Real Time Sensing and Discrimination of Single Chemicals Using Channel of Phi29 DNA Packaging Nanomotor

Abstract

A highly sensitive and reliable method to sense and identify a single chemical at extremely low concentrations and high contamination is important for environmental surveillance, homeland security, athlete drug monitoring, toxin/drug screening, and earlier disease diagnosis. This manuscript reports a method for precise detection of single chemicals. The hub of the bacteriophage phi29 DNA packaging motor is a connector consisting of twelve protein subunits encircled into a 3.6-nm channel as a path for dsDNA to enter during packaging and to exit during infection. The connector has previously been inserted into a lipid bilayer to serve as a membrane-embedded channel. Herein we report the modification of the phi29 channel to develop a class of sensors to detect single chemicals. The Lysine-234 of each protein subunit was mutated to cysteine, generating 12-SH ring lining the channel wall. Chemicals passing through this robust channel and interactions with the SH-group generated extremely reliable, precise, and sensitive current signatures as revealed by single channel conductance assays. Ethane (57 Daltons), thymine (167 Daltons), and benzene (105 Daltons) with reactive thioester moieties were clearly discriminated upon interaction with the available set of cysteine residues. The covalent attachment of each analyte induced discrete step-wise blockage in current signature with a corresponding decrease in conductance due to the physical blocking of the channel. Transient binding of the chemicals also produced characteristic fingerprints that were deduced from the unique blockage amplitude and pattern of the signals. This study shows that the phi29 connector can be used to sense chemicals with reactive thioesters or maleimide using single channel conduction assays based on their distinct fingerprints. The results demonstrated that this channel system could be further developed into very sensitive sensing devices.

Biological systems contain a wide variety of nanomachines and highly-ordered macromolecular structures with diverse biological functions. Double-stranded (ds) DNA viruses package their genome into a preformed protein shell with the aid of a motor.^1–3 Bacterial virus phi29 DNA-packaging nanomotor with its elegant and elaborate channel has inspired its application in nanotechnology. This motor^4,5 uses ATP as energy and is comprised of 1) a central protein core called the connector,^6,7 2) six copies of pRNA (packaging RNA)^8–11 and 3) an ATPase gp16. The connector is composed of 12 copies of protein gp10, which encircle to form a dodecameric channel that acts as a pathway for the translocation of dsDNA. This ATP driven DNA-packaging feature is common in other viruses as well, such as T4,^12–16 λ,^17–20 HK97,²¹ T3,²² T7,^23,24 HSV,^25–29 adenovirus,^30–32 and poxvirus.^33,34 Although the individual building blocks of the connector of different viruses share little sequence homology and exhibit large variations in molecular weight,^7,26,35–39 the connector complexes possess a significant amount of morphological similarity.⁴⁰ In the case of bacteriophage phi29, the structure of the connector has been determined to atomic resolution.^7,35 The connector ring consists of twelve subunits, with a central channel formed by three long α-helices of each subunit. The ring is 13.8 nm at its wide end and 6.6 nm at its narrow end (Fig. 1). The cross-sectional area of the channel is 10 nm² (3.6 nm in diameter) at the narrow end and 28 nm² (6 nm in diameter) at its wider end (Fig. 1).^6,7,35 The wider end of the connector is located within the capsid, while the narrow end partially protrudes out of the capsid.

Figure 1. Structure of phi29 DNA-packaging motor connector showing the location of the cysteine ring and the dimensions of the channel.

Top view (A) and side view (B) of phi29 connector showing the location of the tunnel loop N229 (green) and N246 (red). K234C is located within the tunnel loops shown.

We recently inserted the connector into a lipid bilayer and the resulting system has been shown to exhibit robust properties and generate extremely reliable, precise and sensitive conductance signatures when ions or DNA pass through the channel, as revealed by single channel conductance measurements.^41,42 The connector channel exhibits one-way traffic property for DNA translocation from N-terminal entrance to C-terminal exit with a valve mechanism of DNA packaging.⁴³ Explicit engineering of the phi29 connector^44–50 is possible due to its available crystal structure.^6,7,35 The pore size is nearly identical from sample to sample and chemical conjugations within the large cavity of the pore can therefore be made with relative ease for added functionality. Procedures for large scale production and purification of the connector have already been developed.^{41,44,45,48,51,52} These features make the channel an ideal building block to produce a variety of semi-synthetic receptors with predictable and reliable properties.^48–50

Nanopore-based sensory techniques have been extensively used for the detection of a myriad of biomedical targets, from metal ions, drug compounds, and cellular second messengers to proteins and DNA.^53–70 Engineered transmembrane channels have the potential for stochastic detection, an approach relying on the real-time observation of individual binding events between single substrate molecules and a receptor.^71–73 Protein pores can be selectively functionalized with various probes that can bind individual target molecules with high selectivity and sensitivity.^74–76 The characteristic binding and distinctive current signatures can reveal the identity and concentration of the target analyte.^71,76 In addition, the dynamic interactions between the analyte and the binding sites can be studied in real-time at high resolution using single channel conduction assays.^75,77

The ability to reengineer a biological nanomotor with atomic precision and incorporate it into lipid membranes will provide additional functionality as well as durability. Herein we report on the chemical modification of the phi29 channel as the first step in developing a class of sensors to detect single molecules. We introduced a cysteine ring within the channel wall and conjugated receptor modules within the channel wall.

RESULTS

Introduction of cysteine residues within the channel wall

The phi29 connector is composed of 12 copies of the protein gp10 with the central channel formed by three long helices of each subunit.^{6,7,35,47,79–81} The connector has been inserted into a lipid bilayer and the resulting system has been shown to exhibit robust properties and generate extremely reliable, precise and sensitive conductance signatures when ions or DNA pass through the channel, as revealed by single channel conductance measurements.^41,42 The pore size is near identical (Fig. 2A) from sample to sample and chemical conjugations within the large cavity of the pore can therefore be made with relative ease for added functionality. The native connector contains two cysteine residues (C76 and C275) per gp10 protein that are buried within the wall and therefore are not accessible for conjugation of analytes or recognition moieties. Hence, accessible cysteines were introduced by mutagenesis within the inner wall of the connector, namely the K234C mutant, where the lysine (K) was replaced with a cysteine (C) at the amino acid 234 position, located within the tunnel loop (Fig. 1 A, B). After assembly, the internal mutation generates a cysteine ring, composed of 12 evenly spaced residues in the same plane within the dodecameric connector channel (Fig. 1).

Figure 2.

(A) A current trace showing continuous insertions of single K234C connectors into BLM. (B) Histogram of conductance measurements for single insertion under a specific voltage -75 mV (Red: Wild type connector; N =500. Black: K234C connectors; N=275). (C) Current -Voltage trace under a ramp voltage from −100 mV to 100 mV (2.2 mV/s). All experiments were performed in presence of 5 mM Tris (pH 7.8) with 1 M NaCl under −75 mV.

Characterization of cysteine mutant connector channels in the lipid membrane

The K234C mutant connectors were first reconstituted into liposomes, followed by insertion into planar lipid membranes via vesicle fusion of liposome/connector complexes,⁴¹ thereby forming a membrane-embedded nanopore system. The insertion of the K234C connector channels resulted in step-wise increase of the conductance, as shown in the continuous current trace (Fig. 2A). The internal mutation did not affect the membrane insertion efficiency, signal stability, current homogeneity, membrane durability, and voltage gating properties.^41,42 However, the C-his K234C channels had a smaller cross-sectional area (conductance: 2.2 ± 0.17 nS) compared to the C-his connectors without the mutation (conductance: 3.2 ± 0.2 nS) (Fig. 2B).^41,42 The conductance was calculated using the ratio of the measured current jump induced by a discrete step to the applied voltage. The step size of the connector channels was homogenous (Fig. 2B) and the channels exhibited equal conductance under both positive and negative trans-membrane potentials. The channel conductance is uniform and demonstrates a perfectly linear I–V relationship without displaying any voltage gating phenomenon under the reported conditions of ±100mV (Fig. 2C).

Sensing and discrimination of ethane, thymine, and benzene thioester moieties

Common methods to selectively conjugate to the sulfhydryl side chain of cysteine residues include maleimides and iodoacetamides that form C-S bounds and sulfhydryls that form disulfides under oxidizing conditions. To determine the extent to which functional moieties can recognize, we adopted the well-known methodologies utilizing disulfide linkage or transthioesterification reactions⁷⁸ for conjugating various recognition groups to the cysteine residues exposed at the surface of the inner channel wall (Fig. 3C).

Figure 3. Conjugation of chemical ligands to channel wall resulted in the reduction of channel size as indicated by uniform stepwise blockage of channel current.

(A) Schematic of ligand binding to the inner wall of the connector pore. (B) Synthesis of thioesters with ethane, thymine and benzene moieties. (C) Transthioesterification method, showing that cysteines are readily modified with nucleobases when exposed to thioesters, using thymine thioester as an example; a second reaction via disulfide linkage results in the binding of the methyl thioglycolate by-product.

The thioexchange reaction is highly chemoselective, rapid, and occurs in buffered aqueous solutions (pH 7), which appeared to be ideal for derivatizing the phi29 channel under conduction assay conditions. Binding events are characterized by monitoring the changes in the intensity of the current signal in a BLM setup (Fig. 3A). The binding of chemicals sequentially to each cysteine probe induces step-wise blocks (a single molecule per block) in current with a corresponding decrease in conductance due to the physical blocking of the channel (Fig. 4). The characteristic signature of the captured analyte is given by the unique blockage amplitude and pattern of the signals. The number of molecules bound can be counted from the step-wise blocks and the concentration can be deduced from the number of binding events to a single pore per unit time.

Figure 4.

Data showing discrete blocking steps due to the binding of thioesters groups containing ethane (A), thymine (B) and (C) benzene. Insert: Magnified current trace showing the transient and permanent current blockage events.

The derivatized thioesters served as models to demonstrate the utility of the connector pore in capture and sensing (Fig. 3A). The phenyl and thyminyl groups are similar in size, but the phenyl ring is hydrophobic and the thyminyl ring forms hydrogen bonds with water. The much smaller ethyl group truly tests the ability of the channel to sense small molecules. Two classes of current blockage signals were observed. The first class is the permanent binding event, which occurs through covalent bond formation between a thioester and a cysteine side chain of the channel via the transesterification reaction (Fig. 3C). Permanent binding is displayed as discrete step-wise augment of current blockage (Fig. 4, 5C). The second class is the transient binding events induced by temporary binding of chemicals with the cysteines, as they pass through the channel and appear as the recoverable blockage signals (Fig. 4, insert and Fig. 5A, B).

Figure 5. Analysis of current blockage events induced by thioesters.

Histogram of transient current blockage events (A) with corresponding dwell time distributions (B) and permanent binding events (C) for the binding of thioesters groups containing ethane, thymine, and benzene respectively. The dwell time data was fitted with a single exponential decay function (blue).

We found that the permanent binding events distinguished the ethane (C₃H₅O:57 Daltons), thymine (C₇H₇N₂O₃: 167 Daltons), and benzene (C₇H₅O: 105 Daltons) thioester moieties that remain covalently attached to the channel wall (Fig. 5C, Table 1B), with a characteristic channel current blockage of 33.5 ± 0.5%, 36.3 ± 1.2%, and 38.2 ± 2%, respectively (p-value <0.001). These percentages represent the ratio of the open-pore current and the current reduction step induced by the individual binding events. The high permanent percent blockage value for the ethane moiety shows that this large channel is very sensitive to the presence of small molecules. The higher permanent percent blockage observed for the smaller sized benzene as compared to thymine shows that the shape or polarity of the molecule, besides the molecular weight, also contributed to the current blockage signature.

Table 1.

Analysis of current blockage events

A. Thioesters: Transient Binding Events
Thioesters	Transient Binding Events		Difference (%)		Dwell Time # (ms)
	Component (Cys-X)	Component (Cys-X + Cys-Y)	Components [(Cys-X + Cys-Y) – (Cys-X)] = (Cys-Y)	p-value*
Ethane	16.4 ± 2.0 %	29.6 ± 1.4 %	13.2 ± 2.4 %	<0.822 (Thymine); <0.189 (Benzene)	23.8 ± 3.4
Thymine	18.9 ± 2.6 %	31.7 ± 2.2 %	12.8 ± 3.4 %	<0.822 (Ethane); <0.184 (Benzene)	30.5 ± 5.2
Benzene	19.5 ± 6.2 %	37.3 ± 3.7 %	17.8 ± 7.2 %	<0.189 (Ethane); <0.184 (Thymine)	23.0 ± 4.5

B. Thioesters: Permanent Binding Events
Thioesters	Transient Binding Events	Permanent Binding Events		Ratio of Permanent to Transient Events
	(Cys-X)	(Cys-X + Cys-X)	p-value**	$\frac{(\mathit{Cys}-X+\mathit{Cys}-X)}{(\mathit{Cys}-X)}$	p-value**
Ethane	16.4 ± 2.0 %	33.5 ± 0.5 % (N = 63)	<0.001(Thymine); <0.001 (Benzene)	2.04 ± 0.12	<0.001 (Thymine); <0.065 (Benzene)
Thymine	18.9 ± 2.6 %	36.3 ± 1.2 % (N = 44)	<0.001 (Ethane); <0.001 (Benzene)	1.92 ± 0.14	<0.001 (Ethane); <0.437 (Benzene)
Benzene	19.5 ± 6.2 %	38.4 ± 2.0 % (N = 66)	<0.001 (Ethane); <0.001 (Thymine)	1.96 ± 0.32	<0.065(Ethane); <0.437 (Thymine)

C. Maleimide Binding Events
	Transient Binding Events	Permanent Binding Events	Ratio of Permanent to Transient Events	Dwell Time## (ms)
Maleimide	17.0 ± 2.0 %	38.1 ± 2.1 % (N = 7)	2.2 ± 0.12	24.0 ± 4.8

^*The p-value is obtained from 6 independent experiments.

^#Derived from single exponential decay fit.

^**The p-value is obtained using the number of covalent binding events for each thioesters.

^##Derived from single exponential decay fit.

Although the transient signal represents the hit-and-run behavior (temporary binding interactions) of the chemicals, they can also be used to identify the ethyl, benzyl, or thyminyl groups (Fig. 5 A). Two transient signals, denoted components (Cys-X) and (Cys-X + Cys-Y) are observed for each thioester. Each set of signals is unique for a particular thioester (Figure 5A, Table 1A), and all transient percent blockage values are smaller than observed for the corresponding permanent percent blockage value (Figure 5C, Table 1A, B). This result is surprising considering that the thioexchange reaction used to attach the compounds to the channel reduces the size of the molecule upon the cleavage and release of the thioester end (Fig. 3C). One possible explanation is that the released thioester end (Y) binds concomitantly to the channel through disulfide bond formation (Cys-Y) during the same observable time frame as the thioexchange reaction, giving two molecules bound per channel (Cys-X + Cys-Y) for each step change in conductance (Fig. 5A). We were unable, however, to attach the thioester directly through disulfide bond formation under the same reaction conditions. We further observed that the dwell time for all the transient events for the three thioesters were ~ 25 ms (Table 1A). This relatively high dwell time implies that the transient events are indeed due to temporary binding of chemicals to the cysteine residues, rather than simple translocation through the pore. We cannot distinguish differences in dwell time between the two transient components (Cys-X) and (Cys-X + Cys-Y), implying that the disulfide bond formation (Cys-Y) occurs during the same observable time frame as the thioexchange reaction (Cys-X).

Two consistent findings were observed. The ratio of the permanent percent blockage signal and transient component (Cys-X) is ~2, implying that there are two thioesters bound per binding steps. Another interesting finding is that the difference between the percent blockage of the transient signals [Component (Cys-X) - Component (Cys-X + Cys-Y)] of ~13% for each set of transient signals reflect the temporary binding of the acid end of a thioester, which underwent hydrolysis in the solution phase. Accordingly, the ~13% would probably represent the loss of the thioester end, which is the consistent piece for all thioesters. We are currently designing experiments to assign the transient signals. These consistencies demonstrates that the channel is operating the same for each thioester, and the unique properties of a thioester control the set of percent blockage values. This means that, for these thioesters, three signals or fingerprints can be used singularly or in some combination to identify their presence within the channel.

Sensing of maleimde revealed a consistent blockage ratio of 2 between permanent and transient binding events

Maleimide is a common reagent to link desired functional groups to cysteines. Accordingly, upon addition of maleimide to the conducting buffer, our single channel current measurements revealed only one type of transient binding events with a percent blockage of ~17% (Figure 6 B,C). The permanent binding events with percent blockage of ~ 38% are ~2 times the transient events, which implies that two maleimide binding events occur per step (Fig. 6, Table 1C). Similar dwell time of ~24 ms were observed for the transient binding events of maleimide (Table 1C).

Figure 6. Conjugation of maleimide to channel wall.

(A) Reaction scheme of maleimide-cysteine reaction. (B) Data showing discrete blocking steps due to the binding of maleimide. Insert: Magnified current trace showing the transient and permanent current blockage events. Histogram of transient current blockage events for maleimide (C) with corresponding dwell time distributions (D). The dwell time data was fitted with a single exponential decay function (blue).

The consistent observation was the ratio between permanent and transient signal (Table 1B), which ends up to be ~ 2 for all three thioesters investigated as well as for maleimide. Since the percent blockage values are different for each thioester and maleimide, it is unlikely due to the conformational change of the connector channel upon binding of the chemicals. Instead, a strong possibility is the simultaneous binding of two identical chemicals to neighboring cysteine residues, which probably stabilizes the housing of chemicals within the channel. Each channel can therefore accommodate an even number (2, 4 and maximum 6) thioesters and maleimide moieties based on the percent blockage values.

DISCUSSION

Development of a highly sensitive detection system is important in many areas, including, but not limited to pathogen identification, disease diagnosis, drug screening, environmental monitoring and national security. The detection of pathogens or toxins is analogous to the detection of criminals. The first step is the capture of a suspect by the police. The next step is the gathering of evidence, such as fingerprints, to serve as identification for court prosecution. The concept of ‘capture’ has been used extensively in many sensing, detection and diagnostic technologies that are currently available, such as antigen/antibody reactions,^82–84 biotin/streptavidin interactions,^85–89 ELISA,^90,91 microarray based technologies,^92–94 have been extensively used in practical applications. All of these techniques are well developed and very useful; however, they possess a disadvantage in cases where the target substrate is at a very low concentration or when impurities are present at high concentrations. As a result, due to dissociation by a kinetic process, it is not possible to detect a single substrate molecule on a slide, regardless of how strong the signal of a single antibody is. This is because, at a very low concentration, the background noise will over-ride the specific signal from the Antigen/Antibody complex, and the Antibody/complex will dissociate upon washing. The sensitivity issue is especially important in the early diagnosis of diseases and in environmental surveillance. When a small amount of chemical is spread on to the earth, the massive amount of materials in the environment will make it difficult to detect. The concept of ‘fingerprinting’ has also been used extensively in the detection and diagnosis process. Examples include the spectrum pattern from spectroscopy,^88,95–104 fluorescence, CD, NMR, and electrophysiological measurements of a single pore. The same issue exists in that the impurity will result in a nonspecific signal and hinder the sensitivity and specificity due to the high background noise.

The results herein demonstrate that the phi29 channel is a sensor of very small compounds (the area of the ethylthioester is approximately 60 Ǻ²) even though it's interior volume is very large (approximately 1,000 Ǻ²). Compounds are also differentiated based on their physical properties as well as their size. Sensing through the covalent derivatization of the channel produced an even larger percent blockage signal. This study showed that the phi29 can be used to sense chemically active thioesters in single channel conduction assays by observing a total of three readily observable fingerprints. Future experiments include using the permanently attached compounds within the channel to further fine tune the phi29 channel sensor. Because the phi29 channel contains a large interior volume, attachment of a large variety of recognition domains is possible. Some domains could require the addition of two or more functional groups. We are currently developing methods to control the number of functional groups added to a channel and the pattern that they form. The proof-of-principle studies described here will be extended in the future to include multiple probes that can be constructed by engineering within a single pore for concurrent detection of multiple targets.

EXPERIMENTAL METHODS

Materials

Organic solvents (n-Decane, hexane, Choloroform, methylene chloride, and DMSO) were purchased from Sigma-Aldrich or Fisher. Moisture sensitive reactions were carried out under positive argon pressure. ¹H NMR and ¹³C NMR spectra were obtained in CDCl₃ or DMSO-d₆ using a Bruker AMX400 spectrometer operating at 400.14 MHz for proton and 100.23 MHz for carbon nuclei. Chemical shifts are in ppm and are referenced using an internal TMS standard. Mass spectra were obtained using Micromass Q-TOF-2 spectrometer for ESI.

The phospholipids 1,2-diphytanoyl-sn glycerol-3-phosphocholine (DPhPC) and 1,2- Dioleoyl-sn-Glycero-3-Phosphocholine (DOPC) were purchased from Avanti Polar Lipids.

Mutation K234C and purification

Gp10-K234C plasmids were constructed from vector pET-21a(+) (Novagen) with a two-step PCR. First primer pair F1-R1 was used to amplify gp10 gene from phi29 genome DNA-gp3, and the first PCR product was used as the template for following PCR with primer pair F2-R2 incorporating restriction sites and affinity tags. The second PCR product was digested with Nde I-Xho I, and ligated into the Nde I-Xho I sites of the vector pET-21a(+).⁴⁵

The plasmids constructed previously were transformed into the E. coli strain HMS174 (DE3) for protein expression. 10 ml overnight cell culture were grown at 37°C in Luria-Bertani (LB) medium containing 100 μg/ml ampicillin with shaking at 250 rpm. One percent inoculation was used for 500 ml culture and 0.5 mM IPTG induction was applied when OD₆₀₀ reached 0.5–0.6. Cells were harvested 3 hours post induction by centrifugation at 5000 ×g for 20 min in a Beckman JS-7.5 rotor, followed by French Press, and digested with DNase I and RNase A on ice for half an hour. Cell lysate was clarified by centrifugation and filtration.

Purification of mutant K234C-His tagged connector proteins was conducted with one-step immobilized metal affinity chromatography (IMAC) under native conditions. Ni-NTA His-bind resin (Novagen, San Diego, CA) was packed to a column and regenerated with charge buffer (50 mM NiCl₂). Cell pellets were resuspended in His-Bind buffer (15% glycerol, 500 mM NaCl, 10 mM imidazole and 100 mM Tris-HCl, pH 8.0). Cell lysate was clarified by centrifugation and filtration followed by loading onto His-bind resin by gravity flow. Five column volumes (CV) of His-Wash buffer (15% glycerol, 500 mM NaCl, 50 mM imidazole and 100 mM Tris-HCl, pH 8.0) was used to remove contaminant proteins. Three to five CV of His-Elute buffer (15% glycerol, 500 mM NaCl, 500 mM imidazole and 100 mM Tris-HCl, pH 8.0) was applied to elute His-tagged gp10 protein.

Preparation of lipid vesicles containing the mutant K234C connector channels

The preparation of connector reconstituted liposomes has been described previously.⁴¹ Briefly, 1 ml of 1 mg/ml DOPC in chloroform was syringed in a round bottomed flask. The chloroform was removed under vacuum using the Rotary Evaporator (Buchi). The lipid film was then rehydrated with 1 ml of connector protein solution in a buffer containing 10 mM Tris/pH 7.9, 1 M NaCl, and 250 mM sucrose to bud off vesicles into the solution. The lipid solution was then extruded through a polycarbonate membrane filter (100 nm or 400 nm) to generate unilamellar lipid vesicles. A final molar ratio of lipid vs. connector was established at 4000:1 to 16000:1.^41,42,43

Insertion of the connector into planar bilayer lipid membrane

The insertion of the connector reconstituted liposomes into a lipid bilayer has been described previously.⁴¹ Briefly, a standard BLM cell was utilized to form a free standing lipid bilayer. An aperture of 100 μm or 200 μm in diameter in a thin Teflon partition separated the cell into cis and trans compartments. After the aperture was pre-painted with 0.5 μl 3% (w/v) DPhPC n-decane solution twice to ensure the complete coating of the entire edge of the aperture, these compartments were filled with conducting buffers (10 mM Tris/pH 7.9, 1 M NaCl). After confirming the formation of the lipid bilayer, the connector reconstituted lipid vesicles (0.5–2 μl) was directly incubated to fuse with the planar lipid membrane.

Measurements of current for each membrane embedded channel

A pair of Ag/AgCl electrodes connected directly to both the compartments was used to measure the current traces across the bilayer lipid membrane. The current trace was recorded using an Axopatch 200B patch clamp amplifier coupled with the Axon DigiData 1322A analog-digital converter (Axon Instruments) or the BLM workstation (Warner Instruments). All voltages reported were those of the trans-compartment. Data was low band-pass filtered at a frequency of 5 kHz or 1 KHz and acquired at a sampling frequency of 20 or 200 KHz. The PClamp 9.1 software (Axon Instruments) was used to collect the data, and the software Origin Pro 8.0 was used for data analysis.

For all the binding experiments, the thioesters (10 nM - 1 μM final concentration) were premixed in the conducting buffer (10 mM Tris/pH 7.9, 1 M NaCl). Upon formation of a stable bilayer, the current traces were then recorded over a period of 2–4 hours.

Synthesis of various thioesters

The synthesis of the thioesters was accomplished using a slight modification of the route developed by Ghadiri et al.⁷⁸ Synthesis of the ethylthioester will be used as an example. A 0.30 gram (40 mmol) sample of propanoic acid and a 0.66 gram (41 mmol) sample of CDI were dissolved in 20 mL of freshly distilled CHCl₃. The reaction mixture was refluxed for three hours before it was allowed to come to room temperature. Once the solution was cooled, 0.430 gram (40 mmol) sample of methylthioglycolate was added. This reaction mixture was left overnight. The solvent was removed under reduced pressure and partitioned between a 0.01 molar sodium hydroxide solution and dichloromethane. The organic layer was collected and dried over sodium sulfate. After removing the solvent under vacuum, crude material was separated via flash chromatography with CH₂Cl₂:hexane 80:20 (v/v) as eluent. Ethylthioester was obtained as a colorless oil (0.30 gram,18 mmol) in a 46% yield.

Chemical characterization of various thioesters

¹H-NMR (CHCl₃-d): δ 3.69 (3H, s), 3.65 (2H s), 2.58 (2H q, J=7.6 Hz), 1.12 (3H t, J=7.6 Hz) ¹³C-NMR (CHCl₃-d): δ 198.1, 169.3, 52.7, 37.0, 30.9, 9.4. MS: calculated mass 185.0248 for C₆H₁₀O₃SNa⁺ [M + Na⁺] found 185.0268. For thymine (14% yield) and phenyl (95% yield) thioesters, chemical analysis matched literature results. Provided also is the chemical analysis of the phenyl ester, which is not provided in the literature. ¹H-NMR (CHCl₃-d): δ 7.91 (2H d, J=7.6 Hz), 7.54 (1H t, J=7.6 Hz), 7.39 (2H t, J=7.6 Hz), 3.84 (2H, s), 3.70 (3H s). ¹³C-NMR (CHCl₃-d): δ 183.9, 172.2, 133.8, 130.1, 129.3, 128.5, 53.0, 31.4. MS: calculated mass 233.0248 for C₁₀H₁₀O₃SNa⁺ [M + Na⁺] found 233.0110.