NEW YORK (GenomeWeb) – A team led by University of Fribourg researchers has demonstrated the ability to use nanopores for protein identification and quantitation.
Detailed in a paper published this week in Nature Nanotechnology, the work indicates that nanopores could prove capable of detecting proteins at the single-molecule level, said Michael Mayer, chair of biophysics at the University of Fribourg's Adolphe Merkle Institute and corresponding author on the study. Progressing the technology to this point will require overcoming a number of technological and scientific hurdles, but the new study suggests that it could be possible, he added.
In the study, Mayer and his colleagues used a nanopore to distinguish between a glucose-6-phosphate dehydrogenase protein and a G6PDH protein complexed with an IgG antibody. Mayer said that while distinguishing between two proteins might not seem particularly momentous, the fact that the researchers were able to do it with data from just one translocation through the nanopore was significant.
Typically, experiments using nanopores to detect proteins do so by combining signal from many translocations of copies of the same protein fro a pure solution, he said. "People don't record just a single signal. They record hundreds and thousands of them, and use a distribution of those measurements" to characterize their target.
"What we have shown here is that from a single translocation of just this one protein you can get enough information to distinguish it from another," Mayer said. "And that is critical if you want to eventually be able to quantify what you have in a mixture. You have to be able to do it using a single [translocation event.] When you are looking at a mixture you can't do this whole distribution thing anymore."
Such an approach faces significant hurdles in terms of robustness and reproducibility, though, Mayer noted. "It is very, very challenging to get the exact correct value of the protein volume, shape, or dipole, for instance, each time because there are a number of things that can happen [when translocating a target protein through a nanopore] that will give you, on a single-molecule basis, relatively large variability, whereas combing many translocations averages this [variability] out."
Mayer said he and his colleagues are attacking this problem from two angles: increasing the number of parameters they measure for each protein, and improving the reproducibility of the nanopores and nanopore-based measurements themselves.
In the Nature Nanotechnology study, the researchers measured the proteins along five dimensions: shape, volume, charge, rotational diffusion coefficient, and dipole moment. The more parameters they are able to measure, the better they can identify specific proteins as they move through the nanopore, Mayer said, noting that he would like to add two or three more parameters to their measurements.
At the same time, improving the nanopores themselves will allow the researchers to measure these parameters more accurately and reproducibly, which will also boost the platform's performance.
One of Mayer's main areas of focus is increasing signal to noise while also increasing the platform's bandwidth. Currently, his group uses mobile ligands attached to the pores to anchor the proteins of interest as they pass through, slowing them down. Otherwise, he said, the proteins would move through the pore too quickly for the platform to collect enough identifying data.
Were they able to improve the measurement's time resolution, they might be able to do away with this tethering system, he said. Increased recording bandwidth comes at the cost of signal to noise, though, meaning that to up the device's performance the researchers "have to improve [the system] on multiple fronts," Mayer said.
To do this, Mayer and his team are experimenting with glass or quartz substrates, as opposed to the silicon substrates commonly used for such work. The different conductivity of glass and quartz allows for a significant increase in bandwidth over silicon-based devices, while maintaining existing levels of signal-to-noise.
The issue, Mayer said, is that while nanofabrication techniques using silicon are very well established, they are less so for glass and quartz. "The semiconductor industry is not good at fabricating in glass," he said. However, he added, it is "absolutely technically doable. We just need to do it or contract someone to do it, and that is currently an ongoing effort."
He predicted that in five years a significant proportion of nanopore chips will be glass- or quartz-based.
In addition to these sorts of engineering challenges, there is room for better scientific understandings to improve nanopore-based protein analysis, Mayer said. In particular, he said, more sophisticated approaches to modeling protein shape would be useful.
"We currently describe proteins as shaped like, say, a rugby ball or a lentil," he said. "We haven't made any attempt yet at describing the particle much more realistically. It's of course not a lentil, not a rugby ball shape. It has fine structure with little corners, protrusions, and valleys here and there."
"You can imagine describing these shapes much more realistically," he added. "For example, instead of describing it as one spheroid, use six beads arranged in some three-dimensional shape. And the volume, the dipole moment measurement — everything will become more accurate once we describe the shape more accurately."
Initially, Mayer said, nanopores will most likely be used for narrowly targeted protein analyses, though he noted that even this remains a ways off.
"There's no way you can throw a mixture of 10,000 proteins onto this thing and hope that you're going to figure it out — not right now," he said. "I think there is some hope that we will be able to fish out proteins of interest and then quantify and characterize them, which could be, I think, pretty helpful."
In the long run, though, Mayer said he hoped nanopore-based analysis could be applied to complex mixtures of proteins.
"My dream would be that every life science lab has this little benchtop device where you stick a little cartridge in, drop on a microliter of your sample, and it starts counting and basically calculates what proteins are in there and their concentration," he said.
Mayer noted that significant hurdles stood between existing technology and this vision of the future, but he said that the development of nanopore-based DNA sequencing by Oxford Nanopore (which provides funding for his protein work) had made him "very careful of what I predict can be done or not."
"I won't be the one anymore who goes out and says this is impossible, because that's what I said about sequencing," he said. "One reason I never went into DNA work was because I thought it wasn't possible. And in 20 years, they went from this crazy idea of sequencing DNA [via nanopore] to making it a reality."
In fact, Oxford Nanopore has itself done work applying nanopores to protein work. The company has explored using its GridION and MinION platforms for protein analysis by linking them to aptamers, which, in theory, would bind the target analytes with the nanopores then detect the binding event.
This is similar to a technique demonstrated in 2012 by researchers at the Technical University of Munich, who published a paper in Nature Nanotechnology in which they functionalized a solid-state nanopore with recombinant his-tagged proteins to sense target analytes.
Tacking a different tack, in February 2013 researchers from the University of California, Santa Cruz, published a paper in Nature Biotechnology in which they used the protein unfoldase ClpX to unfold three differentially modified Smt3 proteins and pull them through an α-HL nanopore.
Additionally, in a paper published in 2014 in Nature Biotechnology, Oxford Nanopore co-founder Hagan Bayley (working independently of the company) demonstrated the ability of a nanopore sensor to distinguish between differentially phosphorylated forms of the protein thioredoxin.