b Department of Laboratory Medicine, West China Hospital, Sichuan University and Collaborative Innovation Center, Chengdu 610041, China
Human telomeric G-quadruplex has been attracting intense research attention as drug targets to halt the function of telomerase and regulate gene expression because induction/ stabilization of the G-quadruplex form directly affects the activity of telomerase [1, 2]. Molecular ligands that selectively stabilize Gquadruplex foldings/topologies can be potential therapeutic agents for anticancer treatment [3, 4]. Therefore, figuring out the interactions between G-quadruplex ligands and human telomere sequence are important for the design of new drugs and sensors. Thioflavin T (ThT) is a commercial G-quadruplex ligand that can bind G-quadruplex with a fluorescent light-up signal change and high specificity against DNA duplex [5]. Fluorescence spectra [6], circular dichroism (CD) spectroscopy [7], melting temperature (Tm) experiments [5] and electrospray ionization mass spectrometry (ESI-MS) [8] analyses have been carried out to characterize the binding behaviors. The reported stoichiometric ratio of Gquadruplex to ThT was 1:2 [5]. Shao et al. demonstrated that the ThT binding does not disturb native G-quadruplex structures preformed in Na+ or K+ solutions, but the binding model is strongly dependent on the G-quadruplex structures [9]. When binding to the G-quadruplex possessing hybrid structures, the excited state of ThT has a significant fluorescence enhancement because of the rotation restriction of benzothiazole (BZT) and dimethylaminobenzene (DMAB) rings, which can provide aid in the structure identification of G-quadruplexes between hybrid and other structures (parallel and antiparallel). However, other groups reported that ThT is an efficient inducer for G-quadruplex. For example, Jyotirmayee Mohanty's group used fluorescence spectra and Tm experiments to support that ThT induced an apparent switch over of the hybrid G-quadruplex structure to the antiparallel form [5]. Gabelica et al. carried out fluorescent titrations and ESI-MS analyses for demonstrating that the dissociation constants are sensitive to the base sequence and to the G-quadruplex structure [8]. Although various techniques of bulk studies have been used to investigate the interactions between G-quadruplex and ThT, single-molecule details concerning G-quadruplex structural changes and binding behaviors are still desired.
Nanopore technology utilizing a molecular-scale pore structure for single-molecule study is very sensitive for detecting nucleic acid sequence and its interactions with binding ligands [10-14]. Compared with other single-molecule technologies such as singlemolecule fluorescence resonance energy transfer (FRET) spectroscopy [15] and optical tweezers [16], the nanopore technology overcomes the limitations of time-consuming, costly fluorescent labeling and costly chemical modification. α-Hemolysin (α-HL) nanopore can be used to directly investigate the folding and unfolding of a G-quadruplex [17, 18]. The bacterial protein α-HL can self-assemble across a lipid bilayer membrane and form a mushroom-shaped nanopore with the smallest diameter of ~1.4 nm [19]. When nucleic acid molecule is electrophoretically driven through the pore, ionic current will be interrupted and this perturbation can be recorded as characteristic electric signals, which can reveal structural information of the individual nucleic acid molecule. The α-HL nanopore has exhibited many advantages in single-molecule characterization of various G-quadruplex structures induced by small molecules such as PDS [20], ATP [21], Pb2+ and Ba2+ [22] based on the pore blockage properties. For example, a previous research successfully utilizes the electrochemical confined space of nanopore, which could efficiently convert single molecule characteristics into measurable electrochemical signals with high signal-to-noise ratio [23, 24], to achieve the discrimination of the conformational changes of the Gquadruplex structure [21]. Therefore, extending this strategy for further studying the interactions of G-quadruplex with ThT, a highly selective G-quadruplex ligand, as well as the structural change of G-quadruplex induced by ThT is of great importance.
In this work, we utilized the α-HL nanopore to analyze the interactions between human telomere sequence and G-quadruplex ligand ThT at the nanometer scale. Besides the electrolyte solution containing 50 mmol/L KCl and 450 mmol/L NaCl, the effects of metal ion (K+ and Na+) and pH on the translocation behaviors were studied. By analyzing the translocation event and translocation time, the interactions between G-quadruplex and ThT at the single-molecule level were studied without labeling, amplification, or ligand/receptor titration.
In order to characterize the single-molecule interaction behaviors of G-quadruplex with ThT, the signal properties of ThT, Tel DNA and Tel DNA/ThT complex interacting with the α-HL nanopore in the electrolyte solution containing 50 mmol/L KCl and 450 mmol/L NaCl were captured, respectively (Fig. 1). A human telomere sequence 5'-AGGG(TTAGGG)3-3' linked with a poly(dA)25 single-stranded overhang at the 5' end is designed as the Tel DNA. The 22-nt human telomere sequence in the Tel DNA has been reported to have the capability to fold into a three-tetrad, twocation G-quadruplex in the presence of Na+ [25] or/and K+ [26]. The addition of 25-mer poly-dA nucleotide tail could facilitate both the entrance of the G-quadruplex to the vestibule of α-HL nanopore and the translocation process in the nanopore [14, 21]. Considering the concentration difference between K+ and Na+ in the physiological condition inside cells, the 50 mmol/L Tris-HCl buffer (pH 7.2) containing 50 mmol/L KCl and 450 mmol/L NaCl was used as the electrolyte solution for single-molecule study.
|
Download:
|
| Fig. 1. The characterization of G-quadruplex binding with ThT through α-HL nanopore in the electrolyte solution containing 50 mmol/L KCl and 450 mmol/L NaCl. [A–C] (a) Schematic illustration of ThT, Tel DNA and Tel DNA/ThT complex. (b) Illustration of α-HL nanopore inserted in a lipid bilayer. [A–C] (c) Representative single-channel current traces of the translocation of ThT, Tel DNA and Tel DNA/ThT complex. All experiments were conducted at the transmembrane potential of +160 mV. The concentrations of Tel DNA and ThT were 0.25mmol/L and 2.0 mmol/L, respectively. Each experiment was performed three times independently | |
Single-channel recordings revealed that ThT had no observable ionic current blockade at 160 mV (Fig. 1A-c). Because of carrying positive charges, ThT would be driven away from the nanopore side to the bulk solution in the cis chamber under a positive voltage. When the Tel DNA was added in the electrolyte solution without ThT, a burst of translocation events were observed with an average dwell time of 1708.7 ± 56.7 ms (Fig. 1B-c). However, the linear DNA strand has a very short dwell time less than about 6 ms [17, 18, 27]. This long dwell time can be attributed to the G-quadruplex folding models of the Tel DNA in the presence of KCl or/and NaCl that have been previously reported [28, 29]. Furthermore, in the presence of both ThT and Tel DNA in the electrolyte solution, prolonged translocation events were observed in the current traces (Fig. 1C-c). The dwell time 2966.8 ± 313.6 ms was about 1.7 times longer than that of Tel DNA under the same transmembrane voltage. The drastically prolonged dwell time indicates that the unraveling process needs more time to overcome energy barrier, which suggests that the Tel DNA does bind with ThT to form stabilized Tel DNA/ThT complex [5].
To further evaluate the formation of Tel DNA/ThT complex, CD measurements of Tel DNA and Tel DNA/ThT complex were performed in the electrolyte solution containing 50 mmol/L KCl and 450 mmol/L NaCl (Fig. 3C). The Tel DNA displayed a positive peak at ~295 nm and a negative peak at ~265 nm in the electrolyte solution, suggesting the formation of antiparallel G-quadruplex [5]. After adding ThT, the positive peak intensity at ~295 nm increased while the negative peak intensity at ~265 nm slightly decreased, indicating the interactions of Tel DNA with ThT. These results are consistent with the G-quadruplex translocations where the dwell time of Tel DNA/ThT complex is much longer than that of Tel DNA.
|
Download:
|
| Fig. 3. (A) Histogram of dwell time under various metal ion concentrations for the Tel DNA and Tel DNA/ThT complex at 160 mV. The concentrations of Tel DNA and ThT were 0.25 mmol/L and 2.0 mmol/L, respectively. (B-D) CD spectra recorded for the Tel DNA and Tel DNA/ThT complex at different metal ion concentrations. The concentrations of Tel DNA and ThT were 6.25 mmol/L and 50 mmol/L, respectively. All experiments were conducted in 50 mmol/L Tris-HCl buffer (pH 7.2) containing (B) 500 mmol/L NaCl, (C) 50 mmol/L KCl and 450 mmol/L NaCl, (D) 500 mmol/L KCl, respectively. Each experiment was performed three times independently | |
To further study the translocation behaviors of the Tel DNA/ThT complex, different transmembrane potentials of 140, 160, 180, 200 and 220 mV were applied. As shown in Fig. 2 and Fig. S1 in Supporting information, the corresponding dwell time was apparently voltage-dependent. With the increase of the transmembrane potential from 140 mV to 220 mV, the dwell time was significantly shortened from 3523.3± 267.8 ms to 907.8 ± 76.5 ms. Obviously, the Tel DNA/ThT complex was captured by the vestibule and then translocated through the β-barrel to exit on the trans side. Compared to Tel DNA, the Tel DNA/ThT complex has a longer dwell time but similar current blockade under the given transmembrane potential (Table S1 in Supporting information). The slow unraveling process indicates that ThT possesses the capability to increase mechanical stability of the telomeric Gquadruplex structures. This result is consistent with previously reported Tm experiments that the Tm of the telomeric Gquadruplex increased upon the addition of ThT [5]. No change of current blockades could be attributed to that the ThT binding cannot alter the dimension of G-quadruplex in the electrolyte solution containing 50 mmol/L KCl and 450 mmol/L NaCl [20].
|
Download:
|
| Fig. 2. The effect of transmembrane potential on the dwell time of the G-quadruplex in the electrolyte solution containing 50 mmol/L KCl and 450 mmol/L NaCl. (A) Representative single-channel current traces for Tel DNA/ThT complex translocating through α-HL nanopore with the transmembrane potential of 140, 160, 180, 200 and 220 mV. (B) Histogram of dwell time under various applied potentials (trans vs. cis) for the Tel DNA and Tel DNA/ThT complex. The concentrations of Tel DNA and ThT were 0.25 mmol/L and 2.0 mmol/L, respectively. Each experiment was performed three times independently | |
Since metal ions such as K+ and Na+ are the most abundant cations in cells, the studies of their effects on the dwell time are desired. First, high concentration NaCl solution (500 mmol/L) was tested. As shown in Fig. 3A and Fig. S2 in Supporting information, the Tel DNA in the NaCl solution containing ThT exhibited the average dwell time of 976.2 ± 65.9 ms and current blockade I/I0 of 80.2% ± 0.3% (Table S2 in Supporting information). This dwell time is similar to that of the Tel DNA without ThT (1032.9 ± 71.3 ms), suggesting a low binding possibility between Tel DNA and ThT in 500 mmol/L NaCl solution. In addition, the presence of ThT does not change the CD curve (Fig. 3B) of Tel DNA or fluorescence intensity (Fig. S4 in Supporting information) of ThT, which further supports that Tel DNA can not bind with ThT in 500 mmo/L NaCl solution.
In contrast, in 500 mmol/L KCl solution, the dwell time of Tel DNA was significantly enhanced from 9742.1 ±1221.2 ms to 13734.2 ± 2221.9 ms after adding ThT (Fig. 3A and Fig. S2). And its corresponding current blockade changed from 76.6% ± 0.7% to 79.4% ± 1.9% (Table S2). This indicates that ThT can not only bind with Tel DNA but also alter the dimension of the G-quadruplex structure. CD spectra support the binding of Tel DNA with ThT (Fig. 3D). In 500 mmol/L KCl solution, the G-quadruplex, having a positive CD peak at ~295 nm and a multi-shouldered band at around ~265 nm, keeps its hybrid structure [5]. For the Tel DNA/ ThT complex, the broad multi-shouldered peak disappeared but a characteristic CD peak at ~295 nm presented as well as a negative peak at ~265 nm did, which are corresponded to an antiparallel topology [5]. Herein, the presence of K+ plays an important role in promoting the binding of ThT with Tel DNA and the enhanced structural stability of the Tel DNA/ThT complex. The unraveling process of the Tel DNA/ThT complex within nanopore needs more time to overcome energy barrier caused by different G-quadruplex structure, which was also confirmed by the CD spectra change (Fig. 3C) and the fluorescence intensity enhancement (Fig. S4).
To explore the interaction mechanism of the Tel DNA with ThT, the effect of pH on the dwell time was studied in detail. First, it was demonstrated that the Tel DNA had a relative long dwell time in neutral or alkaline buffer. As shown in Fig. 4A and Fig. S3 in Supporting information, as the pH values of Tris-HCl buffers were changed from 8.0 to 7.2 and to 6.0, the corresponding dwell time of Tel DNA were 1787.3 ±10.6 ms, 1708.7 ± 56.7 ms and 966.7 ± 20.7 ms, respectively. The possible reason is that the increased proton numbers in weak acidic buffer promote the protonation of the carbonyl oxygen of guanine, thus weakens the electrostatic interaction between central carbonyl-lined channel of Tel DNA and monovalent cation, and then leads to a less stable Gquadruplex structure [30]. The current blockades of Tel DNA were 81.0% ± 1.6% (pH 8.0), 83.8% ± 0.5% (pH 7.2) and 86.1% ± 1.9% (pH 6.0), respectively (Fig. 4B), which indicate that the structure form of Tel DNA could also be altered by the change of pH values.
|
Download:
|
| Fig. 4. The effect of pH on the dwell time (A) and current blockade (B) of the Tel DNA and Tel DNA/ThT complex at 160 mV. All experiments were conducted in 50 mmol/L Tris-HCl buffer containing 50 mmol/L KCl and 450 mmol/L NaCl, with different pH values of 8.0, 7.2 or 6.0. The concentrations of Tel DNA and ThT were 0.25 mmol/L and 2.0 mmol/L, respectively. Each experiment was performed three times independently | |
After adding ThT, the longest dwell time of 2966.8 ± 313.6 ms as well as the largest dwell time increase was observed in the solution containing 50 mmol/L KCl and 450 mmol/L NaCl (pH 7.2). In control experiments, the dwell time value had a little increase at pH 8.0 and had no increase at pH 6.0, respectively. It is concluded that pH 7.2 is the best condition for the Tel DNA binding with ThT and having a stabilized structural form. Current blockade value reflects the dimension of nucleic acid structure inside the nanometer scale vestibule [20, 31]. As shown in Fig. 4B, the current blockade has no change at pH 7.2 but increases significantly at either pH 8.0 or pH 6.0. These indicate that Tel DNA can bind with ThT having a high stability but keeping its structural model unchanged at pH 7.2. In addition, although the structural stability of Tel DNA was not increased by the addition of ThT at pH 8.0 or at pH 6.0, its structural model was changed. The specific translocation details are shown in Table S3 in Supporting information. Consequently, ThT can bind with Tel DNA having various G-quadruplex forms, but its corresponding binding models are different. Although fluorescence spectra could provide information on the binding differences (Fig. S5 in Supporting information), this nanopore strategy is the first report of the direct proof of the G-quadruplex model change by ThT.
In conclusion, the interactions between human telomere sequence and a typical highly selective G-quadruplex ligand ThT were studied at the single-molecule level through α-HL protein nanopore. The effects of voltage, metal ion and pH on the translocation behaviors of G-quadruplex with or without ThT were investigated. As data shown, K+ plays an important role in promoting the binding of Tel DNA with ThT. In the presence of K+, the prolonged dwell time and the enhanced current blockade indicate the change of G-quadruplex structure to a stabilized form. Control experiments were carried out and further confirmed that the binding models between Tel DNA and ThT were different in acidic or alkaline buffer. In summary, the single-molecule nanopore platform provides great potential for studying the interactions between G-quadruplex and small molecule ligands, without the need of labeling, amplification, or ligand/receptor titration. There will be wide applications for the design of new drugs and sensors.
AcknowledgmentsThis work was financially supported by the National Natural Science Foundation of China (No. 21475091) and the Science and Technology Department of Sichuan Province (No. 2015GZ0301). We thank Professor Yi-Tao Long's group from East China University of Science and Technology for their supports of nanopore detection device and kindly providing the data analysis software (http://people.bath.ac.uk/yl505/nanoporeanalysis.html). We also appreciate the support of Professor Cheng Yang from Sichuan University for CD measurements.
Appendix A. Supplementary dataSupplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cclet.2017.09.010.
| [1] |
S. Balasubramanian, S. Neidle, Curr. Opin. Chem. Biol. 13 (2009) 345-353. DOI:10.1016/j.cbpa.2009.04.637 |
| [2] |
N.W. Kim, M.A. Piatyszek, K.R. Prowse, et al., Science 266 (1994) 2011-2015. DOI:10.1126/science.7605428 |
| [3] |
A. Marchand, A. Granzhan, K. Iida, et al., J. Am. Chem. Soc. 137 (2015) 750-756. DOI:10.1021/ja5099403 |
| [4] |
J.M. Nicoludis, S.T. Miller, P.D. Jeffrey, et al., J. Am. Chem. Soc. 134 (2012) 20446-20456. DOI:10.1021/ja3088746 |
| [5] |
J. Mohanty, N. Barooah, V. Dhamodharan, et al., J. Am. Chem. Soc. 135 (2012) 367-376. |
| [6] |
A.R. de la Faverie, A. Guedin, A. Bedrat, L.A. Yatsunyk, J.L. Mergny, Nucleic Acids Res. 42 (2014) 1-8. DOI:10.1093/nar/gkt1324 |
| [7] |
D. Zhao, X.W. Dong, N. Jiang, D. Zhang, C.L. Liu, Nucleic Acids Res. 42 (2014) 11612-11621. DOI:10.1093/nar/gku833 |
| [8] |
V. Gabelica, R. Maeda, T. Fujimoto, et al., Biochemistry 52 (2013) 5620-5628. DOI:10.1021/bi4006072 |
| [9] |
L.L. Liu, Y. Shao, J. Peng, et al., Anal. Chem. 86 (2014) 1622-1631. DOI:10.1021/ac403326m |
| [10] |
C. Cao, Y.L. Ying, Z.L. Hu, et al., Nat. Nanotechnol. 11 (2016) 713-718. DOI:10.1038/nnano.2016.66 |
| [11] |
Y.L. Ying, J.J. Zhang, R. Gao, Y.T. Long, Angew. Chem. Int. Ed. 52 (2013) 13154-13161. DOI:10.1002/anie.201303529 |
| [12] |
J. Geng, S. Wang, H. Fang, P.X. Guo, ACS Nano 7 (2013) 3315-3323. DOI:10.1021/nn400020z |
| [13] |
J. Geng, K. Kim, J. Zhang, et al., Nature 514 (2014) 612-615. DOI:10.1038/nature13817 |
| [14] |
N. An, A.M. Fleming, C.J. Burrows, J. Am. Chem. Soc. 135 (2013) 8562-8570. DOI:10.1021/ja400973m |
| [15] |
S. Hohng, S. Lee, J. Lee, M.H. Jo, Chem. Soc. Rev. 43 (2014) 1007-1013. DOI:10.1039/C3CS60184F |
| [16] |
D. Koirala, S. Dhakal, B. Ashbridge, et al., Nat. Chem. 3 (2011) 782-787. DOI:10.1038/nchem.1126 |
| [17] |
J.W. Shim, L.Q. Gu, J. Phys. Chem. B 112 (2008) 8354-8360. DOI:10.1021/jp0775911 |
| [18] |
J.W. Shim, Q.L. Tan, L.Q. Gu, Nucleic Acids Res. 37 (2009) 972-982. DOI:10.1093/nar/gkn968 |
| [19] |
L.Z. Song, M.R. Hobaugh, C. Shustak, et al., Science 274 (1996) 1859-1865. DOI:10.1126/science.274.5294.1859 |
| [20] |
L. Zhang, K.X. Zhang, S. Rauf, et al., Anal. Chem. 88 (2016) 4533-4540. DOI:10.1021/acs.analchem.6b00555 |
| [21] |
Y.L. Ying, H.Y. Wang, T.C. Sutherl, Y.T. Long, Small 7 (2011) 87-94. DOI:10.1002/smll.v7.1 |
| [22] |
C. Yang, L. Liu, T. Zeng, et al., Anal. Chem. 85 (2013) 7302-7307. DOI:10.1021/ac401198d |
| [23] |
Y.L. Ying, Y.T. Long, Sci. China Chem. 60 (2017) 1187-1190. DOI:10.1007/s11426-017-9082-5 |
| [24] |
R. Gao, Y.L. Ying, Y.X. Hu, Y.J. Li, Y.T. Long, Anal. Chem. 89 (2017) 7382-7387. DOI:10.1021/acs.analchem.7b00729 |
| [25] |
Y. Wang, D.J. Patel, Structure 1 (1993) 263-282. DOI:10.1016/0969-2126(93)90015-9 |
| [26] |
J.X. Dai, C. Punchihewa, A. Ambrus, et al., Nucleic Acids Res. 35 (2007) 2440-2450. DOI:10.1093/nar/gkm009 |
| [27] |
Y.L. Ying, D.W. Li, Y. Li, J.S. Lee, Y.T. Long, Chem. Commun. 47 (2011) 5690-5692. DOI:10.1039/c0cc05787h |
| [28] |
K.W. Lim, S. Amrane, S. Bouaziz, et al., J. Am. Chem. Soc. 131 (2009) 4301-4309. DOI:10.1021/ja807503g |
| [29] |
A. Ambrus, D. Chen, J. Dai, et al., Nucleic Acids Res. 34 (2006) 2723-2735. DOI:10.1093/nar/gkl348 |
| [30] |
P. Murat, Y. Singh, E. Defrancq, Chem. Soc. Rev. 40 (2011) 5293-5307. DOI:10.1039/c1cs15117g |
| [31] |
N. An, A.M. Fleming, E.G. Middleton, C.J. Burrows, Proc. Natl. Acad. Sci. U. S. A. 111 (2014) 14325-14331. DOI:10.1073/pnas.1415944111 |
2018, Vol. 29 

