Synthetically Diversified Protein Nanopores: Resolving Click Reaction Mechanisms

Nanopores are emerging as a powerful tool for the investigation of nanoscale processes at the single-molecule level. Here, we demonstrate the methionine-selective synthetic diversification of α-hemolysin (α-HL) protein nanopores and their exploitation as a platform for investigating reaction mechanisms. A wide range of functionalities, including azides, alkynes, nucleotides, and single-stranded DNA, were incorporated into individual pores in a divergent fashion. The ion currents flowing through the modified pores were used to observe the trajectory of a range of azide-alkyne click reactions and revealed several short-lived intermediates in Cu(I)-catalyzed azide-alkyne [3 + 2] cycloadditions (CuAAC) at the single-molecule level. Analysis of ion-current fluctuations enabled the populations of species involved in rapidly exchanging equilibria to be determined, facilitating the resolution of several transient intermediates in the CuAAC reaction mechanism. The versatile pore-modification chemistry offers a useful approach for enabling future physical organic investigations of reaction mechanisms at the single-molecule level.

[3][4] Although less physically robust than solid-state nanopores, 5,6 biological nanopores offer the advantage of having atomically precise structures, thereby providing superior precision and reproducibility in their applications. 2,3,7[26] Compared to solid-state nanopores, 6,[27][28][29][30] the possibilities for tuning the properties of biological protein pores via chemical modification are more limited.For example, the dimensions, charge and chemical reactivity of the widely used heptameric α-hemolysin nanopore (α-HL) can be modified by altering the constituent monomer units through site-directed mutagenesis. 9,20,21,31,32A single modification, typically a cysteine, is then introduced through the stochastic self-assembly of one engineered protein monomer with six wild-type monomers.More recently, non-natural amino acids have been introduced into -HL nanopores through native chemical ligation of synthetic peptide fragments. 18,19The introduction of reactive handles through either method has enabled further site-specific decoration of nanopores, and even permitted the observation of reaction intermediates in chemistries popular for bioorthogonal functionalization. 18,20ere we have exploited the specific reactivity and low abundance of methionine residues to synthetically diversify individual membrane-embedded wild-type α-HL nanopores with a range of synthetic functional groups (Figure 1).In situ functionalization of the nanopore with alkyne and azide linkers (Figure 2) enabled the further diversification of the pores via bioorthogonal click reactions (Figure 3).Finally, the utility of this pore-modification approach to enable physical organic mechanistic investigations at the single-molecule level was demonstrated by application to CuAAC reactions (Figure 4).The high spatio-temporal resolution of the method enabled the assignment of several transient and dynamically exchanging reaction intermediates along the pathways of CuAAC reactions (Figures 5-8).

In Situ Chemical Functionalization of Methionine Residues
The α-hemolysin (α-HL) protein nanopore, which self-assembles spontaneously into heptameric membrane-spanning channels, is one of the most widely exploited biological nanopores (Figure 1A). 33The popularity of this nanopore is likely due to its robustness and well-established structure.][14][15]22,23,34 Furthermore, cysteine mutants may sometimes present purification or misfolding issues arising from disulfide formation.Recently, fully assembled wild-type α-HL nanopores have been modified in situ through direct chemical functionalization of lysine residues. 34While this approach provided rapid access to nanopores modified within the channel, such residues are both abundant within protein structures and often solvent-exposed, and thus, the modification of other lysine residues could not be fully controlled.Therefore, we sought an alternative, more site-selective method for synthetic modification.
Methionine is one of the least abundant of the 20 canonical amino acids, while its hydrophobic nature means that it is rarely exposed on the surface of proteins. 35Moreover, only seven methionine residues are exposed within the transmembrane channel of -HL (one on each subunit, yellow, Figures 1A and S1).We therefore reasoned that the recently developed redoxactivated chemical tagging (ReACT) protocol 35 might be utilized for the chemoselective functionalization of -HL (Figure 1B).
We prepared oxaziridine reagents 1-4 (Figure 1B) based on the published procedures (see Supporting Information). 35Each of these oxaziridine reagents were then introduced to either side of a single membrane-embedded α-HL nanopore (Figures S2-S4) under an applied voltage of +100 mV (Figure 2A).In the presence of 1.7 mM of ethyl derivative 1, six stepwise decreases in the ion current flowing through the pore were observed over a few minutes (Figure 2B and S8).
These steps were attributed to steric blockage of ion current arising from the successive modification of six out of the seven methionine residues within the -HL channel (steps in Figures 2B, S5, and S6). 36Similarly, treatment of a single nanopore with the propargyl reagent 2 (Figures 1B, 2C and S9-S11) resulted in complete functionalization at all seven methionine residues.The ability to monitor the modification state of the nanopore in real-time enabled the reaction to be quenched after the incorporation of only a single linker molecule by the addition of 20 eq. of methionine (Figures 2D, S8, S12 and S13).Nanopores were also easily modified using azide reagent 3 and butynyl reagent 4 (Figures 1B, 2E and S14-S17).Modifications were found to occur irrespective of whether the oxaziridine reagent was added to the cis or the trans side of the nanopore (Figure S13).Thus, we demonstrated the modification of individual nanopores in a site-selective fashion with chemical handles that would enable further synthetic diversification.Modification with 2 (1.7 mM) resulted in full functionalization of all seven methionine residues.(D) Introduction of a single alkyne modification through reaction with 2 (0.17 mM) and quenching by addition of methionine (3.4 mM, 20 equiv.).(E) Introduction of a single azide modification with 3 (0.17 mM) and methionine quenching.All traces were recorded at 298 K, +100 mV in 1 M KCl, 30 mM potassium phosphate buffer (pH 8.0).Small current drops following quenching were due to temporary localized dilution of the electrolyte solution.Free pore currents are indicated in orange.See the Supporting Information for complete experimental traces.

Click Diversification of Nanopores
Having established the ability to control the modification of the methionine residues within the nanopore, we next sought to explore divergent elaboration of the installed linker functionalities.
[39] Strain-promoted azide-alkyne cycloaddition (SPAAC) was performed on individual nanopores modified with a single azide linker, as indicated by the irreversible drop in the ion current upon addition of cyclooctyne 5 (Figure 3A, top and S18-S20).A larger, but intrinsically more reactive dibenzocyclooctyne was found not to react within the nanopore on a reasonable timescale. 20Similarly, ion current traces also indicated that in the presence of CuSO4 and sodium ascorbate, a Cu(I)-catalyzed azide-alkyne [3+2] cycloaddition (CuAAC) was successful with alkyne 6 (Figure 3A, bottom and S21-S23).
Encouraged by these results, we moved on to examine CuAAC reactions within nanopores modified with an alkyne linker.Alkyne-bearing nanopores were allowed to react with a range of water-soluble azides (Figure 3B and S24-S40) in the presence of CuSO4 and sodium ascorbate.
Gratifyingly, facile CuAAC reactions were observed between the alkyne-bearing nanopore and azides bearing either cationic (7) or anionic (8) functional groups, as indicated by the ion current changes.1][42] The nanopore could even be functionalized with a d(T)10 oligonucleotide carrying an azide at the 3ʹ-terminus (10; see SI section 3.2 for detailed structure), giving rise to a hybrid protein-DNA conjugate.
While the irreversible step-changes in the ion current (and I/V traces, see Supporting Information) are a clear indicator of synthetic diversification of the nanopore, our attention was also drawn to the finer details of the ion current traces.For example, in the CuAAC reactions, reversible blockages were observed that might be attributed to Cu(I)-alkyne complexation (yellow, Figure 3B).In addition, such states were always seen to precede the subsequent irreversible formation of the triazole product via a second, short-lived intermediate (yellow to green to red, Figure 3B).These observations provided a tantalizing indication that our modified nanopores might provide the opportunity for gaining mechanistic insights into CuAAC reactions at the single-molecule level.

Single-molecule Investigation of the CuAAC Reaction using Alkyne-modified Nanopores
4][45][46][47][48][49][50][51][52] It is generally accepted that a reversible reaction between a terminal alkyne (12, Figure 4A) and a copper(I) salt leads to formation of a Cu-alkyne -complex (12′, blue), which is subsequently converted into a copper acetylide (13, yellow).From here, the reaction has been proposed to proceed through a mono-Cu pathway (left-hand cycle) by reaction with azide (14) to form a copper triazolide (15, green)   before liberating the product triazole (16, red).Alternatively, reaction via a two-copper mechanism has also been proposed (right-hand cycle), in which the terminal mono-copper species (13, yellow left) binds a second copper ion to form a productive bis-Cu acetylide complex (13′, yellow right).For instance, an elegant recent study isolated terminal mono-and bis-Cu complexes and putative intermediates that were stabilized by N-heterocyclic carbene (NHC) ligands, and suggested that both mono-and bis-Cu catalytic cycles were in operation, with the latter being kinetically favored. 43While previous studies have supported a bis-Cu mechanism, such mechanistic investigations have usually required the use of ligands to stabilize key intermediates, and therefore may not be representative of typical CuAAC synthetic protocols.Thus, the ability to monitor such reactions on the single-molecule level provides the opportunity to resolve short-lived intermediates that would be otherwise difficult to detect via ensemble analyses.Indeed, Bayley and co-workers have previously observed CuAAC reactions on the singlemolecule level using semi-synthetic nanopores.These nanopores were assembled from semisynthetic protein monomers, which were themselves assembled from an alkyne-functionalized peptide and two flanking recombinant polypeptides using native chemical ligation. 18The authors found that the addition of Cu(I) caused no significant change in the ion current, and thus proposed that if a copper acetylide ( 13) is formed, then it was too short-lived to be resolved in that system (<ms).They also observed one intermediate in the CuAAC reaction with a relatively long lifetime (~4.5 s), which was formed irreversibly, and tentatively assigned as a copper triazolide (15 in Figure 4A).In contrast to this previous study, the methionine-modified -HL nanopores developed in the present work yielded data with a higher signal-to-noise ratio, which enabled key transient intermediates in the CuAAC reaction to be resolved (vide infra).In a nanopore modified with a terminal alkyne, introduction of Cu(I) caused increased noise (black) due to short-lived events which are not fully resolved, as well as longer-duration, well-resolved stepwise blockages (yellow).The frequency and duration of the resolved events was dependent on the concentration of Cu(I).(C) A nanopore modified with an internal alkyne also displayed increased noise in the presence of Cu(I), but the longer-duration (yellow) events were absent.All traces were recorded at 298 K, +100 mV in 1 M KCl, 30 mM potassium phosphate buffer (pH 8.0).Data were filtered using a 2 kHz hardware Bezel filter, no additional software filter was applied.
Initially, we sought to examine the reversible blockages observed in alkyne-modified nanopores (Figure 3B, yellow) by varying the concentration of Cu(I) (Figure 5 and S41-S44).
Due to the ion current being dominated by the background of 1 M KCl electrolyte solution in which the experiments were performed, the addition of up to 10 mM of Cu(I) had a minimal effect on the magnitude of the recorded ion currents.However, increasing the concentration of Cu(I) in the presence of an alkyne-functionalized nanopore led to an increase in noise, while the frequency and duration of the reversible, stepped current blockages previously identified in Figure 3B also increased (yellow, Figure 5B).The copper-dependent increase in noise (shallow events with <ms durations) was not observed in wild-type -HL nanopores (Figure 5B cf.5A).
In contrast, a nanopore functionalized with the methyl-terminated internal alkyne 4 also displayed increased noise in the presence of Cu(I), but none of the longer-duration events (no yellow state in Figure 5C and S45).Since this internal alkyne is unable to form a terminal -type copper acetylide (resembling 13), then the noise observed in the traces shown in Figure 5B can be attributed to the reversible formation of the cationic copper -complex of both the terminal and internal alkynes (12′ and 11′ in Figures 4A and 4B, respectively).
Having assigned the blue levels in Figure 5B as incorporating 12 and its copper -complex 12′, we then turned to assigning the reversible current blockages (yellow in Figure 5B).The increase in the population of these stepped current blockages on increasing the Cu(I) concentration is also consistent with the involvement of copper.The increased duration of these events with increasing Cu(I) concentration means that such events cannot be attributed to the simple bimolecular dissociation of 13′ → 13 + Cu + , since the lifetime of the bis-Cu complex 13′ should be independent of the Cu(I) concentration.Instead, such a copper-dependent increase in event duration would be consistent with the noisy yellow events being attributed to both alkyne species in the 13 ⇌ 13′ equilibrium, since increasing the copper concentration shifts this equilibrium to the right and therefore slow down the reverse reaction 13 → 12′ (yellow to blue in Figure 5B).
We reasoned that a more detailed analysis might enable resolution of the two species contributing to each of the noisy 12 ⇌ 12′ (blue) and 13 ⇌ 13′ (yellow) states.Pleasingly, allpoints histogram analysis (see SI section 2.8.1) of the complete ion current traces exemplified in Figure 5 revealed that both the yellow and blue states each consisted of two Gaussian distributions of sub-states (Figure 6, S52 and Tables S2-S5).In addition, the proportions of both sub-populations were found to shift as the concentration of Cu(I) was varied.At [Cu(I)] = 6.6 mM, the blue sub-populations assigned to 12 ⇌ 12′ had equal integrals, indicating that the corresponding equilibrium constant K1 ≈ 150 M −1 at +100 mV (Figure 6C).Similarly, at [Cu(I)] = 3.3 mM, the yellow sub-populations assigned to 13 ⇌ 13′ had equal integrals, indicating that the associated equilibrium constant K3 ≈ 300 M −1 at +100 mV.Indeed, complete modelling of these  S6 and Figure S56).Data obtained at +100 mV in 1 M KCl, 30 mM potassium phosphate buffer (pH 8.0).graphical fitting, and K1 = 168 M −1 , K2 = 0.12, K3 = 292 M −1 by iterative fitting approaches (section 2.8 SI).Moreover, it was also possible to construct the associated speciation plots revealing the expected population of states as the concentration as Cu(I) was varied (Figures 6E   and S56).The difference between the copper -complexation equilibrium constants K1 and K3 is consistent with the increased electron density of the -system in the Cu-terminal alkyne 13 compared to the H-terminal alkyne 12, and with the increased covalency of the bis-copper complex 13′ compared to the H-terminal alkyne complex 12′ (Figure S56).Furthermore, our observation of copper -complexation in 12 ⇌ 12′ and 13 ⇌ 13′ being rapid (and hence noisy) compared to the kinetically well-resolved rearrangement in 12′ ⇌ 13 is consistent with previous mechanistic investigations. 43,46,474][55][56][57][58] Indeed, the populations presented in Figure 6 were determined at a potential of +100 mV, and therefore likely to differ from the non-field perturbed conditions under which chemical reactions are normally performed.Thus, we next set out to examine how lowering the applied voltage would influence the population of states.Lowering the potential lowers the ion current flowing through the pore and therefore the signal-to-noise ratio, making analysis of the type presented in Figure 6 more challenging.Thus, we instead performed a simpler analysis of the voltage-dependency of exchange between the well-resolved 12 ⇌ 12′ (blue) and 13 ⇌ 13′ (yellow) states at a constant concentration of Cu(I) (Figure 7 and S46-S50).The observed voltage-dependencies of the relative populations (Kobs) and rates of exchange between the blue and yellow states (1/on and 1/off) were found to be consistent with the respective assignments to the 12 ⇌ 12′ and 13 ⇌ 13′ equilibria.The application of a positive voltage shifts both the 12 ⇌ 12′ and 13 ⇌ 13′ equilibria to left due to destabilization of the cationic copper -complexes.Thus, increasing the voltage decreases proportion of time spent in the 12′ state, thereby decreasing the rate of the forward reaction of 12′ → 13 (1/on, blue to yellow in Figure 7).Conversely, the increased proportion of time spent in the 13 state at higher voltages increases the rate of the backwards reaction of 13 → 12′ (1/off, yellow to blue in Figure 7).As a result of these changes in the rates of the forward and reverse reactions, there is a corresponding decrease in the apparent equilibrium ratio Kobs between the 12 ⇌ 12′ (blue) and 13 ⇌ 13′ states (yellow) as the voltage is increased.For instance, the preference for the 13 ⇌ 13′ level is decreased by ~three orders of magnitude at +100 mV compared to non-field perturbed conditions (V = 0, Figure 7C).It follows that the population of the cationic 13' sub-population should also be favored compared to conditions in which a (destabilizing) positive potential is applied.Indeed, the application of negative potentials to an alkyne-modified pore in the presence of Cu(I) gave rise to long-lived blockage events with durations of several seconds (Figure S51).It should be noted that the rates and equilibrium constants under these conditions are not directly comparable to the experiments presented in Figure 7 (due to the requirement to add reagents to the opposite side of the nanopore under negative potentials).Nonetheless, only a single, low-noise state was observed at −60 mV, which was consistent with electrostatic trapping of 13′ (Figure S51C).The decrease in noise on increasing the magnitude of the applied negative potential was also commensurate with minimal fluctuation between the stabilized cationic copper -complex 13′ and copper acetylide 13.
Having fully assigned the identity of the reversible intermediates in the sequential equilibria 12 ⇌ 12′ ⇌ 13 ⇌ 13′, our attention turned to assigning the intermediate (green) level, which was always seen to proceed the irreversible formation of the (red) triazole product (Figures 3B and 8).Indeed, such an ordering of events was seen across all substrates examined (Figure 3B and Supporting Information).It is known from detailed mechanistic investigations that the proposed reactive intermediates between the bis-Cu acetylide (13′, yellow) and the Cutriazolide (15, green) are extremely short lived, even in the presence of stabilizing ligands, 43,46 and it seems reasonable that the state (green in Figures 3B and 8) immediately preceding the final triazole product (red in Figures 3B and 8) can be attributed to Cu-triazolide 15.Such an assignment is in agreement with the previous assignment of the intermediate previously observed in CuAAC reactions at the single-molecule level. 18It also seems reasonable to attribute the noisy character of this green intermediate to 15 ⇌ 15′ copper -complexation, even though its shortlived and irreversible nature precludes analysis of the type presented in Figure 6.Indeed, like the copper-acetylide 13, Cu-triazolide 15 also possesses increased electron density (and basicity) compared to its non-cuprated equivalent, the final triazole product 16.
Our single-molecule analysis of the rapidly equilibrating copper-associated intermediates in the CuAAC mechanism are consistent the hypothesis emerging from ensemble investigations that most CuAAC reactions proceed via a bis-Cu mechanism. 43,46,52While previous investigations of transient intermediates in the CuAAC reaction have required the use of stabilizing ligands, this study demonstrates the ability of synthetically functionalized nanopores to observe short-lived and dynamically exchanging reaction intermediates under typical reaction conditions.

CONCLUSION
We have demonstrated the use of methionine-modification as a site-selective method to enable the synthetic diversification of -HL protein nanopores.The installation of reactive linkers within the channel enabled nanopores to be synthetically diversified via bio-orthogonal click reactions with a range of small molecules, nucleotides and nucleic acids.Notably, this method of nanopore modification complements alternative approaches that use site-directed mutagenesis or native chemical ligation to install new functionalities within nanopores, while also offering superior site selectivity compared to other in situ synthetic pore modification protocols. 34In particular, the ability to use both the azide and alkyne oxaziridine linkers expands the potential for orthogonal functionalization using existing libraries of commercially available alkyne and azide coupling partners.
An unexpected outcome of the in situ structural diversification was the ability to resolve the trajectory of reactions at the single-molecule level, by observing transient intermediates that are hard to detect using ensemble techniques.Moreover, it was even possible to resolve species involved in rapidly exchanging (sub-millisecond) equilibria, and to determine individual equilibrium constants and the associated changes in speciation as the reaction conditions were varied.Indeed, the modification chemistry introduced here is highly versatile, rapid and accessible, and may provide a useful platform for implementing physical organic mechanistic investigation at the single-molecule level.

Single Channel Experiment Setup
Single channel experiments were performed in a custom-built cell (Figure S2 and associated text).A hanging drop of hexadecane in n-pentane (5 µL, 10%, v/v) was touched on each side of the Teflon sheet containing an aperture and allowed to dry for 1 minute.KCl/potassium phosphate buffer (600 µL) was added to the well each side of the aperture.Lipid (1,2diphytanoyl-sn-glycero-3-phosphocholine) (approx.8 µL, 10 g L −1 in n-pentane) was added to each side of the well, and left for approx.5 mins to allow the pentane to evaporate.The cell was subsequently placed into a Faraday cage and Ag/AgCl electrodes (Warner), connected to a patch clamp amplifier (Axopatch 200B, Molecular Devices), were suspended either side of the Teflon sheet.The buffer solution on both sides of the Teflon sheet was aspirated and dispensed using a Hamilton syringe to 'paint' a phospholipid bilayer across the aperture.A ±1 mV pulse was applied at 1333 Hz to determine when a bilayer was obtained (capacitance >30 pF).The membrane was characterized with successive 2 second sweeps under an applied potential ranging from +100 to −100 mM.The membrane was deemed acceptable if the range of current flow across the membrane measured <1 pA.A gel loading tip fitted to a 20 µL pipette was introduced into an aqueous solution of α-HL (approx.250 µM (5 µL of the storage solution diluted with 45 µL of water)), without aspirating, such that a tiny amount (<1 µL) of the solution remained on the tip.Under an applied voltage of +100 mV, the α-HL was then fired at the aperture.If after at least 1 minute no channel insertion was observed, then this process was repeated until a single channel arose.The single nanopore was characterized by at least four 2-second I/V sweeps (−100 to +100 mV, Figure S4), showing the characteristic asymmetry at positive vs negative voltages.

Methionine Modification of Protein Nanopore
A fresh solution of oxaziridine reagent 1-4 was made by dissolving the reagent in DMSO to a concentration of 100 mM.Under an applied voltage (+100 mV), oxaziridine reagent solution (1 µL) was added to either the cis or trans well of a single-channel experiment (see Figure S3; final well concentration of reagent ~0.17 mM).The contents of the well were briefly mixed by aspirating and plunging with a 50 µL pipette three times.Upon observing a ~3% drop in the pore ion current, indicating a single methionine modification, a solution of methionine (250 mM, 8 µL, 20 equiv.) was immediately added to the same well, aiming the micropipette tip towards the location of the aperture.The contents of the well were mixed by repeated micropipette aspiration and plunging (5 × 50 µL).If no further quantized drops in the ion current (indicating further methionine modifications) were observed after 5 min, the singly modified pore was characterized by at least four I/V sweeps (−100 to +100 mV).

CuAAC Reaction of Alkyne-modified Protein Nanopore
Under an applied voltage (+100 mV), a solution of CuSO4 (1 µL, 1 M, final concentration ~1.7 mM) was added to the trans well of a single-channel experiment with an alkyne-modified protein nanopore.A freshly made solution of sodium ascorbate (2 µL, 1 M, final concentration ~3.4 mM) was added, and the contents of the well were mixed by micropipette aspirating and plunging (10 × 50 µL).Complete reduction to Cu(I) was evident by disappearance of the blue colour of the buffer.When reversible formation of a copper acetylide became apparent in the current trace, a solution of azide (1-2 µL, 1 M) was added to the trans well (if carrying a positive charge at pH 8) or the cis well (if carrying a negative charge at pH 8).The contents of the well was carefully mixed by micropipette aspirating and plunging (1-2 × 50 µL), or the azide was left to diffuse across the well.The experiment was left until a permanent drop in the pore current and cessation of reversible events indicated a CuAAC reaction had taken place.The modified pore was characterized by at least four I/V sweeps (−100 to +100 mV).

CuAAC Reaction of Azide-modified Protein Nanopore
Under an applied voltage (+100 mV), a solution of CuSO4 (1 µL, 1 M, final concentration ~1.7 mM) was added to the trans well (if using an alkyne reagent carrying a positive charge at pH 8) or the cis well (if using an alkyne carrying a negative charge at pH 8) of a single-channel experiment with an azide-modified protein nanopore.A freshly made solution of sodium ascorbate (2 µL, 1 M, final concentration ~3.4 mM) was added to the same well, and the contents of the well were mixed by micropipette aspirating and plunging (10 × 50 µL).Complete reduction to Cu(I) was evident by disappearance of the blue colour of the buffer.A solution of alkyne reagent (2 µL, 1 M) was added to the same well, and the contents of the well was carefully mixed by micropipette aspirating and plunging (1-2 × 50 µL).The experiment was left until a permanent drop in the pore current indicated a CuAAC reaction had taken place.The modified pore was characterized by at least four I/V sweeps (−100 to +100 mV).

SPAAC Reaction of Azide-modified Nanopore
Under an applied voltage (+100 mV), a solution of cyclooctyne 10 (10 µL, 100 mM, final concentration ~1.7 mM) was added to both wells of a single-channel experiment with an azidemodified protein nanopore.The contents of the well were mixed by micropipette aspirating and plunging (5 × 50 µL).The experiment was left until a permanent drop in the pore current indicated a SPAAC reaction had taken place.The modified pore was characterized by at least four I/V sweeps (−100 to +100 mV).

Figure 1 .
Figure 1.Synthetic diversification of protein nanopores.(A) Experimental setup in which an ion current arises from the flow of ions through a transmembrane α-HL nanopore under an applied voltage.(B) A single methionine residue (yellow) could be modified using oxaziridine reagents (1-4).(C) The installed linkers were further functionalized in a divergent fashion using popular bioorthogonal click reactions.

Figure 2 .
Figure 2. (A) Oxaziridine modification of methionine residues in an α-HL nanopore.(B) Pore modification reaction with 1 (1.7 mM) resulted in six stepwise blockages in the flow of ions though the nanopore indicating six sequential modification reactions (n = 0 to 6).(C)

Figure 4 .
Figure 4. (A) Proposed mechanistic cycles for the CuAAC reaction, via either a mono-Cu or bis-Cu pathway. 43(B) Cu(I) -complexation with an internal alkyne.

Figure 5 .
Figure 5. Influence of Cu(I) on ion currents in wild-type and alkyne-modified nanopores.(A) No change was seen upon addition of Cu(I) to a wild-type -HL nanopore.(B) In a nanopore modified with a terminal alkyne, introduction of Cu(I) caused increased noise (black) due to short-lived events which are not fully resolved, as well as longer-duration, well-resolved stepwise blockages (yellow).The frequency and duration of the resolved events was dependent on

Figure 6 .
Figure 6.(A-D) All-points histogram analysis of the blue and yellow ion current levels resolved two Cu(I)-dependent Gaussian distributions of states corresponding to alkyne species 12, 12′, 13 and 13′ at varying copper concentrations.(E) The populations of 12, 12′, 13 and 13′ states as the concentration of Cu(I) was varied (circles) was used to determine the equilibrium constants K1, K2, and K3 and the corresponding modelled speciation diagram (lines).Fitting details are provided in Section 2.8 in the SI; both graphical and iterative fitting of the populations yielded similar fits (TableS6and FigureS56).Data obtained at +100 mV in 1 M KCl, 30 mM potassium phosphate buffer (pH 8.0).

Figure 7 .
Figure 7. (A) Experimental evaluation of the effect of the applied electric field on exchange between the 12 ⇌ 12′ (blue) and 13 ⇌ 13′ (yellow) copper-dependent intermediates in alkynemodified nanopores.(B) Sample ion current traces are shown at different voltages (more data are provided in the SI).(C) The voltage dependence of on -1 , off -1 , and Kobs is shown in the colored graph insets.Conditions: CuSO4 (0.82 mM), sodium ascorbate (8.2 mM), 1 M KCl, 30 potassium phosphate buffer (pH 8.0) at 298 K. Histogram data and data fitting are provided in Figure S50.

Figure 8 .
Figure 8. Expanded view of ion current changes during a CuAAC reaction taking place between an alkyne-bearing nanopore and azide 7. Data obtained at +100 mV in 1 M KCl, 30 mM potassium phosphate buffer (pH 8.0) at 298 K. See the SI for the full trace, other examples and detailed experimental conditions.
ASSOCIATED CONTENT Supporting Information.The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.XXXX.Experimental setup; synthetic procedures; full nanopore current traces; experimental details; detailed data analysis; Figures S1−S74; Tables S1−S6.