Skip to main content
Premium Trial:

Request an Annual Quote

New Protein Reporters for MinIon Sequencer Point Toward Route for Single-Molecule Proteomics


NEW YORK – Researchers at the University of Washington have developed protein reporter tags that can be read out using Oxford Nanopore's MinIon nanopore sequencing platform.

Detailed in a paper published this week in Nature Biotechnology, the tags offer a potential alternative to commonly used fluorescence-based reporters and could prove useful in a variety of research areas, including synthetic biology and studies of cell signaling pathways, said Jeff Nivala, research assistant professor at UW and senior author on the study.

The study is also among the first instances of protein analysis using a commercially available nanopore sequencing device and provides an example of the technology's potential to enable single-molecule protein measurements.

The method uses engineered proteins, called NanoporeTERs, or NTERs, which are designed to be easily captured and read by nanopore sensors. These NTERs consist of three main parts: a negatively charged C-terminal domain that is easily captured by the nanopore; a folded domain that prevents the full protein from passing through the nanopore; and a secretion domain that causes the protein to be secreted out of the cell being studied and into an extracellular medium after it is expressed. Peptide barcodes are incorporated into the C-terminal portion of the NTER, with the nanopore reading these barcodes and using them to identify and quantify the levels of different NTERs.

The NTERs can be genetically encoded and expressed as part of target proteins in much the same way as traditional fluorescent reporters. They have a number of potential advantages compared to traditional reporters, though, Nivala said.

"The potential barcode space is much larger than what you can get with fluorescent proteins," he said, noting that this allows for a much higher level of multiplexing.

"You also have a lot more flexibility in the structure of the protein," he said. "Do you want it to be secreted, not secreted? Does it need to be as small as possible? Does it need to be very hydrophilic? Because all we really need to do is put this [C-terminal] tail onto [the target protein] so it gets captured by the pore and then some arbitrary peptide barcode into that tail."

Fluorescent protein reporters, on the other hand, are typically less adjustable, Nivala said. "You can mutate them and engineer them and things like that, but to a limited degree. You're trying to work with what nature gave you and build on it from there, so it's a little more constrained."

It can also be difficult to collect rigorously quantitative data with traditional reporters, he said. "They form dimers. They can form aggregates. A lot of them will express more or less depending on the different variants. So it can be hard to be absolutely quantitative."

This is especially true when multiplexing, he said. "You have to be careful about comparing their intensities and inferring relative expression levels from these different variants."

One appealing feature of traditional fluorescence-based reporters is their convenience, Nivala noted. "There are a lot of different ways to read them … you don't have to do any purification because the light is just being emitted from the sample."

He said that he and his colleagues tried to make the NTERs similarly easy to use.

"We wanted to have a similar approach where you need only very minimal sample processing," he said. "We engineered our NanoporeTERs to have a secretion domain, so when they are expressed in the cell, they get excreted into the media and then we can just take that media and load it directly into a nanopore flow cell … and we can just start capturing proteins directly from that sample of media and start reading the barcodes."

In the Nature Biotechnology study, the UW team multiplexed nine different NTERs, and Nivala said the researchers have thus far produced around 20 different barcodes.

In the NTER approach, the nanopores are not sequencing each amino acid of the peptide barcodes but are generating a signal based on the behavior of the full length of the roughly 17 amino acids that fit within the pore. The UW researchers use a convolutional neural network to analyze and distinguish between the signals produced by the different barcodes.

While this approach has worked well to enable the production of a relatively small set of distinct NTERs, scaling is a challenge, Nivala said.

"It's a supervised learning approach, so we need training data for every possible barcode," he said. "We're hoping in the future to have some generalizable model where, given any particular peptide sequence, you could predict what signal it will give you, or vice versa — for a signal you are seeing for the first time, you could predict what is the likely peptide sequence that is generating that signal."

Nivala said he is in discussions with outside parties about potentially licensing the technology. He said the NTERs would likely most immediately see uptake in synthetic biology applications where they could help assess whether synthetic genetic circuits are working as intended to produce the desired proteins and in the desired quantities.

Synthetic biology researchers typically "have [a system] designed to a certain specification, and then nine times out of 10, the circuit doesn't quite behave how they wanted it to," he said. NTERs could help speed up the debugging process, which could in turn help speed overall development of synthetic systems.

"Even having 20 different reporters could be game changing for [synthetic biology researchers] as far as letting them more efficiently figure out what is going on with the circuits they are designing," Nivala said.

Because the NTER barcodes can be coded to include post-translational modification sites, and the presence of PTMs can be detected by the nanopore sensors, the reagents could also be useful for looking at groups of proteins connected in signaling pathways by PTMs like phosphorylation, allowing researchers to tease out the structures of these pathways and how they change in response to various perturbations.

Nivala said he and his colleagues are also working on an expanded version of the reporters that would combine multiple barcodes, moving them through the nanopore, which he projected could expand multiplexing into the millions of proteins.

"You can imagine these sorts of massively parallel assays," he said.

Moving the peptide barcodes through the nanopore requires more sophisticated molecular machinery than the approach demonstrated in the Nature Biotechnology paper, Nivala noted. While he was a graduate student in the lab of University of California, Santa Cruz researcher Mark Akeson, Nivala published a method using the protein ClpXP to unfold a target protein and translocate it through a nanopore.

Other researchers have also demonstrated methods for translocating proteins through nanopores. Last month, a team led by Cees Dekker, a professor at the Delft University of Technology, published a bioRxiv preprint demonstrating the use of a DNA helicase to pull a peptide linked to a DNA molecule through a nanopore one amino acid at a time. Additionally, the system is able to pull the peptide through the pore an unlimited number of times, which the authors suggest could allow researchers to collect enough signal to distinguish between the individual amino acids.

Also last month, a team of researchers at Nanjing University published on a similar approach in Nano Letters.

Collecting and interpreting the signal from peptides passing through a nanopore in a way that allows for identification of individual amino acids is one of the primary challenges to nanopore-based protein sequencing, and one that the UW team's barcoding approach could potentially circumvent.

The last year has seen significant interest in proteomics and in single-molecule protein sequencing specifically, with two firms, Quantum-Si and Nautilus, going public with hundreds of millions of dollars in investment behind them, and several others, including Encodia, Erisyon, and Glyphic Biotechnologies, working on single-molecule protein sequencing systems.

Little of this wave of investment has gone to firms developing nanopore-based protein sequencing technologies, despite the fact that nanopore-based approaches have been one of the most prominent areas of single-molecular protein sequencing research. 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.

Oxford Nanopore is perhaps the most established commercial entity pursuing nanopore-based protein analysis. During a presentation at its 2019 Nanopore Community Meeting, Clive Brown, Oxford Nanopore's chief technology officer, said the company had built a breadboard for a droplet-based protein sequencing system using a version of the translocation approach original published by Nivala, and that it was generating "significant numbers of fairly consistent protein [signals]" that the company was using to train algorithms against purified protein libraries.

Contacted this week for an update on this work, Oxford Nanopore did not respond to emailed questions by press time.

"I definitely think Oxford Nanopore is interested in getting into the protein sequencing stuff," Nivala said, noting that he has been consulting with the company on nanopore protein sequencing since his Ph.D. days.

More generally, Nivala said he expected to see more commercial activity on the nanopore protein analysis front, "especially with all the money you see [invested in] these other companies."

"I think you'll start to see it pick up from Oxford and from others," he said.

"Nanopores are a great contender [for protein sequencing] because of the ability to linearly read a peptide, [their] incredible sensitivity, the proven single-molecule abilities, [the] potential for scaling in parallel readouts — as in MinIon," Dekker said. "Given the early stage, it is hard to predict the winning technology in the long run."