nature nanotechnology 3 Article https://doi.org/10.1038/s41565-023-01462-8 Enzyme-less nanopore detection of post-translational modifications within long polypeptides Received: 17 January 2023 Pablo Martin-Baniandres1,4, Wei-Hsuan Lan1,4, Stephanie Board©2,3, Mercedes Romero-Ruiz©1, Sergi Garcia-Manyes©2,3, Yuj'ia Qing©'E3& Accepted: 15 June 2023 ,. „ . Ä1K-, _ HaganBayley©1 Published online: 27 July 2023 ~%\ Check for updates Means to analyse cellular proteins and their millions of variants at the single-molecule level would uncover substantial information previously unknown to biology. Nanopore technology, which underpins long-read DNA and RNA sequencing, holds potential for full-length proteoform identification. We use electro-osmosis in an engineered charge-selective nanopore for the non-enzymatic capture, unfolding and translocation of individual polypeptides of more than 1,200 residues. Unlabelled thioredoxin polyproteins undergo transport through the nanopore, with directional co-translocational unfolding occurring unit by unit from either the C or N terminus. Chaotropic reagents at non-denaturing concentrations accelerate the analysis. By monitoring the ionic current flowing through the nanopore, we locate post-translational modifications deep within the polypeptide chains, laying the groundwork for compiling inventories of the proteoforms in cells and tissues. Single-molecule nanopore proteomics is gaining momentum1. Nanopore sequencing of ultralong DNA and RNA has enabled applications in basic science and medicine that challenge short-read technologies . Modulation of the ioniccurrent passing through a nanopore might also be used to distinguish and count the millionsof proteoforms expressed from the 20,000 or so protein-encoding human genes. In this way, inventories would beobtained of variations such as post-translational modifications (PTMs) and the outcome of alternative RNA splicing, which are often present at multiple locations throughout a polypeptide chain3. Although folded proteins have been translocated through solid-state nanopores45 or protein nanopores of large sizes6,7, this approach has yet to be shown to locate modifications within a polypeptide sequence. Recently, PTMs have been detected within short peptides810. However, single-molecule proteoform identification requires the knowledgeof the architecture of long polypeptide chains. Obtaining such information encounters two main roadblocks. First, proteins must be linearized for sequential PTM readout during nanopore translocation. Second, unlike DNAor RNA, polypeptides have a low-density and heterogeneous distribution of charge along their chains, which renders electrophoresis inapplicable as a means of translocation. One solution is to incorporate charged leader sequences, such as a single-stranded DNA1113, to implement the electrophoretic capture and unfolding of proteins. However, as soon as the leader sequence exits the pore, a directional force is no longer present, and if the remaining polypeptide unfolds, it will diffuse through the pore in a partially extended conformation1113. In another approach, unfoldases (for example, CIpX (refs. 14,15) or VATAN (ref. 16)) have been employed to drive the translocation of tagged polypeptides through nanoporesafterelectrophoreticthread-ing. However, unlike the ratcheting enzymes used for DNA and RNA nanopore sequencing1719, these unfoldases are not capable of the residue-by-residue translocation of polypeptides. Further, whether large PTMs will betolerated during enzymatictranslocationisunclear. 'Department of Chemistry, University of Oxford, Oxford, UK. department of Physics, Randall Centre for Cell and Molecular Biophysics and London Centre for Nanotechnology, King's College London, London, UK. 3Single Molecule Mechanobiology Laboratory, The Francis Crick Institute, London, UK. "These authors contributed equally: Pablo Martin-Baniandres, Wei-Hsuan Lan. Ele-mail: yujia.qing@chem.ox.ac.uk; hagan.bayley@chem.ox.ac.uk Nature Nanotechnology | Volume 18 | November 2023 11335-1340 1335 Article https://doi.org/10.1038/s41565-023-01462-8 <, rn 'WWW1 4 muttm- CG(SDKIIHLTDDSFDTDVLK GADAILVDFWAEWSGPSK ~ MIAPILDEIADEYQGKLT VAKLNIDQNPGTAPKYGIR GIPTLLLFKNGEVAATKVGAL SKGQLKEFLDANLAGSAG SAGSAGSAGSAGSAGSAG SAGSAGR)„S n = 2, dimer 4, tetramer 6, hexamer 8, octamer (without the C-terminal linker) Ml NMWl C —N x2 < i a o--- 200 ms 200 ms 0 pA 0 pA 200 ms 500 ms Fig. 11 Electro-osmosis-driven translocation of Trx-linker concatemers through a protein nanopore. a, EOF in a charge-selective aHL nanopore (NN-113R)7 drives the sequential co-translocational unfolding of Trx units within a polyprotein of >1,000 aa. b, A sodium dodecyl sulfate-polyacrylamide gel showing the Trx-linker dimer (28 kDa), tetramer (55 kDa), hexamer (83 kDa) and octamer (110 kDa). c, Current recordings for the C-terminus-first translocation of a dimer, a tetramer, a hexamer and an octamer without post-acquisition filtering. A1 20PAI m A2 A3 A1 20pA| m.^WWW^M 10 ms Feature A 0 pA The repeating features Aare indicated by orange and blue bars. d. Zoomed-in view of the repeating feature A boxed in blue in c without post-acquisition filtering. Three levels are assigned as follows: Al, a linker within the pore; A2 and A3, different segments of partly unfolded Trx within the pore. Conditions in c and d are as follows: 750 mM GdnHCl, 10 mM HEPES, 5 mM TCEP at pH 7.2, Trx-linker concatemers (cis) (dimer: 2.23 uM; tetramer: 0.63 uM; hexamer: 0.25 uM; octamer: 0.81 uM), +140 mV (trans), 24 ± 1 °C. Therefore, we aimed to establish a general non-enzymatic means to map modifications within full-length polypeptidechains and eventually to inventory thecollection of proteoforms in individual cells rather than perform an ensemble analysis of peptide fragments. Previously, we distinguished C-terminal PTMs (phosphorylated serines) in a model protein using the DNA leader approach12. However, the detection by nanopores of PTMs deep within a long polypeptide chain has remained a challenge. As an alternative to electrophoresis, electro-osmotic flow (EOF) within a charge-selective protein nanopore has been shown to modulate the binding and dissociation of neutral small molecules20,andpromotethetrapping of short peptides or folded proteins21,22. Recently, the unidirectional translocation of polypeptides through the wild-type (WT) staphylococcal a-hemolysin (aHL) pore was reported in the presence of a high concentration of guanidinium chloride (GdnHCl) (ref. 23). Computational analysis suggested that guanidinium cations line the interior of the pore and therefore alter the charge selectivity of the pore, generating EOF23. Here we use strong electro-osmosis directly attributable to the charged side chains in an engineered aHL pore to capture long underivatized polypeptides and detect modifications within them as they are unfolded and translocated. In addition, to aid in theco-translocational unfoldingof protein analytes, chaotropic reagents (for example, GdnHCl and urea) are employed, which might contribute to the EOF through the pore. Electro-osmotic translocation of protein concatemers We constructed dimers, tetramers, hexamers and octamers of thiore-doxin(Trx) (Fig. la,b,Supplementary Tableland Supplementary Fig. 1). Nature Nanotechnology | Volume 18 | November 2023 11335-1340 1336 Article https://doi.org/10.1038/s41565-023-01462-8 K-t 0 NH, AX A i Cl" HjN-^NHj H2N NH (NN-113R)7 aHL 0 pA L 2s 750 mM CI (NN-113R)7aHL > V (NN-113R)7 aHL 99999 «9999999999«» EOF 1 2s \ .nnn 2s 750 mM CI ipn p 750 mM KCl + 2 M un 1 750 mM GdnHCI Fig. 21 Chaotrope-facilitated electro-osmotic translocation of the Trx-linker octamers through a nanopore. a, Translocation of Trx-linker octamers through a weakly anion-selective WT aHL was not observed in the absence of a chaotrope. b-d, Current traces showing the translocation of Trx-linker octamers through the electro-osmotically active nanopore (NN-113R)7 in the presence of750 mM KCl (b), 750 mM KCl and 2 M urea (c) or 750 mM GdnHCI (d) with 2 kHz post- acquisition filtering. The use of non-denaturing concentrations of chaotropic agents (urea and GdnHCI) accelerated the co-translocational unfolding of the Trx units. Conditions: 10 mM HEPES at pH 7.2,0.81 uM Trx-linker octamer (cis), +140 mV (trans), 24 ± 1 °C with 750 mM KCl (a and b); 2 M urea and 750 mM KCl (c); 750 mM GdnHCI (d). Trx (108 amino acids (aa)) had the two catalytic cysteines removed (Trx: C32S/C35S)11. The Trx monomers were connected by 29 aa linkers, capa-bleof spanning the 10-nm-long lumen of the aHL nanopore when fully extended (0.35 nm per aa). An exception was that the Trx-linker octamer had no C-terminal linker (Supplementary Tablel). We used a previously characterized anion-selective aHL mutant (NN-113R)7 (PNa+/PCi- = 0.33 (ref. 20)) to generate electro-osmosis. All the four Trx-linker concate-mers were captured by (NN-113R)7 in the presence of 750 mM GdnHCI (Fig. lc) at a capture rate -25 times faster than that of the WTaHL pore (PNa+/Pci- = 0.78 (ref. 20)) (£(octamer) = 2.50 s"1 U.M-1 with (NN-113R)7 versus -0.11 s"1 uM1 with (WT)7). The recording conditions were 750 mM GdnHCI, 10 mM HEPES at pH 7.2, +140 mV (trans), 24 ± 1 °C. Electro-osmosis-driven concatemer translocation produced current patterns containing repeating features (Fig. lc and Extended Data Fig. 1). The most abundant feature, A, consisted of three levels (Al, A2 and A3) (Fig. lc,d). The percentage residual current (/res%) for each level in feature A was consistent across all such events for each polypeptide translocation and between all the individual concatemers observed with the same or different pores (Supplementary Table 2). A spike to -0 pA was seen at the beginning of almost all the translocation events and was speculated to represent the rapid unfolding and translocation of the first Trx-linker unit. The spike was followed by up to n -1 repeats of the three-step feature A (n is the number of Trx-linker units in the concatemer), which unambiguously demonstrated the stepwise translocation of entire polypeptide chains one unit at a time. Less often, a d ifferent repeati ngelementBwas recorded (Extended Data Fig. la and Supplementary Table 3). Further, when two identical concatemers were linked by a disulfide bond between the N-terminal cysteines, feature B occurred only after feature A within each translocation event (Extended Data Fig. lb). Therefore, we assigned these two features as C-terminus-first (A) and N-terminus-first (B) translocation events. In the presence of a C-terminal linker (for example, Trx tetramer; Fig. lb), the ratio of C-terminus-first versus N-terminus-first translocation events was -2:1 at +140 mV (Supplementary Table 3); in the absence of the C-terminal leader sequence (for example, Trx octamer; Fig. lb), the C-terminus-first pattern dominated (C versus N = -10:1 at +140 mV) (Supplementary Table 3). The repeatingfeature A was lost at a GdnHCI concentration of 3 M (Supplementary Fig. 2). At 750 mM GdnHCI, -12% of the translocated octamers produced a maximum of seven repeats of featu re Afollowing the initial spike (Supplementary Table 4); a kinetic analysis revealed two populations of A3: one had a mean dwell time -500 times longer than the other at +140 mV ( = 320 ± 60 ms versus 0.69 ± 0.04 ms) (Supplementary Table5). The longer-lived A3 (rA3 > 10 ms) was seen in 25% of the final features A recorded as the translocation of an octamer was completed, but only in 3% of the preceding features A. Tentatively, we assign level Al as a threaded linker preceding the C terminus of a folded Trx unit; level A2as the C-terminal portion of a partially unfolded Trx unit extended into the nanopore; and level A3 as the spontaneous unfolding and passage of the remaining Trx polypeptide through the nanopore (Fig. Id). The dominant absence of a multilevel feature for the first unit and an extended duration for the last unit suggest that the unfolding kinetics of Trx units differ when the polypeptide chain is unable to fully span the lumen of the nanopore. Similar repeatingfeatureswerealsoseen in theabsenceof GdnHCI (recording conditions: 750 mM KCl, 10 mM HEPES at pH 7.2, +140 mV (frans),24±l°C)orinthepresenceofa non-denaturing concentration of urea (recording conditions: 2 M urea, 750 mM KCl, 10 mM HEPES at pH 7.2, +140 mV (trans), 24 ± 1 °C). Under these conditions, the translocation kinetics were slower (Fig. 2), and these options were not pursued further. Detection of PTMs during electro-osmotic translocation To determine whether PTMs near the middle of a long polypeptide chain could be located during electro-osmosis-driven translocation, we constructed Trx-linker nonamers containing a modification site (RRASAC) at two different positions in the central linker (Supplementary Table land Supplementary Fig. 3) for serine phosphorylation (14S-Por24S-P)orcysteine-directedglutathionylationorglycosylation (16C-GSH, 26C-GSH, 16C-SLN or 26C-SLN) (Fig. 3a). In the presence of a phosphate group (P) or glutathione (GSH) or 6'-sialyllactosamine (SLN), level Alfor the modified units exhibited a smaller /res% value and higher root-mean-square noise (/r.m.s.) than that of the unmodified segments Nature Nanotechnology | Volume 18 | November 2023 11335-1340 1337 Article https://doi.org/10.1038/s41565-023-01462-8 0 pA "| C—i p 24S-P ] i ^ J......Mi| ^ ^ r; 0 pA I16C-GSH C^—x 9 n 26C-GSH ■ (16C-SLN ___________j_____l__^^Tr_w^i*«_ _ _ UJ_ _ 1______j___ I26C-SLN __T_j_ __fl*¥-__j____l_ __1__j_____ t Fig. 31 Detection of PTMs in protein concatemers traversing a nanopore driven by EOF. a, Trx-linker nonamers tested with a charge-selective nanopore ((NN-113R)7) containing an RRASAC sequence within the central linker, which was post-translationallyphosphorylated (purple), S-glutathionylated (green) or glycosylated (yellow), b, Recordings of C-terminus-first translocation events of Trx-linker nonamers (left), showing a distinct level Al (boxed in purple, green or yellow) in the presence of a PTM compared with the level Al of unmodified units (orange dash). Traces have been filtered at 2 kHz; transient A3 levels were truncated by filtering and therefore deviate from -0 pA. The A3 level produced < lo 3.0 J° < 2.0 50 ms CL E 1.0 0 pA 0 < 3.0 100 ms < 2.0 E 1.0 0 pA 0 3.0 < Jo < 2.0 200 ms E 1.0 0 pA 0 < 3.0 1 a ° < a 2.0 100 ms E 1.0 0 pA 0 0 10.0 20.0 0 10.0 20.0 Al (26C-GSH) 0 10.0 20.0 0 10.0 20.0 by the translocation of an unmodified unit before the modified linker is indicated with a blue arrow and each of the features A is indicated by orange and blue bars. The number of repeats of feature A within the polypeptide translocation event shown is specified. Scatter plots of/r m s and A/res% for individual polypeptide translocation events (right), where A/res% = - /resx(Al, Trx-linker + PTM), where is the mean/res% value of the remaining Al levels for unmodified repeat units within an individual translocation event. Conditions, 375 mM GdnHCl, 375 mM KC1,10 mM HEPES at pH 7.2,1.2 uM Trx-linker nonamer (cis), +140 mV (trans), 24 ± 1 °C. within an individual polypeptide (Fig. 3b and Supplementary Table 6). Furthermore, the average increment in the current blockade was roughly proportional to the mass of the PTM, with the phosphate giving the smallest increment and the trisaccharide giving the largest (Supplementary Table 6); however, there was substantial overlap between the 14S-P/24S-P and 16C-GSH/26C-GSH populations (Fig. 3b and Supplementary Fig. 4). Mixed nonamerscontainingdifferent PTMs at the same site (26C-GSH and 26C-SLN) were discriminated by the same Nature Nanotechnology | Volume 18 | November 2023 11335-1340 1338 Article https://doi.org/10.1038/s41565-023-01462-8 pore (Supplementary Fig. 5). All the three PTMs tested caused a smaller current blockade at serine 14 (14S) or cysteine 16 (16C) than at serine 24 (24S) or cysteine 26 (26C) (Fig. 3b). Given that 14S/16C must be closer to the dsopening of the aHL pore than 24S/26C in a C-terminus-first threading configuration, it is probable that the central constriction of the pore is located closer to 24S/26C (Fig. 3a and Supplementary Fig. 6). The findings also suggest that the polypeptide might not be fully extended under the EOF (Supplementary Fig. 6 shows further analysis), which corroborates the force spectroscopy data for polypeptides under forces of <20 pN (refs. 24,25). Conclusions Here we have established that electro-osmotically active nanopores cancaptureand unfold individual proteinscomprisinglong(>l,200aa) polypeptide chains for PTM identification and localization. To a first approximation, the electro-osmotic force acting on a polypeptide remains constant during translocation, creating a unidirectional bias desirablefor placing PTMs in sequence. In contrast, theoverall time for unforced polypeptide translocation scales roughly as the square of its length, because the polypeptide chain can move backand forth before diffusingout of the pore26. This is the case within an electro-osmotically inactive nanopore after the exit of a charged leader sequence11 or immediately after a protein domain has unfolded during movement propelled by a motor protein14, which is not ideal for the sequential detection of modification sites within individual polypeptide chains. Furthermore, as a label-free method, our approach circumvents the need to derivatize proteins at either the N or C terminus for electro-phoretic translocation, which could be problematic for eukaryotic proteins due to the widespread presence of N-acetylation and the lack of efficient N- or C-terminus-specific chemistries. Here we have located PTMs in linkers deep within long polyprotein chains by exploiting stepwise unfolding. Our encouraging results lay the groundwork for building inventories of underivatized full-length proteoforms from cells and tissues. The detection of PTMs in freely translocating domains require the slowing and stretching of polypep-tidechains, which might be produced by physical effects (for example, heat and voltage ramps), nanopores with different internal geometries and surface charge and the use of weakly bound ligands. Recording at megahertz acquisition rates27 might also prove advantageous. These endeavours are beyond the scope of the present work and indeed depend on the advances we report here. Our strategy will be readily transferable to nanopore sequencing devices (for example, the MinlON) for highly parallel PTM profiling, which will be useful for producing inventories of full-length human proteoforms, which are -500 aa in median length28. To achieve the complete characterization of the proteoforms in individual cells, our approach faces challenges. For example, because proteins vary in their resistance to unfolding, it will be problematic to establish universal conditions for the translocation of all the protein components of a cell. Similarly, the capture rate will probably differ between proteins. To this end, voltage sweeps might be used in combination with dena-turants to promote protein capture and facilitate co-translocational unfolding. Further, compact PTMs (for example, methylation) might be challenging to directly detect. Ligand-assisted detection might be performed with antibodies or chemical binders. In summary, our enzyme-less approach, targetingfull-length proteins, presentsa viable nanopore technology, ultimately allowing comprehensive proteoform inventories to be established for tissues and single cells. These massive sets of information will extend beyond what is recognized from DNA and RNA sequencing and potentially unveil yet-unknown aspects of the biology of cells and tissues. Online content Any methods, additional references, Nature Portfolio reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; detailsof author contri-butionsandcompetinginterests;and statements of data and codeavail-ability are available at https://doi.org/10.1038/s41565-023-01462-8. References 1. Restrepo-Perez, L, Joo, C. & Dekker, C. Paving the way to single-molecule protein sequencing. Nat. Nanotechnoi 13, 786-796 (2018). 2. Nurk, S. et a I. The complete sequence of a human genome. Science 376,44-53 (2022). 3. Sharma, K. et a I. Ultradeep human phosphoproteome reveals a distinct regulatory nature of Tyr and Ser/Thr-based signaling. Cell Rep. 8,1583-1594 (2014). 4. Yusko, E. C. et al. Real-time shape approximation and fingerprinting of single proteins using a nanopore. Nat. Nanotechnoi. 12,360-367 (2017). 5. Wang, X. et al. Nanopore detection using supercharged polypeptide molecular carriers. J. Am. Chem. Soc. 145, 6371-6382 (2023). 6. Huang, G. et al. PlyAB nanopores detect single amino acid differences in folded haemoglobin from blood. Angew. Chem. Int. Ed. 61, e202206227 (2022). 7. Biesemans, A., Soskine, M. & Maglia, G. A protein rotaxane controls the translocation of proteins across a ClyA nanopore. Nano Lett. 15, 6076-6081 (2015). 8. Ensslen, T., Sarthak, K., Aksimentiev, A. & Behrends, J. C. Resolving isomeric posttranslational modifications using a biological nanopore as a sensor of molecular shape. J. Am. Chem. Soc. 144, 16060-16068(2022). 9. Nova, I. C. et al. Detection of phosphorylation post-translational modifications along single peptides with nanopores. Nat. Biotechnol. https://doi.org/10.1038/s41587-023-01839-z (2023) 10. Versloot, R. C. A. et al. Quantification of protein glycosylation using nanopores. Nano Lett. 22, 5357-5364 (2022). 11. Rodriguez-Larrea, D. & Bayley, H. Multistep protein unfolding during nanopore translocation. Nat. Nanotechnoi. 8,288-295 (2013). 12. Rosen, C. B., Rodriguez-Larrea, D. & Bayley, H. Single-molecule site-specific detection of protein phosphorylation with a nanopore. Nat. Biotechnol. 32,179-181 (2014). 13. Rodriguez-Larrea, D. & Bayley, H. Protein co-translocational unfolding depends on the direction of pulling. Nat. Commun. 5, 4841 (2014). 14. Nivala, J., Marks, D. B. & Akeson, M. Unfoldase-mediated protein translocation through an a-hemolysin nanopore. Nat. Biotechnol. 31,247-250 (2013). 15. Cardozo, N. et al. Multiplexed direct detection of barcoded protein reporters on a nanopore array. Nat. Biotechnol. 40,42-46 (2022). 16. Zhang, S. et al. Bottom-up fabrication of a proteasome-nanopore that unravels and processes single proteins. Nat. Chem. 13, 1192-1199 (2021). 17. Cockroft, S. L., Chu, J., Amorin, M. & Ghadiri, M. R. A single-molecule nanopore device detects DNA polymerase activity with single-nucleotide resolution. J. Am. Chem. Soc. 130, 818-820(2008). 18. Cherf, G. M. et al. Automated forward and reverse ratcheting of DNA in a nanopore at 5-A precision. Nat. Biotechnol. 30,344-348 (2012). 19. Garalde, D. R. et al. Highly parallel direct RNA sequencing on an array of nanopores. Nat. Methods 15,201-206 (2018). 20. Gu, L.-Q. L. Q., Cheley, S. & Bayley, H. Electroosmotic enhancement of the binding of a neutral molecule to a transmembrane pore. Proc. Natl Acad. Sci. USA 100,15498-15503 (2003). Nature Nanotechnology | Volume 18 | November 2023 11335-1340 1339 Article 21. Huang, G., Willems, K., Soskine, M., Wloka, C. & Maglia, G. Electro-osmotic capture and ionic discrimination of peptide and protein biomarkers with FraC nanopores. Nat. Commun. 8, 935 (2017). 22. Asandei, A. et a I. Electroosmotic trap against the electrophoretic force near a protein nanopore reveals peptide dynamics during capture and translocation. ACS Appi Mater. Interfaces 8,13166-13179(2016). 23. Yu, L. etal. Unidirectional single-file transport of full-length proteins through a nanopore. Nat. Biotechnol. https://doi. org/10.1038/s41587-022-01598-3 (2023). 24. Winardhi, R. S., Tang, Q., Chen, J., Yao, M. & Yan, J. Probing small molecule binding to unfolded polyprotein based on its elasticity and refolding. Biophys. J. 111,2349-2357 (2016). 25. Shrestha, P. et al. Single-molecule mechanical fingerprinting with DNA nanoswitch calipers. Nat. Nanotechnoi 16,1362-1370 (2021). 26. Palyulin, V. V., Ala-Nissila, T. & Metzler, R. Polymer translocation: the first two decades and the recent diversification. Soft Matter 10, 9016-9037(2014). 27. Lin, C.-Y. et al. Ultrafast polymer dynamics through a nanopore. Nano Lett. 22, 8719-8727 (2022). https://doi.org/10.1038/s41565-023-01462-8 28. Brocchieri, L. & Karlin, S. Protein length in eukaryotic and prokaryotic proteomes. Nucleic Acids Res. 33,3390-3400 (2005). Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. © The Author(s) 2023 Nature Nanotechnology | Volume 18 | November 2023 11335-1340 1340 Article https://doi.org/10.1038/s41565-023-01462-8 Methods Construction of Trx-Iinker concatemer genes All the reagents were purchased from New England Biolabs (NEB) and DNAoligonucleotides were obtained from Integrated DNATechnologies, unless otherwise indicated. Trx-Iinker concatemer genes were prepared as previously described29. Briefly, the Trx-Iinker monomer gene was amplified with a 5' primer containing a BamHI restriction site and a 3' primer containing a Bglll restriction site, which permitted the in-framecloning of the monomer into the vector pQE30(QIAGEN). Synthetic genes encoding the concatemers were then constructed by the iterative cloning of monomer into monomer, dimer into dimer and tetramer into tetramer. To aid purification, an N-terminal SUMO tag was inserted between the His6 tag and the first monomer unit. In addition, N-terminal cysteine-glycinecodons were included to give the final concatemer constructs: His6-SUM0-CysGly-(Trx-linker)„ (n = 2, 4,6) and His6-SUMO-CysGly-(Trx-linker)7Trx. To produce Trx-Iinker nonamers (His6-SUM0-(Trx-linker)„, n = 9) containing a modification site, the N-terminal cysteine-glycine codons were removed from the tetramer gene and a DNA cassette was designed to contain two terminal restriction sites (BamHI and Bglll) and two internal restriction sites (Kpnl and Avrll) (5'-p GATCCGGTACCGGCGGTCCTAGG AGATCTGGCGGTA-3' and 5'-p GCCATGGCCGCCAGGATCCTCTAGACCGCCATTCGA-3'). Using the interactive cloning strategy described above, a 'cloneable' Trx-Iinker octamer gene was assembled with the DNA cassette as the middle unit flanked by two Trx-Iinker tetramer genes (that is, the final construct is His6-SUMO-(Trx-linker)4-Kpnl-Avrll-(Trx-linker)4). A Trx-Iinker monomer mutant gene encoding an RRASAC peptide motif was created by site-directed insertion (forward primer, 5'- AGCGCCTGCGCGGGTTCTGCTGGTTCC-3'; reverse primer, 5'CGCACGGCG GCTCCCTGCACTTCCGGC-3') and subsequently cloned in between the Kpnl and Avrll sites within the Trx-Iinker octamer to give (Trx-linker)4-Trx-linker(RRASAC)-(Trx-linker)4. The placement of a single correctly oriented insert was confirmed by sequencing using primers targeting the Kpnl and Avrll ligation sites (forward primer, 5'-TGCGAGCGCCTGCGGTGG-3'; reverse primer, 5'-ACGCTCGCGGACGCCACC-3'). Expression and purification of Trx-Iinker concatemers Genes encoding the N-terminal His6-SUM0-tagged concatemers of Trx were cloned into the pOP3SU plasmid (kindly provided by M. Hyvonen). BLR(DE3) competent cells (Novagen) were transformed with the plasmids and grown in a Luria broth medium supplemented with ampicillin (100 \ig ml1) at 37 °C with continuous shaking (250 r.p.m.). Protein expression was induced in the exponential growth phase (OD600 = 0.6) with isopropyip-D-l-thiogalactopyranoside (0.5 mM final concentration). After 8 h, the cells were harvested by centrifugation (10 min, 5,000xg), resuspended in a binding buffer (30 mM Tris HCI, 250 mM NaCI, 25 mM imidazole at pH 7.2) supplemented with a protease inhibitor cocktail (complete, EDTA free; Roche) and lysed by sonication. Cell debris was removed by centrifugation at 20,000*gfor 45 min, and the supernatant loaded ontoaHis-Trap HP column (5 ml, Cytiva) at 0.2 ml min-1. The column was washed with 50 ml of the binding buffer before single-step elution with 15 mL of 30 mMTrisHCI,250 mM NaCI,300 mM imidazoleatpH 7.2. Asingle peak containing the almost pure protein was collected and dialysed (Slide-A-Lyzer G2 Dialysis Cassette, 10,000 molecular weight cutoff, 30 ml; Thermo Fisher) for 3 h against 41 of dialysis buffer (50 mMTris HCI,250mMNaCI,2mMl,4-dithio-D-threitol(DTT)atpH8.0),at4°C with continuous stirring, to remove excess imidazole. After injecting His6-tagged Ulpl protease into the dialysis cassette at a molar concentration ratio of1:200 (Ulpl:Trx-linker concatemer), the mixture was transferred into a fresh dialysis buffer overnight for SUMO-tag cleavage. The cassette was then transferred one last time into fresh dialysis buffer without DTT for 4 h. The dialysed protein was loaded onto a column packed with HisPur Ni-NTA Agarose Resin (5 ml, Thermo Fisher) equilibrated with a binding buffer (50 mM Tris HCI, 250 mM NaCI at pH 8.0) and the flow through was reapplied five more times. The final flow through containing the His6-SUMO-free protein was aliquoted and flash frozen for storage at-80°C. Expression and purification of SUMO protease Ulpl The Pfgetl9 Ulpl plasmid (Addgene) containing a His6-tagged Ulpl gene was transformed into T7 Express competent cells (NEB) and grown in a Luria broth medium supplemented with kanamycin (100 \ig mf1) at 37 °C with shaking (250 r.p.m.). Expression was induced at OD600 = 0.5 with isopropyip-D-l-thiogalactopyranoside (0.5 mM). Cells were harvested after 3 h by centrifugation, resuspended in lysis buffer (4 ml g-1; 50 mM Tris HCI, 300 mM NaCI, 10 mM imidazole at pH 7.5) supplemented with lysozyme (1 mg ml4) and incubated on ice for 30 min before sonication. The lysate was spun at 20,000 r.p.m. for 45 min to remove the cell debris and the supernatant was applied to a column packed with HisPur Ni-NTA Agarose Resin (5 ml, Thermo Fisher) and equilibrated with a binding buffer (50 mM Tris HCI, 300 mM NaCI at pH 7.5). The column was washed with 10 column volumes of wash buffer (50 mM Tris HCI, 300 mM NaCI, 20 mM imidazole at pH 7.5) and the protein was eluted with 10 ml of elution buffer (50 mM Tris HCI, 300 mM NaCI, 300 mM imidazole at pH 7.5). The eluted protein was dialysed against a storage buffer (50 mM Tris HCI, 200 mM NaCI, 2 mM 2-mercaptoethanol) overnight, aliquoted and flash frozen as a 50% stock in glycerol. Phosphorylation of Trx-Iinker concatemers Trx-Iinker concatemers (1 mg ml"1) were incubated with 50,000 units of the catalytic subunit of cAMP-dependent protein kinase (NEB)-which recognizes the RRAS motif within the central linker of the Trx-Iinker nonamer-in a protein kinase buffer (50.0 mM Tris HCI at pH 7.5,10.0 mM MgCI2, 0.1 mM EDTA, 4.0 mM DTT, 0.01% Brij 35 and 2.0 mM ATP) (NEB) at 30 °C for 1 h. The solution was then supplemented with additional ATP at a final concentration of 2 mM and DTT at a final concentration of 2 mM before overnight incubation at 30 °C. Trx-Iinker concatemers were purified and concentrated usingcentrifugal filters (Amicon Ultra 0.5 ml, 100 K), aliquoted and flash frozen for storage at -20 °C (10 mM HEPES at pH 7.2 and 750 mM KCI). Phosphorylation of the Trx-Iinker concatemers at a single site was verified by liquid chro-matography-mass spectrometry. Modification of cysteines on Trx-Iinker concatemers Reagents were purchased from Sigma-Aldrich, unless otherwise indicated. Trx-Iinker nonamer was first treated with tris(2-carboxyethyl) phosphine (TCEP) (70 to 100 eq) at 32 °C for 2 h in a protein storage buffer (50 mM Tris HCI, 250 mM NaCI at pH 8.0). Excess TCEP was removed by a desalting column (PD MiniTrap G-25 column, Cytiva). To glutathionylate the Trx-Iinker nonamer, the reduced protein was reacted with oxidized glutathione (100 eq) at 32 °C overnight in a protein storage buffer (50 mM Tris HCI, 250 mM NaCI at pH 8.0) before desalting to remove the excess reagent. The modified protein was aliquoted and flash frozen for storage at -20 °C. To glycosylate the Trx-Iinker nonamers, the reduced protein was reacted first with 2,2'dithiodipyridine (20 eq) at 32 °C overnight in the protein storage buffer (50 mM Tris HCI, 250 mM NaCI at pH 8.0). After the removal of excess 2,2'dithiodipyridine with a desaltingcolumn, the activated nonamer was reacted with the 6'sialyllactosamine derivative (NeuAca(2-6) LacNAc-PEG3-Thiol, 5 eq; Sussex Research Laboratories) overnight at 32 °C in a protein storage buffer (50 mM Tris HCI, 250 mM NaCI at pH 8.0). Modified nonamers were desalted (PD MiniTrap G-25 column, Cytiva), aliquoted and flash frozen for storage at -20 °C. The occurrence of glutathionylation or glycosylation at single sites was verified by liquid chromatography-mass spectrometry. Nature Nanotechnology Article https://doi.org/10.1038/s41565-023-01462-8 Single-channel recording Planar lipid bilayers of l,2-diphytanoyl-sn-glycero-3-phosphocholine (Avanti Polar Lipids) were formed by using the Miiller-Montal method on a 50-n.m-diameter aperture made in a Teflon film (25 n.m thick, Goodfellow) separating two 500 \i\ compartments (cis and trans) of the recordingchamber.Eachcompartmentwasfilledwitharecordingbuffer (750 mM GdnHCI,1.5 M GdnHCI,3.0 M GdnHCI,2.0 M urea/750 mM KCI or 750 mM KCI, 10 mM HEPES, 5 mM TCEP at pH 7.2 for Trx-linker dimer, tetramer, hexamer and octamer; 375 mM GdnHCI/375 mM KCI, 10 mM HEPES at pH 7.2 for Trx-linker nonamers). To record with Trx-linker dimer, tetramer, hexamer or octamer and ensure a reduced N-terminal cysteine, pretreatment of the protein samples with 5 mM TCEP was carried out for 10 min at room temperature. This pretreatment was not carried out with nonamers which lacked the N-terminal cysteine residue. Trx-linker concatemers were added to the els compartment (dimer, 2.20 uM; tetramer, 0.63 uM; hexamer, 0.25 uM; octamer, 0.81 uM; nonamer, 1.20 \iM). Ionic currents were measured at 24 ± 1 °C by using Ag/AgCI electrodes connected to the headstage of an Axopatch 200B amplifier. After a single (NN-113R)7 pore had inserted into the bilayer, the solution was replaced with a fresh buffer by manual pipetting, to prevent further insertions. Signals were low-pass filtered at 10 kHz and sampled at 50 kHz with a Digidata 1440A digitizer (Molecular Devices). Data analysis To establish the current signatures for the stepwise co-translocational unfoldingofTrxconcatemers,currenttraceswereanalysedusingClamp-fit 10.7 (Molecular Devices). The remaining current as a percentage of the open-pore current (/res%) was calculated for each step in individual A or B features (for example, /res%(Al) = /A1//open x 100%). The standard deviations were derived from data for Trx-linker units collected using separate pores. Trx-linker units that produced a level A3 or B3 with a dwell time of <1 ms were excluded from the/res% analysis due to possible undersampling. Root-mean-square noise values (/r.m.J for each current level were measured from current traces after the application of a post-recording filter of 2 kHz. Unless otherwise stated, the noise of the open pore was subtracted as follows: /rms 2 = /r m s (Al)2 - /r m s (open pore)2. Toobtain the stepwise kinetic profilesoftheco-translocational unfolding of Trx concatemers, current traces were idealized using Clampfit 10.7. The dwell-time analysis was performed by using the maximum interval likelihood algorithm ofQUB2.0 software (https://qub.mandelics.com/). Data availability Source data are provided with this paper. All other data pertaining to this study are available from the corresponding authors upon reasonable request. Reference 29. Carrion-Vazquez, M. et al. Mechanical and chemical unfolding of a single protein: a comparison. Proc. Natl Acad. Sci. USA 96, 3694-3699(1999). Acknowledgements This research was supported by the Wellcome Leap Delta Tissue Program, Oxford Nanopore Technologies and a European Research Council Advanced Grant (SYNTISU). The work was supported in part by the Francis Crick Institute that receives its core funding from Cancer Research UK (FC001002), the UK Medical Research Council (FC001002) and the Wellcome Trust (FC001002). W.-H.L. is funded by an Oxford-Taiwan Graduate Studentship in partnership with a Department of Chemistry Scholarship, University of Oxford. S.G.-M. is supported by a Leverhulme Trust Research Leadership Award (RL 2016-015), a Wellcome Trust Investigator Award (212218/Z/18/Z) and a Royal Society Wolfson Fellowship (RSWF/R3/183006). Y.Q. was supported by a Glasstone Research Fellowship and a Fellowship by Examination, Magdalen College, Oxford. For the purpose of open access, the authors have applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission. Author contributions P.M.-B., W.-H.L., M.R.-R., S.G.-M., Y.Q. and H.B. designed the study. P.M.-B., W.-H.L., S.B., M.R.-R. and Y.Q. performed the experiments. P.M.-B., W.-H.L., Y.Q. and H.B. wrote the manuscript. Competing interests H.B. is the founder of, a consultant for and a shareholder of Oxford Nanopore Technologies, a company engaged in the development of nanopore sensing and sequencing technologies. P.M.-B., W.-H.L., Y.Q., M.R.-R., H.B. and S.G.-M. have filed patents describing the electro-osmotically active nanopores and their applications in proteoform characterization. S.B. was listed as a contributor to the intellectual property. Additional information Extended data is available for this paper at https://doi.org/10.1038/s41565-023-01462-8. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s41565-023-01462-8. Correspondence and requests for materials should be addressed to Yujia Qing or Hagan Bayley. Peer review information Nature Nanotechnofogythanksthe anonymous reviewers for their contribution to the peer review of this work. Reprints and permissions information is available at www.nature.com/reprints. Nature Nanotechnology Article https://doi.org/10.1038/s41565-023-01462-8 N terminus C terminus W N terminus C terminus 200 ms 7xC a1 a2 a1 a3—^ j I ^ EOF 50 ms Cx5 b1 b2 b1 b3—^ J OpA Feature a Feature b C terminus a disulfide linkage between n-terminal cysteines C terminus 6xC ■ N Ni -C x7 a1 a2 a11 b1 b2 a3 b3-^ ^ OpA _I CM Feature a Extended Data Fig. 11 Repeating currentfeatures recorded during electroosmosis-drivenconcatemer translocation through a nanopore. a, Two repeating current features, A or B, were recorded with a charge-selective nanopore ((NN-113R)7) and Trx-linker octamers pre-treated with 5 mM tris(2-carboxyethyl)phosphine (TCEP) for 10 min before their addition to the cis compartment of the recording chamber. Conditions: 750 mM GdnHCl, 10 mM HEPES, 5 mM TCEP, pH 7.2,0.81 uM Trx-linker octamer (cis), +140 mV (trans), 24 ± 1 °C. b, Without the TCEP pre-treatment, features A were always seen before features B when they occurred together within a single translocation event. The first two levels (Bl and B2) in features B have larger noise and higher Ires% compared with Al and A2 recorded within a single polypeptide translocation event by the same pore (Al: Ires%= 35 ± 1 %, Ir m s = 1.1 ± 0.1 pA, N = 25; A2: Ires% = 21 ± 1%, Ir.m.s. = 1.5 ± 0.2 pA, N = 25; Bl: Ires%= 38 ± 1 %, Irms. = 1.7 ± 0.4 pA, Feature b 50 ms N = 39; B2: Ires% = 32 ± 1%, Ir.ms. = 2.0 ± 0.5 pA, N = 39; Ir.ms. values for each level were reported without subtraction of the noise of the pore; number of individual levels from multiple polypeptide translocation events recorded by the same pore are specified). The translocating molecules, which gave sequential A and B features, were assigned as dimers of octamers linked by a disulfide bond between the two N-terminal cysteines. Therefore, in the unlinked molecules (see 'a'), C terminus-first translocation occurred when features A were observed and N terminus-first translocation occurred when features B were observed. The recorded repeating features are indicated by orange and blue bars. Conditions: 750 mM GdnHCl, 10 mM HEPES, pH 7.2,0.81 uM Trx-linker octamer (cis), +140 mV (trans), 24 ± 1 °C. All traces were filtered at 2 kHz for clarity; transient A3 levels were truncated by filtering and therefore deviated from -0 pA. Nature Nanotechnology