6.5–44^30,32-34	42.8 ± 6.8 (23)
Terfenadine	35.9 (321)	84	7–204^30,32,35-37	9.7 ± 1.0 (25)

Shown are IC₅₀ values (number of cells shown in parentheses), success rate for completed experiments (with quality control criteria described in Fig. 7 legend), and the expected literature IC₅₀ values. In the final column, IC₅₀ values recorded on the Patchliner are provided for comparison (IC₅₀ given as mean ± SEM). All IC₅₀ values recorded on the SyncroPatch 384PE agree well with the literature values.^29-37 n.d. = not done. The IC₅₀ values recorded on the SyncroPatch 384PE were calculated across the whole plate (single concentration of compound per well) and the IC₅₀ taken from the average concentration–response curve; therefore, SEM values cannot be meaningfully calculated. IC₅₀ values for the Patchliner were calculated using cumulative compound addition (all concentrations on all wells) and the mean IC₅₀ calculated from the individual concentration–response curves.

In order to test external fluidic exchange times on the SyncroPatch 384PE, the inward rectifying potassium channel, Kir2.1, endogenously expressed in RBL cells (ATCC, Manassas, VA) was used. The external solution was exchanged from normal (low) K⁺ to an external solution containing high K⁺, and recordings were made at a constant holding potential of −80 mV. Figure 9 shows a recording from an example cell. The exchange time was approximately 50 ms. This exchange time is fast enough to reliably and reproducibly record GABA_A-mediated responses shown in Figure 10 . Experiments were performed using a stacked solutions approach where wash solution was first aspirated into the pipette followed by ligand so that the ligand was rapidly washed away after application to minimize exposure time and desensitization. This approach has already proven successful with the Patchliner^18,19 and SyncroPatch 96¹⁹ for minimizing exposure time by avoiding lengthy solution pickup times. The patch clamp chip utilized a multihole per well format with four holes per well, as opposed to one hole per well. Figure 10A shows the PatchControl 384 software where HEK cells expressing the GABA_A receptor subunit combination α₁β₂γ₂ were activated three consecutive times with 5 µM GABA. The traces overlay almost exactly. Figure 10B shows an example of a cumulative concentration–response curve for GABA, and Figure 10C shows the concentration–response curve for GABA activation for an average of 328 cells recorded in one experiment revealing an EC₅₀ for GABA activation of 3.5 ± 0.2 µM (n = 328), in good agreement with values reported in the literature.³⁸ A double stacked solutions approach was also used to investigate modulation of the GABA_A receptor–mediated current. In the experiment shown in Figure 10D , 100 µM pentobarbital plus GABA (1 µM) was first aspirated into the pipette, followed by GABA alone (1 µM). In this way, receptors were first activated by the low concentration of GABA (1 µM), and this response was enhanced by application of the potentiator.

Figure 9.

Solution exchange time measured on the SyncroPatch 384PE using Kir2.1 expressed in RBL cells. The external solution was exchanged to a high K⁺-containing solution and then back into normal external recording solution; the holding potential remained constant at −80 mV. Shown is the current response of an example cell. The solution exchange time for wash in, that is, the time to peak, was 50 ms.

Figure 10.

GABA_A receptor–mediated responses on the SyncroPatch 384PE. (A) GABA_A receptors (α₁β₂γ₂) expressed in HEK cells were activated by 5 µM GABA using a constant holding potential of −80 mV. Chips with four holes per well were used for the experiments and solutions were stacked inside the pipette; currents were activated by GABA and the ligand was quickly washed away to minimize exposure time. The GABA_A receptor–mediated currents could be reproducibly activated. Shown are current responses to three repetitive applications of 5 µM GABA. The three current responses overlay almost exactly. (B) Representative traces showing a 4-point concentration curve to GABA performed on a single cell. GABA was applied for 100 ms in increasing concentrations from 100 nM to 10 µM. (C) Average concentration–response curve for GABA; the EC₅₀ calculated from the average of the individual concentration–response curves was 3.5 ± 0.2 µM (n = 328). (D) Positive allosteric modulation of GABA_A receptor–mediated responses. Shown are current responses to 1 µM GABA followed by coapplication with 100 µM pentobarbital. The GABA-activated response is increased approximately sixfold by the modulator. Multihole chips with four holes per well were used for the experiment.

By using a stacked solutions approach coupled with aspiration of small ligand volumes and a fast pipetting speed, exposure time is low enough to reliably and reproducibly record fast-desensitizing ligand-gated receptors such as glutamate receptors. Figure 11 shows GluA2 receptors expressed in HEK cells (University of Sussex, UK) and repetitive activation by glutamate. Raw data traces and corresponding online analysis of the peak amplitude over time of an exemplary cell are shown in Figure 11A and B , respectively. Receptors were activated with 100 µM Na-glutamate repeated three times, which resulted in inward currents of similar peak amplitudes. The current onset time was approximately 10 ms. Figure 11 (C,D) shows an example of a cumulative concentration–response curve for Na-glutamate (in mM: 0.1, 0.3, and 1).

Figure 11.

Fast solution exchange on the SyncroPatch 384PE exemplified by recordings of GluA2 receptor expressing HEK cells. (A–D) Recordings were performed using single hole chips. GluA2 receptor–mediated currents were evoked by Na-glutamate using the stacked solutions procedure. Cells were held at a constant voltage of −80 mV. (A,B) Raw current traces and time plot of one exemplar cell are shown. The current responses were highly reproducible in response to three repetitive applications of 100 µM Na-glutamate. (C) Representative traces showing current responses to increasing concentrations from 100 µM to 1 mM Na-glutamate performed on a single cell. (D) Enlargement of current traces shown in C reveals a fast current onset. The rapid exchange time to Na-glutamate (15 ms) facilitates the activation of these fast-desensitizing receptors.

In addition to cell lines that have been successfully used on the SyncroPatch 384PE, human iPSCs (iCell cardiomyocytes², Cellular Dynamics International; Cor.4U, Axiogenesis) have also been recorded. Figure 12 shows a recording of iCells and Cor.4U cells on the SyncroPatch 384PE. In this experiment, half the plate (columns 1–12) received iCell cardiomyocytes² and the second half of the plate (rows 13–24) received Cor.4U cells. Shown are current responses to a voltage step protocol in Na⁺- and Ca²⁺-containing external solution. In these preliminary experiments, the success rate was approximately 50% for cells with a seal resistance of >200 MΩ at the start of the experiment. In most cells, both a fast and a slower component can be seen. The fast component is most likely mediated by the Na_V1.5 channel and the slower component mediated by the voltage-gated L-type calcium channel. Optimization of cell culture and harvesting techniques is required to improve the success rate (based on seal resistance) before these cells can be routinely used in an HTS setting.

Figure 12.

Voltage-gated Na⁺ and Ca²⁺ currents recorded from human iPSCs on the SyncroPatch 384PE. Screenshot of the PatchControl 384 software is shown during the recording of iCell cardiomyocytes² (Cellular Dynamics International; rows 1–12) and Cor.4U cells (Axiogenesis; rows 13–24). Currents were elicited using a voltage step protocol from −80 to 50 mV increasing in 10 mV steps (intersweep holding potential −100 mV). Shown is the current response at 30 mV. In this case, the color coding is based on seal resistance where all cells with a seal resistance of >200 MΩ are shown in green, cells with a seal resistance between 100 and 200 MΩ are shown in blue, cells with a seal resistance of <100 MΩ are shown in light blue, and wells that are disabled are shown in gray. In this experiment, more than 50% of cells had a seal resistance of >200 MΩ regardless of cell type (iCell cardiomyocytes² or Cor.4U). Sixteen wells are highlighted and are shown on the right. One cell is highlighted and shown in the bottom middle of the screen, and the statistical analysis parameters are shown at the bottom left of the screen. Recordings were made in a Na⁺- and Ca²⁺-containing solution, and in most cells, both a fast inactivating inward Na⁺ current and a slower inactivating Ca²⁺ current could be observed.

In conclusion, the SyncroPatch 384PE is a new modular system for automated patch clamp recordings that can be incorporated into different pipetting robotic systems for high-throughput recordings of ion channels. The system is capable of recording voltage-gated ion channels expressed in cell lines where a success rate of more than 80% for completed experiments (with defined quality control criteria) has been achieved. Additionally, external fluidic exchange is fast (<50 ms), and this coupled with low exposure times due to a stacked solutions approach ensures that ligand-gated ion channels expressed in cell lines can also be reliably recorded. The module contains 384-channel digital amplifier channels and a 384 pipettor head, ensuring that 384 cells are recorded truly in parallel. This, in combination with success rates for completed experiments of routinely more than 80% with strict quality control criteria and fast turnaround of experiments (each experiment took approximately 20 min to complete), results in an anticipated throughput of around 20,000 data points per 8 h day. Due to its modular design, two SyncroPatch 348PE modules may be incorporated into a single pipetting robot, thus almost doubling throughput. Since the SyncroPatch 384PE module is not restricted to a particular type of pipetting robot, it can be integrated into existing, preferred pipetting robots, making integration into HTS pathways considerably easier than is currently possible with available stand-alone automated patch clamp systems. The data quality, throughput, and flexibility of the system make it highly compatible with screening large compound libraries on ion channels.

Footnotes

Acknowledgements

We thank Anaxon for providing the CHO hNa_V1.7 cell line, Merck Millipore for providing the HEK hERG cell line (for Patchliner experiments), Cellular Dynamics International and Axiogenesis for the human iPSCs (iCell cardiomyocytes², Cor.4U), and John Atack of the University of Sussex, Brighton, UK, for the HEK GluA2 cell line. We also thank Bayer AG and Alfred George, Northwestern University, Chicago, for useful input in development of the software.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Neher

Sakmann

Single Channel Currents Recorded from Membrane of Denervated Frog Muscle Fibres. Nature 1976, 260, 799–802.

Asmild

Oswald

Krzywkowski

K. M.

. Upscaling and Automation of Electrophysiology: Toward High Throughput Screening in Ion Channel Drug Discovery. Recept. Channels 2003, 9 (1), 49–58.

Lepple-Wienhues

Ferlinz

Seeger

. Flip the Tip: An Automated, High Quality, Cost-Effective Patch Clamp Screen. Recept. Channels 2003, 9 (1), 13–17.

Brüggemann

George

Klau

. High Quality Ion Channel Analysis on a Chip with the NPC-Technology. Assay Drug Dev. Technol. 2003, 1, 665–673.

Brüggemann

George

Klau

. Ion Channel Drug Discovery and Research: The Automated Nano-Patch-Clamp Technology. Curr. Drug Discov. Technol. 2004, 1, 91–96.

Brüggemann

Stoelzle

George

. Microchip Technology for Automated and Parallel Patch Clamp Recording. Small 2006, 2, 840–846.

Mathes

Friis

Finley

. QPatch: The Missing Link between HTS and Ion Channel Drug Discovery. Comb. Chem. High Throughput Screen. 2009, 12, 78–95.

Tao

Santa Ana

Guia

. Automated Tight Seal Electrophysiology for Assessing the Potential hERG Liability of Pharmaceutical Compounds. Assay Drug Dev. Technol. 2004, 2, 497–506.

Guia

Rothwarf

. A Benchmark Study with Sealchip Planar Patch-Clamp Technology. Assay Drug Dev. Technol. 2003, 1, 675–684.

10.

Stoelzle

Obergrussberger

Brüggemann

. State-of-the-Art Automated Patch Clamp Devices: Heat Activation, Action Potentials, and High Throughput in Ion Channel Screening. Front. Pharmacol. 2011, 2, Article 76. DOI: 10.3389/fphar.2011.00076. http://journal.frontiersin.org/article/10.3389/fphar.2011.00076.

11.

Schroeder

Neagle

Trezise

D. J.

. Ionworks HT: A New High-Throughput Electrophysiology Measurement Platform. J. Biomol. Screen. 2003, 8, 50–64.

12.

Finkel

Wittel

Yang

. Population Patch Clamp Improves Data Consistency and Success Rates in the Measurement of Ionic Currents. J. Biomol. Screening 2006, 11, 488–496.

13.

Gillie

D. J.

Novick

S. J.

Donovan

D. T.

. Development of a High-Throughput Electrophysiological Assay for the Human Ether-à-go-go Related Potassium Channel hERG. J. Pharmacol. Toxicol. Methods 2013, 67 (1), 33–44.

14.

Kuryshev

Y. A.

Brown

A. M.

Duzic

. Evaluating State Dependence and Subtype Selectivity of Calcium Channel Modulators in Automated Electrophysiology Assays. Assay Drug Dev. Technol. 2014, 12 (2), 110–119.

15.

Scheel

Himmel

Rascher-Eggstein

. Introduction of a Modular Automated Voltage-Clamp Platform and Its Correlation with Manual Human Ether-à-go-go Related Gene Voltage-Clamp Data. Assay Drug Dev. Technol. 2011, 9 (6), 600–607.

16.

Spencer

C. I.

Nianzehn

Johnson

. Ion Channel Pharmacology under Flow: Automation via Well-Plate Microfluidics. Assay Drug Dev. Technol. 2012, 10 (4), 313–324.

17.

Dunlop

Bowlby

Peri

. High-Throughput Electrophysiology: An Emerging Paradigm for Ion-Channel Screening and Physiology. Nat. Rev. Drug Discov. 2008, 7, 358–368.

18.

Farre

Haythornthwaite

Haarmann

. Port-a-Patch and Patchliner: High Fidelity Electrophysiology for Secondary Screening and Safety Pharmacology. Comb. Chem. High Throughput Screening 2009, 12, 24–37.

19.

Obergrussberger

Haarmann

Rinke

. Automated Patch Clamp Analysis of nAChα7 and Na_V1.7 Channels. Curr. Protoc. Pharmacol. 2014, 65, 11.13.1–11.13.48. DOI: 10.1002/0471141755.ph1113s65.

20.

Brüggemann

Farre

Haarmann

. Planar Patch Clamp—Advances in Electrophysiology. In Methods in Molecular Biology—Potassium Channels; Lippiat

J. D.

, Ed.; Humana Press, a part of Springer Science and Business Media: Totowa, NJ, 2008; Vol. 491, pp 165–176.

21.

Becker

Stoelzle

Göpel

. Minimized Cell Usage for Stem Cell-Derived and Primary Cells on an Automated Patch Clamp System. J. Pharmacol. Toxicol. Methods 2013, 68, 82–87.

22.

Sangameswaran

Fish

L. M.

Koch

B. D.

. A Novel Tetrodotoxin-Sensitive, Voltage-Gated Sodium Channel Expressed in Rat and Human Dorsal Root Ganglia. J. Biol. Chem. 1997, 272 (23), 14805–14809.

23.

Vijayaragavan

O’Leary

M. E.

Chahine

Gating Properties of Nav1.7 and Nav1.8 Peripheral Nerve Sodium Channels. J. Neurosci. 2001, 21 (20), 7909–7918.

24.

Cummins

T. R.

Dib-Hajj

S. D.

Waxman

S. G.

Electrophysiological Properties of Mutant Nav1.7 Sodium Channels in a Painful Inherited Neuropathy. J. Neurosci. 2004, 24 (38), 8232–8236.

25.

Bean

B. P.

Cohen

C. J.

Tsien

R. W.

Lidocaine Block of Sodium Channels. J. Gen. Physiol. 1983, 81, 613–642.

26.

Kaczorowski

G. J.

Garcia

M. L.

Bode

. The Importance of Being Profiled: Improving Drug Candidate Safety and Efficacy Using Ion Channel Profiling. Front. Pharmacol. 2011, 2, Article 78. DOI: 10.3389/fphar.2011.00078. http://dx.doi.org/10.3389/fphar.2011.00078.

27.

Sanguinetti

M. C.

Jiang

Curran

M. E.

. A Mechanistic Link between an Inherited and an Acquired Cardiac Arrhythmia: HERG Encodes the IKr Potassium Channel. Cell 1995, 81, 299–307.

28.

Curran

M. E.

Splawski

Timothy

K. W.

. A Molecular Basis for Cardiac Arrhythmia: HERG Mutations Cause Long QT Syndrome. Cell 1995, 80, 795–803.

29.

Zhou

Vorperian

V. R.

Gong

. Block of HERG Potassium Channels by the Antihistamine Astemizole and Its Metabolites Desmethylastemizole and Norastemizole. J. Cardiovasc. Electrophysiol. 1999, 10 (6), 836–843.

30.

Wang

Della Penna

Wang

. Functional and Pharmacological Properties of Canine ERG Potassium Channels. Am. J. Physiol. Heart Circ. Physiol. 2003, 284, H256–H267.

31.

Chiu

P. J. S.

Marcoe

K. F.

Bounds

S. E.

. Validation of a [3H]Astemizole Binding Assay in HEK293 Cells Expressing HERG K+ Channels. J. Pharmacol. Sci. (Tokyo, Jpn.) 2004, 95, 311–319.

32.

Kirsch

G. E.

Trepakova

E. S.

Brimecombe

J. C.

. Variability in the Measurement of hERG Potassium Channel Inhibition: Effects of Temperature and Stimulus Pattern. J. Pharmacol. Toxicol. Methods 2004, 50, 93–101.

33.

Kang

Wang

Cai

. High Affinity Blockade of the HERG Cardiac K(+) Channel by the Neuroleptic Pimozide. Eur. J. Pharmacol. 2000, 392 (3), 137–140.

34.

Walker

B. D.

Singleton

C. B.

Bursill

J. A.

. Inhibition of the Human Ether-a-go-go-Related gene (HERG) Potassium Channel by Cisapride: Affinity for Open and Inactivated States. Br. J. Pharmacol. 1999, 128, 444–450.

35.

Crumb

W. J.

Jr.

Loratadine Blockade of K1 Channels in Human Heart: Comparison with Terfenadine under Physiological Conditions. J. Pharmacol. Exp. Ther. 2000, 292 (1), 261–264.

36.

Lacerda

A. E.

Kramer

Shen

K.-Z.

. Comparison of Block among Cloned Cardiac Potassium Channels by Non-Antiarrhythmic Drugs. Eur. Heart J. Suppl. 2001, 3 (Suppl. K), K23–K30.

37.

Davie

Pierre-Valentin

Pollard

. Comparative Pharmacology of Guinea Pig Cardiac Myocyte and Cloned hERG (I(Kr)) Channel. J. Cardiovasc. Electrophysiol. 2004, 15 (11), 1302–1309.

38.

Ebert

Thompson

S. A.

Saounatsou

. Differences in Agonist/Antagonist Binding Affinity and Receptor Transduction Using Recombinant Human g-Aminobutyric Acid Type A Receptors. Mol. Pharmacol. 1997, 52, 1150–1156.

Automated Patch Clamp Meets High-Throughput Screening

Abstract

Keywords

Introduction

Materials and Methods

System Components

Cell Culture and Harvesting

Standard cell lines

Stem cell–derived cardiomyocytes

Patch clamp solutions

Electrophysiology

Results and Discussion

Footnotes

Acknowledgements

Declaration of Conflicting Interests

Funding

References