Jump to content
[dismiss]
Molecupedia is a parody encyclopedia written by Molecular Reality Corporation. It is not Wikipedia and is not affiliated with the Wikimedia Foundation. About

Nanopore protein sequencing

Page protected
From Molecupedia, the encyclopedia of very small things
This is the current revision of this page. The present address (URL) is a permanent link to this version.
Attempts to read the amino acid sequence of peptides and proteins with nanopores

Nanopore protein sequencing is the attempt to determine the order of amino acids in peptides and proteins by passing single molecules through nanopores and reading the ionic current. It borrows the method of nanopore sequencing of DNA and runs into four differences. The alphabet has 20 letters where DNA has 4. The chain has no uniform charge for an electric field to pull on. It folds. And it cannot be copied, so every molecule has to be read as it is. Since 2013 groups have moved proteins through pores with unfoldase enzymes, read a single peptide 117 times, identified single amino acid substitutions with an error rate under one in a million after repeated reads, and in 2026 read a peptide that an enzyme had shortened one residue at a time.[1][2][3]

A nanopore instrument that sequences an unknown protein from end to end is not yet available. The group that achieved the 117 rereads wrote that turning its method into a technology for de novo sequencing "remains a substantial challenge".[2] A paper of 2023 states that direct sequencing of protein with nanopores "has not yet been realized".[4] Portal Biotech and Glyphic Biotechnologies are among the companies working toward instruments.

Why proteins are harder

A DNA reader has four letters and a read head a few bases long. For the R9.4 pore of Oxford Nanopore Technologies the table of expected current levels has 4,096 entries, one for each six-base word.[5] For a protein, with 20 amino acids at each position and a read head about 8 residues wide, the group of 2021 wrote that "the number of measurements required to build an ion-current-to-amino-acid map is impractically large". The table would hold 208, about 2.6 × 1010, entries. The authors point out that many proteomics applications work from known candidate sequences: identifying or "fingerprinting" proteins in mixtures, mapping modifications and measuring small samples, all by comparing single-molecule measurements with reference signals of known proteins.[2]

The second difficulty is the force. DNA carries one negative charge per base along its length, so the field pulls on it uniformly and a strand threads into a pore by itself. In the words of the 2024 paper that engineered a pore to cope, "proteins, unlike DNA, do not have a uniform charge", and so "the electrophoretic force cannot in general be used to translocate or linearize them".[6] Third, a protein is folded, and the narrowest part of the pores used for sequencing, about 1 nm, admits only an unfolded chain, so the protein must be unfolded by an enzyme or a chemical denaturant. Fourth, there is no amplification. A DNA molecule can be copied before it is read; a single protein molecule has to be read as it is, with its modifications, which is why rereading the same molecule matters so much. The variety to be read is large. A paper of 2023 speaks of "cellular proteins and their millions of variants".[7]

The older methods include Edman degradation, which removes and identifies the N-terminal amino acid of a peptide one residue at a time and was described by Pehr Edman and colleagues in 1950.[8] Studies of the proteome, a 2022 paper said, "would benefit greatly from methods to directly sequence and digitally quantify proteins and detect posttranslational modifications with single-molecule sensitivity".[9]

Approaches

Four drawings of a protein pore with a peptide or protein threading through it, driven by a helicase on a DNA leash, an unfoldase, electro-osmotic flow, and a peptidase
Four ways of moving a peptide or protein through a nanopore: a helicase on a DNA leash, an unfoldase, electro-osmotic flow, and an enzyme that shortens the chain.

Motors on a DNA leash

Enzymes that ratchet DNA are good at it, so one approach attaches the peptide to DNA. In the experiment of 2021 an 80-nucleotide DNA strand was joined by click chemistry to a peptide of 26 residues and pulled through the MspA pore by Hel308, a DNA helicase that moves DNA in steps of about 0.33 nm, close to the length of one amino acid.[2] The helicase took the peptide through the narrowest part of the pore in observable steps, and a classifier told apart three variants that differed by a single residue with an average accuracy of 87% on single reads. A second helicase queued behind the first could take over when the first fell off, which rewound the peptide for another read. The record was 117 rereads of one molecule, and the error in calling the variant fell below one in a million once more than about 30 rereads were combined. Only 45 of the 117 rewinds went back far enough to cover the variant site.[2] A group in China reported ratcheting of peptide-oligonucleotide conjugates through MspA in the same year.[10] A later study drew peptides that carried one or two closely spaced phosphate groups through the sensing region of a nanopore and discriminated them with 95% accuracy on individual reads.[11]

Unfoldases

A cell has its own machines for pulling proteins apart. In 2013 the AAA+ unfoldase ClpX moved engineered proteins through an alpha-hemolysin pore and gave sequence-dependent features in the current,[1] in the year that another study showed proteins unfolding in several steps as they cross such a pore.[12] In 2024 the same enzyme ratcheted proteins through CsgG on a commercial sensor array, and the authors showed that ClpX moves its substrate in steps of two residues. The strands were synthetic and hundreds of residues long, single substitutions could be told apart, a phosphorylation was mapped, and rereads raised the accuracy. A folded protein domain was also read from end to end.[13]

Flow and denaturant

Three groups did without enzymes. In 2023 a bath of 2 M guanidinium chloride, which unfolds proteins and drives an electro-osmotic flow through alpha-hemolysin, carried proteins through single file and in one direction. Each residue spent about 10 µs in the pore, much slower than DNA transport, and proteins of up to 764 residues crossed. A classifier named the protein and its orientation from one event with accuracy above 90%.[14] Another group engineered sets of charges about 1 nm apart in the lumen of the CytK pore, which created a flow strong enough to pull unstructured natural polypeptides through against the electric field, with a force of about 20 pN at -100 mV in simulations, similar to the electric force on single-stranded DNA.[6] A third group used an engineered charge-selective nanopore to unfold polyproteins of more than 1,200 residues one domain at a time and to locate modifications deep inside the chain.[7]

Single amino acids and stepwise degradation

A different line identifies amino acids one at a time. With a short polycationic carrier, an aerolysin pore identified 13 of the 20 natural amino acids by their current blockades.[15] An MspA pore with a single nickel ion at its constriction separated all 20 amino acids and four modified ones with an accuracy of 98.6% by machine learning, and a pore with a copper ion reached a validation accuracy of 99.1% with 30.9% of the signals recovered.[4][16]

The 2026 result joins the two ideas. A peptide was tethered to an MspA pore that carried a nickel-bound adapter, which let the N terminus of the peptide be recognized by multiple rereadings, and a cholesterol-tagged aminopeptidase shortened the peptide by one residue at a time. Each cut moved the signal in a stepwise way, and the authors reported single-amino-acid resolution, with single substitutions, modifications and unnatural amino acids identified.[3] Sequential removal of the N-terminal residue is also the principle of Edman degradation.

Fingerprinting, trapping and scanning

Some methods identify a protein without reading it in order. Short DNA tags clicked onto the cysteines of denatured proteins raised the capture rate of solid-state pores tenfold, slowed the passage by more than 20 times and left a pattern of pulses that a classifier used to identify proteins from a few events.[17] A DNA-origami sphere docked on a solid-state pore held single proteins for hours, long enough to watch the chaperone Hsp90 change shape.[18] A glass nanopore on a positioner scanned tethered molecules at a chosen speed and number of times, and averaging thousands of readings of one molecule raised the signal-to-noise ratio by two orders of magnitude over free translocations.[19]

Designed pores

Nanopores used to be borrowed from bacteria. In 2024 transmembrane β-barrel pores were designed on a computer with chosen sizes, and their conductances, from 110 to 430 pS, correlated with diameters of about 0.5 to 1.1 nm. The paper reported conduction and structure and no analyte sensing.[20]

Selected results

Selected results, 2013 to 2026
Year Result Pore Source
2013 An unfoldase moves engineered proteins through a pore Alpha-hemolysin [1]
2020 13 of the 20 amino acids identified by their blockades Aerolysin [15]
2021 One peptide reread 117 times; single-residue variant called with error below 1 in 106 MspA [2]
2023 Proteins of up to 764 residues carried through by denaturant and flow, without an enzyme Alpha-hemolysin [14]
2023 Polypeptides of more than 1,200 residues unfolded and moved through, modifications located Engineered nanopore [7]
2024 Strands hundreds of residues long moved in two-residue steps by an unfoldase CsgG [13]
2024 Unstructured polypeptides pulled through against the electric field CytK [6]
2024 All 20 amino acids and four modified ones told apart, 98.6% MspA with nickel [4]
2026 Peptide read as an enzyme shortens it one residue at a time MspA [3]

Instruments and companies

Portal Biotech says its technology evolved from "the pioneering work from the Maglia Lab at the University of Groningen", and describes a protein product that analyzes "intact proteins from end-to-end, identifying the protein and the location of any modifications".[21] Glyphic Biotechnologies of Berkeley, California, founded in 2021, describes a method it calls Protein Sequencing by Expansion, in which proteins and peptides receive an initiating linker, each amino acid is "sequentially added and uniformly spaced along the linker", and the assembled molecule "passes through a nanopore, where distinct electrical signatures from each amino acid reveal the protein's sequence".[22]

The first commercial single-molecule protein sequencer was optical. Quantum-Si announced the commercial availability of its Platinum platform in December 2022.[23] A paper by the company's scientists the same year described identifying peptides on a semiconductor chip by dye-labeled recognizers of N-terminal amino acids while an aminopeptidase cleaved them one at a time, with the readout in fluorescence and no pore.[9]

Signal analysis

Protein signals are many-level and noisy, and classification leans on machine learning. A self-supervised network, NanoSSL, learned from masked fragments of nanopore events and was tested on variants of the amyloid-beta peptide.[24] A voltage-matrix analysis, which tabulates classification accuracy across several voltages, was applied to two tumor markers and their mixtures in solid-state pores, and once the features were chosen to avoid overfitting the classification no longer depended on voltage.[25] See Machine learning on nanopore signals.

At Molecular Reality

Molecular Reality Corporation does not claim to read protein sequences. Its guide says of DNA that "reading requires slowing the molecule down", and of its present work that "we work on detection, the art of recognizing a molecule's shadow", closer to "spotting a friend by the patch on their jacket than to reading the jacket's every thread".[26] The founder has named as a long-term goal an analytically universal single device, one that senses molecules of any kind (see Analytically universal single device).[27]

See also

Notes


References

  1. ^ a b c Nivala, J.; Marks, D. B.; Akeson, M. (2013). "Unfoldase-mediated protein translocation through an α-hemolysin nanopore". Nature Biotechnology 31 (3): 247-250. doi:10.1038/nbt.2503. PMID 23376966.
  2. ^ a b c d e f Brinkerhoff, H.; Kang, A. S. W.; Liu, J.; Aksimentiev, A.; Dekker, C. (2021). "Multiple rereads of single proteins at single-amino acid resolution using nanopores". Science 374 (6574): 1509-1513. doi:10.1126/science.abl4381. PMID 34735217.
  3. ^ a b c Chen, J.; et al. (2026). "Sequential reading of a stepwise-shortened peptide immobilized on nanopore". Nature 658 (8134): 260-268. doi:10.1038/s41586-026-10881-1. PMID 42527597.
  4. ^ a b c Wang, K.; et al. (2024). "Unambiguous discrimination of all 20 proteinogenic amino acids and their modifications by nanopore". Nature Methods 21 (1): 92-101. doi:10.1038/s41592-023-02021-8. PMID 37749214.
  5. ^ "kmer_models". GitHub. Oxford Nanopore Technologies. Retrieved October 25, 2026.
  6. ^ a b c Sauciuc, A.; et al. (2024). "Translocation of linearized full-length proteins through an engineered nanopore under opposing electrophoretic force". Nature Biotechnology 42 (8): 1275-1281. doi:10.1038/s41587-023-01954-x. PMID 37723268.
  7. ^ a b c Martin-Baniandres, P.; et al. (2023). "Enzyme-less nanopore detection of post-translational modifications within long polypeptides". Nature Nanotechnology 18 (11): 1335-1340. doi:10.1038/s41565-023-01462-8. PMID 37500774.
  8. ^ Edman, P.; et al. (1950). "Method for determination of the amino acid sequence in peptides". Acta Chemica Scandinavica 4: 283-293. doi:10.3891/acta.chem.scand.04-0283.
  9. ^ a b Reed, B. D.; et al. (2022). "Real-time dynamic single-molecule protein sequencing on an integrated semiconductor device". Science 378 (6616): 186-192. doi:10.1126/science.abo7651. PMID 36227977.
  10. ^ Yan, S.; et al. (2021). "Single molecule ratcheting motion of peptides in a Mycobacterium smegmatis porin A (MspA) nanopore". Nano Letters 21 (15): 6703-6710. doi:10.1021/acs.nanolett.1c02371.
  11. ^ Nova, I. C.; et al. (2024). "Detection of phosphorylation post-translational modifications along single peptides with nanopores". Nature Biotechnology 42 (5): 710-714. doi:10.1038/s41587-023-01839-z. PMID 37386295.
  12. ^ Rodriguez-Larrea, D.; Bayley, H. (2013). "Multistep protein unfolding during nanopore translocation". Nature Nanotechnology 8 (4): 288-295. doi:10.1038/nnano.2013.22.
  13. ^ a b Motone, K.; et al. (2024). "Multi-pass, single-molecule nanopore reading of long protein strands". Nature 633 (8030): 662-669. doi:10.1038/s41586-024-07935-7. PMID 39261738.
  14. ^ a b Yu, L.; et al. (2023). "Unidirectional single-file transport of full-length proteins through a nanopore". Nature Biotechnology 41 (8): 1130-1139. doi:10.1038/s41587-022-01598-3. PMID 36624148.
  15. ^ a b Ouldali, H.; et al. (2020). "Electrical recognition of the twenty proteinogenic amino acids using an aerolysin nanopore". Nature Biotechnology 38 (2): 176-181. doi:10.1038/s41587-019-0345-2. PMID 31844293.
  16. ^ Zhang, M.; et al. (2024). "Real-time detection of 20 amino acids and discrimination of pathologically relevant peptides with functionalized nanopore". Nature Methods 21 (4): 609-618. doi:10.1038/s41592-024-02208-7. PMID 38443507.
  17. ^ Soni, N.; et al. (2025). "Full-length protein classification via cysteine fingerprinting in solid-state nanopores". Nature Nanotechnology 20 (10): 1482-1490. doi:10.1038/s41565-025-02016-w. PMID 40993352.
  18. ^ Schmid, S.; et al. (2021). "Nanopore electro-osmotic trap for the label-free study of single proteins and their conformations". Nature Nanotechnology 16 (11): 1244-1250. doi:10.1038/s41565-021-00958-5. PMID 34462599.
  19. ^ Leitao, S. M.; et al. (2023). "Spatially multiplexed single-molecule translocations through a nanopore at controlled speeds". Nature Nanotechnology 18 (9): 1078-1084. doi:10.1038/s41565-023-01412-4. PMID 37337057.
  20. ^ Berhanu, S.; et al. (2024). "Sculpting conducting nanopore size and shape through de novo protein design". Science 385 (6706): 282-288. doi:10.1126/science.adn3796. PMID 39024453.
  21. ^ "Portal Biotech". Portal Biotech. Retrieved October 25, 2026.
  22. ^ "Glyphic Biotechnologies". Glyphic Biotechnologies. Retrieved October 25, 2026.
  23. ^ "Quantum-Si announces commercial availability of Platinum, the world's first next-generation single-molecule protein sequencing platform". Quantum-Si. December 20, 2022. Retrieved October 25, 2026.
  24. ^ Xie, Y.; et al. (2026). "NanoSSL: attention mechanism-based self-supervised learning method for protein identification using nanopores". Bioinformatics 42 (1): btaf657. doi:10.1093/bioinformatics/btaf657. PMID 41499226.
  25. ^ Akita, R.; et al. (2025). "Voltage-matrix nanopore profiling for the discrimination of protein mixtures". Chemical Science 16 (40): 18607-18615. doi:10.1039/d5sc05182g. PMID 41059106.
  26. ^ "Keltar's Field Guide to Molecular Reality". Molecular Reality Corporation. August 2026.
  27. ^ Kemmishtree (March 28, 2026). "Marching orders". EQBio-OS.

Further reading

  • Restrepo-Pérez, L.; Joo, C.; Dekker, C. (2018). "Paving the way to single-molecule protein sequencing". Nature Nanotechnology 13 (9): 786-796. doi:10.1038/s41565-018-0236-6.
  • Ritmejeris, J.; Chen, X.; Dekker, C. (2024). "Single-molecule protein sequencing with nanopores". Nature Reviews Bioengineering 3 (4): 303-316. doi:10.1038/s44222-024-00260-8.
Retrieved from "https://molecularreality.com/wiki/Nanopore_protein_sequencing"
Last edited 3 hours ago by Aperture tube