Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersaline-tolerant sensor using abasic site-engineered monomorphism

Analytica Chimica Acta xxx (2017) 1e9

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersaline-tolerant sensor using abasic site-engineered monomorphism Tao Wu a, 1, Meiyun Ye b, 1, Tianyi Mao b, 1, Fan Lin a, Yuehua Hu a, Ning Gan c, Yong Shao a, * a b c

Institute of Physical Chemistry, Zhejiang Normal University, Jinhua, 321004, Zhejiang, China College of Chemistry and Life Sciences, Zhejiang Normal University, Jinhua, 321004, Zhejiang, China Faculty of Material Science and Chemical Engineering, Ningbo University, Ningbo, 315211, Zhejiang, China

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


Fluorescence discrimination of hybrid-2 G-quadruplex from hybrid1 G-quadruplex in human telomere is achieved.
Abasic site-induced monomorphism can be used as the best sensor platform.
A selective Kþ sensor with a remarkable hypersaline-tolerant capability is developed.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 October 2016 Received in revised form 13 January 2017 Accepted 21 January 2017 Available online xxx

Coexistence of the polymorphic hybrid-1 and hybrid-2 conformers for a given human telomeric Gquadruplex-forming sequence (htG4) complicates its fine structure identification and limits its application as a sensor element. With help from abasic site (AP site)-engineered htG4s serving as the monomorphic representatives of the two typical hybrid conformers, we found that thioflavin T (ThT) can selectively target the hybrid-2 conformer over the hybrid-1 counterpart in monomer and tandem htG4 molecules. The htG4 that solely adopts the monomorphic hybrid-2 conformer engineered by the AP site is most efficient in lighting up ThT fluorescence in Kþ and a selective Kþ sensor is realized with a remarkable hypersaline-tolerant capability that can work even in 30000-fold excess of Naþ. At 600 mM Naþ, the dynamic range for Kþ detection can be extended to 30 mM with the limit of detection of 20 mM. This is the first report on the fluorescence discrimination of these two hybrid conformers of htG4 although they have long been categorized with their characteristically structural topologies. Our work will attract much interest in the development of sensors based on the monomorphic htG4 conformer since such high performance in sensor development has not been previously achieved. © 2017 Elsevier B.V. All rights reserved.

Keywords: Human telomeric G-quadruplex Hybrid-1/Hybrid-2 Abasic site Hypersaline Fluorescent sensor Selectivity

1. Introduction

* Corresponding author. E-mail addresses: [email protected], [email protected] (Y. Shao). 1 These authors equally contributed to this work.

Along with gradual elucidation of unique biological functions of G-quadruplex (G4) structure in living cell [1,2], G4 has also been recognized as an essential element in development of innovative

http://dx.doi.org/10.1016/j.aca.2017.01.041 0003-2670/© 2017 Elsevier B.V. All rights reserved.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

2

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

sensors and devices due to the reliable tunability of its folding and stability [3e11]. Illustratively, some well-developed aptamers for, for example, ATP and thrombin, involve the G4 motif as the binding configuration. The representative human telomeric DNA G4 (htG4) has received much attention due to its importance in chromosome end protection [1] and many efforts have been made in finding ligands to stabilize and sense the variant htG4 structures [12e23]. The htG4 stabilizers are promising in developing into potential therapeutic drugs [24,25]. Concurrently, htG4 has also served successfully as a practical detection platform for telomerase activity [26e28], telomerase inhibitor [29], DNA oxidation [30], anticancer drugs [31], metal ions [16,32e39], and as a novel element in logic gates [40]. During developing htG4-based practices, the polymorphic coexistence for a given htG4 sequence [41] should be seriously considered when evolving towards a specific binding ligand. The strand orientation in htG4 has been resolved with variant conformer categories at least including the basket-type antiparallel topology in Naþ solution, the hybrid topology (hybrid-1 and hybrid-2) in Kþ solution [16,42,43], and the parallel topology in Kþ-containing crystal and molecular crowding condition [44e49]. Even for a given htG4 sequence, these conformers coexist through a sequence-dependent manner [50e56]. For example, the presence of an individual 50 -TA-30 dinucleotide beyond the 50 end of the htG4 tetrad core (TAQ, 50 -TAQ-30 , Q ¼ 50 G3(T2AG3)3-30 , Table S1) gives a G4 blend in Kþ with about 30% hybrid-2 (two successive lateral loops followed by one propeller loop from 50 to 30 ) and 70% hybrid-1 (one propeller loop followed by two successive lateral loops from 50 to 30 ). However, further extending another 50 -TT-30 dinucleotide at the 30 end of TAQ (to give TAQTT, 50 -TAQTT-30 ) just reverses this hybrid-2-to-hybrid-1 ratio [50,53]. More interestingly, the lengthened htG4 thus with ability to form tandem G4s brings occurrence of hybrid-1 and hybrid-2 components in one strand [57e59], although other structure combination recipe has also been reported [48]. Due to the close structure similarity with the hybrid-2 and hybrid-1 conformers, their discrimination with a ligand is a great challenge. In practice, such conformer heterogeneity in a solution should also compromise the performance of htG4 as a sensor element.
r and Vorlí
[60] recently found that introduction of Sklena ckova an abasic site (AP site) into 50 -AQ-30 (AQ) in replace of the adenine base can extremely narrow the structure distribution to a specifically monomorphic G4 conformer dependent on the position of the AP site. The replacement occurring at the adenine of the first loop (A7, counting from 50 to 30 end) favors the monomorphic hybrid-1 conformer in Kþ (thus named AP-H1, Fig. 1), while the replacement occurring at the adenine of the third loop (A19) supports the monomorphic hybrid-2 conformer (AP-H2, Fig. 1). Furthermore, the parallel G4 (AP-P, Fig. 1) is the dominant conformer when the three loop adenines (A7, A13, and A19, Fig. 1) are all replaced with AP sites. The parent AQ without any AP site, otherwise, forms polymorphic G4s in solution [60,61]. These findings inspire us to identify a ligand that can favorably target any of these hybrid conformers in order to improve the htG4-based sensor performance. In this work, we found that thioflavin T (ThT) can selectively bind with the hybrid-2 htG4 with an affinity-induced fluorescence brighter than binding with the hybrid-1 htG4. As a proof-of-principle application, we first develop a monomorphic hybrid htG4-based Kþ sensor that can even work at an extreme condition of highly concentrated salt. This sensor can endure 30000-fold excess of Naþ for 20 mM Kþ detection. Our work demonstrates competence of htG4 in sensor development by engineering it towards a monomorphic conformer.

Fig. 1. Dependence of ThT fluorescence (0.5 mM) at 488 nm on the htG4 sequences (1 mM) in 0.1 M Kþ (pH 7.5). The sequences given above are the typical htG4s showing the position of adenine replacement with the AP site. The percentage labeled at abscissa is the population of hybrid-2 conformer acquired by previous NMR experiments. The dotted line reflects the fluorescence level for AP-H1 with solely the hybrid-1 conformer. Inset: dependence of ThT fluorescence on the reported population of the hybrid-2 conformer.

2. Experimental section 2.1. Materials and reagents DNA oligonucleitides (Table S1) were synthesized by TaKaRa Biotechnology Co., Ltd (Dalian, China) and purified by HPLC. The nucleic acid concentrations (in strand unit) were measured by first dissolving DNA in pure water and detecting the UV absorbance at 260 nm using extinction coefficients calculated by nearest neighbor analysis. Buffers, metal ions (nitrate salt), NH4Ac, and EDTA were of analytical grade (Sigma Chemical Co., St. Louis, USA) and used without any purification. NaNO3 at the highest commercially available purity was used as the hypersaline condition. Thioflavin T (ThT, ultrapure grade) was obtained from AAT Bioquest, Inc. (California, USA) and used as received. For the abasic site-containing APH1 and AP-H2, tetrahydrofuran residue was used as the chemically stable abasic site analogue for replacement of the naturally tautomeric deoxyribose structure, as used in Ref. [60]. Milli-Q water (18.2 mU; Millipore Co, Billerica, USA) was used throughout the experiments. 2.2. Fluorescence measurements Fluorescence spectra were acquired with a FLSP920 spectrofluorometer (Edinburgh Instruments Ltd., Livingston, UK) at 20 ± 1  C, which was equipped with a temperature-controlled circulator (Julabo Labortechnik GmbH, Seelbach, Germany). Fluorescence was measured in a quartz cell with path length of 1 cm. If not specified, in order to prepare the htG4 solution with stable conformation, the nucleic acid strand was annealed in a thermocycler (first at 92  C, then slowly cooled to room temperature) and stored at 4  C overnight. ThT at the specified concentration was added into the nucleic acid solution, and the resulting solutions allowed incubation for 30 min before fluorescence measurements. Tris buffer (pH 7.5) containing the corresponding metal salt was used in the htG4 targeting investigation.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

Fluorescence lifetime was measured on a time-correlated single photon counting FLSP920 system, with excitation at 450 nm and detection at the ThT emission maxima. The excitation source was a nanosecond laser (450 nm) and a ludox solution was used as the scatter for the instrument response. The data were fitted with a single exponential decay and c2 was less than 1.1. The htG4 binding-induced turn-on fluorescence provided a chance for determination of the ThT binding affinity via fluorescence technique. ThT (0.5 mM) was titrated with nucleic acid and the resultant fluorescence intensity at 488 nm was plotted as a function of the nucleic acid concentration. KaleidaGraph (Synergy Software, PA) was used to fit the binding constant. The stoichiometric ratio of ThT binding with htG4 was determined by the Job's plot analysis. The total concentration of Hyp and htG4 was maintained at 2 mM and the ThT-to-htG4 concentration ratio was sequentially varied. 2.3. DNA melting temperature (Tm) measurements The melting temperature (Tm) of htG4 was determined using a UV2550 spectrophotometer (Shimadzu Corp., Kyoto, Japan), equipped with a TMSPC-8 Tm analysis system. The absorbance of htG4 at 295 nm as a function of solution temperature between 5  C and 90  C was collected in 0.5  C increments, with a 30-s equilibration time applied after each temperature increment. For the effect of ThT binding on the htG4 melting behavior, the experiments were carried out in phosphate buffer (pH 5.3) in order to avoid ThT hydrolysis at the used high temperature during Tm measurements [16]. 2.4. ITC measurements Isothermal titration calorimetry (ITC) measurements were carried out at 20  C using an ITC200 microcalorimeter (MicroCal, LLC, Northampton, MA). DNA (30 mM) in the sample cell was droply titrated with 2 mL of ThT (the first 0.4 mL injection was followed by 19 injections of 2 mL with 20s duration at 150s time intervals). The titration of ThT in the syringe into the identical buffer solution in the sample cell without DNA were used as a control to obtain the dilution heat. Origin (version 7.0) software was used for data analysis. The area of each injection peak was automatically integrated. A binding isotherm curve was obtained by plotting the total heat per injection (kcal mol1 of injectant) as a function of the molar ratio of ThT to DNA.

3

not been identified. A molecular rotor mechanism has been proposed by us [16] as the rationale for the turn-on ThT fluorescence response upon binding with htG4 and the ThT binding does not change the G4 conformation in Kþ [16,65]. We first checked the dependence of ThT fluorescence on the htG4 sequence not containing the AP site in Kþ solution. As shown in Fig. 1, the resultant ThT fluorescence responses to the polymorphic conformerscoexisting htG4s of TTAQTT, TAQTT, TAQT, and TAQ (see Table S1 for the detailed sequences) just follow a linear dependence on the hybrid-2 populations that have been numerically reported using previous NMR experiments (about 75%, 70%, 50%, and 30%, respectively) [50e54]. Furthermore, AP-H2 at the monomorphic hybrid-2 conformer as a result of the AP site engineering [60] indeed gives the highest ThT fluorescence response, while AP-H1 at the monomorphic hybrid-1 conformer brings about a threefold lower ThT fluorescence. According to such linear dependence of fluorescence on the hybrid-2 population (Inset of Fig. 1. 1 mM htG4s were all incubated with 0.5 mM ThT), nearly 100% hybrid-2 population is predicted with AP-H2. This suggests competence of AP-H2 in achieving a monomorphic hybrid-2 conformer by engi
r neering an AP site into the loop sequence, as confirmed by Sklena
[60]. These sequence-dependent results demonand Vorlí ckova strate the ThT binding preference to the hybrid-2 htG4 conformer. Moreover, Q and QT not containing the 50 nucleotides and AP-P result in very weak ThT fluorescence, since they adopt structures other than the hybrid (2-tetrad basket antipapallel [66,67] and papallel [60], respectively). The fluorescence titration experiments carried out at 100 mM Kþ (Fig. 2) predict the binding affinities of 6.0  105 and 1.0  105 M1 for AP-H2 and AP-H1, respectively, assuming a 1:1 binding mode [16,62]. This binding mode for these two AP site-containing htG4s is also roughly estimated using the Job's plot analysis (Fig. S1). Obviously, these results also demonstrate the favored hybrid-2-over-hybrid-1 binding of ThT. It is widely accepted that the 30 and 50 terminal nucleotides are very crucial in defining the hybrid-1 and hybrid-2 conformers via forming substructures over the quadruplex tetrad (for example, base-triplet, base-pair, etc) [42e55]. Thus, there are at least two contributors that will define the ThT binding with htG4: the htG4 topology-determined tetrad (the detailed tetrad structure is determined by syn- and anti-guanines) and the terminal substructures. It has been theoretically reported that ThT interacts mainly with the hybrid-1 conformer at the 30 end via sandwich stacking with the tetrad and the terminal substructures (Scheme 1) [65]. Thus, the hybid-1 binding observed here should make ThT contact with guanines, resulting in fewer enhancements in

2.5. Analysis of urine samples Human urine was used as the sample to confirm the feasibility of our sensor. The urine samples collected from healthy volunteers were centrifuged at 12000 g at room temperature for 15 min. The supernatants were filtered through 0.22 mm membranes. Then the urine samples were appropriately diluted with 25 mM Tris-HNO3 buffer (pH ¼ 7.5) containing 600 mM Naþ, in which Kþ standard was added and then the added Kþ content was measured for the recovery calculation. 3. Result and discussion 3.1. Fluorescently targeting hybrid-2 population over hybrid-1 in monomer htG4 Herein, ThT was selected as the potential hybrid-2-selective porbe. Although several previous experimental [15,16,62e64] and theoretical [15,65] investigations have demonstrated ThT binding with htG4s, the ThT preference to the hybrid-2 htG4 conformer has

Fig. 2. Typical fluorescence titration curves of 0.5 mM ThT with G4 in 25 mM Tris buffer (pH 7.5) containing 100 mM Kþ in the absence (A) and presence (B) of 600 mM Naþ. The solid lines suggest a 1:1 binding mode via the curve fitting.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

4

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

Scheme 1. Binding modes of Kþ and ThT with (a) AP-H1 and (b) AP-H2. The red-colored spheres represent the AP site, while the blue-colored spheres represent the bound Kþ (the larger ones for inter-tetrad binding and the smaller one for external coordination binding occurring to only AP-H1). The dotted lines in (a) show the external coordination of Kþ at the edge of the 50 tetrad with A1, T18, G20, and the deoxyribose and phosphate backbone from G20 and G21, according to Ref. [71]. Besides the indicated nucleotides, the others (including the nucleotide substructures beyond the 30 tetrad) were omitted for clarity. However, at high coexisted Naþ concentration, this external coordination site would be also possibly occupied by Naþ.

fluorescence because of partial quenching caused by the guanine contact, as previously confirmed by us [68]. However, ThT should mainly bind to the hybrid-2 conformer through a substructure (most likely involving the A1/T12 base pair, see Fig. 1 for numbering) formed on top of the 50 tetrad (Scheme 1) [16,60,65]. It has been reported that such a base-pair substructure can effectively occur only for the hybrid-2 conformer [60,65]. Accordingly, this hybrid-2 binding mode involving the A1/T12 base pair thus makes ThT much brighter [68]. In order to evaluate the role of the G4 conformer in defining the ThT binding, we further examined a control sequence of AAAQAA. AAAQAA in Kþ finally gives the hybrid-1 conformer, but at the moment of Kþ addition, accompanying the hybrid-1 formation, a significant amount of the hybrid-2 conformer is also formed and slowly converts to the hybrid-1 conformer at room temperature, as observed with NMR [54]. According to such folding uniqueness of this sequence, indeed, we observed a 38% decrease in the ThT fluorescence (Fig. S2) after overnight incubation in Kþ relative to that at the moment of Kþ addition, which is in good agreement with the NMR-specified hybrid-2 amount at the initial stage of Kþ addition (about 40%) [54]. We monitored the time evolution of ThT fluorescence during the initial stage of Kþ addition and the slow conversion was also observed (Fig. S2). Note that the overnight Kþ-aged AAAQAA only caused a slightly higher ThT fluorescence than AP-H1 (Fig. 1), in good agreement with the preferred hybrid-1 conformer for the Kþaged AAAQAA [54]. Thus, ThT can be used to follow the conformer conversion during the folding process of htG4s.

conformers of the tandem htG4s. We then compared the lengthened htG4s with the rightly-folded monomer ones. According to the hybrid conformer pattern for (TTAQ)2TT [57,58] and the 50% hybrid-2 population for TAQT [54], when the (TTAQ)2TT concentration was half of the TAQT concentration, these two DNA solutions would contain the same hybrid-1 and hybrid-2 conformer concentrations, respectively. As shown in Inset of Fig. 3, indeed, 0.5 mM (TTAQ)2TT causes a fluorescence response of ThT (0.5 mM) similar to 1 mM TAQT, suggesting that these two DNAs have the same distribution in hybrid-1 and hybrid-2 conformers. This is in good agreement with the proposed distribution of the involved hybrid conformers in (TTAQ)2TT [57,58]. Moreover, we found that 0.33 mM (TTAQ)3TT leads to a ThT fluorescence about two folds higher than 1 mM TAQ, although these DNAs can give almost the sameconcentrated G4 individuals at the used strand concentrations. Considering the 30% hybrid-2 population for TAQ [50,52], we thus expect that (TTAQ)3TT should contain two hybrid-2 conformers. This is also reflected by the slightly higher fluorescence for

3.2. Fluorescently sensing the hybrid-2 and hybrid-1 distribution in lengthened htG4 The adopted hybrid conformer preference for an individual G4 component in a lengthened htG4 molecule, that can form tandem hybrid htG4s [49], is strongly dependent on the htG4 length and location. For example, (TTAQ)2TT (Table S1) can form two contiguous G4s: one hybrid-1 conformer at the 50 end and one hybrid-2 conformer at the 30 end [57,58]. However, most (>90%) of (TTAQ)3TT with three G4-forming sequences (Table S1) can fold into three contiguous G4s [59]. It is expected that, the ThT binding capacity toward an individual G4 with the corresponding hybrid type in the lengthened htG4 shouldn't be affected by the context conformers, as evidenced by the Mao's and Chaires's experiments that no [55] or weak [57,58] interaction was found between inter-

Fig. 3. Dependence of ThT fluorescence on the length of htG4 for estimating the hybrid-2 population. ThT: 0.5 mM; buffer: 25 mM Tris (pH 7.5) containing 100 mM Kþ. Inset: comparison of emission spectra of 0.5 mM ThT in the presence of TAQT (1 mM), TAQ (1 mM), (TTAQ)2TT (0.5 mM), and (TTAQ)3TT (0.33 mM), respectively.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

(TTAQ)3TT (as expected by us with 0.66 mM hybrid-2 involved) in comparison with (TTAQ)2TT (as modeled by Chaires and et al. [57,58] with 0.5 mM hybrid-2 involved). Several reports have confirmed the hybrid-2 preference at the 30 end in the lengthened htG4s. As observed by Petraccone and Chaires [59], among the involved three G4s in (TTAQ)3TT, there are two resolved G4 individuals that are more stable than the remained one. Accordingly, the hybrid-2 conformer is found to be more stable than the hybrid1 conformer [60]. The preferable formation of hybrid-2 structure at the internal and the 30 end of a lengthened htG4 has been € tsch et al. [48]. Very recently, Luo and confirmed by H€ ansel and Do Mu also theoretically predicted that the most stable conformer in the lengthened htG4 was the hybrid-2 structure [69]. As occurred for monomer htG4s, our experiments also confirm the preferable binding of ThT with the hybrid-2 G4 individuals in the tandem htG4 molecules. Therefore, ThT could serve as a useful fluorescent ligand to identify the hybrid conformers adopted by the lengthened htG4s, which is different from the previously developed ligands with the inter-quadruplex interfaces as the binding fields [21,70]. The involved hybrid-2 individuals in (TTAQ)2TT and (TTAQ)3TT were roughly estimated in this work by the extrapolated m value, which is the number of hybrid-2 conformers involved in these lengthened htG4s. Using TTAQTT as a control, its m value is 0.75 [54]. Thus, at the same m[DNA] value ([DNA] is the strand concentration), the hybrid-2 concentration (also the hybrid-1 concentration) in solutions of TTAQTT, (TTAQ)2TT, and (TTAQ)3TT will be the same, which should result in the same ThT fluorescence. As shown in Fig. 3 by plotting the resultant ThT fluorescence as a function of m[DNA], the fluorescence responses for (TTAQ)2TT and (TTAQ)3TT predict the m values of 1.20 and 2.40, respectively. These values are in good agreement with the proposed hybrid structure preferences [48,57e60,69].

5

triplex intermediates within about tens of minutes [74]. Such difference in the AP-H2 and AP-H1 folding to the Kþ-stabilized states is also substantiated by fluorescence kinetics experiments (Fig. S3B). These conformer features in Kþ will regulate their binding with ThT. The monomorphic hybrid-1 and hybrid-2 conformers occurring for AP-H1 and AP-H2 sequences provide the possibility to investigate their ThT binding specificities using the ITC methodology. For the ThT binding with their Kþ-stabilized hybrid conformers, fitting to the ITC titration data at comparable ThT and DNA concentrations gave good results for the strong DNA binding site (Fig. 4), and the weak nonspecific binding that could occur at the high ThT concentration [65] was not attempted to fit in this work. The fitted binding constants from ITC experiments (Table S2) are different from those obtained using the fluorescence method as observed in Fig. 2, because of the much higher reagent concentrations used in the ITC experiments [75,76]. The AP-H2 binding is more 1:1 stoichiometric (with the binding site n ¼ 1.12) than the AP-H1 binding (n ¼ 0.70). Additionally, both of the bindings are exothermic in enthalpy. The ThT binding enthalpy with AP-H2 is 1.5 times higher than that with AP-H1 (Table S2). These results predict a more specific binding of ThT with AP-H2, as predicted by Luo and Mu via the 50 end A1/T12 stacking interaction, as opposed to the 30 end sandwich stacking for AP-H1 (Scheme 1. The lowest substructure layer formed from T11, T12, and A13 for the sandwich binding is omitted for clarity) [65]. The DNA melting experiments (Fig. S4) confirm the efficient stacking of ThT with AP-H2 by the observation of more significant increases in the melting temperature (Tm) upon ThT binding. For the AP-H1 binding, the ThT sandwich interaction mode would experience a more steric hindrance, resulting in a less increase in the ThT binding-induced stability.

3.3. Analysis of the binding mode of ThT with AP-H1 and AP-H2 Isothermal titration calorimetry (ITC) was used to evaluate the binding behaviors of Kþ with the monomorphic htG4s of AP-H1 and AP-H2. Kþ titration into their Naþ solutions (using Naþ preincubation to avoid most of the energy contribution from electrostatic binding with DNA backbone during Kþ titration) results in endothermic exchanges of Naþ with Kþ for the two DNAs (Fig. S3A). This suggests structure conversions with AP-H1 and AP-H2 from the chair conformers [16] to the hybrid ones. AP-H2 requires up to two times less energy gain to complete this exchange than AP-H1, suggesting the more favorable Kþ binding with AP-H2. Chaires et al. found that besides the two channel Kþ-binding sites existing in inter-tetrads in both hybrid conformers, the hybrid-1 conformer has another specific Kþ-binding site at the 50 end to form a tertiary substructure involving the first 50 adenine (Scheme 1) [71]. Such external coordination Kþ-binding site for the hybrid-1 conformer has also been confirmed by Lah and Vesnaver et al. considering the thermally activated process with this Kþ binding step [72]. Therefore, based on our results, we expect that AP-H1 should hold three Kþ-binding sites; different from AP-H2 holding two Kþ-binding sites (Scheme 1). Furthermore, prior to Kþ binding saturation, APH1 causes an increase in the baseline of energy gain after each titration, very different from AP-H2 with a complete decay to the initial baseline after a prompt increase of the required energy gain upon Kþ titration. This is in good agreement with the previous observation that the hybrid-2 structure folds faster than hybrid-1 [73]. Luo and Mu also theoretically predicted that hairpin and triplex intermediates were found for htG4 folding to the hybrid-1 conformer, whereas no intermediate was observed during the hybrid-2 folding [69]. The hybrid-1 conformer formation in Kþ from the Naþ-preformed structure was expected through hairpin and

Fig. 4. Binding of ThT with AP-DNA evaluated by ITC. 30 mM AP-DNA was titrated with 300 mM ThT in 25 mM Tris buffer (pH 7.5) containing 100 mM Kþ at 20  C. The curves are the fitting results.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

6

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

Since ThT is positively charged, the electrostatic interaction with htG4 should be considered. We then carried out fluorescence titration experiments in 100 mM Kþ in coexistence with 600 mM Naþ. In comparison with 100 mM Kþ alone, the difference in ThT fluorescence for the AP-H2 and AP-H1 bindings is further increased with the addition of 600 mM Naþ (Fig. 2. The fitted binding constants are 2.0  105 and 7.5  104 M1 for AP-H2 and AP-H1, respectively). Thus, the AP-H1 binding should be featured with more electrostatic contribution in comparison with the AP-H2 binding. The differences in the binding environments and binding forces are further confirmed with a longer excited state lifetime and a less ion-strength effect of the lifetime for the AP-H2 binding with respect to the AP-H1 binding (Table S3).

In combining the detailed structures [60,71], the Kþ binding specificities [71,72], and the theoretically simulated ThT binding modes [65] of the hybrid-1 and hybrid-2 htG4s with our experimental results, the Kþ and ThT bindings with AP-H1 and AP-H2 are summarily proposed as shown in Scheme 1. 3.4. A hypersaline-tolerant sensor with a remarkable selectivity using the monomorphic AP-H2 as the most efficient selector In order to confirm the favorable binding of ThT with the hybrid2 conformer and potential applications, the Kþ-dependent fluorescence response was checked under variant Naþ concentrations. As shown in Fig. 5, among the three htG4s of AP-H2, AP-H1, and AQ

Fig. 5. Dependence of fluorescence intensity of 0.5 mM ThT at 488 nm on the Kþ concentration (htG4: 1 mM) in 25 mM Tris buffer (pH 7.5) containing (A) 100, (B) 400, (C) 600, (D) 1000 mM Naþ. Inset: photographs of 2 mM solutions of ThT and AP-H2/-H1 under UV illumination with Kþ and Naþ coexisted at the indicated concentrations.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

with the monomorphic hybrid-2, hybrid-1 conformers, and the polymorphic hybrid blend, respectively, AP-H2 is always the most efficient htG4 to light up ThT in a Kþ concentration-dependent manner even with the coexisted Naþ concentration increasing up to 1000 mM. This suggests that the specific binding of ThT with APH2 is still kept in such high ion-strength solutions. The hybrid structure of AP-H2 should be still kept at the used condition, as confirmed by Macgregor and Helmy et al. for htG4s in Kþ with highly concentrated Naþ salt [77]. The previouly observed preferable emitting of ThT in binding with the hybrid htG4 conformers [16] means that AP-H2 remains to favor the hybrid-2 conformer in the conditions tested here. The DNA melting experiments (Fig. S4) also support the stability of the hybrid structure even with the coexisted Naþ concentration increasing up to 1000 mM. As shown in Fig. S5, the possibility that the hybrid-2 structure might to convert to parallel conformer for AP-H2 at highly concentrated Naþ salt is ruled out with the observation of the resultant ThT fluorescence still higher than that obtained with the same sequence but containing three AP sites (AP-P, Fig. 1), since AP-P adopts a parallel G4 structure in Kþ [60]. In fact, AP-P causes ThT to glow, however, at the same level as AP-H1 and AQ even at 1000 mM Naþ (Fig. S5). Expectedly, such hypersaline-tolerant performance of AP-H2 in response to Kþ can be also expanded to Liþ salt (Fig. S6) due to its similar inability as Naþ to bring a stable G4 hybrid structure. The results in Fig. 5 predict a practice for the ThT/AP-H2 system as a fluorescence Kþ sensing platform that can operate at hypersaline condition. Up to now, the sensor that can work with an excellent salt-tolerant performance is still a challenge [16,78], since it requires high selectivity for the targets of interest. The sensor possessing this feature will find wide applications in identifying purity of salt reagents and water analysis with a low cost because the complex desalting and separation procedures are unnecessary. In practice, unlike heavy-metal ions, alkali metal ions are difficult to separate using a simple method and a straightforward sensor is required for, for example, direct detection of Kþ in a Naþ salt. In order to practice our findings, variant metal ions (each in 2 mM for Liþ, Naþ, Kþ, Rbþ, Csþ, Ca2þ, Mg2þ, Ba2þ, Sr2þ, Al3þ, Cd2þ, Co2þ, Cr3þ, 2þ 2þ 2þ Cu2þ, Fe3þ, Mn2þ, NHþ 4 , Ni , Pb , and Zn ) were tested by examining their effects on the fluorescent response of 10 mM ThT in the presence of 1 mM AP-H2, AP-H1, and AQ, respectively. EDTA (10 mM) was added to avoid hydrolysis of some of the heavy metal ions and to get rid of interference from binding of some metal ions with DNA backbone and/or G4 structures [79,80]. An extra addition of 600 mM Naþ (thus 620 mM Naþ in total, considering the Naþ from EDTA) was utilized to give the hypersaline condition. As shown in Fig. 6A, a substantial increase in ThT fluorescence occurs only for Kþ using AP-H2 as the sensing element, suggesting the practicability of our sensor in Kþ assay under the hypersaline condition. Note that this excellent performance in selectivity cannot be achieved using AP-H1 and AQ as the sensing elements (Fig. S7). Especially, besides Kþ, AP-H1 and AQ also get significant responses for Ba2þ and Sr2þ since these two metal ions have been reported to stabilize G4 due to the similar size as Kþ [81,82]. In comparison with AP-H1, AQ experiences a slightly better tolerance on Ba2þ and Sr2þ because of the presence of minor hybrid-2 species in the AQ solution [54]. These facts suggest that the EDTA binding of these two metal ions cannot completely compete with their binding with AP-H1 and AQ. This excellent selectivity was also confirmed by the experiments of Kþ coexisting with Ba2þ and Sr2þ. In the presence of 10 mM Ba2þ and Sr2þ, AP-H2 can also efficiently sense Kþ even with its concentration being less than 1 mM in comparison to those obtained with AP-H1 (Fig. S8). In the previously developed G4-based Kþ sensors [16,78], interference from Ba2þ and Sr2þ was not considered because of unavailability of an ideal sensing platform at that time. In this work, however,

7

Fig. 6. (A) Kþ-selective response of ThT/AP-H2 in the presence of highly concentrated Naþ salt. 2 mM of individual metal ion was added into 25 mM Tris buffer (pH 7.5) containing 600 mM Naþ, 1 mM AP-H2, 10 mM ThT, and 10 mM EDTA. F0 and F respectively stand for the fluorescent intensities in the absence and presence of the metal ions (from 1 to 20: Liþ, Naþ, Kþ, Rbþ, Csþ, Ca2þ, Mg2þ, Ba2þ, Sr2þ, Al3þ, Cd2þ, 2þ 2þ 2þ Co2þ, Cr3þ, Cu2þ, Fe3þ, Mn2þ, NHþ 4 , Ni , Pb , and Zn ). Inset: the corresponding emission spectra. (B) Dependence of fluorescence intensity (F) of 10 mM ThT (1 mM htG4) at 488 nm on the addition of low concentration Kþ in 25 mM Tris buffer (pH 7.5) containing 600 mM Naþ. F0 is the fluorescence in the absence of Kþ with the corresponding htG4.

significantly less interference from Ba2þ and Sr2þ is achieved at 620 mM Naþ using AP-H2 as the sensing element (Fig. 6A), again substantiating the strong binding specificity of the hybrid-2 htG4 structure with Kþ, as demonstrated in the ITC experiments (Fig. S3A) and Ref. [71]. The results in Fig. 5 also predict a more sensitive response to Kþ using AP-H2 as the best sensing element and an almost linear response with Kþ concentration at several to tens of mM range prior to fluorescence saturation. This preference of AP-H2 over APH1 and AQ is also kept at low Kþ concentration, as observed in Fig. 6B. The ThT fluorescence change in the presence of AP-H2 is logarithmically dependent on the Kþ concentration even at mM level, as opposed to the negligible changes using AP-H1 and AQ as the sensing elements. The Kþ concentration as low as 20 mM can be detectable in the presence of 620 mM Naþ that is about 30000-fold excess of Kþ (Fig. 6B). In our point of view, this is the first report on such high selectivity of Kþ over Naþ among the developed G4based sensors [16,78,83]. In order to further examine the practice of our sesnor, we finally applied our sensor to Kþ ion detection in human urine samples, in which significant amount of Naþ and other ions also exist. 600 mM Naþ was added to level off the effect of these ions. The analytical results for the urine samples were listed in Table S4. The recoveries of spiked urine samples ranged from 94% to 114%, demonstrating the feasibility of our sensor in practical sample analysis.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

8

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9

4. Conclusions In conclusion, using the AP site-engineered htG4s as the monomorphic representatives of the two typical hybrid-1 and hybrid-2 conformers, we found that ThT, as a molecular rotor probe, can selectively target the hybrid-2 conformer with respect to the hybrid-1 counterpart in monomer and tandem htG4 molecules. This is the first report on the fluorescence discrimination of these two hybrid structures of htG4s although they have long been categorized according to their characteristically structural topologies. AP-H2 that solely adopts the hybrid-2 conformer is the most efficient htG4 in fluorescently lighting up ThT in Kþ and a selective Kþ sensor can be developed with a remarkable capability that can work even in a hypersaline condition of 30000-fold excess of Naþ. Such high performance in sensor development has not been previously achieved. Our work will inspire much interest in the development of sensors based on the monomorphic htG4 structures. Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant No. 21675142 and 21545009) and the National Undergraduate Training Programs for Innovation and Entrepreneurship (Grant No. 201610345011). Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.aca.2017.01.041. References [1] R.E. Verdun, J. Karlseder, Replication and protection of telomeres, Nature 447 (2007) 924e931. [2] S. Balasubramanian, L.H. Hurley, S. Neidle, Targeting G-quadruplex in gene promoters: a novel anticancer strategy, Nat. Rev. 10 (2011) 261e275. [3] F. Wang, X. Liu, I. Willner, DNA switches: from principles to applications, Angew. Chem. Int. Ed. 54 (2015) 1098e1129. [4] Y. Zhao, F. Chen, Q. Li, L. Wang, C. Fan, Isothermal amplification of nucleic acids, Chem. Rev. 115 (2015) 12491e12545. [5] D.-L. Ma, H.-Z. He, K.-H. Leung, H.-J. Zhong, D.S.H. Chan, C.-H. Leung, Label free luminescent oligonucleotide-based probes, Chem. Soc. Rev. 42 (2013) 3427e3440. [6] L.A. Yatsunyk, O. Mendoza, J.L. Mergny, “Nano-oddities”: unusual nucleic acid assemblies for DNA-based nanostructures and nanodevices, Acc. Chem. Res. 47 (2014) 1836e1844. [7] J. Ren, T. Wang, E. Wang, J. Wang, Versatile G-quadruplex-mediated strategies in label-free biosensors and logic systems, Analyst 104 (2015) 2556e2572. [8] J. Liu, Z. Cao, Y. Lu, Functional nucleic acid sensors, Chem. Rev. 109 (2009) 1948e1998. [9] Y. Du, B. Li, E. Wang, “Fitting” makes “sensing” simple: label-free detection strategies based on nucleic acid aptamers, Acc. Chem. Res. 46 (2013) 203e213. [10] F. Wang, C.H. Lu, I. Willner, From cascaded catalytic nucleic acids to enzymeeDNA nanostructures: controlling reactivity, sensing, logic operations, and assembly of complex structures, Chem. Rev. 114 (2014) 2881e2941. [11] L. Wu, X. Qu, Cancer biomarker detection: recent achievements and challenges, Chem. Soc. Rev. 44 (2015) 2963e2997. [12] Y. Guo, L. Xu, S. Hong, Q. Sun, W. Yao, R. Pei, Label-free DNA-based biosensors using structure-selective light-up dyes, Analyst 141 (2016) 6481e6489. [13] Z. Crees, J. Girard, Z. Rios, G.M. Botting, K. Harrington, C. Shearrow, L. Wojdyla, A.L. Stone, S.B. Uppada, J.T. Devito, Oligonucleotides and G-quadruplex stabilizers: targeting telomeres and telomerase in cancer therapy, Curr. Pharm. Des. 20 (2014) 6422e6437. [14] A. Arola, R. Vilar, Stabilisation of G-quadruplex DNA by small molecules, Curr. Top. Med. Chem. 8 (2008) 1405e1415. [15] J. Mohanty, N. Barooah, V. Dhamodharan, S. Harikrishna, P.I. Pradeepkumar, A.C. Bhasikuttan, Thioflavin T as an efficient inducer and selective fluorescent sensor for the human telomeric G-quadruplex DNA, J. Am. Chem. Soc. 135 (2013) 367e376. [16] L. Liu, Y. Shao, J. Peng, C. Huang, H. Liu, L. Zhang, Molecular rotor-based fluorescent probe for selective recognition of hybrid G-quadruplex and as a Kþ sensor, Anal. Chem. 86 (2014) 1622e1631. [17] C.C. Chang, I.C. Kuo, I.F. Ling, C.T. Chen, H.C. Chen, P.J. Lou, J.J. Lin, T.C. Chang, Detection of quadruplex DNA structures in human telomeres by a fluorescent

carbazole derivative, Anal. Chem. 76 (2004) 4490e4494. [18] B.R. Vummidi, J. Alzeer, N.W. Luedtke, Fluorescent probes for G-quadruplex structures, ChemBioChem 14 (2013) 540e558. [19] A.C. Bhasikuttan, J. Mohanty, Targeting G-quadruplex structures with extrinsic fluorogenic dyes: promising fluorescence sensors, Chem. Commun. 51 (2015) 7581e7597. [20] A.R.O. Cousins, D. Ritson, P. Sharma, M.F.G. Stevens, J.E. Moses, M.S. Searle, Ligand selectivity in stabilising tandem parallel folded G-quadruplex motifs in human telomeric DNA sequences, Chem. Commun. 50 (2014) 15202e15205. [21] C. Zhao, L. Wu, J. Ren, Y. Xu, X. Qu, Targeting human telomeric higher-order DNA: dimeric G-quadruplex units serve as preferred binding site, J. Am. Chem. Soc. 135 (2013) 18786e18789. [22] L. Zhang, K. Zhang, S. Rauf, D. Dong, Y. Liu, J. Li, Single-molecule analysis of human telomere sequence interactions with G-quadruplex ligand, Anal. Chem. 88 (2016) 4533e4540. [23] D.-L. Ma, Z. Zhang, M. Wang, L. Lu, H.-J. Zhong, C.-H. Leung, Recent developments in G-quadruplex probes, Chem. Biol. 22 (2015) 812e828. [24] Z. Tan, J. Tang, Z. Kan, H. Hao, Telomere G-quadruplex as a potential target to accelerate telomere shortening by expanding the incomplete end-replication of telomere DNA, Curr. Top. Med. Chem. 15 (2015) 1940e1946. [25] T.M. Ou, Y.J. Lu, J.H. Tan, Z.S. Huang, K.Y. Wong, L.Q. Gu, G-quadruplexes: targets in anticancer drug design, ChemMedChem. 3 (2008) 690e713. [26] Z. Zhang, E. Sharon, R. Freeman, X. Liu, I. Willner, Fluorescence detection of DNA, adenosine-50 -triphosphate (ATP), and telomerase activity by zinc(II)protoporphyrin IX/G-quadruplex labels, Anal. Chem. 84 (2012) 4789e4797. [27] L. Wu, J. Wang, L. Feng, J. Ren, W. Wei, X. Qu, Label-free ultrasensitive detection of human telomerase activity using porphyrin-functionalized graphene and electrochemiluminescence technique, Adv. Mater. 24 (2012) 2447e2452. [28] L. Tian, Y. Weizmann, Real-time detection of telomerase activity using the exponential isothermal amplification of telomere repeat assay, J. Am. Chem. Soc. 135 (2013) 1661e1664. [29] D. Gomez, J.L. Mergny, J.F. Riou, Detection of telomerase inhibitors based on Gquadruplex ligands by a modified telomeric repeat amplification protocol assay, Cancer Res. 62 (2002) 3365e3368. [30] N. An, A.M. Fleming, H.S. White, C.J. Burrows, Nanopore detection of 8oxoguanine in the human telomere repeat sequence, ACS Nano 9 (2015) 4296e4307. [31] Z. Bagheryan, J.B. Raoof, R. Ojani, P. Rezaei, A human telomeric G-quadruplexbased electronic nanoswitch for the detection of anticancer drugs tamoxifen (Tam), Analyst 140 (2015) 4068e4075. [32] L. Xu, X. Shen, S. Hong, J. Wang, Y. Zhang, H. Wang, J. Zhang, R. Pei, Turn-on and label-free fluorescence detection of lead ions based on target-induced Gquadruplex formation, Chem. Commun. 51 (2015) 8165e8168. [33] T. Li, B. Li, E. Wang, S. Dong, G-quadruplex-based DNAzyme for sensitive mercury detection with the naked eye, Chem. Commun. 24 (2009) 3551e3553. [34] M. Liu, Z. Wang, L. Pan, Y. Cui, Y. Liu, A SERS/fluorescence dual-mode nanosensor based on the human telomeric G-quadruplex DNA:application to mercury(II) detection, Biosens. Bioelectron. 69 (2015) 142e147. [35] K.G. Qu, C.Q. Zhao, J.S. Ren, X.G. Qu, Human telomeric G-quadruplex formation and highly selective fluorescence detection of toxic strontium ions, Mol. Biosyst. 8 (2012) 779e782. [36] H.X. Sun, J.F. Xiang, W. Gai, Q. Shang, Q. Li, A.J. Guan, Q.F. Yang, Y. Liu, Y.L. Tang, G.Z. Xu, Visual detection of potassium by a cyanine dye supramolecular aggregate responsive to G-quadruplex motif transition, Analyst 137 (2012) 5713e5715. [37] Y.W. Hua, D.X. Han, Q.X. Zhang, T.S. Wu, F.H. Li, L. Niu, Perylene ligand wrapping G-quadruplex DNA for label-free fluorescence potassium recognition, Biosens. Bioelectron. 38 (2012) 396e401. [38] J. Zhang, J.H. Chen, R.C. Chen, G.N. Chen, F.F. Fu, An electrochemical biosensor for ultratrace terbium based on Tb3þ promoted conformational change of human telomeric G-quadruplex, Biosens. Bioelectron. 25 (2009) 378e382. [39] Y. Xiang, Y. Lu, DNA as sensors and imaging agents for metal ions, Inorg. Chem. 53 (2014) 1925e1942. [40] Y. Guo, W. Yao, Y. Xie, X. Zhou, J. Hu, R. Pei, Logic gates based on G-quadruplexes: principles and sensor applications, Microchim. Acta 183 (2016) 21e34. [41] D. Renciuk, I. Kejnovska, P. Skolakova, K. Bednarova, J. Motlova, M. Vorlickova, Arrangements of human telomere DNA quadruplex in physiologically relevant Kþ solutions, Nucleic Acids Res. 37 (2009) 6625e6634. [42] J. Dai, C. Punchihewa, A. Ambrus, D. Chen, R.A. Jones, D. Yang, Structure of the intramolecular human telomeric G-quadruplex in potassium solution: a novel adenine triple formation, Nucleic Acids Res. 35 (2007) 2440e2450. [43] J. Dai, M. Carver, D. Yang, Polymorphism of human telomeric quadruplex structures, Biochimie 90 (2008) 1172e1183. [44] Y. Xue, Z.Y. Kan, Q. Wang, Y. Yao, J. Liu, Y.H. Hao, Z. Tan, Human telomeric DNA forms parallel-stranded intramolecular G-quadruplex in Kþ solution under molecular crowding condition, J. Am. Chem. Soc. 129 (2007) 11185e11191. [45] H. Yu, X. Gu, S.I. Nakano, D. Miyoshi, N. Sugimoto, Beads-on-a-string structure of long telomeric DNAs under molecular crowding conditions, J. Am. Chem. Soc. 134 (2012) 20060e20069. [46] B. Heddi, A.T. Phan, Structure of human telomeric DNA in crowded solution, J. Am. Chem. Soc. 133 (2011) 9824e9833. [47] G.N. Parkinson, M.P. Lee, S. Neidle, Crystal structure of parallel quadruplexes

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041

T. Wu et al. / Analytica Chimica Acta xxx (2017) 1e9 from human telomeric DNA, Nature 417 (2002) 876e880. €nsel, F. Lo €hr, L. Trantirek, V. Do €tsch, High-resolution insight into G[48] R. Ha overhang architecture, J. Am. Chem. Soc. 135 (2013) 2816e2824. [49] V. Singh, M. Azarkh, M. Drescher, J.S. Hartig, Conformations of individual quadruplex units studied in the context of extended human telomeric DNA, Chem. Commun. 48 (2012) 8258e8260. [50] A. Ambrus, D. Chen, J. Dai, T. Bialis, R.A. Jones, D. Yang, Human telomeric sequence forms a hybrid-type intramolecular G-quadruplex structure with mixed parallel/antiparallel strands in potassium solution, Nucleic Acids Res. 34 (2006) 2723e2735. [51] K.N. Luu, A.T. Phan, V. Kuryavyi, L. Lacroix, D.J. Patel, Structure of the human telomere in Kþ solution: an intramolecular (3þ1) G-quadruplex scaffold, J. Am. Chem. Soc. 128 (2006) 9963e9970. [52] A.T. Phan, V. Kuryavyi, K.N. Luu, D.J. Patel, Structure of two intramolecular Gquadruplexes formed by natural human telomere sequences in Kþ solution, Nucleic Acids Res. 35 (2007) 6517e6525. [53] A.T. Phan, K.N. Luu, D.J. Patel, Different loop arrangements of intramolecular human telomeric (3þ1) G- quadruplexes in Kþ solution, Nucleic Acids Res. 34 (2006) 5715e5719. [54] J. Dai, M. Carver, C. Punchihewa, R.A. Jones, D. Yang, Structure of the hybrid-2 type intramolecular human telomeric G-quadruplex in Kþ solution: insights into structure polymorphism of the human telomeric sequence, Nucleic Acids Res. 35 (2007) 4927e4940. [55] J.A. Punnoose, Y. Cui, D. Koirala, P.M. Yangyuoru, C. Ghimire, P. Shrestha, H. Mao, Interaction of G-quadruplexes in the full-Length 30 human telomeric overhang, J. Am. Chem. Soc. 136 (2014) 18062e18069. [56] N. An, A.M. Fleming, C.J. Burrows, Human telomere G-quadruplexes with five repeats accommodate 8-oxo-7,8-dihydroguanine by looping out the DNA damage, ACS Chem. Biol. 11 (2016) 500e507. [57] L. Petraccone, J.O. Trent, J.B. Chaires, The tail of the telomere, J. Am. Chem. Soc. 130 (2008) 16530e16532. [58] L. Petraccone, N.C. Garbett, J.B. Chaires, J.O. Trent, An integrated molecular dynamics (MD) and experimental study of higher order human telomeric quadruplexes, Biopolymers 93 (2010) 533e548. [59] L. Petraccone, C. Spink, J.O. Trent, N.C. Garbett, C.S. Mekmaysy, C. Giancola, J.B. Chaires, Structure and stability of higher-order human telomeric quadruplexes, J. Am. Chem. Soc. 133 (2011) 20951e20961.
, K. Bedna
rova
, R. Marek, J. Sagi, V. Sklena
r, [60] M. Babinský, R. Fiala, I. Kejnovska M. Vorlí ckov
a, Loss of loop adenines alters human telomere d[AG3(TTAG3)3] quadruplex folding, Nucleic Acids Res. 42 (2014) 14031e14041. [61] V. Singh, M. Azarkh, T.E. Exner, J.S. Hartig, M. Drescher, Human telomeric quadruplex conformations studied by pulsed EPR, Angew. Chem. Int. Ed. 48 (2009) 9728e9730. [62] V. Gabelica, R. Maeda, T. Fujimoto, H. Yaku, T. Murashima, N. Sugimoto, D. Miyoshi, Multiple and cooperative binding of fluorescence light-up probe thioflavin T with human telomere DNA G-quadruplex, Biochemistry 52 (2013) 5620e5628. [63] Y. Kataoka, H. Fujita, Y. Kasahara, T. Yoshihara, S. Tobita, M. Kuwahara, Minimal thioflavin T modifications improve visual discrimination of guaninequadruplex topologies and alter compound-induced topological structures, Anal. Chem. 86 (2014) 12078e12084. [64] A.R. de la Faverie, A. Guedin, A. Bedrat, L.A. Yatsunyk, J.L. Mergny, Thioflavin T as a fluorescence light-up probe for G4 formation, Nucleic Acids Res. 42 (2014) e65. [65] D. Luo, Y. Mu, All-atomic simulations on human telomeric G-quadruplex DNA

9

binding with thioflavin T, J. Phys. Chem. B 119 (2015) 4955e4967. [66] D. Renciuk, I. Kejnovska, P. Skolakova, K. Bednarova, J. Motlova, M. Vorlickova, Arrangements of human telomere DNA quadruplex in physiologically relevant Kþ solutions, Nucleic Acids Res. 37 (2009) 6625e6634. [67] K.W. Lim, S. Amrane, S. Bouaziz, W. Xu, Y. Mu, D.J. Patel, K.N. Luu, A.T. Phan, Structure of the human telomere in Kþ solution: a stable basket-type Gquadruplex with only two G-tetrad layers, J. Am. Chem. Soc. 131 (2009) 4301e4309. [68] L. Liu, Y. Shao, J. Peng, H. Liu, L. Zhang, Selective recognition of ds-DNA cavities by a molecular rotor: switched fluorescence of thioflavin T, Mol. Biosyst. 9 (2013) 2512e2519. [69] D. Luo, Y. Mu, Computational insights into the stability and folding pathways of human telomeric DNA G-quadruplexes, J. Phys. Chem. B 120 (2016) 4912e4926. [70] A.R.O. Cousins, D. Ritson, P. Sharma, M.F.G. Stevens, J.E. Moses, M.S. Searle, Ligand selectivity in stabilising tandem parallel folded G-quadruplex motifs in human telomeric DNA sequences, Chem. Commun. 50 (2014) 15202e15205. [71] R.D. Gray, L. Petraccone, J.O. Trent, J.B. Chaires, Characterization of a Kþinduced conformational switch in a human telomeric DNA oligonucleotide using 2-aminopurine fluorescence, Biochemistry 49 (2010) 179e194. [72] M. Bon cina, J. Lah, I. Prislan, G. Vesnaver, Energetic basis of human telomeric DNA folding into G-quadruplex structures, J. Am. Chem. Soc. 134 (2012) 9657e9663. [73] I. Bessi, H.R.A. Jonker, C. Richter, H. Schwalbe, Involvement of long-lived intermediate states in the complex folding pathway of the human telomeric Gquadruplex, Angew. Chem. Int. Ed. 54 (2015) 8444e8448. [74] R.D. Gray, J. Li, J.B. Chaires, Energetics and kinetics of a conformational switch in G-quadruplex DNA, J. Phys. Chem. B 113 (2009) 2676e2683. [75] A. Arora, C. Balasubramanian, N. Kumar, S. Agrawal, R.P. Ojha, S. Maiti, Binding of berberine to human telomeric quadruplexespectroscopic, calorimetric and molecular modeling studies, FEBS J. 275 (2008) 3971e3983. [76] Y. Sato, S. Nishizawa, K. Yoshimoto, T. Seino, T. Ichihashi, K. Morita, N. Teramae, Influence of substituent modifications on the binding of 2-amino1,8-naphthyridines to cytosine opposite an AP site in DNA duplexes: thermodynamic characterization, Nucleic Acids Res. 37 (2009) 1411e1422. [77] R.M. Abu-Ghazalah, S. Rutledge, L.W.Y. Lau, D.N. Dubins, R.B. Macgregor, A.S. Helmy, Concentration-dependent structural transitions of human telomeric DNA sequences, Biochemistry 51 (2012) 7357e7366. [78] H. Qin, J. Ren, J. Wang, N.W. Luedtke, E. Wang, G-quadruplex-modulated fluorescence detection of potassium in the presence of a 3500-fold excess of sodium ions, Anal. Chem. 82 (2010) 8356e8360. [79] L. Berti, G.A. Burley, Nucleic acid and nucleotide-mediated synthesis of inorganic nanoparticles, Nat. Nanotechnol. 3 (2008) 81e87. [80] T. Li, E. Wang, S. Dong, Lead(II)-induced allosteric G-quadruplex DNAzyme as a colorimetric and chemiluminescence sensor for highly sensitive and selective Pb2þ detection, Anal. Chem. 82 (2010) 1515e1520. [81] C. Yang, L. Liu, T. Zeng, D. Yang, Z. Yao, Y. Zhao, H.C. Wu, Highly sensitive simultaneous detection of lead(II) and barium(II) with G-quadruplex DNA in a-hemolysin nanopore, Anal. Chem. 85 (2013) 7302e7307. [82] J. Seo, E.S. Hong, H.J. Yoon, S.K. Shin, Specific and nonspecific bindings of alkaline-earth metal ions to guanine-quadruplex thrombin-binding aptamer DNA, Int. J. Mass Spectrom. 330 (2012) 262e270. [83] J. Yin, Y. Hu, J. Yoon, Fluorescent probes and bioimaging: alkali metals, alkaline earth metals and pH, Chem. Soc. Rev. 44 (2015) 4619e4644.

Please cite this article in press as: T. Wu, et al., Human telomeric hybrid-2-over-hybrid-1 G-quadruplex targeting and a selective hypersalinetolerant sensor using abasic site-engineered monomorphism, Analytica Chimica Acta (2017), http://dx.doi.org/10.1016/j.aca.2017.01.041