NEW YORK – Researchers from the University of Washington have developed a new approach that promises to achieve long-range, single-molecule sequencing of intact protein strands using the commercially available platform from Oxford Nanopore Technologies.
Described in a BioRxiv preprint published in October, the proof-of-concept study, which was partially supported by Oxford Nanopore, opens the door for further potential proteomics applications on the company's sequencers.
"The big breakthrough with DNA and RNA [nanopore] sequencing was using the motor to pull the DNA and RNA strands through the pore," said Jeff Nivala, a molecular engineering professor at the University of Washington and the senior author of the study. "How you would do this for proteins was really the challenge."
According to Nivala, one major hurdle to nanopore sequencing of full-length proteins is that the molecules are typically heterogeneously charged, making it challenging to rely on the electrophoretic force to translocate the protein strands through the pores in a controllable fashion.
In addition, Nivala said his team sought to utilize the commercially available Oxford Nanopore devices given that "they are out of the box and ready to go." However, the challenge associated with that is the Oxford Nanopore flow cells do not allow users access to the trans side — opposite of protein loading — of the pore, where unfoldase motor proteins are normally deployed to facilitate protein translocation and sequencing.
To overcome these challenges, Nivala's team devised a two-step approach using the Oxford Nanopore MinIon device and R9.4 flow cell. First, the protein substrate is transported through the nanopore via electrophoretic force, while a blocking domain is affixed to the C-terminal of the protein to prevent the complete translocation of the molecule. Subsequently, ClpX — an ATP-powered protein unfoldase — is added to the cis side of the array to pull the analyte back out of the pore in a controllable fashion.
While the initial threading of the protein through the pore by electrophoretic force happens too fast to result in any reliable signals, Nivala said, the unfoldase-mediated translocation of proteins back out of the pore can produce slow, reproducible ionic current signals.
In addition, Nivala said the team has designed a "slippery" amino acid sequence near the N-terminal of the protein strand to allow the molecule to temporarily slip off the ClpX protein. This allows the analyte to thread through the pore again via electrophoretic force, followed by ClpX-mediated reverse translocation for repeated sequencing readout.
In their study, Nivala and collaborators used synthetic proteins to benchmark the performance of the method. Overall, the study showed single-amino acid reading sensitivity and demonstrated the capability to analyze all 20 different single-amino acid residues within synthetic analytes in a static background.
In addition, Nivala said the method illustrated its capability to process folded full-length protein strands over 100 amino acids in length, setting the stage for analyzing natural protein molecules using the approach.
Furthermore, the UW researchers built a biophysical model to help predict nanopore ionic current signals directly from protein sequences. This model would potentially enable a "lookup table" style of approach, similar to mass spectrometry, to help identify and fingerprint full-length, single-molecule proteins, the authors noted.
"I think this will be really important going forward for protein fingerprinting-type approaches," Nivala said. "If you know what the theoretical nanopore squiggles look like when [the samples] go through a nanopore, you can then identify the full-length protein based on that match to your model protein."
While other nanopore-based fingerprinting approaches or traditional mass spectrometry typically look at small peptides, a unique advantage of their model is that it can achieve analysis of full-length proteins, Nivala noted.
"It's a pretty impressive amount of work," said Liviu Movileanu, a professor at Syracuse University who was not involved in the study. "What is remarkable is that [the authors] went beyond just measuring the currents; they analyzed the signals and produced modeling … that in a sense can be expanded to [protein] fingerprinting in the future."
Such fingerprinting models can be potentially deployed for disease detection or biomarker profiling, Movileanu added.
Additionally, he applauded the authors' efforts to design a method using the commercially available Oxford Nanopore flow cells, which have already demonstrated "an amazing ability for high throughput" in DNA sequencing.
Furthermore, Movileanu said the analysis of the unfolding and translocation behaviors of different proteins by the unfoldase in the study also serves as a knowledgebase for the field to investigate and utilize the enzyme for further applications.
Despite the promises, Nivala noted that the method presented in the current paper is "not at the final version" and has some limitations. For one, he pointed out that the two-step flow cell loading process, where protein substrates and the unfoldase enzymes are loaded separately, can limit the throughput of data collection during each cycle.
To that end, he said the team is working to develop a workflow where the unfoldase can be prebound to the analyte and only become activated when the protein strand is captured in the pore, similar to the mechanism adopted in nanopore DNA sequencing.
Another potential limitation of the method is that a ligation step will be required to affix the synthetic N- and C-terminal sequences to the analyte when analyzing natural proteins, said Nivala.
Nivala said the near-term goal for the team is to continue enhancing the predictive models for protein fingerprinting by collecting more data on natural protein sequences.
In the long run, he said the goal is to eventually achieve single amino acid de novo sequencing using nanopores. "I think de novo sequencing of individual amino acids of natural proteins is still going to be a big challenge, and there are hurdles to go," Nivala said. "We're exploring different alternative pores that will help provide better resolution."
Nivala said UW has filed provisional patents covering aspects of this study, and there are "ongoing considerations" regarding potential commercialization plans of the method. In addition, Nivala noted that his team has an open collaboration with Oxford Nanopore, which partially supported the study. As such, "there are considerations that go along with that as far as the IP," he said.
An Oxford Nanopore spokesperson declined to comment on whether the company has any interest in commercializing the method described in the study or provide details on how the approach fits into the company's overall protein sequencing R&D.
"Protein sequencing compared to even RNA sequencing is still a baby," Nivala said. "It's still early days, but [it's] really exciting to see where it can go."