A Columbia University-led team recently secured nearly $1.5 million in National Human Genome Research Institute funding to develop integrated sensor systems that can be coupled with both solid-state and protein-based nanopore sequencing systems — work that they hope will speed up the detection of signals in these systems, while also collecting signals from multiple pores at once.
"The goal of this work is to build high-bandwidth electronic interfaces to these nanopore transducers, both solid-state and biological," Columbia University electrical engineering researcher Kenneth Shepard told In Sequence. "High bandwidth, here, means high temporal resolution: the ability to resolve and see things that are occurring very fast."
The group is also continuing to improve on its amplifier design to reduce the noise associated with multiplexed nanopore systems and is looking at ways of adding more amplifier channels without compromising on electronics quality and bandwidth.
"If you want to try to make these amplifiers very small, you're often … forced to tolerate a higher noise amplifier that you have to filter to shorter bandwidths or lower bandwidths because of the additional noise that you're picking up," Shepard explained.
"So our goal is to be able to try to make very high quality amplifiers, but also do it in a way that we can scale it down and make them smaller and still be able to preserve that high quality," he said.
Shepard is the lead investigator on the three-year project, which last month was awarded $1.47 million through NHGRI's Advanced DNA Sequencing Technology Program (see GWDN 9/14/2012). Also collaborating on the effort is Marija Drndic, a physics researcher from the University of Pennsylvania, who brings her group's expertise on solid-state nanopore sequencing to the mix (IS 8/3/2010).
The effort builds on the complementary metal-oxide semiconductor, or CMOS, current preamplifier strategy that Shepard, Drndic, and others described in Nature Methods this spring (>IS 3/27/2012). That study focused on integrating CMOS electronics with solid-state nanopores.
The team is now working on similar electronic systems that are compatible with nanopore sequencing systems regardless of whether solid-state or protein pores are used as the transducers.
"The goals of integrating the nanopore transducers with CMOS electronics are two-fold: the first is to bring multiplexing capability in so you can parallelize the platform and the other important goal is to improve the quality of the amplification," Shepard said.
In particular, he explained, the group is interested in coming up with an amplifier with a good signal-to-noise ratio that operates at very high bandwidth frequencies. At these higher bandwidths, it's possible to get better temporal resolution of the signals generated by each base as DNA moves through the pore, allowing for faster sequencing.
"Traditionally, these measurements are done using very low bandwidth electronics — basically electrophysiological amplifiers — that have, at best, maybe 50 to 100 kilohertz of bandwidth, typically more like 50 [kilohertz]," Shepard said, "which really doesn't let you resolve very much temporally."
"We're talking about expanding that to a megahertz or even more — maybe up to 10 megahertz bandwidth, which would let you resolve things at the microsecond or sub-microsecond scales," he explained.
Ideally, the researchers are keen to come up with electronics that are fast enough to allow the current associated with each nucleotide to be measured in real time as DNA moves through a nanopore transducer.
Shepard cautioned that it's still not clear whether it will be possible to reliably detect signals fast enough to sequence unfettered DNA moving through nanopores, though "it's not inconceivable."
In the meantime, various teams having been coming up with methods to slow DNA's transit through the nanopore so that the current signals associated with individual bases in the molecule can be read, generally using approaches that ratchet DNA through the pore (IS 3/27/2012).
"Right now, most of the approaches that are out there in trying to use nanopore transducers for sequencing really rely on slowing down the reaction, the translocation, very significantly," Shepard noted. "Because they don't have the electronics to keep up with anything that isn't slowed down very considerably."
Even in nanopore sequencing systems where the speed of DNA translocation is kept in check using some sort of ratcheting system, improvements on the electronics side are expected to offer an edge in terms of sequencing speed and accuracy.
That's because many of these DNA-slowing systems rely on enzymes that operate stochastically, meaning some reactions may occur more quickly or slowly than others, Shepard explained. If a nanopore system is coupled with electronics that are not fast enough to capture the most rapid of these DNA translocation events, signals from bases passing through the pore during those speedy translocations would be missed.
"Even in the context in which you're able to slow things down, there's still a tremendous value to have higher bandwidth electronics," Shepard said.
"And ultimately if the higher bandwidth electronics let you run the reactions nominally faster," he added, "that has advantages for throughput and everything else."
The electronics work stemmed from developments being described on the sequencing technology side, Shepard explained, where many have expressed interest in coming up with platforms capable of quickly sequencing single molecules of DNA.
That move has helped spawn a flurry of research into nanopore sequencing systems that rely on electrochemical signals rather than the optical detection of fluorescently labeled bases, he noted, since the signal-to-noise ratios associated with a single fluorophore are "very severe," making optical signals tricky to detect and to scale up in a single-molecule setting.
In contrast, studies have demonstrated that it's possible to produce signals for bases in single strands of DNA that are many orders of magnitude higher using nanopore-based electrochemical approaches, Shepard said.
"It's those higher signal levels from the electronic approaches that really win the day," he explained. "You can use that additional signal to do all sorts of things: improve signal-to-noise, improve your bandwidth for detection, and lots of other stuff."
But taking advantage of that additional signal level and capturing it in a multiplexed system that reads many nanopores in parallel inevitably presents electronics challenges. And that's where CMOS integration comes in, Shepard said.
Whereas measurements in early nanopore experiments have often relied on relatively low bandwidth amplifier systems initially designed for electrophysiological experiments, he and his team are attempting to optimize the electronics for nanopore sequencing to harness as much available signal as possible.
As described previously, the team has been working on an integrated CMOS preamplifier that bumps up the signal-to-noise ratio, in part by dialing the noise associated with a process called parasitic capacitance, which can occur when ions or electrons build up on either side of a nanopore membrane.
Now, the researchers are using a similar CMOS-based strategy to design a modular system that can be easily paired with either solid-state or protein-based nanopores.
For the most part, solid-state and protein nanopores present similar challenges, Shepard said, though the current levels associated with each are somewhat different in their present iteration. In particular, he said, the existing protein nanopore systems tend to have significantly lower current levels than those based on solid-state nanopores.
Since the goal is to create a system that allows users to swap back and forth between solid-state and biological pores using the same amplifier — or even to mix and match the two — researchers are focused on developing a front-end amplifier that can be reconfigured "on the fly" to deal with the specific characteristics of the nanopore transducers that it interfaces with, Shepard explained.
"We've made the amplifiers very modular," he said, "so that they can adjust to whatever the electrical characteristics are of the transducer that they're mating to."
Shepard did not comment on whether his group is currently collaborating with any of the companies that are working on developing commercial nanopore systems. In March, the team told IS that it had patented its CMOS current preamplifier through Columbia University and was in talks with parties who had expressed interest in licensing this technology.