Kinetic and Thermodynamic Characterization of Telomeric G

Aug 12, 2008 - Telomeres, the ends of eukaryotic chromosomes, are special. DNA-protein assemblies that ... degradation.1 Human telomeres comprise telo...
0 downloads 0 Views 521KB Size
Download PDF
Anal. Chem. 2008, 80, 6935–6941

Kinetic and Thermodynamic Characterization of Telomeric G-Quadruplex by Nonequilibrium Capillary Electrophoresis: Application to G-Quadruplex/Duplex Competition Youzhi Xu, Xiaojun Feng, Wei Du, Xin Liu, Qingming Luo, and Bi-Feng Liu* The Key Laboratory of Biomedical Photonics of MOE s Hubei Bioinformatics and Molecular Imaging Key Laboratory s Division of Biomedical Photonics at Wuhan National Laboratory for Optoelectronics, Department of Systems Biology, College of Life Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, China In this paper, nonequilibrium capillary electrophoresis (NECE) was attempted for the first time to investigate a dual equilibrium system, where the intramolecular Gquadruplex folding was in competition with the intermolecular duplex formation. Samples of an equilibrium mixture of human telomeric DNA and its complementary strands were separated in capillaries under nonequilibrium conditions without K+. Polyethylene oxide was added to the running buffer facilitating the separation of single-stranded DNA, duplex, and G-quadruplex. Thus, the folding/unfolding rate constants of the G-quadruplex and the association/dissociation constants of the duplex could be simultaneously derived from the same experiment. Results indicated that the duplex formation induced minimal influence on the G-quadruplex folding. On the basis of the kinetic characterization of the G-quadruplex at varying temperatures, the thermodynamic parameters of the G-quadruplex could also be determined. Thus, the NECE method provided a new avenue for studying the kinetics and thermodynamics of nucleic acids within dual equilibrium systems with significant advantages of extremelow sample cost (∼10-18 mol) and high repeatability. Telomeres, the ends of eukaryotic chromosomes, are special DNA-protein assemblies that protect cells from recombination and degradation.1 Human telomeres comprise telomeric binding proteins and noncoding repeat sequences of guanine-rich DNA,2,3 which can readily fold into a four-stranded structure known as the G-quadruplex through a Hoogsteen hydrogen bond in the presence of monovalent salts such as K+, Na+, Li+, and NH4+.4,5 * Corresponding author. Address: The Key Laboratory of Biomedical Photonics of Ministry of Education, Huazhong University of Science and Technology, Wuhan 430074, China. E-mail: [email protected]. Phone: +86-27-87792203. Fax: +86-27-87792170. (1) Blackburn, E. H. Cell 2001, 106, 661–673. (2) Moyzis, R. K.; Buckingham, J. M.; Cram, L. S.; Dani, M.; Deaven, L. L.; Jones, M. D.; Meyne, J.; Ratliff, R. L.; Wu, J. R. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 6622–6626. (3) de Lange, T. Genes Dev. 2005, 19, 2100–2110. (4) Williamson, J. R.; Raghuraman, M. K.; Cech, T. R. Cell 1989, 59, 871– 880. (5) Parkinson, G. N.; Lee, M. P.; Neidle, S. Nature 2002, 417, 876–880. 10.1021/ac801335y CCC: $40.75  2008 American Chemical Society Published on Web 08/12/2008

The G-quadruplex structures are found essential for maintaining the telomere ends.6,7 Disruption of this secondary structure leads to DNA degradation and eventual cell death, which can be explored for potential therapeutic applications, such as cancer treatment.8,9 Moreover, similar G-quadruplex forming sequences have also been found in many other vital regions of the chromosome such as the promoter of oncogenes,10,11 the immunoglobulin switch regions,12 and the regulatory regions of insulin13 indicating the involvement of G-quadruplex in various biological processes. To study the function of G-quadruplex, it is important to understand how fast it assembles. The kinetics of G-quadruplex formation can be quantitatively described by its folding/unfolding rate constants. However, measurement of these kinetic parameters is usually difficult especially in the case of fast intramolecular conformation interconversion. In addition, the presence of complementary strands may induce the competition between the formation of the DNA duplex and G-quadruplex, generating further complexity for such measurement. Traditional methods like polyacrylamide gel electrophoresis (PAGE),14,15 nuclear magnetic resonance (NMR),16 and fluorescence resonance energy transfer (FRET)17 are relatively laborious and time-consuming. Recently, Zhao and co-workers18 demonstrated the use of surface plasmon resonance (SPR) for fast determination of all the kinetic param(6) Zahler, A. M.; Williamson, J. R.; Cech, T. R.; Prescott, D. M. Nature 1991, 350, 718–720. (7) Fletcher, T. M.; Sun, D.; Salazar, M.; Hurley, L. H. Biochemistry 1998, 37, 5536–5541. (8) Neidle, S.; Parkinson, G. Nat. Rev. Drug Discovery 2002, 1, 383–393. (9) Pendino, F.; Tarkanyi, I.; Dudognon, C.; Hillion, J.; Lanotte, M.; Aradi, J.; Segal-Bendirdjian, E. Curr. Cancer Drug Targets 2006, 6, 147–180. (10) Simonsson, T.; Pecinka, P.; Kubista, M. Nucleic Acids Res. 1998, 26, 1167– 1172. (11) Siddiqui-Jain, A.; Grand, C. L.; Bearss, D. J.; Hurley, L. H. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 11593–11598. (12) Sen, D.; Gilbert, W. Nature 1988, 334, 364–366. (13) Hammond-Kosack, M. C.; Dobrinski, B.; Lurz, R.; Docherty, K.; Kilpatrick, M. W. Nucleic Acids Res. 1992, 20, 231–236. (14) Boles, T. C.; Hogan, M. E. Biochemistry 1987, 26, 367–376. (15) De Cian, A.; Mergny, J. L. Nucleic Acids Res. 2007, 35, 2483–2493. (16) Kato, Y.; Ohyama, T.; Mita, H.; Yamamoto, Y. J. Am. Chem. Soc. 2005, 127, 9980–9981. (17) Green, J. J.; Ying, L.; Klenerman, D.; Balasubramanian, S. J. Am. Chem. Soc. 2003, 125, 3763–3767. (18) Zhao, Y.; Kan, Z. Y.; Zeng, Z. X.; Hao, Y. H.; Chen, H.; Tan, Z. J. Am. Chem. Soc. 2004, 126, 13255–13264.

Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

6935

eters of the human telomeric G-quadruplex. In this method, singlestranded human telomeric sequences (TTAGGG)4 were immobilized on an optical sensor chip surface where complementary strands (CCCTAA)4 were introduced for hybridization reactions. With the time course of duplex formation being monitored, both the folding/unfolding rate constants of the G-quadruplex and the association/dissociation rate constants of the duplex could be derived at the same time. Although the SPR-based method was conducted at the physiological concentration of K+, the immobilization of telomeric sequences on sensor chip surface could not mimic the actual situation of free telomeric DNA in the liquid intracellular environment. Furthermore, the nonspecific binding of targets and degradation of immobilized probes also limited its application entailing inventions of new analytical approaches. As one of the most versatile techniques, capillary electrophoresis has already been playing an essential role in nucleic acid analysis. For instance, capillary array electrophoresis19-23 has been the golden standard for genome sequencing purposes in the past decade. Affinity capillary electrophoresis has been widely used for DNA-protein interaction studies under equilibrium conditions.24-29 Recently, a new approach of nonequilibrium capillary electrophoresis (NECE) was introduced by the research group of Krylov30 to investigate the kinetics of DNA-protein interactions. In their method, a mixture of DNA and proteins was injected into a capillary for separation based on different massto-charge ratios under nonequilibrium conditions using a running buffer without the equilibrium mixture. The DNA-protein complex decayed during the NECE separation, resulting in electropherograms containing characteristic features. Thus, the association/dissociation constants of the DNA-protein complex could be consequently derived from the electropherogram data. Several more examples of NECE for DNA-protein interaction studies31-34 could also be found lately. In NECE, target molecules were separated in liquid solution without immobilization, which represented an experimental condition closer to the in vivo environment of natural occurring biological macromolecules in comparison to the SPR-based approach. Therefore, NECE is potentially a feasible method for enhanced kinetic characterization of nucleic acids such as human telomeric sequences. (19) Mathies, R. A.; Huang, X. C. Nature 1992, 359, 167–169. (20) Huang, X. C.; Quesada, M. A.; Mathies, R. A. Anal. Chem. 1992, 64, 2149– 2154. (21) Woolley, A. T.; Mathies, R. A. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 11348– 11352. (22) Kheterpal, I.; Mathies, R. A. Anal. Chem. 1999, 71, 31A–37A. (23) Dovichi, N. J.; Zhang, J. Z. Angew. Chem., Int. Ed. 2000, 39, 4463–4468. (24) Charles, J. A. M.; McGown, L. B. Electrophoresis 2002, 23, 1599–1604. (25) Deng, Q.; German, I.; Buchanan, D.; Kennedy, R. T. Anal. Chem. 2001, 73, 5415–5421. (26) German, I.; Buchanan, D. D.; Kennedy, R. T. Anal. Chem. 1998, 70, 4540– 4545. (27) Kotia, R. B.; Li, L. J.; McGown, L. B. Anal. Chem. 2000, 72, 827–831. (28) Wan, Q. H.; Le, X. C. Anal. Chem. 2000, 72, 5583–5589. (29) Wang, H. L.; Xing, J.; Tan, W.; Lam, M.; Carnelley, T.; Weinfeld, M.; Le, X. C. Anal. Chem. 2002, 74, 3714–3719. (30) Berezovski, M.; Krylov, S. N. J. Am. Chem. Soc. 2002, 124, 13674–13675. (31) Berezovski, M.; Nutiu, R.; Li, Y.; Krylov, S. N. Anal. Chem. 2003, 75, 1382– 1386. (32) Berezovski, M.; Drabovich, A.; Krylova, S. M.; Musheev, M.; Okhonin, V.; Petrov, A.; Krylov, S. N. J. Am. Chem. Soc. 2005, 127, 3165–3171. (33) Berezovski, M.; Krylov, S. N. Anal. Chem. 2005, 77, 1526–1529. (34) Berezovski, M.; Musheev, M.; Drabovich, A.; Krylov, S. N. J. Am. Chem. Soc. 2006, 128, 1410–1411.

6936

Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

Since the kinetic study of telomeric G-quadruplex formation involves only complementary DNA strands and intramolecular conformation interconversion, it is difficult to apply separation based on mass-to-charge ratios. Besides, two equilibriums exist in the quadruplex/duplex competition scenario instead of one in the previous DNA-protein interactions studies.30-34 Here, we successfully addressed above issues by using capillary nongel electrophoresis as the separation mechanism and demonstrated for the first time the use of NECE for investigating the kinetics and thermodynamics of the human telomeric G-quadruplex within the dual equilibrium system of quadruplex/duplex competition. In our method, equilibrium mixtures of telomeric DNA (TTAGGG)4 and its complementary strands (CCCTAA)4 were injected into capillaries for separation under nonequilibrium conditions (without K+) using a running buffer containing polyethylene oxide (PEO). The presence of PEO facilitated the separation among single-stranded DNA (ssDNA), duplex, and G-quadruplex, resulting in electropherograms with characteristic features. Thus, the folding/unfolding rate constants of the Gquadruplex and the association/dissociation constants of the duplex could be derived simultaneously from the same experiment. Results indicated that the duplex formation induced minimal influence on the G-quadruplex folding. By investigation of the kinetics of the G-quadruplex at varying temperatures in the absence of complementary DNA strands, both the kinetic and thermodynamic parameters of the G-quadruplex could be determined using NECE. Moreover, a DNA sample amount as low as ∼10-18 mol was sufficient for analysis. EXPERIMENTAL DETAILS Chemicals and Oligonucleotides. Tris(hydroxymethyl)aminomethane (Tris), NaB4O3 · 10H2O, HBO3, HCl, NaOH, acrylamide, potassium chloride, ethylenediaminetetraacetic acid (EDTA), formaldehyde, and fluorescein were purchased from Tianjing Chemical Co. Ltd. (Tianjing, China). Polyethylene oxide (PEO, MWav 8M), γ-methacryloxypropyltrimethoxy silane (MAPS, a bifunctional-group reagent), N,N,N′,N′-tetramethylethylenediamine (TEMED), and ammonium persulfate were purchased from SigmaAldrich (MO). All reagents were of analytical grade unless specified otherwise. Water was purified by the Millipore-Q system (Millipore) before use for the preparation of all solutions. Samples and all buffer solutions were autoclaved (121 °C, 0.12 MPa) and filtrated (0.45 µm microporous membrane filtration) before experiments. Oligodeoxynucleotides, 5′-TTAGGGTTAGGGTTAGGGTTAGGG-3′ (ssDNAG) and 5′-CCCTAACCCTAACCCTAACCCTAA-3′ (ssDNAC), were synthesized and purified (HPLC grade) by TaKaRa Biotechnology Co. Ltd. (Dalian, China). ssDNAG was 5′tagged with 6-carboxyfluorescein (6-FAM) for laser-induced fluorescence detection. ssDNA stock solutions were prepared with pure water. Equipments and Capillary Preparation. Separations were carried out on the PACE 5500 system (Beckman-Coulter, Fullerton, CA) equipped with a high-voltage power supply (0-30 kV), a laser-induced fluorescence detector (LIF), an argon ion laser source (30 mW), and a temperature control system. Uncoated fused-silica capillaries (i.d. 75 µm, o.d. 375 µm) were purchased from Yongnian Optical Fiber Factory (Hebei Province, China) and

Figure 2. Schematic illustration of NECE separation of equilibrium mixtures containing four equilibrium fractions: free ssDNAG, free ssDNAC, duplex, and G-quadruplex. The top panel shows the spatial distribution of the separated components in the capillary at different times (t0 ) 0, t2 > t1 > t0). The graph at the bottom shows a schematic electropherogram with the detected fluorescence intensity of the separated components as functions of the position in the capillary at time t2. Only ssDNAG is labeled with a fluorescence tag, thus the migration of ssDNAC does not contribute to the electropherogram. A1-A5 indicates the five characteristic features on the electropherogram corresponding to the G-quadruplex, ssDNAG unfolded from the G-quadruplex, free ssDNAG, ssDNAG dissociated from the duplex, and duplex, respectively. Figure 1. (A) NECE electropherograms of four equilibrium mixtures: (a) 3.5 × 10-7 M ssDNAG, 1.4 × 10-7 M ssDNAC, and 150 mM K+; (b) 3.5 × 10-7 M ssDNAG and 150 mM K+; (c) 3.5 × 10-7 M ssDNAG; (d) 3.5 × 10-7 M ssDNAG and 3.5 × 10-7 M ssDNAC. Separation conditions are described in the Experimental Details section. (B) Enlarged view of the electropherogram (a) in part A. Numbers indicate the five characteristic features on the electropherogram, which are identified as G-quadruplex (A1), ssDNAG unfolded from the Gquadruplex (A2), free ssDNAG (A3), ssDNAG dissociated from duplex (A4), and duplex (A5).

treated following the protocol introduced by S. Hjerte´n et al.35 before use. In brief: (a) Capillaries were first cleaned by sequentially perfusing 0.1 M NaOH, pure water, and methanol (anhydrous) for 120 min each and blow dried with N2. (b) MAPS solution (1:1 in methanol) was then filled into the capillaries (both ends of the capillary were sealed with Teflon tubes) for derivatization at room temperature for 12 h. (c) Capillaries were then derivatized with a polyacrylamide layer by filling with 4% acrylamide solution containing 0.1% (w/v) APS and 0.1% (v/v) TEMED overnight. (d) After the capillaries were rinsed thoroughly with pure water for 120 min to remove the excessive polyacrylamide, they were treated with 37% formaldehyde solution (pH 10.0) for 12 h to cross-link the polyacrylaminde layer. (e) Finally, the capillaries were soaked in pure water for storage after rinsing and cleaning. Sample Preparation and NECE. DNA equilibrium mixtures were prepared with different volume ratios of ssDNAG and ssDNAC in 1× TE buffer solution (pH 7.9); ssDNAG (5 × 10-5 M stock solution) was first diluted in 250 µL 1× TE buffer and denaturized (35) Hjerten, S. J. Chromatogr., A 1985, 347, 191–198.

Table 1. Formulas for Calculating Equilibrium Constants and Kinetic Parametersa G-quadruplex KF ) A1 + A2/A3 ku ) ln (A1 + A2/A1)/tQ kf ) KF(ku)

duplex KA ) (A4 + A5)/A3[ssDNAC0 a(A4 + A5)] kd ) ln (A4 + A5/A5)/tD ka ) KA(kd)

a kf/ku are the folding/unfolding rate constants of the G-quadruplex; ka/kd are the association/dissociation constants of the duplex; KF and KA are the equilibrium constants of G-quadruplex and duplex, respectively; A1-A5 are areas of the characteristic features in the electropherogram of Figure 2; tQ and tD are the migration time of the G-quadruplex and duplex, respectively; a is the constant determined by the ratio of ssDNAG concentration and the total areas of the characteristic features.

in a 95 °C water bath for 15 min after adding 3.0 M KCl solution (final concentration 150 mM). After slow cooling down to room temperature, ssDNAC (5 × 10-5 M stock solution) was added to the above solution, and 1× TE buffer was used to bring the total volume to 500 µL. After equilibrium at the electrophoretic separation temperature in a water bath for 4 h, samples were ready for experiments. Fused-silica capillaries precoated with polyacrylamide were sequentially rinsed with pure water and 1× TBE (with 0.32% PEO, pH 7.9) running buffer for 5 min each and then equilibrated at the separating voltage of -9.4 kV for 5 min before the experiments. For NECE, DNA equilibrium mixtures (∼40 pL) were injected into the capillary (40/47 cm, effective length/total length) by a pressure pulse of 5 s × 3450 Pa, separated in 1× TBE running buffer (without K+) at -9.4 kV (200 V/cm) using negative polarity mode, and detected by a laser-induced fluorescence detector Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

6937

Figure 3. NECE electropherograms of equilibrium mixtures of 3.5 × 10-7 M ssDNAG and 150 mM K+ with varying concentrations of ssDNAC (0, 3.5 × 10-8, 7.0 × 10-8, 1.4 × 10-7, 2.1 × 10-7, and 2.8 × 10-7 M). Separation conditions were described in the Experimental Details section. A1-A5 indicates the five characteristic features on the electropherogram corresponding to the G-quadruplex, ssDNAG unfolded from the G-quadruplex, free ssDNAG, ssDNAG dissociated from the duplex, and duplex, respectively.

(excitation at 488 nm, emission at 520 nm). Each sample was analyzed five times. To ensure reproducibility, capillaries were sequentially rinsed with pure water and 1× TBE running buffer for 5 min between each run of sample and rinsed with pure water for 60 min everyday before use. Data were collected, processed, and analyzed by PACE Station Software (Beckman). The peak areas of the resulting electropherograms were determined by curve fitting using MATLAB 7.0 (see Supporting Information for details). The kinetic and thermodynamic parameters of the G-quadruplex formation were calculated according to the formulas described in the Results and Discussion section. RESULTS AND DISCUSSION NECE of Equilibrium Mixtures. NECE was first attempted for studying the dual equilibrium system of quadruplex/duplex competition by separating equilibrium mixtures of ssDNAG (3.5 × 10-7 M, 5′-tagged with 6-FAM) and ssDNAC (1.4 × 10-7 M). The presence of K+ in the equilibrium mixture induced the G-quadruplex formation of ssDNAG, resulting in the following two equilibriums: kf

R y\z Q

(1)

ku

ka

R y\z D

(2)

kd

where, R, C, D, and Q represent the ssDNAG, ssDNAC, duplex, and G-quadruplex, respectively; kf and ku are the folding/unfolding rate constants of the G-quadruplex; ka and kd are the association/ dissociation constants of the duplex. Thus, the equilibrium mixtures contained four different components: free ssDNAG, free ssDNAC, duplex, and G-quadruplex. Since ssDNAG was 5′-tagged with 6-FAM, only free ssDNAG, duplex, and G-quadruplex were detectable under laser-induced fluorescence detection. PEO was added to the running buffer facilitating the separation of different 6938

Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

components. The PEO additive did not induce the formation or refolding of the G-quadruplex in the absence of K+, which was confirmed by circular dichroism experiments (data not shown). Picoliters of equilibrium mixtures (∼10-18 mol of DNA) were then injected into the capillary for separation under nonequilibrium conditions without K+, resulting in the electropherogram (a) illustrated in Figure 1A,B with five essential features: peaks A1, A3, and A5 and curves A2 and A4. To reveal the identities of these characteristic features, further NECE experiments were carried out with equilibrium mixtures containing different ingredients. As summarized in Figure 1B, ssDNAG with K+ resulted in peaks A1 and A3 and curve A2 (b). Without K+ in the equilibrium mixture, ssDNAG alone resulted in only peak A3 (c) and ssDNAG with ssDNAC resulted in peak A5 (d). Therefore, identities of the five electropherogram features from A1 to A5 could be determined as the G-quadruplex, ssDNAG unfolded from G-quadruplex, free ssDNAG, ssDNAG dissociated from the duplex, and duplex, respectively. Theory for Kinetic Characterization Using NECE. The experiments above demonstrated successful separation of equilibrium mixtures using NECE. However, it is necessary to understand the fate of each component in the NECE scheme in order to quantitatively characterize the quadruplex/duplex competition system. As illustrated in Figure 2, equilibrium mixtures of ssDNAG (5′-tagged with 6-FAM) and ssDNAC are injected into the capillary for separation at a high voltage. Because of the presence of PEO in the running buffer, the migration mobility of each component differs from one another significantly. The equilibrium fraction of ssDNAG migrates as a single electrophoretic zone and results in peak area A3. The migration of ssDNAC does not contribute to the electropherogram due to the lack of a fluorescent label. Since the separation of equilibrium mixtures are under nonequilibrium conditions without K+, equilibriums 1 and 2 are no longer maintained in the electric field. Thus, G-quadruplex and duplex continuously decay during NECE separation, generating nonequilibrium fractions of ssDNAG. The rate of ssDNAG production from the unfolding of G-quadruplex and the dissociation of duplex reduces exponentially during separation, resulting in the exponential curve areas A2 and A4. The remaining fractions of intact G-quadruplex and duplex both migrate as single zones and result in peak areas A1 and A5. It must be noted that the characteristic features of the electropherogram in Figure 2 correspond to the specific positions of equilibrium fractions in the capillary at time t2. The actual experimental electropherogram will have the same features but with an opposite ordering, because fractions with higher migration mobility are detected earlier in time. The five feature areas of the electropherogram (Figure 2) correspond to the equilibrium fractions of the G-quadruplex (A1), ssDNAG unfolded from G-quadruplex (A2), free ssDNAG (A3), ssDNAG dissociated from the duplex (A4), and duplex (A5). Under the assumption that the fluorescent yield of the fluorescently labeled DNA does not change among the single-stranded unfolded, duplex, and G-quadruplex states, the detected fluorescence intensity is proportional to the concentration of each component. Thus, the equilibrium constants and kinetic parameters of Gquadruplex and duplex can be calculated according to the formulas

Table 2. Equilibrium Constants and Kinetic Parameters of the G-Quadruplex and Duplexa 0M -1

ku (s ) kf (s-1) KF tf1/2 (s) tu1/2 (s) kd (M-1 s-1) ka (s-1) KA (M-1)

1.21 × 10 (±0.12%) 1.09 × 10-2 8.99 (±0.33%) 573 63.6 -3

3.5 × 10-8 M

7.0 × 10-8 M

1.4 × 10-7 M

2.1 × 10-7 M

2.8 × 10-7 M

1.24 × 10 (±0.30%) 1.11 × 10-2 8.94 (±0.27%) 559 62.4 1.65 × 10-5 (±0.28%) 1.32 × 106 8.03 × 1010 (±0.21%)

1.23 × 10 (±0.25%) 1.11 × 10-2 9.02 (±0.19%) 564 62.4 1.62 × 10-5 (±0.26%) 1.29 × 106 7.96 × 1010 (±0.18%)

1.25 × 10 (±0.22%) 1.12 × 10-2 8.92 (±0.24%) 554 61.9 1.63 × 10-5 (±0.20%) 1.30 × 106 7.98 × 1010 (±0.14%)

1.23 × 10 (±0.29%) 1.10 × 10-2 8.94 (±0.30%) 564 63.0 1.61 × 10-5 (±0.31%) 1.27 × 106 7.91 × 1010 (±0.16%)

1.22 × 10-3 (±0.17%) 1.10 × 10-2 9.01 (±0.23%) 568 63.0 1.63 × 10-5 (±0.15%) 1.31 × 106 8.02 × 1010 (±0.20%)

-3

-3

-3

-3

average value 1.22 × 10-3 (±1.15%) 1.10 × 10-2 (±0.95%) 8.97 (±0.47%) 564 (±1.18%) 62.7 (±0.96%) 1.63 × 10-5 (±0.91%) 1.30 × 106 (±1.35%) 7.98 × 1010 (±0.61%)

a Electropherograms were obtained from equilibrium mixtures of 3.5 × 10-7 M ssDNAG and 150 mM K+ with varying concentrations of ssDNAC. Equilibrium constants and kinetic parameters were calculated using formulas listed in Table 1. The half-lives of the folded (tf1/2) and unfolded (tu1/2) G-quadruplex were calculated using tf1/2 ) ln 2/ku and tu1/2 ) ln 2/kf, respectively. Numbers in the parentheses are relative standard deviations.

Figure 4. NECE electropherograms of equilibrium mixtures of 3.5 × 10-7 M ssDNAG and 150 mM K+ at varying temperatures (278, 283, 288, 293, and 298 K). Separation conditions were described in the Experimental Details section. A1-A3 indicate the three characteristic features on the electropherogram corresponding to the G-quadruplex, ssDNAG unfolded from the G-quadruplex, and free ssDNAG, respectively.

listed in Table 1 (see Supporting Information for details of mathematical derivation). In principle, by measuring the areas of the characteristic features on the electropherograms, we are then able to experimentally determine the kinetics of both the Gquadruplex and duplex. Kinetic Characterization of G-quadruplex and Duplex. On the basis of the theory for kinetic characterization using NECE, further experiments were carried out with equilibrium mixtures of 3.5 × 10-7 M ssDNAG and varying concentrations of ssDNAC. Resulting electropherograms are shown in Figure 3. As the concentration of ssDNAC rises, areas A1-A3 decrease while areas A4 and A5 increase accordingly due to the increase of the duplex fraction in the equilibrium mixture. Furthermore, A4 and A5 is missing from the electropherogram of 0 M ssDNAC but A1-A3 remains, which indicated that equilibrium 2 can be efficiently separated by NECE alone., i.e., the kinetic characterization of G-quadruplex can be achieved without the presence of complementary strands. With the measurement of the feature areas of the electropherograms obtained from the experiments above, the equilibrium

constants and kinetic parameters of both the G-quadruplex and duplex could be calculated using the formulas listed in Table 1. Results are summarized in Table 2. Experiments were repeated five times at each concentration of ssDNAC. The kinetic values listed in Table 2 are averages of five experiments with relative standard deviations (RSD) given in the parentheses. As noted, the RSD values of all the kinetic parameters obtained at each ssDNAC concentration are less than 0.5%, suggesting the high repeatability of the NECE method. The averages of all the kinetic constants have a RSD lower than 1.5%, which indicated that ssDNAC concentration did not vary the kinetics of the Gquadruplex or duplex, i.e., the intermolecular duplex formation induced minimal influence on the intramolecular G-quadruplex formation. Given a random concentration of ssDNAC, we are then able to determine all the kinetic constants of the G-quadruplex and duplex simultaneously. Furthermore, in the absence of ssDNAC (ssDNAC concentration 0 M, Figure 3), the kinetics of the G-quadruplex formation can also be characterized by the NECE method alone. In contrast, complementary strands were required for the kinetic characterization of the G-quadruplex in the previously reported SPR-based method.18 Thus, NECE turns out to be a superior method with higher simplicity and flexibility. Thermodynamic Characterization of the G-Quadruplex. The unfolding rate constant (ku) of the G-quadruplex is related to the Kelvin temperature (T) according to the Arrhenius equation: ku ) A exp(-Ea ⁄ RT)

(3)

A is a pre-exponential factor; Ea is the activation energy of the reaction; R is the gas constant (8.314). The equilibrium constant (KF) of the G-quadruplex is also related to the Kelvin temperature according to the Gibbs equation: ∆G ) ∆H - T∆S ) -RT ln KF

(4)

∆G, ∆H, and ∆S are the Gibbs free energy, enthalpy, and entropy of the reaction, respectively. Therefore, the thermodynamic parameters of the G-quadruplex (Ea, A, ∆H, and ∆S) can be determined using eqs 3 and 4based on the kinetic characterization of ku and KF at varying temperatures. Since ssDNAC was not required for the kinetic characterization of the G-quadruplex, NECE separations were then carried out with 3.5 × 10-7 M Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

6939

Table 3. Kinetic and Thermodynamic Parameters of the G-Quadruplex at Varying Temperatures ku (s-1)a kf (s-1)a KF a tf1/2 (s)a tu1/2 (s)a Ea (kJ mol-1)b Ab ∆H (kJ mol-1)b ∆S (J mol-1 K-1)b

278 K

283 K

288 K

293 K

298 K

1.87 × 10-4 (±0.20%) 6.96 × 10-3 37.23 (±0.15%) 3706 99.6

3.05 × 10-4 (±0.17%) 8.02 × 10-3 26.30 (±0.16%) 2273 86.4

4.93 × 10-4 (±0.18%) 8.98 × 10-3 18.22 (±0.11%) 1406 77.4 64.87(±0.79) 2.88 × 108(±4.87 × 106) -48.56(±1.06) -144.50(±3.67)

7.99 × 10-4 (±0.14%) 9.88 × 10-3 13.36 (±0.19%) 859 64.7

12.1 × 10-3 (±0.12%) 1.09 × 10-2 8.99 (±0.23%) 573 63.6

a Electropherograms were obtained from equilibrium mixtures of 3.5 × 10-7 M ssDNAG and 150 mM K+ at varying temperatures. Equilibrium constants and kinetic parameters were calculated using formulas listed in Table 1. The half-lives of the folded (tf1/2) and unfolded (tu1/2) G-quadruplex were calculated using tf1/2 ) ln 2/ku and tu1/2 ) ln 2/kf, respectively. Numbers in the parentheses are relative standard deviations. b Ea, A, ∆H, and ∆S were determined using eqs 5 and 6. Numbers in the parentheses are standard deviations.

Table 4. Comparison of Kinetic Characterization Results Using Different Methods Kinetic Parameters G-quadruplex -1

ssDNA sequence

ku (s )

kf (s )

1.21 × 10 ∼10-5 6.0 × 10-3 1.3 × 10-3 7.9 × 10-3

1.10 × 10 1.7 × 10-3 2.4 × 10-2 1.2 × 10-2 1.6 × 10-2

-3

a

(TTAGGG)4 (T4G4)4b (G3TTA)3G3c (TTAGGG)4a (G4AG3T)2AAGGTG4a

duplex -1

-1

kd (M

ka (s-1)

method

ref

-5

1.30 × 106

NECE PAGE FRET SPR SPR

this paper 14 17 18 32

s )

1.63 × 10

-2

-1

1.6 × 10-5 5.0 × 10-5

1.3 × 105 1.4 × 105

Thermodynamic Parametersd ssDNA Sequence

∆H (kJ mol-1)

∆S (J mol-1 K-1)

(TTAGGG)4 TTAGGG

-48.56 (±1.06) -37.2 (±5)

-144.50 (±3.67) -111 (±10)

KF

method

reference

8.99 4.9

NECE NMR

this paper 16

a In the presence of 150 mM K+ at 25 °C. b In the presence of 50 mM Na+ at 37 °C. c ku was experimentally determined from kf by FRET, and KF was determined using UV-melting in the presence of 100 mM Na+ at 37 °C. d Numbers in the parentheses are standard deviations.

ssDNAG at five different temperatures as illustrated in Figure 4. As temperature increases, peak area A1 decreases while areas A2 and A3 increase accordingly. Peak areas A1 and A3 correspond to the G-quadruplex and ssDNAG, respectively. The kinetic parameters of the G-quadruplex were then calculated using the formulas listed in Table 1. The values of ku and KF are averaged from five repeated experiments at each temperature and summarized in Table 3 with the RSD shown in the parentheses. Deriving from eqs 3 and 4, we can get the following two equations:

ln KF )

( ) Ea RT

(5)

∆S ∆H R RT

(6)

ln ku ) ln A -

Both ln ku and ln KF are in linear relation to 1/T. With calculation of the slopes and intercepts of eqs 5 and 6, the thermodynamic parameters of the G-quadruplex (Ea, A, ∆H, and ∆S) can be determined (Table 3). Theoretically, kinetic characterizations of the G-quadruplex at two different temperatures are sufficient for determining all the thermodynamic constants. In our experiments, results from five temperatures were used for enhanced accuracy. 6940

Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

On the basis of the kinetic characterization of the G-quadruplex at varying temperatures, both the kinetic and thermodynamic parameters could be derived simultaneously, which were further found highly comparable with those previously reported using different methods such as SPR, PAGE, and FRET (Table 4). CONCLUSIONS In this paper, we have for the first time successfully used a NECE-based method to investigate the dual equilibrium system of quadruplex/duplex competition. With the use of capillary nongel electrophoresis as the separation mechanism, equilibrium mixtures of telomeric DNA and its complementary strands were efficiently separated under nonequilibrium conditions. The resulting electropherograms allowed us to calculate the kinetic parameters of the G-quadruplex and duplex simultaneously. Results indicated that the duplex formation induced minimal influence on the G-quadruplex folding. By investigation of the kinetics of the G-quadruplex at varying temperatures, both the kinetic and thermodynamic parameters of the G-quadruplex could be determined using the NECE method. The presence of complementary DNA strands was essential in previous kinetic studies of human telomeric sequences using a SPRbased method. However, the complementary strand was not required in our experiments, suggesting the simplicity and flexibility of NECE.

The RSD of all the kinetic constants obtained from five repeated experiments was typically less than 0.5%, which indicated the robustness and high repeatability of the NECE approach. Furthermore, a DNA sample amount as low as ∼10-18 mol was sufficient for analysis. Thus, NECE provided a superior platform for investigating the kinetics and thermodynamics of human telomeric sequences compared to previously reported methods. Undoubtedly, the novel NECE method will provide a new avenue for studying the kinetics and thermodynamics of nucleic acids within dual equilibrium systems with significant advantages of simplicity, flexibility, extreme-low sample cost, and high repeatability.

30570468, and 30600134), National Basic Research Program of China (Grants 2007CB914203 and 2007CB714507), and Program for New Century Excellent Talents in University (Grant NCET05-0644).

ACKNOWLEDGMENT The authors gratefully acknowledge the financial support from National Natural Science Foundation of China (Grants 20405006,

Received for review March 26, 2008. Accepted July 14, 2008.

SUPPORTING INFORMATION AVAILABLE Mathematical derivation of equations listed in Table 1. This material is available free of charge via the Internet at http:// pubs.acs.org.

AC801335Y

Analytical Chemistry, Vol. 80, No. 18, September 15, 2008

6941