Tetramethylammonium-Filled Protein Nanopore for Single-Molecule

Sep 4, 2015 - Nanopore technology, as the simplest and most inexpensive single-molecule tool, is being intensively developed. In nanopore stochastic ...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/ac

Tetramethylammonium-Filled Protein Nanopore for Single-Molecule Analysis Ying Wang, Fujun Yao, and Xiao-feng Kang* Key Laboratory of Synthetic and Natural Functional Molecular Chemistry, College of Chemistry & Materials Science, Northwest University, Xi’an 710069, P. R. China S Supporting Information *

ABSTRACT: Nanopore technology, as the simplest and most inexpensive single-molecule tool, is being intensively developed. In nanopore stochastic sensing, KCl and NaCl have traditionally been employed as pore-filled electrolytes for recording the change of ion conductance in nanopores triggered by analyte translocation through the pore. However, some challenges limit its further advance. Here we used tetramethylammonium (TMA) chloride, instead of KCl, as a novel analysis system for nanopores. Some unique nanopore characteristics were observed: (1) The stability of the planar lipid bilayer for embedding the protein pores was elevated at least 6 times. (2) The TMA-Cl system could effectively reduce the noise of single-channel recording. (3) It was easy to control the insertion of protein pores into the lipid bilayer, and the formed single nanopore could last for a sufficiently long time. (4) TMA-Cl could be used as a DNA speed bump in the nanopore to slow DNA translocation speed. (5) The capacity of the nanopore capture of DNA (capture rate) increased significantly and simultaneously increased the translocation time of DNA in the pore. (6) The TMA-filled nanopore could discriminate between various polynucleotides.

N

translocation speed through the nanopore is of utmost importance for DNA sequencing. For unconstrained ssDNA molecules, translocation speed is about 1−20 μs/base under standard experimental conditions.28−30 Under such a high speed, the nanopore is incapable of identifying each base in a DNA strand under the maximum 50 kHz bandwidth of current commercial instruments. In addition, how to effectively capture analyte molecules into the nanopore from bulk solution (capture rate) is still a crucial factor for detecting sensitivity and for obtaining accurate measurement and analytical time. The current nanopore technology for KCl is extremely inefficient because only a small number of DNA molecules can go through the nanopore. Some methods have been proposed to improve its efficiency, such as controlling the solution viscosity31 and pH,30 as well as introducing additional positive charges and DNA-binding biomolecules into the pore32,33 for slowing the DNA velocity, a gel-encapsulated nanopore for stability,27 a salt gradient for increasing capture rate,34,35 and fluorinated fos-choline for controlling protein insertion.36 Although these approaches have resulted in some improvements, they have not yet been explored systematically. In this paper, we propose a simple, effective, and systematic

anopore sensors can be used to analyze and characterize individual biomolecules by recording the changes in the ion current of KCl/NaCl-filled nanopores. This ability to detect single-molecule events endows nanopore sensors with inherent merits such as simplicity, label-free detection, extraordinary sensitivity, high selectivity, and short detection time.1,2 Nanopore-based single-molecule technology has been employed as a molecular identifier and detector for genetic sequencing,3−7 protein sequencing,8,9 protein−DNA/RNA interaction,10,11 and sizing biopolymers,12,13 as well as various analytical detections including DNA/RNA,14−16 metal ions,17,18 small organic molecules,19−21 peptides,22 proteins,23−25 and biomarkers.26 An ultimate goal in the use of nanopores is to sequence DNA, in which the ideal nanopore sequencer should effectively capture DNA into a pore, identify individual nucleic acid bases in the DNA strand, and produce a direct readout of the base sequence. Despite the number of diverse applications, some challenges still exist and limit its further development for both basic research and practical analytical devices. First, there are the common issues of the unstable lipid bilayer, higher system noise, and the shorter lifetime of a single protein nanopore. The free-standing fluid lipid bilayer used for embedding the protein pores is fragile and incurs noise easily from mechanical disturbance.27 It typically lasts only dozens of minutes under optimal conditions. Furthermore, during data collection, the pore needs to last for a long time, in particular a single protein pore rather than multipores, which requires special skills and knacks. Moreover, regulating molecular © 2015 American Chemical Society

Received: July 11, 2015 Accepted: September 4, 2015 Published: September 4, 2015 9991

DOI: 10.1021/acs.analchem.5b02611 Anal. Chem. 2015, 87, 9991−9997

Article

Analytical Chemistry

capacitance measurements, TMA at a concentration range of 1 to 4 M not only did not result in the leakage of the planar lipid bilayer, but it also lasted for a longer lifetime. The average lifetimes of the DphPC planar lipid bilayer were 9.0 ± 1.2 h (n = 15) and 12.0 ± 2.7 h (n = 15) for 1 and 4 M TMA-Cl, respectively, which were at least 6 times that of 1 M KCl (1.5 ± 0.5 h, n = 15). Experimental results displayed that TMA-Cl can elevate the stability of bilayer, and a higher concentration resulted in a superior performance. In addition, we were also able to re-form the bilayer in TMA-Cl solution which was similar to KCl. α-HL is a typical pore-forming protein for stochastic sensing. Comparing TMA-Cl with KCl electrolyte, we did not observe obvious distinctions for the insertion of preformed α-HL heptamer in DphPC planar lipid bilayer. However, under a controlled concentration of 0.05 ng/mL αHL, the formed single channel in TMA-Cl lasted for sufficiently long times: 7.0 ± 1.4 h in 1 M TMA-Cl; 12.0 ± 2.6 h in 4 M TMA-Cl. The latter lasted almost as long as the lipid bilayer’s lifetime. In this process, no further insertion was observed. In contrast, during the bilayer lifetime, a single αHL channel in 1 M KCl only existed for 0.5 ± 0.1 h (n = 15). Noise in the single nanopore recording is an important factor for nanopore single-molecule measurement and detection accuracy. Surprisingly, the TMA-Cl system can also effectively reduce recording noise as shown in Figure 1 for the current

strategy to simultaneously address these problems using tetramethylammonium chloride.



EXPERIMENTAL SECTION Reagents and Materials. Oligonucleotides including poly(dA)25, poly(dC)25, and poly(dT)25 were synthesized and purified by electrophoresis (Integrated DNA Technologies). Tetramethylammonium chloride (98%), tetraethylammonium chloride (98%), and tetrapropylammonium chloride (98%) were used as received from Sigma-Aldrich. All DNA solutions were prepared using DNase-free water (100 μM as stock) and stored at 4 °C. Before nanopore experiments, these DNA samples were heated to 95 °C for 5 min and then gradually cooled to room temperature. 2-Diphytanoylphosphatidylcholine (DphPC) lipid was obtained from Avanti Polar Lipids. The thickness of Teflon film (Goodfellow) was 25 μm. WT-αHL Monomers and Homoheptamer Pores. The wild-type (WT) α-HL monomers were expressed in Escherichia coli BL-21 (DE3) pLysS and purified by size exclusion chromatography. The assembly and purification of heptametrical protein pores were carried out as reported previously.37 Single-Channel Recording and Data Analysis. A bilayer of 2-diphytanoylphosphatidylcholine was formed over a 120− 160 μm orifice in a Teflon septum that divided a planar bilayer chamber into cis and trans compartments. The formation of the bilayer was achieved using the Montal−Mueller method.38 Both solutions contained organic ammonium salts or KCl at a desired concentration and were buffered with 10 mM Tris (pH 8.5). α-HL protein and DNA polymers were added to the cis compartment, which was connected to a “ground”. The final concentration of the α-HL proteins used for the single-channel insertion was 0.05−0.2 ng/mL. Currents were recorded with a patch clamp amplifier (Axopatch 200B, Molecular Devices, Sunnyvale, CA), filtered with a built-in four-pole Bessel filter at 5 kHz, sampled at 20 kHz by a computer equipped with a Digidata 1440 A/D converter (Molecular Devices), and acquired with Clampex 10.3 software (Molecular Devices). Single-channel event amplitude and duration were analyzed using Clampfit 10.3 (Molecular Devices) and origin 8.0 (Microcal, Northampton, MA) software. Mean dwell time values were obtained from the dwell histograms after the peaks were fitted to single exponential functions. The standard deviation of open pore current was obtained from singlechannel current baseline histograms by fitting the distributions to Gaussian functions. The values of mean signal amplitude and I/I0 were obtained from signal amplitude and I/I0 histograms by fitting the distributions to Gaussian functions.



Figure 1. Single-channel current traces in 1 M KCl, 1 M TMA-Cl, and 4 M TMA-Cl (a) and their corresponding histograms (b). The traces and histograms of α-HL (wild-type) protein nanopore clearly show different current noises. All histograms contained a 20 000 sampling number. The recordings were made at 20 kHz, 25 °C, and the applied potential was +120 mV (a positive potential on the trans side, the cis side at ground). The buffer used was 10 mM Tris-HCl (pH 8.5).

traces in TMA-Cl and KCl. The noise was dependent on the TMA-Cl concentration. The standard deviations of the open pore current baseline were 2.49 pA, 1.34 pA, and 1.01 pA for 1 M KCl, 1 M TMA-Cl, and 4 M TMA-Cl at +120 mV, and the half-peak widths of the opening current histogram were 4.89 ± 0.01 pA, 2.63 ± 0.01 pA, and 1.96 ± 0.01 pA for 1 M KCl, 1 M TMA-Cl, and 4 M TMA-Cl (Figure 1). Then we calculated the signal-to-noise ratio (SNR) which was the ratio of the mean signal amplitude to the standard deviation of the open pore current in different electrolytes.41,42 The calculated SNR values of poly(dA)25 were 39.20, 46.57, and 115.25 for 1 M KCl, 1 M TMA-Cl, and 4 M TMA-Cl. The results suggested that the use of TMA-Cl instead of KCl electrolyte could improve the signalto-noise ratio, and higher TMA-Cl concentrations resulted in an even better signal-to-noise ratio. TMA-Cl is an ideal electrolyte solution in the nanopore detection system because of its low-noise characteristics and stability of the single channel and should be favorable for DNA

RESULTS AND DISCUSSION

TMA-Stabilized Lipid Bilayer for Low-Noise SingleNanopore Recording. Tetramethylammonium (TMA or Me4N+) is the simplest quaternary ammonium cation, consisting of four methyl groups attached to a central nitrogen atom. TMA salt is often used as a supporting electrolyte in organic electrochemistry.39 A recent study of the TMA ion indicated that it has unexpectedly hydrophilic properties, despite the presence of four hydrophobic methyl groups in its structure.40 The TMA amphiphiles could be very useful for a lipid bilayer-based nanopore system. In this work, we first examined the effects of TMA on planar lipid bilayer stability and membrane protein insertion. As monitored by bilayer 9992

DOI: 10.1021/acs.analchem.5b02611 Anal. Chem. 2015, 87, 9991−9997

Article

Analytical Chemistry detection. These characteristics should be ascribed to the interaction between TMA cation and lipid bilayer. Previous work found that phosphatidylcholine cell membranes contain alkyltrimethylammonium,43 which can play an important role in lipid membrane functions and stability.44 The cationic TMA amphiphile molecule, similar to fluorinated fos-choline,36 could aggregate on the planar lipid bilayer to stabilize it and prevent the insertion of more protein nanopores. It should also be noted that the nanopore opening current values were different in both TMA-Cl and KCl. With the change of the electrolyte from 1 M KCl to 1 M TMA-Cl, the opening current of a single α-HL channel decreased from 110 ± 10 pA to 65 ± 5 pA at +120 mV. When the concentrations (0.5 M to 4 M) of TMA-Cl rose, the opening current accordingly increased from 40 ± 2 pA to 120 ± 10 pA at +120 mV. The change of open channel current value was in agreement with the measured conductance (Figure S1, Supporting Information). TMA-Cl as a DNA Speed Bump in the Nanopore. To evaluate the effect of TMA-Cl on ssDNA translocation speed, we primarily tested the translocations of different homopolymers such as poly(dA)25, poly(dT)25, and poly(dC)25 through WT α-HL nanopore in 1 M KCl and 1 M TMA-Cl. After changing KCl to TMA-Cl, we clearly observed that the translocation speed of ssDNA slowed (Figure 2a,b). Since the translocation speed of ssDNA molecules is about 1−20 μs/base in KCl/NaCl at +120 mV, poly(dA)25, poly(dT)25, and poly(dC)25 translocate through α-HL nanopores within 500 μs (0.5 ms) under standard experimental conditions. So we divided the signals into two parts according to the translocation time of ssDNA events: long events (≥0.5 ms) and short events (