NEW YORK — Researchers at Delft University of Technology and the University of Illinois at Urbana-Champaign have developed a nanopore-based approach to protein analysis capable of identifying single amino acid changes in peptides.
Detailed in a paper published last week in Science, the work advances efforts to sequence proteins using nanopores by enabling more controlled translocation of peptides through pores and allowing for repeated reads of individual peptides, said Cees Dekker, a Delft professor and senior author on the study.
Over the last decade, nanopore sequencing of nucleic acids has become a reality, with UK-based Oxford Nanopore bringing several such platforms to market. A number of researchers have also been working on nanopore-based sequencing of proteins, and if that proves technically feasible, it could enable single-molecule level detection and, potentially, de novo sequencing of peptides and proteins, along with detection of post-translational modifications.
Nanopore-based protein analysis has been a hard problem, though, facing challenges ranging from the difficulty of moving peptides through pores in a consistent fashion to that of correlating the signals produced by that movement to the presence of a particular amino acid.
For years, researchers have been using different enzymes to pull target peptides through nanopores, noted Henry Brinkerhoff, a post-doctoral researcher in Dekker's lab and first author on the Science study.
"The real problem there was, though, that these enzymes that pull on proteins don't take nice, neat steps," he said. "They tend to stall a lot and fall off, and there are just a lot of practical problems to getting these things to work."
This has left the field looking for ways to better translocate proteins and peptides through nanopores, Brinkerhoff said. "How exactly do we pull a protein through a pore in a controlled, step-by-step fashion?"
In their recent study, Dekker, Brinkerhoff, and their colleagues took the approach of linking peptides of interest to pieces of DNA and then pulling the molecules by their DNA attachments using a DNA helicase developed for DNA sequencing.
This helicase takes a consistent half-base-pair step per ATP cycle, a distance that happens to be more or less equivalent to a single amino acid, Dekker said. "So it is very convenient [for protein analysis]."
Perhaps even more notable is the fact that the helicase-based system allowed the researchers to pass the same peptide through a nanopore multiple times, which let them improve the quality of the data they used to identify particular amino acids.
In the system presented by the Delft researchers, the helicase moves along the DNA attachment pulling both the attachment and the peptide through the nanopore. When the helicase hits the peptide portion of the molecule it falls off.
"That could be the end of the experiment," Dekker said. "But at high helicase concentrations, a new helicase is basically lined up behind the first one, and as soon as [the initial helicase] falls off, the new one zips right in and starts pulling the [DNA-linked peptide] through the nanopore again."
Dekker said that the approach allowed researchers to repeat the read of the same peptide hundreds of times. That, in turns, lets them collect data across multiple analyses of the peptide, which improves confidence in the identification of particular amino acids.
"With a single read you have an error rate of maybe 15 percent in identifying these different peptide variants that are changed by a single amino acid," Brinkerhoff said. "But when you can make this non-destructive, repetitive measurement, you can get results where you can almost definitively say what you are seeing."
Even with the controlled translocation and multiple reads enabled by the helicase-based system, actually sequencing peptides de novo remains out of reach at the moment, Dekker noted. One major challenge facing nanopore-based protein sequencing is that while the helicase is able to draw peptides through the pore one amino acid at a time, the changes in nanopore signal produced by those amino acids are influenced by the amino acids ahead and behind them in the chain. This means that the same amino acid will produce different signals depending on what other amino acids surround it.
Whether this difficulty can be surmounted remains to be seen, Brinkerhoff said.
"There is some theoretical development that needs to happen," he said, noting that while nanopore-based DNA sequencing mainly relies on signal stemming from the electrophoretic behavior of DNA passing through the pore, proteins and peptides produce more complicated signals.
"You have some electro-osmotic force, where it is like a water flow that is kind of pulling everything in," he said. "Then on top of that things might fluctuate where you have positive charges counteracting the electrophoretic forces. There are a lot of open questions about what kinds of signals we should even expect."
Even so, Dekker said he believed the approach could prove useful for identifying amino acid variants and post-translationally modified forms of proteins or peptides of interest using a peptide fingerprinting approach similar to those employed by mass spectrometry-based proteomic workflows.
Brinkerhoff highlighted analysis of major histocompatibility complex antigen peptides as an area where the method could carve out an early niche.
"You can't really sequence [MHC peptides] very well with mass spec because they are very heterogenous, they are always in mixtures, they are in low copy numbers," he said. "They are just like the perfect test case for this."
Delft has filed for a patent covering the approach, and Dekker said his lab has been in discussions with different companies about the technology. In 2018, Dekker and his Delft colleague Chirlmin Joo launched a company called Bluemics to commercialize nanopore-based protein detection technology developed in their labs, but they shut down the company after struggling to find investors and currently have no plans to restart it, Dekker said.
This summer, a team of researchers at Nanjing University published in Nano Letters on an approach that similarly links oligonucleotides to peptides to control their passage through nanopores.
Brinkerhoff said that Oxford Nanopore, which controls much of the IP around nanopore sequencing of nucleic acids and proteins, filed for patents related to linking DNA to proteins for nanopore sequencing around the same time Delft filed for patents covering the Dekker lab's work. He noted, though, that "there is a lot of divergence between" the two applications.
Oxford Nanopore declined to comment on whether it believed any portion of the method described in the Science paper infringed its IP.