Skip to main content
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2026 Sep 15;20(38):25993–26005. doi: 10.1021/acsnano.5c15163

Liposome-Supplemented Cell-free Synthesis of a Peptide that Forms β‑barrel Nanopores

Shoko Fujita 1, Mana Sato 1, Ryuji Kawano 1,*
PMCID: PMC13629683  PMID: 42915753

Abstract

Rational design of nanopore-forming proteins and peptides is a promising approach for single-molecule analysis by using custom-designed nanopores tailored to specific target molecules. One challenge in this approach is the expression or synthesis of the designed nanopores, due to their high hydrophobicity. We sought to streamline the synthesis of an artificial and highly aggregative nanopore-forming peptide using liposome-supplemented cell-free protein synthesis. By optimizing the liposomal lipid composition, we recovered yields of the expressed aggregative peptide that were suitable for functional measurements. Channel current measurements involving peptide-embedded liposome fusion to a planar lipid bilayer confirmed the pore formation and single-molecule detection capabilities of the expressed peptide nanopore. Through the exploration of various lipid compositions, it was found that the incorporation of anionic lipids and a sufficient acyl chain length were crucial for efficient peptide insertion and the reproduction of native oligomeric states. Such modulation of liposomal lipid composition alongside cell-free protein synthesis has the capacity to alleviate the aggregation propensity constraints in the design of nanopores and other membranous peptides.

Keywords: nanopores, cell-free protein synthesis, liposomes, channel current measurement, de novo design


graphic file with name nn5c15163_0008.webp


graphic file with name nn5c15163_0006.webp

Introduction

Nanopore sensing is a fast, low-cost, and label-free method for electrical single-molecule detection. Nanopore DNA sequencers are established as commercial applications of this technology, − while the implementation of nanopores in single-molecule protein sequencing and identification is underway. − Promising avenues in the detection of proteins, which are more structurally complex than DNA, include the use of not only biological nanopores, derived from naturally occurring pore-forming proteins or toxins, but also designed artificial nanopores that are tailored to the specific protein target. As such, artificially- or computationally designed (de novo designed) nanopores, transmembrane β-barrel proteins, − α-helical peptide bundles, − and β-hairpin peptide assemblies , have been reported.

The de novo design of such nanopores typically involves the expression and subsequent testing of multiple designed candidates, but their expression in E. coli or through solid-phase peptide synthesis remains challenging due to their toxicity and high hydrophobicity. Functional expression and folding of pore-forming proteins in E. coli requires precise sequence design and minimal cytotoxicity, , while the solid-phase synthesis of peptide nanopores necessitates carefully optimized purification conditions to prevent aggregation. , SVG28, a 28-residue β-barrel peptide nanopore we previously reported (Figure A), was also highly hydrophobic, tended to aggregate, and could not be purified after conventional solid-phase synthesis. It was eventually synthesized by incorporating the pro-form of the peptide bond, known as the O-acyl isopeptide. To overcome such complexities in conventional synthesis, we have attempted to utilize a cell-free protein synthesis (CFPS) system. CFPS synthesizes proteins in vitro utilizing the transcription and translation machineries of cells and enables facile and parallel synthesis. We chose the PURE (protein synthesis using recombinant elements) system among CFPS platforms. It utilizes individually purified synthesis machineries, does not contain proteases, and is thus applicable to peptide synthesis.

1.

1

Synthesis of SVG28 and its variants with hydrophilic arginine tags via the PURE system. A. The amino acid sequence of SVG28 (upper left) and the positions of the hydrophilic mutations in SVG28-D4 (upper right). Schematic indicating the pore structure of 7-mer β-barrel (bottom) B–C. Amino acid sequences of SVG28-R4 and R2-SVG28-R2 (B) and SVG28 variant-R4 and R2-SVG28 variant-R2 (C). D–G. MALDI-TOF/MS mass spectra of SVG28-R4 (D), R2-SVG28-R2 (E), SVG28 variant-R4 (F), and R2-SVG28 variant-R2 (G). Obs. mass and Calc. mass stand for observed and calculated [M + H]+ values of the peptides, respectively. The superscript f in f[M + H]+ represents the formylation of N-terminal methionine. Peaks 1 ∼ 4 shown in Figure G correspond to the following: (1) MRRRGSYSVGVSVSYSDSDG Calc. mass 2194.3 Obs. mass 2194.7, (2) MRRRGSYSVGVSVSYSDSDGS Calc. mass 2281.4 Obs. mass 2283.1, (3) MRRRGSYSVGVSVSYSDSDGSY Calc. mass 2444.6 Obs. mass 2445.1, (4) MRRRGSYSVGVSVSYSDSDGSYSV Calc. mass 2632.4 Obs. mass 2631.3.

One initial strategy for synthesizing hydrophobic transmembrane peptides, such as SVG28, using the PURE system was to insert hydrophilic mutations that do not inhibit the pore-forming ability. As shown in Figure A, SVG28 has alternating Ser (S) and Val (V) sequences in its transmembrane region, a β-turn formed by an Asp-Ser-Asp-Gly (DSDG) sequence, and Gly–Arg (GR) sequences at the terminus. In this study, SVG28 was difficult to obtain in a soluble and useable form using the PURE system because of its strong tendency to aggregate. To improve the solubility of the peptide, aspartic acid residues (D) were introduced just outside the transmembrane region, resulting in the hydrophilic variant SVG28-D4 (Figure A). SVG28-D4 was fortuitously synthesized in the PURE system, but its pore diameter was smaller than that of SVG28, probably due to the reduced transmembrane hydrophobicity required to assemble the pore structure. Such small pores made it difficult to apply this variant for the single-molecule detection of oligopeptides.

In the present study, we tackled the challenge of synthesizing the original SVG28, a highly hydrophobic and aggregative β-hairpin peptide, using the PURE system. We investigated three approaches: the incorporation of a solubilizing tag to improve solubility; the application of an organic solvent-based purification protocol developed for the amyloid-β peptide, which has similar physicochemical properties to SVG28; and the addition of liposomes to the PURE reaction solution to recover the synthesized hydrophobic SVG28 within the liposomal lipid membrane. We also evaluated the pore-forming ability and single-molecule detection capability of the synthesized SVG28 using channel current measurements. Among these strategies, liposome-supplemented CFPS was the most effective. We further demonstrated the applicability of liposome-supplemented CFPS to other pore-forming peptide sequences. To the best of our knowledge, the shortest protein successfully synthesized via liposome-supplemented CFPS reported to date was 77 amino acids. The application of liposome-supplemented CFPS to hydrophobic peptide synthesis opens further possibilities for the use of the PURE system in peptide synthesis.

Results and Discussion

Synthesis of SVG28 by Introduction of Solubilizing Tags

To synthesize SVG28 and its variants (Figure B and C) using the PURE system, we first introduced a polyarginine solubilization tag to either the C-terminus or both the N- and C-terminus of the peptides (Figure B and C). MALDI-TOF/MS analysis revealed that the arginine-tagged constructs either gave lower yields (Figure D and E), failed to be synthesized (Figure F), or underwent premature termination (Figure G). These data imply that extremely hydrophobic or aggregative peptide sequences were poorly synthesized or truncated. The concentration of these peptides was quantified by UV–vis spectroscopy. Based on absorbance at 205 nm, the amounts of peptides are as follows: SVG28-D4 (soluble hydrophilic variant), 2.3 ± 0.02 μg; SVG28-R4, 0.9 ± 0.9 μg; R2-SVG28-R2, 0.2 ± 0.2 μg. SVG28 variant-R4 was below the detection limit, while R2-SVG28 variant-R2 was 0.7 ± 0.1 μg, showing the reduced yield of peptides recovered in the soluble fraction.

To confirm the ability of the PURE system to express the hydrophobic SVG28 sequence, we fused superfolder green fluorescent protein (sfGFP) to the C-terminus of SVG28 and used sfGFP fluorescence to confirm translation of the SVG28-containing fusion protein (Figure A, second from the left). Compared to the hydrophilic variant of SVG28 with sfGFP (SVG28-D4-sfGFP), SVG28-sfGFP showed similar brightness but formed large precipitates at the bottom of the test tube (Figure A, second from the left). SVG28-sfGFP aggregates were visible under fluorescence but not under visible light, probably due to the low synthesis yield. This result shows that while the highly hydrophobic SVG28 can be synthesized using the PURE system, it cannot be purified using conventional methods such as reverse-phase column purification due to its strong propensity to aggregate.

2.

2

Confirmation of SVG28 synthesis and results of the DOPC-supplemented PURE system. A. Visual observation of sfGFP fluorescence in the synthesis of SVG28-sfGFP and SVG28-D4-sfGFP in the absence and presence of DOPC liposomes. B. MALDI-TOF/MS mass spectrum of SVG28 after solubilization. C. In liposome-supplemented CFPS, nascent translated chains with high hydrophobicity are spontaneously inserted into the liposomal membranes during synthesis. D. Comparison of the fluorescence intensity of the supernatant after SVG28-sfGFP synthesis in the absence and presence of DOPC liposomes. Error bars represent standard error. E. SDS-PAGE analysis after centrifugation purification. P indicates the pellet and S the supernatant fraction. SDS-PAGE bands corresponding to the molecular weight of sfGFP (28 kDa) were observed for SVG28-sfGFP and SVG28-D4-sfGFP. Since sfGFP alone does not interact with liposomes (data not shown), these bands could be attributed to the termination of translation in the sfGFP region of SVG28-or SVG28-D4-sfGFP. The absence of this byproduct in the sfGFP synthesis system was shown in Figure S3F. F. Typical current–time trace of DOPC liposomes purified from the PURE reaction solution of SVG28.

Solubilization of SVG28 Aggregates Postsynthesis

We next tried to solubilize SVG28 aggregates using a previously reported solubilization protocol for amyloid-β aggregates. This pathogenic β-hairpin peptide exhibits handling difficulties in vitro due to its low solubility and high aggregation propensity in the aqueous phase. A two-step treatment with hexafluoro-2-propanol (HFIP), which degenerates secondary structures through hydrophobic interactions, followed by dimethyl sulfoxide (DMSO), which disrupts the hydrogen-bonding network as a proton acceptor, has been reported to solubilize amyloid-β aggregates. The total PURE reaction solution of SVG28 (without sfGFP) was directly treated with HFIP and DMSO, and SVG28 was crudely fractionated into <5000 Da species using a gel filtration column (see Methods for details). The strong MALDI-TOF/MS peak corresponding to the molecular weight of SVG28 confirmed the full-length cell-free synthesis without premature termination (Figure B). However, the amount of purified SVG28 from 20 μL of the PURE reaction solution was less than 150 ng (Figure S1), compared to 2.3 μg of SVG28-D4, which was insufficient for channel current measurements. The low recovery of CFPS-synthesized SVG28 using these small reaction volumes is in line with observations from the previous study on amyloid-β, in which an approximately 20% loss from the initial 500 μg of peptide was reported during the disaggregation process.

Supplementing the PURE System with DOPC Liposomes

We next investigated peptide synthesis using the PURE system with incorporation of lipid components. The addition of membranous lipid particles such as liposomes (bilayer vesicles) − or nanodiscs/bicelles (membrane discs) − to CFPS has been shown to facilitate the spontaneous cotranslational insertion of ribosomally synthesized hydrophobic nascent chains into the lipid bilayers, thereby preventing hydrophobic aggregation (Figure C). This approach has yielded several functional membrane proteins ,,− and de novo α-helical membrane peptides. ,, To apply this approach to SVG28, we first introduced DOPC liposomes extruded through a 1 μm membrane into the PURE solution of SVG28-sfGFP. DOPC is a lipid that has previously been shown to be compatible with lipid-containing PURE systems. , Although SVG28-sfGFP precipitates remained apparent at the bottom of the test tube in the presence of DOPC liposomes (Figure A, rightmost), the fluorescence intensity of the DOPC liposome-containing supernatant increased (Figure D). The increase in supernatant fluorescence suggests that although some SVG28-sfGFP precipitated into aggregates, some of it was also reconstituted into the liposomes. The DOPC liposomes contained 500 mM glucose inside and were purified through two rounds of centrifugation and resuspension. The purification was confirmed by SDS-PAGE, which showed a 32 kDa band corresponding to SVG28-sfGFP (Figure E).

SVG28 without sfGFP was then synthesized under the same conditions. Turbidity caused by the DOPC liposomes made it difficult to visually confirm the presence or absence of SVG28 aggregates under visible light. DOPC liposomes purified from the PURE reaction solution by centrifugation were applied to channel current measurements to observe pore-forming signals through the spontaneous fusion of the liposomes with a planar DOPC:cholesterol (DOPC:chol., 2:1 mol:mol) bilayer. − However, no current signals indicative of pore formation were observed (Figure F). Three possible explanations could account for this result: (1) SVG28 was not synthesized, (2) SVG28 was synthesized but did not insert into the DOPC liposomes, or (3) SVG28 successfully inserted but the SVG28-inserted liposomes failed to fuse with the planar bilayer. (1) was excluded by our earlier confirmation of SVG28 synthesis (Figure B), and (3) was excluded by control experiments in which DOPC liposomes from the PURE reaction solution of α-hemolysin (αHL) showed pore-forming signals (Figure S2A). We therefore concluded that SVG28 failed to insert effectively into the DOPC liposomes.

Peptide Reconstitution into DOPE/DOPG Liposomes

To enhance the insertion efficiency of CFPS-synthesized SVG28 into the liposomes, we explored alternative liposomal lipid compositions with greater affinity for SVG28 than DOPC. The N-terminus of SVG28 contains the positively charged terminus and cationic amino acid arginine (NH2-MRGSY···), and so we hypothesized that anionic lipids such as DOPG would enhance liposome-amino acid interactions. Kamat reported that the N-terminal five residues of the membrane protein contribute to membrane affinity in liposome-supplemented CFPS, further supporting this hypothesis. Based on previous studies that reported efficient protein insertion with CFPS using DOPE/DOPG liposomes to mimic the inner membrane composition of E. coli, we tested DOPE:DOPG (PE:PG, 1:1 mol:mol) liposomes. ,, The PE:PG liposomes purified from the liposome-supplemented PURE reaction solution of SVG28 were fused to a planar PC:chol. bilayer in channel current measurements. In contrast to DOPC liposomes, pore-forming signals of SVG28 were successfully observed when using PE:PG liposomes (Figure A).

3.

3

Analysis of current signals obtained from PE:PG liposomes purified from the PURE reaction solution of SVG28. A. Typical current–time traces for PE:PG liposomes purified from the liposome-supplemented PURE reaction solution of SVG28. B, C. The ratio of duration (B) and the count (C) of stable pore formation and unstable pore formation taken from the channel current analysis. For the solid-phase synthesis data, the raw data originally reported in Shimizu et al. (2022) was reanalyzed.

To compare the pore-forming tendencies of SVG28 synthesized via the PURE system and via solid-phase synthesis, the observed current signals were classified into either stable pore formation or unstable pore formation based on their pore current noise , (detailed definitions are provided in the Methods and Figure S4). For nanopore sensing, it is necessary to have a high proportion of stable pore formation. We measured the occurrences and duration of the open pore state to estimate the probability of its appearance. We found that the proportion of stable pore formation of SVG28 synthesized via the PURE system was significantly lower than that of SVG28 synthesized via solid-phase synthesis, in terms of both stable pore duration and count (Figures B,C and S5D,E). The pore size and the number of monomers in pore assembly can be estimated from the current conductance of pore-forming signals. SVG28 synthesized via solid-phase synthesis typically forms a 7-mer pore with a conductance of approximately 1.1 nS in 1 M KCl. In contrast, SVG28 formed via the PE:PG liposome-supplemented PURE system predominantly resulted in smaller pores, with a conductance of approximately 0.5 nS (Figure S5F).

Therefore, although the SVG28 sequence synthesized via the PURE system is the same as that obtained via solid-phase synthesis, the ratio of stable pore formation and the resulting pore sizes was seen to differ. Furthermore, the stable pore ratio and pore size also varied depending on the liposomal PE/PG ratio (3:1 vs 1:1 mol:mol, Figure S5). These results suggest that SVG28 synthesized via the PURE system may form different β-barrel assemblies due to the influence of the surrounding liposomal lipid composition, and the morphology of the assembled β-barrel structures may be retained even after the liposomes fuse with the excess PC:chol. planar lipid bilayer.

Optimization of Liposomal Lipid Composition for the Stable Pore Formation of SVG28 and Other Aggregative Pore-forming Sequences

The improper assembly of CFPS-synthesized SVG28 could be attributed to its short hydrophobic transmembrane length relative to the bilayer lipid membrane. Efficient insertion of membrane proteins and peptides into lipid bilayers requires that the hydrophobic length of the peptide transmembrane regions matches the thickness of the hydrophobic core of the lipid bilayer. ,, Therefore, we further explored the impact of the liposomal lipid composition by varying the lipid acyl-chain length. In this experiment, PC lipid was used instead of PE since PE lipids with varied acyl chain lengths are not commercially available. We tested three different types of PC:PG bilayers: 18:1 DOPC (18PC), 16:1 DYPC (16PC), and 14:1 DRPC (14PC), mixed with DOPG at 1:1 molar ratio (18PC:PG, 16PC:PG, 14PC:PG). All-atom MD simulations showed the minimum augmentation of the membrane thickness–due to stretching of the lipid molecules around the SVG28 pore–for 16PC:PG, suggesting that 16PC:PG is the most suitable lipid composition (Figure A,B).

4.

4

Channel current measurements of SVG28 synthesized with PC/DOPG liposomes. A. Snapshots of 7-mer SVG28 in 18PC:PG, 16PC:PG, and 14PC:PG membranes after 200 ns MD simulation. Lipid molecules are shown in green, SVG28 is shown in pink, and the membrane-interface-anchoring tyrosine residues in SVG28 are shown in light blue. For clarity, water and ion molecules are not shown. B. The membrane compression around SVG28 in 18PC:PG, 16PC:PG, and 14PC:PG membranes during the last 5 ns of MD simulation. Error bars represent standard error. C–E. The typical current–time traces of 18PC:PG (C), 16PC:PG (D), and 14PC:PG (E) liposomes. F, G. The ratio of duration (F) and the count (G) of stable pore formation and unstable pore formation. H. A histogram of the current conductance of open levels for 18PC:PG, with peak-fitting curves plotted using the Gaussian function. I. Comparison of pore-formation frequency. Error bars represent standard error. Statistical significance (*p < 0.05, **p < 0.01) was determined using the Mann–Whitney U test. J. CD spectrum of the 18PC:PG liposomes purified from the liposome-supplemented PURE reaction solution of SVG28.

Under the same channel current measurement conditions as before, pore-forming signals were observed for all three types of liposomes (Figure C–E). However, 14PC:PG showed insufficient current signals for reliable comparison. The thinness of the 14PC:PG bilayer may have hindered SVG28 insertion. This contrasts with the ability of α-helical peptides to insert into membranes with insufficient thickness by tilting themselves. , The proportion of stable pores based on pore duration improved when using 18PC:PG (38%) and 16PC:PG (30%) liposomes compared with 18PE:PG liposomes (17%) (Figures F, B, and S6A). The counts of stable pores also improved (18, 11, 7 counts for 18PC:PG, 16PC:PG, and 18PE:PG, respectively) (Figures G, C, and S6B). Although the stable pore duration remained lower than that of SVG28 yielded via solid-phase synthesis (74%), the stable pore count became comparable (14 counts). The pore-forming signals of 18PC:PG liposomes showed pore conductance of around 1.4 nS in 1 M KCl (Figure H), close to the conductance of 7-mer solid-phase SVG28 pores (1.1 nS), whereas 16PC:PG only showed pores with a conductance of 0.8 nS (Figure S6C). Hence, 18PC:PG provided the optimal liposomal lipid composition as it relatively consistently and stably formed pores with conductances similar to those of SVG28 yielded by solid-phase synthesis. In exploration of the optimal electrostatic strength of the liposomes, 18PC:PG (3:1 mol:mol) and 18PC:PG (1:3 mol:mol) liposomes were also tested, but 18PC:PG (1:1 mol:mol) remained optimal owing to its stable pore formation and pore size (Figure S7).

Considering the discrepancy seen between the prediction for 16PC:PG preference in all-atom MD simulations and the preference for 18PC:PG seen from channel current measurements, we calculated the side-chain fluctuation of tyrosine (Tyr), which anchors to the membrane interface, in the MD simulation. The averaged fluctuation of tyrosine was seen to be lowest in the 18PC:PG membrane (Figure S6D). In this system, the Tyr side chains may be stabilized by being embedded in a thick membrane with a suitable depth for anchoring (Figure S6D,E). This result suggests that both the Tyr anchoring effect and hydrophobic matching of the nascent peptide chain influence the stability of pore structures within membranes. Such stability of pore structures within membranes is considered to correlate with liposomal membrane insertion efficiency, as reported in previous reports. ,,

We next characterized SVG28 synthesized via the PURE system in detail, using the optimal lipid composition of 18PC:PG (1:1 mol:mol), which we anticipated would facilitate the most efficient insertion of SVG28. To confirm that the obtained pore-forming signals were derived from the cell-freely synthesized SVG28 properly recovered within liposomes, rather than from artifacts, we performed control measurements using 18PC:PG from the PURE reaction solution without SVG28-coding DNA or with SVG28-coding DNA but with 18PC:PG liposomes added only after the synthesis. A significant increase in the pore frequency between 18PC:PG from the PURE reaction solution with SVG28-coding DNA and these controls (Figure I) shows that the observed pore openings originated from SVG28 synthesized and recovered within liposomes during the PURE reaction.

To verify that changing the liposomal lipid composition from DOPC to DOPC:DOPG (1:1 mol:mol) increased the amount of cell-freely synthesized SVG28 incorporated with liposomes, we evaluated the incorporated peptide amount using the bicinchoninic acid (BCA) assay, which allows peptide quantification even in the presence of lipids. Liposome fractions from the cell-free reaction solution of SVG28 were separated by density gradient ultracentrifugation and analyzed. The amount of peptide recovered with DOPC liposomes was below the detection limit. In contrast, the amount recovered with DOPC:DOPG liposomes (1:1 mol:mol) was 0.4 ± 0.1 μg (n = 3, mean ± s.d.) per 30 μL reaction solution. Assuming a total synthesis yield of approximately 2.3 ± 0.02 μg, as previously confirmed for the cell-free synthesis of the hydrophilic SVG28 variant SVG28-D4 described above, these results suggest that approximately 20% of the synthesized peptide was recovered within the liposomes. In previous studies by Harris, , the insertion efficiency of natural membrane proteins was reported to be 20–40%, relatively close to this result.

To confirm whether the interference from the presence of liposomes on the cell-free reaction solution caused the lack of incorporation of SVG28 with DOPC liposomes, the synthesis efficiency was briefly evaluated by the SDS-PAGE analysis of SVG28-sfGFP. The synthesis yield estimated by band intensity of SVG28-sfGFP was 0.96-fold with DOPC liposomes and unchanged (1.0-fold) with DOPC:DOPG liposomes compared to the SVG28-sfGFP synthesis without liposomes (n = 3). This indicates that the lack of channel current signals of DOPC liposomes is due to the reduced interaction between cell-freely synthesized SVG28 and DOPC liposomes, not the interference of DOPC liposomes in the cell-free synthesis reaction.

To confirm that cell-freely synthesized SVG28 folds in the liposomal lipid bilayer rather than on its surface, we incubated the liposomes purified from the cell-free synthesis reaction solution of SVG28 with proteinase K. After the incubation at 37 °C for 1 h, channel current measurement of the treated liposomes still showed stable pore formation (Figure S8A). Furthermore, we added calcein encapsulated liposomes at self-quenching concentration to the cell-free synthesis reaction solution of SVG28, and measured the fluorescence of calcein released from the liposomes upon pore formation of SVG28 in liposomal membranes. A significant increase in fluorescence intensity upon the cell-free synthesis of SVG28 with DOPC:DOPG (1:1 mol:mol) liposomes (Figure S8B) suggests that SVG28 formed pores in the liposomal membrane. In addition, the changing pore-forming behavior in channel current measurements among 18PC:PG, 16PC:PG, and 14PC:PG, which exhibit almost identical surface physical properties, suggests that the cell-free-synthesized SVG28 folds in the liposomal membranes rather than on its surface.

With the cell-free synthesis of SVG28, the liposomal diameter was not changed from 717 ± 78 nm, compared to 695 ± 30 nm (mean ± s.d., n = 3), liposomes from the cell-free synthesis reaction solution without SVG28-coding DNA (Figure S9), showing that SVG28 expression does not induce the fusion of liposomes. Circular dichroism spectroscopy of the 18PC:PG liposomes from the PURE reaction solution of SVG28 showed a negative peak at 227 nm, indicating the β-sheet-like structure of the synthesized SVG28 (Figure J). The red-shifted negative peak corresponds to the previously reported cell-freely synthesized hydrophilic variants of SVG28 and suggests an unordered conformation with fractions of β-turn/β-strand structures.

To discuss the scalability of this system, we synthesized another hydrophobic peptide sequence, SVG28-N2, which had previously been confirmed not to be synthesized in the liquid-phase PURE system. Although the number of observed pore formations was lower than that of SVG28, stable pore formation of SVG28-N2 was observed (Figure S10). This result shows that this system is applicable to other pore-forming sequences. We also applied this system to synthesize and evaluate pore-forming ability of SVG28 that has an additional transmembrane α-helix and could not be adequately synthesized in the liquid-phase PURE system. Furthermore, using this system, synthesis and stable pore formation were confirmed for a substantially different pore-forming sequence that had low sequence identity to SVG28 (57% identity) and was difficult to synthesize by conventional solid-phase synthesis without O-acyl isopeptides despite multiple synthesis attempts (data not shown).

CFPS-Synthesized SVG28 Provides Utility in Nanopore Sensing

The pore frequency calculated above for liposome-supplemented CFPS using 18PC:PG liposomes (1.1 pores per hour) was lower than that for SVG28 prepared via solid-phase synthesis (1 μM, 1.9 pores per hour) (Figure A). To improve this, we tested several strategies to enhance the liposome fusion efficiency. The liposomes prepared using a 1 μm membrane were further resized using a 100 nm membrane (see Methods). The resizing, which lowers the free energy barrier for membrane fusion, , increased the pore frequency from 1.1 to 1.6 pores per hour (Figure A). Dynamic light scattering (DLS) measurement showed the resizing of liposomes from 717 ± 79 nm to 283 ± 12 nm (n = 3, mean ± s.d.) (Figure S11). Malinin and Lentz demonstrated the linear dependency of liposome curvature (1/R) and rate-limiting fusion energy barrier, and based on the measured diameter in the DLS measurement, the theoretically calculated increase in fusion efficiency is 2.5. This was comparable to the 1.4-fold increase in step pore frequency in the channel current measurement. Furthermore, assuming no loss of solution during the extrusion operation, the experimental pore frequency increase after resizing can be corrected to 2.0, corresponding to a 1.8-fold increase.

5.

5

Characterization of SVG28 synthesized via the liposome-supplemented PURE system. A. Pore frequency of 1 μm 18PC:PG liposomes purified from the PURE reaction solution of SVG28 (overlapping with the data in Figure I), 18PC:PG liposomes resized using a 100 nm membrane, and SVG28 via solid-phase synthesis. For solid-phase synthesis data, the raw data originally reported in Shimizu et al. (2022) was reanalyzed. B, C. The ratio of duration (B) and the count (C) of stable pore formation and unstable pore formation of SVG28 prepared via solid-phase synthesis, 18PC:PG liposomes purified from the PURE reaction solution of SVG28 (overlapping with the data in Figure F and G), 18PC:PG liposomes resized using 100 nm membrane, and SVG28-D4, a hydrophilic variant of SVG28 synthesized via the PURE system. For SVG28-D4 data, the raw data originally reported in Fujita et al. (2023) was reanalyzed.D, E. Histograms of the current conductance (D) and normalized RMS (E) of open levels for 18PC:PG liposomes purified from the PURE reaction solution of SVG28 and resized using a 100 nm membrane. Peak-fitting curves plotted using the Gaussian function. F. The current–time traces of cell-free SVG28 with and without 600 nM L-PLL under +100 mV in 1 nS pores. Black lines represent 0.4 nS below the open level of the pores. G. Illustration of a typical blocking current. H, I. Box plots of the blocking event frequency (H) and blocking duration (I) of cell-free SVG28 pores with and without 600 nM L-PLL under +100 mV. Center: median, square: average, bounds of box: IQR 25th and 75th percentiles, whiskers: minimum and maximum within 1.5 IQR, rhombus: outliers. Significance (*p < 0.05) was determined using the two-sample t-test. J–L. Blocking event frequency (J), blocking duration (K), and blocking conductance (L) of 600 nM L-PLL translocating CFPS-synthesized SVG28 pores at +50, +100, and +150 mV. Error bars represent standard error. M. The current–time traces of cell-free SVG28 with 500 nM poly­(dT)50 or 500 nM R7G under +100 mV in 1 nS pores. Black lines represent the 0.4 nS below the open level of the pores. N, O. Box plots of the blocking event frequency (N) and blocking duration (O) of cell-free SVG28 pores with 600 nM L-PLL or 500 nM poly­(dT)50 or 500 nM R7G or without detection target molecules under +100 mV (overlapping with the data in Figure H and I). Center: median, square: average, bounds of box: IQR 25th and 75th percentiles, whiskers: minimum and maximum within 1.5 IQR, rhombus: outliers. Significance (*p < 0.05, **p < 0.01, ***p < 0.001) was determined using a two-sample t-test. All data points are plotted for the blocking duration.

In contrast, neither the addition of cationic lipids (DOTAP) to the planar lipid bilayer, nor the addition of calcium ions to the cis electrolyte solution improved the pore frequency (Figure S12A–C). Notably, liposome resizing also improved the stable pore duration ratio, from 38% for the liposomes before extrusion to 67% for the liposomes after extrusion (Figures B,C; S12D and S12E). This value is comparable to SVG28 prepared via solid-phase synthesis (74%). Since resizing primarily affects the liposome fusion efficiency, the altered fusion dynamics or membrane curvature of smaller liposomes may influence the channel current properties of assembled pore structures. The resulting stable pore duration was superior to that of the previously reported hydrophilic variant of SVG28 (Figure A) synthesized via the PURE system (Figure B and C).

The pore conductance of resized 18PC:PG liposomes was frequently observed as 1.0 nS in 1 M KCl (corresponding to 7-mer, Figure D), which is the same as for SVG28 prepared via solid-phase synthesis (1.1 nS, Figure S13A). We quantified the noise level of CFPS-synthesized SVG28 pores in anticipation of their use in applications for single-molecule analysis. The noise level (normalized RMS noise) was calculated as the square root of the difference between the unbiased variances of the noise after pore opening and the baseline, and normalized according to the pore size (see Methods for details). The proportion of pores with a normalized RMS noise below 4, indicating sufficiently low-noise pores for use in single-molecule detection, was 27% of SVG28 prepared via the PURE system, compared to 32% of SVG28 prepared via solid-phase synthesis (Figures E and S13B). The reduced proportion of low-noise pores prepared via the PURE system may be attributed to the presence of the translation-initiating methionine and the influence of DOPG on the assembly of the SVG28 pore.

Channel current measurement of the resized liposomes was performed using another planar bilayer lipid membrane composition, DOPC:DOPG:cholesterol (PC:PG:chol, 1:1:1 mol:mol:mol). The frequently observed pore conductance of cell-free SVG28 with PC:PG:chol. bilayer was 1.2 nS, and the normalized RMS <4 was 22% (Figure S14). The close agreement in pore conductance between the PC:chol. and PC:PG:chol. membranes suggests that the assembled β-barrel structures of cell-freely synthesized SVG28 in liposomal membranes were retained even after the liposomes fuse with the planar lipid bilayer membrane. The increased pore noise with PC:PG:chol. bilayer is probably due to the changed planar bilayer membrane environment.

Compared to the hydrophilic variant SVG28-D4, SVG28 prepared via the liposome-supplemented PURE system did not exhibit a reduced pore size or a significant increase in pore noise (Figure S13C,D), suggesting its applicability in the single-molecule detection of polypeptides, as previously reported. Then, we validated the applicability of SVG28 prepared via the PURE system (using the 18PC:PG liposomes after resizing through 100 nm membrane) for single-molecule detection using the model polypeptide long poly-l-lysine (Mw: 30,000–70,000; L-PLL). The addition of 600 nM L-PLL to CFPS-synthesized SVG28 pores resulted in transient downward blocking currents (Figures F–H and S15A), indicating the translocation of L-PLL through the pores. A threshold of blocking duration time greater than 10 ms was set to distinguish between pore noise and blocking currents (Figure G and I) with similar wave forms, as shown in Figure F. The frequency of blocking events increased with an increase in applied voltage from +100 to +150 mV above this threshold (Figure J), and their blocking duration was decreased (Figure K). This result confirms the electrophoretic translocation of L-PLL through CFPS-synthesized SVG28 pores (Figure S15B). The blocking conductance (i.e., the amount of blocked ionic current; Figure G) decreased within this voltage range (Figure L), suggesting a stretched conformation of L-PLL, which is attributable to the enhanced water flow through the cell-free SVG28 pores, induced by electroosmosis. SVG28 pores synthesized via CFPS therefore demonstrate the capability for single-molecule detection, just as observed for SVG28 pores prepared via solid-phase synthesis.

To test other detection target molecules, 500 nM poly­(dT)50 or RRRRRRRG (R7G) peptides were added to the channel current measurement of cell-free SVG28 pores. Upon addition of these target molecules, downward current-blocking signals were observed (Figure M). The blocking event frequency was significantly increased compared with the inherent pore noises (Figure N). The event frequency increased with the molecular weight of the detection target (30–70 kDa for L-PLL, 15 kDa for poly­(dT)50, 1.2 kDa for R7G). The duration of blocking events of poly­(dT)50 and R7G was significantly increased compared with the inherent pore noises and tended to be relatively shorter than those observed for L-PLL. However, no statistically significant differences were detected among the L-PLL, poly­(dT)50, and R7G (Figure O).

Conclusion

We implemented the cell-free synthesis and channel current measurement of the highly hydrophobic de novo nanopore-forming β-hairpin peptide SVG28 using a liposome-supplemented PURE system with optimized lipid composition. Initially, fusion of the peptide with a fluorescent protein (sfGFP) and solubilization of peptide aggregates using organic solvents confirmed the synthesis of SVG28 via CFPS and demonstrated the propensity of SVG28 to aggregate. Through iterative exploration of the liposomal lipid composition and measurement conditions, stable pore formation with a conductance of 1.0 nS in 1 M KCl (7-mer) was observed for SVG28 synthesized by CFPS. This reproduced the native oligomeric state of SVG28 prepared via solid-phase peptide synthesis. We found that the key to achieving sufficient incorporation of CFPS-synthesized SVG28 into liposomes was the addition of anionic lipids (DOPG) to the liposomes and sufficient acyl chain length (∼16:1 or 18:1). Having found the optimal liposomal lipid composition of 18PC:PG, we confirmed the translocation of polypeptides through the CFPS-synthesized SVG28 pore. We demonstrated that aggregative peptides that require specialized chemical synthetic methods, such as the O-acyl isopeptide approach, and are difficult to synthesize through commercial peptide synthesis providers can also be synthesized using liposome-supplemented cell-free synthesis.

We show that optimization of the liposomal lipid composition was essential to achieve sufficient incorporation of the peptide into the lipid membrane for functional measurements. Our results suggest that peptides and proteins that were previously considered to aggregate or fail to synthesize using liposomes with neutral charge ,, or without lipid components may be synthesized using liposome-supplemented CFPS with optimized liposomal lipid compositions. Further investigation is required to evaluate how the different liposomal lipid compositions (PE:PG, 16PC:PG, PC:PG (1:3 and 3:1 mol:mol)) influenced the pore sizes of the assembled peptide despite synthesizing the same SVG28 sequence. In addition, other unexplored liposomal lipid conditions, such as DOPS, another negatively charged lipid, and cholesterol, a lipid that contributes to membrane fluidity, as well as variation in measurement temperature, a critical factor influencing membrane fluidity for both liposomes and planar bilayer lipid membranes, are expected to be examined.

Liposome-supplemented CFPS with optimized lipid compositions, as demonstrated herein, will accelerate the custom design of nanopores through alleviation of the aggregation propensity constraints of nanopore sequence design. The simple and parallel CFPS system can be applied to the comprehensive design of various nanopores. Furthermore, this approach could be extended to other membrane peptides, such as antimicrobial peptides, cell-penetrating peptides, and signaling peptides, where their functionality can be similarly evaluated through the fusion of peptide-embedded vesicles with planar bilayers or other vesicles, or via CFPS within vesicles. Given that comprehensive peptide synthesis using the PURE system has recently been reported, this study is expected to further broaden the application of the cell-free protein synthesis system to other peptides.

Methods

Reagents and Chemicals

The following reagents were used: Purefrex2.0 (GeneFrontier, Japan); GL-Tip SDB (GL Sciences, Japan); HiTrap desalting columns (Cytiva, Japan); Water for molecular biology (Sigma-Aldrich, USA); Trifluoroacetic acid (TFA, FUJIFILM Wako Pure Chemical Industries, Japan); Acetonitrile (ACN, FUJIFILM Wako Pure Chemical Industries, Japan); α-cyano-4-hydroxycinnamic acid (CHCA, Tokyo Chemical Industry, Japan); Polyethylene glycol 4000 (PEG4000, Tokyo Chemical Industry, Japan); Sodium trifluoroacetate (Sigma-Aldrich, USA); 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC, Avanti Polar Lipids, USA); Cholesterol (Sigma-Aldrich, USA); n-decane (FUJIFILM Wako Pure Chemical Industries, Japan); 3-morpholinopropane-1-sulfonic acid (MOPS, Nacalai Tesque, Japan); Potassium chloride (KCl, Nacalai Tesque, Japan); Potassium hydroxide (KOH, FUJIFILM Wako Pure Chemical Industries, Japan); d­(+)-Glucose (Kishida Chemical, Japan); Sucrose (Kishida Chemical, Japan); N-(2-Hydroxyethyl)­piperazine-N′-2-ethanesulfonic Acid (HEPES, Nacalai Tesque, Japan); 1,2-dioleoyl-sn-glycero-3-phospho-(1′-rac-glycerol) (DOPG, Avanti Polar Lipids, USA); 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE, Avanti Polar Lipids, USA); 1,2-dipalmitoleoyl-sn-glycero-3-phosphocholine (DYPC, Avanti Polar Lipids, USA); 1,2-dimyristoleoyl-sn-glycero-3-phosphocholine (DRPC, Avanti Polar Lipids, USA); Poly-l-lysine hydrobromide (Mw:30,000–70,000, Sigma-Aldrich, USA); poly­(dT)50 ssDNA (FASMAC Co., LTD, Japan); DNA templates for CFPS were synthetic genes provided by Integrated DNA Technologies, USA. Buffered electrolyte solutions and buffered glucose solutions were prepared from ultrapure water from a Milli-Q system (Millipore, USA). Products of CFPS were dissolved at a concentration of 10 μM in ultrapure water and stored at −30 °C. Prepared liposome solutions were stored at 4 °C. The solid-phase synthesis of R7G and confirmation of synthesis are described in our previous paper.

CFPS of Peptides without Liposomes

Purefrex2.0 (GeneFrontier, Japan) was used for CFPS. Template DNAs were dissolved at 10 or 12 ng/μL in nuclease-free water and stored at 4 °C. Synthesis was performed at 37 °C for 6 h. A reaction mixture containing 1 μL of 10 ng/μL DNA, 10 μL of solution I, 1 μL of solution II, 2 μL of solution III, and 6 μL of nuclease-free water was prepared. After translation, the peptides were desalted and purified with GL-Tip SDB (GL Sciences, Japan). MALDI-TOF/MS analyses were carried out using the MALDI-TOF autoflex speed (Bruker, USA) in linear/positive mode. PEG4000 was used for external mass calibration. The lyophilized peptides from 20 μL of reaction solution were dissolved in 3 μL of TA30 (ACN: 0.1%TFA = 3:7, vol:vol). Ten μg of CHCA dissolved in 100 μL of TA30 and the peptide sample were mixed in a sample-matrix ratio of 1:2 (vol:vol). To quantify the purified peptide concentration, the lyophilized peptides for 20 μL of reaction solution was dissolved with 4 μL of ultrapure water, and their UV–vis spectrum were measured using NanoDrop Ultra (Thermo Fisher Scientific, USA).

Solubilization of SVG28

Following the report by Broersen et al., 200 μL of HFIP was added to 20 μL of the PURE reaction solution of SVG28, vortexed for 1 min, and then dried with nitrogen gas. The residual SVG28/PURE reaction mixture/HFIP film was redissolved in 500 μL of DMSO and vortexed for 1 min. Prior to gel filtration, large precipitates were removed using an Ultrafree-MC centrifugal filter (Millipore, USA) with a pore size of 0.1 μm, by centrifuging at 12,000 g for 4 min. SVG28 was roughly purified from the SVG28/PURE reaction mixture/DMSO solution using a desalting column with an exclusion limit of Mr 5000 (HiTrap desalting columns, Cytiva, USA) that was pre-equilibrated with 0.1% (vol:vol) TFA. 500 μL of the SVG28/proteins in PURE reaction mixture/DMSO solutions was applied to the column, followed by an additional injection of 1 mL of 0.1% (vol:vol) TFA. SVG28 was eluted from the column by an additional injection of 1.5 mL of 0.1% (vol:vol) TFA and subsequently lyophilized.

Liposome-Supplemented CFPS of SVG28

Liposomes were prepared using the gentle hydration method. 50 μL of 270 mg/mL lipids dissolved in chloroform was placed in a glass microtube and evaporated by rotation under a flow of nitrogen gas. To completely remove the chloroform, the microtube was placed in a desiccator under vacuum overnight. The lipid films were hydrated with 500 μL of 0.5 M glucose for 3 h at 40 °C. After hydration, 200 μL of a buffer containing 0.5 M glucose and 100 mM HEPES (pH 7.4) was added to 200 μL of the liposome suspension. The liposome solution was then extruded using an extruder (Avanti Polar Lipids, USA) equipped with a 1.0 μm pore-size polycarbonate membrane (Whatman, UK). Liposome-supplemented CFPS was performed at 37 °C for 6 h. A reaction mixture containing 1 μL of 12 ng/μL DNA, 10 μL of solution I, 1 μL of solution II, 2 μL of solution III, 6 μL of nuclease-free water (Sigma-Aldrich, USA), and 10 μL of 10 mg/mL liposome solution was prepared. After the CFPS, the reaction mixture was left at 4 °C for about 1 week. In the purification, 150 μL of 50 mM HEPES was added to the PURE reaction solution, and then the mixture was centrifuged at 20,000 g for 5 min. After centrifugation, 160 μL of the supernatant was carefully discarded. Another 150 μL of 50 mM HEPES was added to the centrifuged solution and centrifuged again at 20,000 g for 5 min. All supernatant was then removed, and the pellet was resuspended in 10 μL of 50 mM HEPES. For further resizing of purified liposomes, a mixture containing 190 μL of 500 mM glucose and 50 mM HEPES, and 10 μL of liposome solution in 50 mM HEPES was extruded using an extruder (Avanti Polar Lipids, USA) equipped with a 0.1 μm pore-size polycarbonate membrane (Whatman, UK).

Characterization of Cell-freely Synthesized SVG28

The purified liposomes were analyzed by Tricine-SDS-PAGE using a 16.5% polyacrylamide gel (p-PAGEL, ATTO, Japan). The gels were stained using Coomassie Brilliant Blue (CBB) staining (EzStain AQua, ATTO, Japan) or fluorescent staining (ProLite Orange Protein Gel stain 5000X, AAT Bioquest, USA). The fluorescence intensity of SVG28-sfGFP was measured using a spectrometer (SpectraMax iD5, Molecular Devices, USA) with an excitation wavelength of 395 nm and a detection wavelength of 510 nm. CD (circular dichroism) measurements were performed using a J-1100 CD Spectrometer (JASCO, Japan). Purified liposome solutions in 10 μL of 50 mM HEPES were prepared for the microsampling disc with a 0.2 mm path length. The spectra were acquired every 0.5 nm from 215 to 300 nm with a CD scale of 20 mdeg/0.05 dOD, a 1 nm bandwidth, and four times integration.

For cases in which only the liposome fraction is purified for DLS measurements and BCA assays, density gradient ultracentrifugation (DGU) tubes were prepared by first adding 900 μL of 0% sucrose in 50 mM HEPES to the tube. Subsequently, 900 μL of 25% sucrose in 50 mM HEPES was slowly injected into the bottom of the tube using a syringe. Finally, a mixture of 30 μL of the cell-free synthesis reaction solution and 370 μL of 40% sucrose in 50 mM HEPES was slowly injected into the bottom of the tube using a syringe to form a density layer. Then, tubes were centrifuged using a TLS-S5 swinging-bucket rotor (Beckman Coulter, USA) at 40,000 rpm for 3 h at 4 °C in an Optima MAX-XP ultracentrifuge (Beckman Coulter, USA). After centrifugation, 100–200 μL of the liposome fraction was carefully collected. The outer solution of liposomes purified by DGU was exchanged by two rounds of centrifugation: 1000 μL of 50 mM HEPES was added, followed by centrifugation at 20,000 × g for 5 min, and the supernatant was discarded. The liposome pellet was then resuspended with 50 mM HEPES and stored at 4 °C. The average size of liposomes was measured using dynamic light scattering (DLS) using Litesizer DLS 700 (Anton Paar, Australia), measured in backscattering mode (175°). The amount of peptides with liposomes was quantified using the bicinchoninic acid (BCA) assay using the TaKaRa BCA Protein Assay Kit (TaKaRa, Japan). Liposome fractions purified by DGU in 50 mM HEPES were prepared to a final concentration of 2% SDS in a total volume of 100 μL, and mixed with 100 μL of BCA mix and reacted at 37 °C for 1 h in a 96-well microplate. After the reaction, absorbance at 562 nm was measured using a SpectraMax iD5 (Molecular Devices, USA). A cell-free synthesis reaction solution with liposomes and without SVG28-coding DNA was used as a blank, and its absorbance was subtracted to calculate the peptide concentration.

Characterization of the Folding of Cell-freely Synthesized SVG28

To confirm the liposomal membrane incorporation of cell-freely synthesized SVG28, 30 μL of the liposome solution purified from the cell-free synthesis reaction solution of SVG28 by twice centrifugation at 20,000 × g as described above, was incubated with 30 μL of 1 mg/mL proteinase K at 37 °C for 1 h. After the incubation, liposomes were purified by twice centrifugation at 20,000 × g as described above. Finally, the pellet was resuspended in 100 μL of 50 mM HEPES and then subjected to channel current measurements.

In the LUV leakage assay, the LUVs were prepared using the Bangham method. The calcein solution contains 20 mM HEPES, 100 mM potassium glutamate, and 80 mM calcein (pH 7.4). The 2.5× purification buffer contains 50 mM HEPES, 250 mM potassium glutamate, and 2 M sucrose (pH 7.4). Eight mg of DOPC, 4 mg of DOPC and DOPG, and 2 mg of cholesterol in chloroform were prepared in a microtube, and chloroform was evaporated under flowing nitrogen gas and evaporated overnight under vacuum. One mL of the calcein solution was added to the tube and sonicated for 90 min at 50 °C and 42 kHz. The resulting GUVs were resized to 100 nm using an extruder (Avanti Polar Lipids, USA). To remove external calcein, the LUV was purified using an illustra NAP-5 column (Cytiva, USA), and 3 μL of each fraction was mixed with 27 μL of purification buffer in a 384-well plate, and the fluorescence was measured using a SpectraMax iD5 (Molecular Devices, USA). Ten μL of cell-free synthesis reaction solution, including 3 nmol/μL DNA, was mixed with 22 μL of purification buffer and 3 μL of LUV solution. The entire 384-well plate was placed in an air incubator and incubated at 37 °C for 6 h. After 6 h, the fluorescence intensity was measured again.

Bilayer Lipid Membrane Preparation Using a Microfabricated Device

Microdevices for channel current measurement were fabricated by machining a poly­(methyl methacrylate) (PMMA) plate (Mitsubishi Rayon, Japan) using a computer-aided design and a three-axis milling machine (MM-100, Modia Systems, Japan), as shown in Figure S16A and B. Two wells with a diameter of 4.0 mm and a depth of 4.5 mm, along with a groove between the wells, were manufactured on the PMMA plate (Figure S16A and B). Each well had a through-hole at the bottom, with Ag/AgCl electrodes set into these holes (Figure S16A). A polymeric film made of 5 μm-thick parylene C (polychloro-p-xylylene, Parylene Japan, Japan) was patterned with a single pore (100 μm in diameter) using conventional photolithography and then sandwiched between PMMA sheets (Figure S16C). The PMMA sheets enclosing the parylene C film were inserted to separate the two chambers. Bilayer lipid membranes (BLMs) were prepared as follows using the droplet contact method in the microdevices. Two lipid monolayers containing the aqueous droplets came into contact in the small hole punched in the parylene film to form lipid bilayers (Figure S16D).

First, 2.5 μL of 10 mg/mL DOPC:Cholesterol = 2:1 (mol:mol) in n-decane was added to both chambers. Then 23 μL of 1 mg/mL liposome solution, 250 mM glucose, 25 mM HEPES, 1 M KCl, and 10 mM MOPS was added to the chamber on the voltage-applied side, and 23 μL of 1 M KCl and 10 mM MOPS was added to the chamber on the ground side. For the channel current measurement with calcium ions, positively charged lipids or negatively charged lipids, 23 μL of 15 mM calcium chloride (CaCl2, Nacalai Tesque, Japan), 1 mg/mL liposome solution, 250 mM glucose, 25 mM HEPES, 1 M KCl, and 10 mM MOPS was added to the chamber on the voltage applied side, and 2.5 μL of 10 mg/mL DOPC:DOTAP:Cholesterol = 3:1:2 (mol:mol:mol) in n-decane was used to prepare a planar bilayer, or 2.5 μL of 10 mg/mL DOPC:DOPG:Cholesterol = 1:1:1 (mol:mol:mol) in n-decane was used to prepare a planar bilayer, respectively. The channel current was monitored using a JET patch-clamp amplifier (Tecella, USA). Signals were detected with a 4 kHz low-pass filter at a sampling frequency of 20 kHz. A constant voltage of +100 mV was applied to the recording chamber while the chamber on the ground side was grounded. Channel current measurements were performed at room temperature. Data were analyzed using Clampfit 11.3 (Molecular Devices, USA), Excel (Microsoft, USA), and OriginPro 2022 (OriginLab Corporation, USA).

Current Signal Analysis

To quantify the peptide pore-forming ability, current signals were classified into stable pore-formation signals and unstable pore-formation signals. , The “step” signal of stable pore formation–representing stable single pore formation–is classified by a sharp increase in the signal in less than 10 ms, and the 95% confidence interval of the pore current value being less than 10 times the baseline value (Figure S4A). The “square-top” signal of stable pore formation–representing stable pore formation involving insertion and dissociation of monomers–is classified by a sharp increase in the signal in less than 10 ms with staircase-like current–time traces, and the 95% confidence interval of the pore current value being less than 10 times the baseline value for each step (Figure S4B). The “multi-level” signal of unstable pore formation–representing unstable pore formation involving lipid molecules between peptides–is classified by a sharp increase in the signal in less than 10 ms and the 95% confidence interval of the pore current value being more than 10 times the baseline value (Figure S4C). The “erratic” signal of unstable pore formation–representing the random disruption of membranes–is classified by a blunt increase in the signal in more than 10 ms (Figure S4D). Signals that were longer than 5 s and had a conductance of more than 0.2 nS were used in the analysis. Signal distributions were calculated using n = 20 data points for one measurement device. For the estimation of the pore diameter, the relationship between the conductance of natural β-barrel proteins and the reported pore diameters , was used. For the quantification of pore noise, the RMS noise was calculated by subtracting the unbiased variance of the baseline for approximately 1 s before the pore opening from the unbiased variance of the current value for approximately 1 s after the pore opening. The square root of the result was taken and normalized to the RMS when the pore conductance was at 1 nS.

Molecular Dynamics Simulations

The monomeric structure of SVG28 was modeled as reported in our previous study, , using homology modeling. Each monomer was aligned to form a 7-mer structure by aligning each monomer to the 7-mer 3csl 484–511 fragment predicted in AlphaFold2 , using UCSF Chimera software. Each 7-mer structure was embedded in 18PC:PG, 16PC:PG, or 14PC:PG membranes with 1 M KCl at 298 K using the CHARMM-GUI membrane builder. The systems were initially equilibrated using the protocol provided by CHARMM-GUI, followed by an additional 500 ps of equilibration, and subsequently simulated for 200 ns with the CHARMM36m force field. All simulations were executed with GROMACS-2021.2 or 2023.2. Membrane compression was calculated by subtracting the membrane thickness at a distance of 40–45 Å from the protein center from the membrane thickness at a distance of 25–40 Å from the protein center. Membrane thickness was calculated as the average of z-coordinates of phosphate atoms of the lipids, using the Tk Console of VMD version 1.9.

Single-Molecule Detection of L-PLL, poly­(dT)50, R7G

For the single-molecule detection of L-PLL or RRRRRRRG (R7G), 23 μL of solution containing 1 mg/mL resized liposomes using a 100 nm membrane, either 600 nM L-PLL or 500 nM R7G, 250 mM glucose, 25 mM HEPES, 1 M KCl, and 10 mM MOPS was added to the chamber on the cis side, respectively, and 23 μL of 1 M KCl and 10 mM MOPS was added to the ground side. For the single-molecule detection of poly­(dT)50, liposome solution resized using 100 nm membrane, 250 mM glucose, 25 mM HEPES, 1 M KCl, and 10 mM MOPS was added to the chamber on the cis side, and 23 μL of 500 nM poly­(dT)50, 1 M KCl, and 10 mM MOPS was added to the ground side. Constant voltages of +50, +100, and +150 mV were applied to the recording chamber. The blocking current signals were filtered for conductance greater than 0.4 nS, in accordance with our previous study. Noisy pores were intentionally included in the analysis to help account for any pore-derived noise from the SVG28 synthesized using CFPS.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsnano.5c15163.

  • DNA/amino acid sequences of peptides/proteins for CFPS; SDS-PAGE; positive control for channel current measurements; classification of channel current signals; lipid composition-dependent pore formation and conductance analysis; protease protection and calcein leakage assays; DLS measurements; additional data for single-molecule detection; microdevice configuration (PDF) (PDF)

R.K. conceived the original idea. M. S. conducted a LUV calcein leakage assay. S.F. performed all the other experiments and analysis. S.F. and R.K. wrote the entire manuscript.

JSPS KAKENHI Grant Numbers 21H05229, 25H00416, 24KJ1026, and JST-CREST JPMJCR21B2.

The authors declare no competing financial interest.

References

  1. Kasianowicz J. J., Brandin E., Branton D., Deamer D. W.. Characterization of Individual Polynucleotide Molecules Using a Membrane Channel. Proc. Natl. Acad. Sci. U.S.A. 1996;93(24):13770–13773. doi: 10.1073/pnas.93.24.13770. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Deamer D. W., Branton D.. Characterization of Nucleic Acids by Nanopore Analysis. Acc. Chem. Res. 2002;35(10):817–825. doi: 10.1021/ar000138m. [DOI] [PubMed] [Google Scholar]
  3. Derrington I. M., Butler T. Z., Collins M. D., Manrao E., Pavlenok M., Niederweis M., Gundlach J. H.. Nanopore DNA Sequencing with MspA. Proc. Natl. Acad. Sci. U.S.A. 2010;107(37):16060–16065. doi: 10.1073/pnas.1001831107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Jain M., Olsen H. E., Paten B., Akeson M.. The Oxford Nanopore MinION: Delivery of Nanopore Sequencing to the Genomics Community. Genome Biol. 2016;17(1):239. doi: 10.1186/s13059-016-1103-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brinkerhoff H., Kang A. S. W., Liu J., Aksimentiev A., Dekker C.. Multiple Rereads of Single Proteins at Single-Amino Acid Resolution Using Nanopores. Science. 2021;374(6574):1509–1513. doi: 10.1126/science.abl4381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Yan S., Zhang J., Wang Y., Guo W., Zhang S., Liu Y., Cao J., Wang Y., Wang L., Ma F., Zhang P., Chen H. Y., Huang S.. Single Molecule Ratcheting Motion of Peptides in a Mycobacterium Smegmatis Porin A (MspA) Nanopore. Nano Lett. 2021;21(15):6703–6710. doi: 10.1021/acs.nanolett.1c02371. [DOI] [PubMed] [Google Scholar]
  7. Chen Z., Wang Z., Xu Y., Zhang X., Tian B., Bai J.. Controlled Movement of SsDNA Conjugated Peptide through Mycobacterium Smegmatis Porin A (MspA) Nanopore by a Helicase Motor for Peptide Sequencing Application. Chem. Sci. 2021;12(47):15750–15756. doi: 10.1039/D1SC04342K. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Yu L., Kang X., Li F., Mehrafrooz B., Makhamreh A., Fallahi A., Foster J. C., Aksimentiev A., Chen M., Wanunu M.. Unidirectional Single-File Transport of Full-Length Proteins through a Nanopore. Nat. Biotechnol. 2023;41(8):1130–1139. doi: 10.1038/s41587-022-01598-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Sauciuc A., Morozzo della Rocca B., Tadema M. J., Chinappi M., Maglia G.. Translocation of Linearized Full-Length Proteins through an Engineered Nanopore under Opposing Electrophoretic Force. Nat. Biotechnol. 2024;42:1275. doi: 10.1038/s41587-023-01954-x. [DOI] [PubMed] [Google Scholar]
  10. Motone K., Kontogiorgos-Heintz D., Wee J., Kurihara K., Yang S., Roote G., Fox O. E., Fang Y., Queen M., Tolhurst M., Cardozo N., Jain M., Nivala J.. Multi-Pass, Single-Molecule Nanopore Reading of Long Protein Strands. Nature. 2024;633(8030):662–669. doi: 10.1038/s41586-024-07935-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Lu C., Bonini A., Viel J. H., Maglia G.. Toward Single-Molecule Protein Sequencing Using Nanopores. Nat. Biotechnol. 2025;43(3):312–322. doi: 10.1038/s41587-025-02587-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Vorobieva A. A., White P., Liang B., Horne J. E., Bera A. K., Chow C. M., Gerben S., Marx S., Kang A., Stiving A. Q., Harvey S. R., Marx D. C., Khan G. N., Fleming K. G., Wysocki V. H., Brockwell D. J., Tamm L. K., Radford S. E., Baker D.. De Novo Design of Transmembrane β Barrels. Science. 2021;371(6531):eabc8182. doi: 10.1126/science.abc8182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Berhanu S., Majumder S., Müntener T., Whitehouse J., Berner C., Bera A. K., Kang A., Liang B., Khan N., Sankaran B., Tamm L. K., Brockwell D. J., Hiller S., Radford S. E., Baker D., Vorobieva A. A.. Sculpting Conducting Nanopore Size and Shape through de Novo Protein Design. Science. 2024;385(6706):282–288. doi: 10.1126/science.adn3796. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Goverde C. A., Pacesa M., Goldbach N., Dornfeld L. J., Balbi P. E. M., Georgeon S., Rosset S., Kapoor S., Choudhury J., Dauparas J., Schellhaas C., Kozlov S., Baker D., Ovchinnikov S., Vecchio A. J., Correia B. E.. Computational Design of Soluble and Functional Membrane Protein Analogues. Nature. 2024;631(8020):449–458. doi: 10.1038/s41586-024-07601-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Xu C., Lu P., Gamal El-Din T. M., Pei X. Y., Johnson M. C., Uyeda A., Bick M. J., Xu Q., Jiang D., Bai H., Reggiano G., Hsia Y., Brunette T. J., Dou J., Ma D., Lynch E. M., Boyken S. E., Huang P. S., Stewart L., DiMaio F., Kollman J. M., Luisi B. F., Matsuura T., Catterall W. A., Baker D.. Computational Design of Transmembrane Pores. Nature. 2020;585(7823):129–134. doi: 10.1038/s41586-020-2646-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Scott A. J., Niitsu A., Kratochvil H. T., Lang E. J. M., Sengel J. T., Dawson W. M., Mahendran K. R., Mravic M., Thomson A. R., Brady R. L., Liu L., Mulholland A. J., Bayley H., DeGrado W. F., Wallace M. I., Woolfson D. N.. Constructing Ion Channels from Water-Soluble α-Helical Barrels. Nat. Chem. 2021;13(7):643–650. doi: 10.1038/s41557-021-00688-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Niitsu A., Thomson A. R., Scott A. J., Sengel J. T., Jung J., Mahendran K. R., Sodeoka M., Bayley H., Sugita Y., Woolfson D. N., Wallace M. I.. Rational Design Principles for DE Novo α-Helical Peptide Barrels with Dynamic Conductive Channels. J. Am. Chem. Soc. 2025;147(14):11741–11753. doi: 10.1021/jacs.4c13933. [DOI] [PubMed] [Google Scholar]
  18. Shimizu K., Mijiddorj B., Usami M., Mizoguchi I., Yoshida S., Akayama S., Hamada Y., Ohyama A., Usui K., Kawamura I., Kawano R.. De Novo Design of a Nanopore for Single-Molecule Detection That Incorporates a β-Hairpin Peptide. Nat. Nanotechnol. 2022;17(1):67–75. doi: 10.1038/s41565-021-01008-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Fujita S., Kawamura I., Kawano R.. Cell-Free Expression of De Novo Designed Peptides That Form β-Barrel Nanopores. ACS Nano. 2023;17(4):3358–3367. doi: 10.1021/acsnano.2c07970. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Sohma Y., Sasaki M., Hayashi Y., Kimura T., Kiso Y.. Novel and Efficient Synthesis of Difficult Sequence-Containing Peptides through O–N Intramolecular Acyl Migration Reaction of O-Acyl Isopeptides. Chem. Commun. (Camb.) 2004;4(1):124–125. doi: 10.1039/B312129A. [DOI] [PubMed] [Google Scholar]
  21. Shimizu Y., Inoue A., Tomari Y., Suzuki T., Yokogawa T., Nishikawa K., Ueda T.. Cell-Free Translation Reconstituted with Purified Components. Nat. Biotechnol. 2001;19(8):751–755. doi: 10.1038/90802. [DOI] [PubMed] [Google Scholar]
  22. Chou K. C.. Prediction of Beta-Turns. J. Pept. Res. 1997;49(2):120–144. doi: 10.1111/j.1399-3011.1997.tb00608.x. [DOI] [PubMed] [Google Scholar]
  23. Kato A., Maki K., Ebina T., Kuwajima K., Soda K., Kuroda Y.. Mutational Analysis of Protein Solubility Enhancement Using Short Peptide Tags. Biopolymers. 2007;85(1):12–18. doi: 10.1002/bip.20596. [DOI] [PubMed] [Google Scholar]
  24. Steinkühler J., Peruzzi J. A., Krüger A., Villaseñor C. G., Jacobs M. L., Jewett M. C., Kamat N. P.. Improving Cell-Free Expression of Model Membrane Proteins by Tuning Ribosome Cotranslational Membrane Association and Nascent Chain Aggregation. ACS Synth. Biol. 2024;13(1):129–140. doi: 10.1021/acssynbio.3c00357. [DOI] [PubMed] [Google Scholar]
  25. Broersen K., Jonckheere W., Rozenski J., Vandersteen A., Pauwels K., Pastore A., Rousseau F., Schymkowitz J.. A Standardized and Biocompatible Preparation of Aggregate-Free Amyloid Beta Peptide for Biophysical and Biological Studies of Alzheimers Disease. Protein Eng. Des. Sel. 2011;24(9):743–750. doi: 10.1093/protein/gzr020. [DOI] [PubMed] [Google Scholar]
  26. Kalmbach R., Chizhov I., Schumacher M. C., Friedrich T., Bamberg E., Engelhard M.. Functional Cell-Free Synthesis of a Seven Helix Membrane Protein: In Situ Insertion of Bacteriorhodopsin into Liposomes. J. Mol. Biol. 2007;371(3):639–648. doi: 10.1016/j.jmb.2007.05.087. [DOI] [PubMed] [Google Scholar]
  27. Nomura S. I. M., Kondoh S., Asayama W., Asada A., Nishikawa S., Akiyoshi K.. Direct Preparation of Giant Proteo-Liposomes by in Vitro Membrane Protein Synthesis. J. Biotechnol. 2008;133(2):190–195. doi: 10.1016/j.jbiotec.2007.08.023. [DOI] [PubMed] [Google Scholar]
  28. Moritani Y., Nomura S. I. M., Morita I., Akiyoshi K.. Direct Integration of Cell-Free-Synthesized Connexin-43 into Liposomes and Hemichannel Formation. FEBS J. 2010;277(16):3343–3352. doi: 10.1111/j.1742-4658.2010.07736.x. [DOI] [PubMed] [Google Scholar]
  29. Lyukmanova E. N., Shenkarev Z. O., Khabibullina N. F., Kopeina G. S., Shulepko M. A., Paramonov A. S., Mineev K. S., Tikhonov R. V., Shingarova L. N., Petrovskaya L. E., Dolgikh D. A., Arseniev A. S., Kirpichnikov M. P.. Lipid-Protein Nanodiscs for Cell-Free Production of Integral Membrane Proteins in a Soluble and Folded State: Comparison with Detergent Micelles, Bicelles and Liposomes. Biochim. Biophys. Acta Biomembr. 2012;1818(3):349–358. doi: 10.1016/j.bbamem.2011.10.020. [DOI] [PubMed] [Google Scholar]
  30. Shenkarev Z. O., Lyukmanova E. N., Butenko I. O., Petrovskaya L. E., Paramonov A. S., Shulepko M. A., Nekrasova O. V., Kirpichnikov M. P., Arseniev A. S.. Lipid-Protein Nanodiscs Promote in Vitro Folding of Transmembrane Domains of Multi-Helical and Multimeric Membrane Proteins. Biochim. Biophys. Acta Gen. Subj. 2013;1828(2):776–784. doi: 10.1016/j.bbamem.2012.11.005. [DOI] [PubMed] [Google Scholar]
  31. Gessesse B., Nagaike T., Nagata K., Shimizu Y., Ueda T. G.-.. Protein Coupled Receptor Protein Synthesis on a Lipid Bilayer Using a Reconstituted Cell-Free Protein Synthesis System. Life (Basel) 2018;8(4):54. doi: 10.3390/life8040054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Price C. E., Kocer A., Kol S., van der Berg J. P., Driessen A. J. M.. In Vitro Synthesis and Oligomerization of the Mechanosensitive Channel of Large Conductance, MscL, into a Functional Ion Channel. FEBS Lett. 2011;585(1):249–254. doi: 10.1016/j.febslet.2010.11.057. [DOI] [PubMed] [Google Scholar]
  33. Ando M., Akiyama M., Okuno D., Hirano M., Ide T., Sawada S., Sasaki Y., Akiyoshi K.. Liposome Chaperon in Cell-Free Membrane Protein Synthesis: One-Step Preparation of KcsA-Integrated Liposomes and Electrophysiological Analysis by the Planar Bilayer Method. Biomater. Sci. 2016;4(2):258–264. doi: 10.1039/C5BM00285K. [DOI] [PubMed] [Google Scholar]
  34. Kamiya K.. Formation and Function of OmpG or OmpA-Incorporated Liposomes Using an in Vitro Translation System. Sci. Rep. 2022;12(1):2376. doi: 10.1038/s41598-022-06314-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nishiguchi R., Tanaka T., Hayashida J., Nakagita T., Zhou W., Takeda H.. Evaluation of Cell-Free Synthesized Human Channel Proteins for in Vitro Channel Research. Membranes (Basel) 2023;13(1):48. doi: 10.3390/membranes13010048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Meyer C., Arizzi A., Henson T., Aviran S., Longo M. L., Wang A., Tan C.. Designer Artificial Environments for Membrane Protein Synthesis. Nat. Commun. 2025;16(1):4363. doi: 10.1038/s41467-025-59471-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Peruzzi J. A., Steinkühler J., Vu T. Q., Gunnels T. F., Hu V. T., Lu P., Baker D., Kamat N. P.. Hydrophobic Mismatch Drives Self-Organization of Designer Proteins into Synthetic Membranes. Nat. Commun. 2024;15(1):3162. doi: 10.1038/s41467-024-47163-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Okamura K., Matsushita S., Kato Y., Watanabe H., Matsui A., Oka T., Matsuura T.. In Vitro Synthesis of the Human Calcium Transporter Letm1 within Cell-Sized Liposomes and Investigation of Its Lipid Dependency. J. Biosci. Bioeng. 2019;127(5):544–548. doi: 10.1016/j.jbiosc.2018.11.003. [DOI] [PubMed] [Google Scholar]
  39. Harris N. J., Pellowe G. A., Booth P. J.. Cell-Free Expression Tools to Study Co-Translational Folding of Alpha Helical Membrane Transporters. Sci. Rep. 2020;10(1):9125. doi: 10.1038/s41598-020-66097-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Blackholly L. R., Harris N. J., Findlay H. E., Booth P. J.. Cell-Free Expression to Probe Co-Translational Insertion of an Alpha Helical Membrane Protein. Front. Mol. Biosci. 2022;9:795212. doi: 10.3389/fmolb.2022.795212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sekiya Y., Sakashita S., Shimizu K., Usui K., Kawano R.. Channel Current Analysis Estimates the Pore-Formation and the Penetration of Transmembrane Peptides. Analyst. 2018;143(15):3540–3543. doi: 10.1039/C8AN00243F. [DOI] [PubMed] [Google Scholar]
  42. Webb R., East J., Sharma R., Lee A.. Hydrophobic Mismatch and the Incorporation of Peptides into Lipid Bilayers: A Possible Mechanism for Retention in the Golgi. Biochemistry. 1998;37(2):673–679. doi: 10.1021/bi972441+. [DOI] [PubMed] [Google Scholar]
  43. Andersen O. S., Koeppe R. E.. 2nd. Bilayer Thickness and Membrane Protein Function: An Energetic Perspective. Annu. Rev. Biophys. Biomol. Struct. 2007;36(1):107–130. doi: 10.1146/annurev.biophys.36.040306.132643. [DOI] [PubMed] [Google Scholar]
  44. Killian J. A.. Synthetic Peptides as Models for Intrinsic Membrane Proteins. FEBS Lett. 2003;555(1):134–138. doi: 10.1016/S0014-5793(03)01154-2. [DOI] [PubMed] [Google Scholar]
  45. O’Keeffe A. H., East J. M., Lee A. G.. Selectivity in Lipid Binding to the Bacterial Outer Membrane Protein OmpF. Biophys. J. 2000;79(4):2066–2074. doi: 10.1016/S0006-3495(00)76454-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Morton R. E., Evans T. A.. Modification of the Bicinchoninic Acid Protein Assay to Eliminate Lipid Interference in Determining Lipoprotein Protein Content. Anal. Biochem. 1992;204(2):332–334. doi: 10.1016/0003-2697(92)90248-6. [DOI] [PubMed] [Google Scholar]
  47. Fukunaga K., Teramoto T., Nakashima M., Ohtani T., Katsuki R., Matsuura T., Yokobayashi Y., Kakuta Y.. Structural Insights into Lab-Coevolved RNA-RBP Pairs and Applications of Synthetic Riboswitches in Cell-Free System. Nucleic Acids Res. 2025;53(6):gkaf212. doi: 10.1093/nar/gkaf212. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Xia Y., Sun J., Liang D.. Aggregation, Fusion, and Leakage of Liposomes Induced by Peptides. Langmuir. 2014;30(25):7334–7342. doi: 10.1021/la501618f. [DOI] [PubMed] [Google Scholar]
  49. Micsonai A., Wien F., Kernya L., Lee Y.-H., Goto Y., Réfrégiers M., Kardos J.. Accurate Secondary Structure Prediction and Fold Recognition for Circular Dichroism Spectroscopy. Proc. Natl. Acad. Sci. U. S. A. 2015;112(24):E3095–E3103. doi: 10.1073/pnas.1500851112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Abedin F., Kandel N., Tatulian S. A.. Effects of Aβ-Derived Peptide Fragments on Fibrillogenesis of Aβ. Sci. Rep. 2021;11(1):19262. doi: 10.1038/s41598-021-98644-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Malinin V. S., Frederik P., Lentz B. R.. Osmotic and Curvature Stress Affect PEG-Induced Fusion of Lipid Vesicles but Not Mixing of Their Lipids. Biophys. J. 2002;82(4):2090–2100. doi: 10.1016/S0006-3495(02)75556-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Lee J. Y., Schick M.. Calculation of Free Energy Barriers to the Fusion of Small Vesicles. Biophys. J. 2008;94(5):1699–1706. doi: 10.1529/biophysj.107.119511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Anzai K., Masumi M., Kawasaki K., Kirino Y.. Frequent Fusion of Liposomes to a Positively Charged Planar Bilayer without Calcium Ions. J. Biochem. 1993;114(4):487–491. doi: 10.1093/oxfordjournals.jbchem.a124204. [DOI] [PubMed] [Google Scholar]
  54. Akabas M. H., Cohen F. S., Finkelstein A.. Separation of the Osmotically Driven Fusion Event from Vesicle-Planar Membrane Attachment in a Model System for Exocytosis. J. Cell Biol. 1984;98(3):1063–1071. doi: 10.1083/jcb.98.3.1063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wessman P., Morin M., Reijmar K., Edwards K.. Effect of α-Helical Peptides on Liposome Structure: A Comparative Study of Melittin and Alamethicin. J. Colloid Interface Sci. 2010;346(1):127–135. doi: 10.1016/j.jcis.2010.02.032. [DOI] [PubMed] [Google Scholar]
  56. Levadny V., Tsuboi T.-A., Belaya M., Yamazaki M.. Rate Constant of Tension-Induced Pore Formation in Lipid Membranes. Langmuir. 2013;29(12):3848–3852. doi: 10.1021/la304662p. [DOI] [PubMed] [Google Scholar]
  57. Rojko N., Anderluh G.. How Lipid Membranes Affect Pore Forming Toxin Activity. Acc. Chem. Res. 2015;48(12):3073–3079. doi: 10.1021/acs.accounts.5b00403. [DOI] [PubMed] [Google Scholar]
  58. 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. 2017;8(1):935. doi: 10.1038/s41467-017-01006-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Nguyen N. T. H., Fujita S., Takeuchi N., Kawano R., Inoue T.. Self-Assembling Synthetic Peptides Remodel Mitochondrial Properties via Nanopore Formation. bioRxiv. 2025:08.26.672490. doi: 10.1101/2025.08.26.672490. [DOI] [Google Scholar]
  60. Shahidullah K., London E.. Effect of Lipid Composition on the Topography of Membrane-Associated Hydrophobic Helices: Stabilization of Transmembrane Topography by Anionic Lipids. J. Mol. Biol. 2008;379(4):704–718. doi: 10.1016/j.jmb.2008.04.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Pandi A., Adam D., Zare A., Trinh V. T., Schaefer S. L., Burt M., Klabunde B., Bobkova E., Kushwaha M., Foroughijabbari Y., Braun P., Spahn C., Preußer C., Pogge von Strandmann E., Bode H. B., von Buttlar H., Bertrams W., Jung A. L., Abendroth F., Schmeck B., Hummer G., Vázquez O., Erb T. J.. Cell-Free Biosynthesis Combined with Deep Learning Accelerates de Novo-Development of Antimicrobial Peptides. Nat. Commun. 2023;14(1):7197. doi: 10.1038/s41467-023-42434-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Waterhouse A., Bertoni M., Bienert S., Studer G., Tauriello G., Gumienny R., Heer F. T., De Beer T. A. P., Rempfer C., Bordoli L., Lepore R., Schwede T.. SWISS-MODEL: Homology Modelling of Protein Structures and Complexes. Nucleic Acids Res. 2018;46(W1):W296–W303. doi: 10.1093/nar/gky427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Jumper J., Evans R., Pritzel A., Green T., Figurnov M., Ronneberger O., Tunyasuvunakool K., Bates R., Žídek A., Potapenko A., Bridgland A., Meyer C., Kohl S. A. A., Ballard A. J., Cowie A., Romera-Paredes B., Nikolov S., Jain R., Adler J., Back T., Petersen S., Reiman D., Clancy E., Zielinski M., Steinegger M., Pacholska M., Berghammer T., Bodenstein S., Silver D., Vinyals O., Senior A. W., Kavukcuoglu K., Kohli P., Hassabis D.. Highly Accurate Protein Structure Prediction with AlphaFold. Nature. 2021;596(7873):583–589. doi: 10.1038/s41586-021-03819-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Mirdita M., Schütze K., Moriwaki Y., Heo L., Ovchinnikov S., Steinegger M.. ColabFold: Making Protein Folding Accessible to All. Nat. Methods. 2022;19(6):679–682. doi: 10.1038/s41592-022-01488-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Pettersen E. F., Goddard T. D., Huang C. C., Couch G. S., Greenblatt D. M., Meng E. C., Ferrin T. E.. UCSF Chimera - A Visualization System for Exploratory Research and Analysis. J. Comput. Chem. 2004;25(13):1605–1612. doi: 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]
  66. Wu E. L., Cheng X., Jo S., Rui H., Song K. C., Dávila-Contreras E. M., Qi Y., Lee J., Monje-Galvan V., Venable R. M., Klauda J. B., Im W.. CHARMM-GUI Membrane Builder toward Realistic Biological Membrane Simulations. J. Comput. Chem. 2014;35(27):1997–2004. doi: 10.1002/jcc.23702. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Huang J., Rauscher S., Nawrocki G., Ran T., Feig M., De Groot B. L., Grubmüller H., MacKerell A. D.. CHARMM36m: An Improved Force Field for Folded and Intrinsically Disordered Proteins. Nat. Methods. 2017;14(1):71–73. doi: 10.1038/nmeth.4067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Van Der Spoel D., Lindahl E., Hess B., Groenhof G., Mark A. E., Berendsen H. J. C.. GROMACS: Fast, Flexible, and Free. J. Comput. Chem. 2005;26(16):1701–1718. doi: 10.1002/jcc.20291. [DOI] [PubMed] [Google Scholar]
  69. Humphrey W., Dalke A., Schulten K.. VMD: Visual Molecular Dynamics. J. Mol. Graph. 1996;14(1):33–38. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials


Articles from ACS Nano are provided here courtesy of American Chemical Society

RESOURCES