Planar systems employ a flat substrate through which one or several apertures, functioning as microelectrodes, have been bored. This configuration opens planar systems to automation, because those holes can be bored in microtiter plate–like consumables, and cells can be patched via vacuum. The IonWorks Quattro from MDS Analytical Technologies has 64 holes per well of a 384-well PatchPlate, and records the average signal from 64 cells in a strategy the company calls “population patch clamping.” Similarly, the PatchXpress (also from MDS) can measure 16 individual cells at once, and the QPatch HT (from Sophion) can record from 48. Two other systems set to launch in 2010—Nanion's SynchroPatch and Cellectricon's Dynaflow HT—will record 96 cells in parallel.Channeling Drug Discovery Possibility
At the moment, electrophysiology, or “ephys,” is a field in transition. Using automated patch clamp platforms, companies like ChanTest can record from thousands of cells per day, introducing ephys far earlier into the drug development pipeline than ever before. AstraZeneca routinely subjects libraries of thousands of chemical compounds to similar analyses, says Dabrowski.
Yet that's still not enough to make electrophysiology a first-tier screening application, says Dabrowski. “It's about one digit off; a factor of 10 [is] still needed to do high-throughput screening on ephys platforms.”
According to Chris Mathes, vice president and general manager of Sophion, the highest quality data come from gigaohm or “gigaseal” connections, because only the current through the channel can flow to the electrode. Weaker “megaohm” seals are “leaky,” he says, meaning potential loss of signal, baseline fluctuation, and noise.
“There are tricks you can use to get around that,” says Mathes of megaohm's setbacks, “but most people agree that, if you can do gigaseal recording, it's better.”
Sophion's QPatch systems all produce gigaohm seals, as does MDS's Patch X-press and Nanion's SynchroPatch. MDS's IonWorks and Cellectricon's Dynaflow systems generate megaohm seals. The difference is in the planar substrate: silicon and glass can produce a gigaseal, and plastics usually cannot. But plastics are cheaper to manufacture. “If you change the substrate, you lose the gigaseal, but you can gain in throughput,” Mathes observes.
Another distinction among the various high-throughput systems lies in the fluidics. While some systems are well-based, the QPatch and Dynaflow use microfluidics for rapid fluid exchange across the cell. This feature is particularly useful in studies of ligand-gated channels because they often desensitize (that is, stop responding) to a compound, sometimes within milliseconds. As MDS's Yamane puts it, desensitization is like a baseball cap. “When you first put your cap on, you can feel it. Later, you forget it's there.”
In the case of the Dynaflow HT, micro-fluidics enable millisecond-scale solution exchange, which enables data collection before a receptor can desensitize. “We have essentially perfect control of the solution environment around the cell,” says Mattias Karlsson, chief technical officer at Cellectricon.
Developed in conjunction with Astra-Zeneca, the Dynaflow HT is not, per se, planar. Instead, the system integrates electrical and microfluidic circuitry in a microtiter plate format to probe up to six cells in parallel in each of 16 experimental modules. “It is more like a lateral patch clamp system,” says Karlsson.
The Dynaflow HT system can collect up to 10,000 datapoints in 8 to 12 hours, notes Karlsson, compared to 2,000 per 6-hour shift on the IonWorks Quattro. But for Dabrowski, the calculus for determining the best platform is more complicated—a balance of throughput, cost, and reliability. Automated ephys platforms are simply too expensive for standard HTS at a company like AstraZeneca, he says. “Where I see the best fit for Dynaflow is in concentration-response screening.”
At the National Institutes of Health (NIH)–funded Johns Hopkins Ion Channel Center, director Min Li's compound-discovery workflow combines an IonWorks Quattro and Hamamatsu FLIPR (fluorometric imaging plate reading) system with Corning's label-free Epic instrument, which detects morphologic changes resulting from drug treatment. In one case, Li's team screened a 300,000-compound library for modulators of hERG proteins— voltage-gated potassium channels that help maintain cardiac rhythm.
Starting with an indirect fluorescent FLIPR assay on engineered cells expressing only the channel of interest, the team whittled their library down to only a few thousand possibilities. They then profiled those compounds on the IonWorks to understand each compound's pharmacology. Using the Epic device to assess the drugs' activities in native, non-engineered cells, the team ended with a final tally of 30 leads.
If Johns Hopkins can do that, so too can pharmaceutical companies. Indeed, at least one such compound is in phase 2 trials: AstraZeneca's AZD1386, a capsaicin receptor antagonist. From this perspective, then, ion channel drug development is right on schedule: it takes 10 to 15 years to develop a drug, and automated ephys systems only entered the market in 2002. “It's likely that in the next five years or so, we will see new drugs developed based on automated patch clamp,” says Mathes. “For the first time, drug companies are, in a more serious way, looking at ion channel targets as real targets of drug development.”