Spectral Analysis of Interaction between Human Telomeric G-Quadruplex and Liliflorin A, the First Lignan Derivative Interacted with G-Quadruplex DNA

LIU Ting-ting, ZHOU Shuang, JIA Qian-lan, WANG Wen-shu,2*,YAN Xiao-qian, ZHANG Wen-hao, WANG Shuai-qi, JIAO Yu-guo

1. College of Life and Environmental Sciences, Minzu University of China, Beijing 100081, China

2. Beijing Engineering Research Center of Food Environment and Health, Minzu University of China, Beijing 100081, China

3. Center of Biomedical Analysis, Tsinghua University, Beijing 100084, China

Spectral Analysis of Interaction between Human Telomeric G-Quadruplex and Liliflorin A, the First Lignan Derivative Interacted with G-Quadruplex DNA

LIU Ting-ting1, ZHOU Shuang1, JIA Qian-lan1, WANG Wen-shu1,2*,YAN Xiao-qian1, ZHANG Wen-hao3, WANG Shuai-qi1, JIAO Yu-guo1

1. College of Life and Environmental Sciences, Minzu University of China, Beijing 100081, China

2. Beijing Engineering Research Center of Food Environment and Health, Minzu University of China, Beijing 100081, China

3. Center of Biomedical Analysis, Tsinghua University, Beijing 100084, China

Human telomeric G-quadruplex is a four-stranded structure folded by guanines (G) via Hoogsteen hydrogen bonding. The ligands which stabilize the G-quadruplex are often telomerase inhibitors and may become antitumor agents. Here, the interaction between a lignan derivative liliflorin A and human telomeric sequence dGGG(TTAGGG)3G-quadruplex HTG21 were examined by CD, FRET, and NMR spectroscopic methods. In addition, Molecular Docking was used to study the binding of liliflorin A to dTAGGG(TTAGGG)3G-quadruplex HTG23. The CD data showed that liliflorin A enhanced HTG21Tm. TheTmvalue of G-quadruplex was enhanced 3.2 ℃ by 4.0 μmol·L-1liliflorin A in FRET. The NMR spectra of HTG21 showed vivid alteration after reacting with liliflorin A in 3 hours. Molecular Docking suggested liliflorin A bound to the wide groove of HTG23 at G9, G10, G16 and G17. Liliflorin A was the first lignan derivative that could stabilize HTG21 selectively and provided a new candidate for antitumor drug design targeting on human telomeric G-quadruplex.

Liliflorin A; G-quadruplex; Human telomere; Spectral analysis; Interaction

Introduction

In the process of screening bioactive compounds, spectral analysis could give straightforward, vivid and sensitive information for chemical reactions between ligand and biomacromolecule which makes the screening fast and efficiently. Guanine-rich DNA sequencesinvivo, such as telomeric DNA sequence Tel21, Tel26 and oncogene promoter regions (c-myc,bcl-2, or c-kit), can form G-quadruplex via Hoogsteen hydrogen bonding, which plays an important role in many significant bioprocesses[1]. Telomerase is a cancer-specific reverse transcriptase activated in 80%～90% of tumors, and expressed in very low levels or almost undetectable in normal cells[2]. It is reported that when telomeric DNA sequence formed G-quadruplex, it becomes insensitive to the elongation by the telomerase, which is a significant biological process for cells to proliferate[3]. Thus, ligands binding to and stabilizing telomeric G-quadruplex could inhibit the activity of telomerase and induce apoptosis of tumors[4].

However, the low selectivity of the reported ligands over duplex DNA and other DNA folded structures lead to their various bioactivities and inevitable side effects, when they were evaluated as antitumor leads. Accordingly, screening new ligands with high selectivity on human telomeric G-quadruplex is deemed to be an attractive tactic for developing effective antitumor leads[5].

In our previous research, a new lignan named liliflorin A was extracted fromMagnolialilifloraDesr. (Magnoliaceae), and it relieved DNA damages induced by UVB irradiation in rat lymphocyte cells in SCGE assay[6]. It is reported that UVB-irradiation may cause a selective excitation of guanine followed by its oxidative decomposition in the telomeric structure[7]. Thus, it is deduced by us that liliflorin A might interact with G-quadruplex in the telomeric structure, leading to its protective effect in SCGE, which is the motivation of our present study. Herein, due to their sensitivity and efficiency, CD spectra were carried out to investigate whether liliflorin A could stabilize human telomeric G-quadruplex: HTG21 {dGGG(TTAGGG)3}. Furthermore, a series of FRET were recorded to examine the selective binding toward HTG21, the results of which were confirmed by NMR experiments. Finally, Docking was performed to check how and where liliflorin A could interact with HTG23 {dTAGGG(TTAGGG)3} as a confirmation of the spectral analysis. As a result, liliflorin A is the first lignan isolated from plants which can stabilize HTG21 selectively, compared with not only the hairpin loop structure nucleotide F10T, but also the two G-quadruplex formed by oncogene promoter regions sequences c-mycand c-kit. It is a good candidate for antitumor drug design targeting human telomeric G-quadruplex.

1 Materials and methods

1.1 Reagents

DNA (HTG21: 5’-G3(T2AG3)3-3’; F21T: 5’-FAM-G3(T2AG3)3-TAMRA-3’; c-myc2345: 5’-TGAG3TG4-AG3TG4A2-3’; F-myc-T: 5’-FAM-GAG3TG4AG3TG4A2G-TAMRA-3’; c-kit: 5’-AG3AG3CGCTG3AG2AG3-3’; F-kit1: 5’-FAM-G3AG3CGCTG3AG2AG3-TAMRA-3’; ds26: 5’-CA2TCG2ATCGA2T2CGATC2GAT2G-3’; F10T: 5’-FAM-TATAGCTATA-HEG-TATAGCTATA-TAMRA-3’) were purchased from Shanghai Sangon Biotechnology Co. (Shanghai, China), purified by PAGE.

Liliflorin A was abstracted fromMagnolialiliifloraDesr. in our laboratory[6]. Berberine and Quercetin were obtained from National Institute for Food and Drug Control (Beijing, China) and were used without further purification. Deuteriumoxide (D2O) was obtained from Sigma-Aldrich Chemical Co. (Germany). Dimethyl sulfoxide (DMSO) was purchased from Sigma Co. (USA). KCl, NaCl, KH2PO4and K2HPO4were of all analytical reagent grades purchased from Beijing Chem. Co. Tris was purchased from Cambridge Isotope Laboratories, Inc.

1.2 Sample preparation

Liliflorin A, Berberine and Quercetin were initially dissolved as a 50.0 μmol·L-1stock solution in DMSO. The oligomer DNA was heated at 95.0 ℃ for 5 minutes, then slowly cooled to room temperature, and incubated at 4.0 ℃ for 6 hours at least. The ligand-DNA complex were formed by adding small aliquots of compound from 50.0 μmol·L-1solution into the DNA samples in CD and FRET experiments. The solution was equilibrated at room temperature for 24.0 hours before measurements. Final analysis of the CD and FRET data were carried out by Origin 8.0 (OriginLab Corp.).

1.3 CD experiments

The oligomer DNA (HTG21) at a final concentration of 5.0 μmol·L-1was diluted in 10.0 mmol·L-1Tris-HCl buffer (containing 100.0 mmol·L-1NaCl, pH 7.4) to be tested by CD experiments. Experiment was performed at 25.0 ℃ using a Pistar π-180 spectropolarimeter. The scan of the buffer alone was used as the background, which was subtracted from the average scan for each sample. A quartz cuvette with 4 mm path length was used for the spectra recorded over a wavelength range of 230～450 at 1 nm bandwidth, 1 nm step size, and 0.5 s time per point. The CD spectrum data were obtained from 230 to 450 nm[8]. CD-melting experiments were taken at 295 nm and at intervals of 5.0 ℃ over the range 10.0～90.0 ℃, with a constant temperature being maintained for 1s prior to each reading to ensure a stable value[9]. The final data were the average of three measurements.

1.4 FRET experiments

Fluorescence melting curves were determined using a real-time PCR machine (MYIQ2, Bio-rad, USA), with 0.2 μmol·L-1of labeled oligomer DNA (F21T, F10T, F-myc-T, F-kit1) in the 10.0 mmol·L-1Tris-HCl buffer (pH 7.4) containing 60.0 mmol·L-1KCl of a total reaction volume of 20 μL. Fluorescence readings with excitation at 470 nm and detection at 530 nm were taken at intervals of 1.0 ℃ over the range 37.0～99.0 ℃, with a constant temperature being maintained for 30 s prior to each reading to ensure a stable value[9]. As the competitor, a series of double-stranded (ds26) concentration was used by a competitive FRET-melting experiment.

1.5 NMR experiments

The oligonucleotides (HTG21, c-myc2345 and c-kit) were dissolved in 80% phosphate buffer solution (20.0 mmol·L-1KH2PO4/K2HPO4, 70.0 mmol·L-1KCl, 90%H2O/10% D2O, pH 7.4) and 20% DMSO-d6. The known concentration of (0.01 mmol·L-1) Dimethyl-2-silapentane-5-sulfonate (DSS) was used as internal reference. The concentrations of each G-quadruplex recorded in the NMR samples were 1.0 mmol·L-1. As the NMR experiments required relatively high concentration of compounds (1.0 mmol·L-1), but the compounds were insoluble in water at such high concentration, thus, 20% DMSO has been added to enhance the solubility of the compound. Liliflorin A was first dissolved in DMSO-d6 as 200.0 mmol·L-1stock solution. The ligand-quadruplex complex was formed by adding small aliquots of compound from 200.0 mmol·L-1solution into the G-quadruplex samples (HTG21, c-myc2345 and c-kit). The molar ratio of [ligand]/[G-quadruplex] was 1∶1 in the NMR experiments. The solution was equilibrated at room temperature for 24 hours before measurements.1H-NMR spectra of the ligand-quadruplex complex were recorded every one hour.

NMR experiments were performed on a Bruker AVANCE 600 spectrometer equipped with a 5 mm BBI probe capable of delivering z-field gradients. The1H-NMR spectra were recorded by the standard Bruker pulse program p3919gp that applies 3-9-19 pulses with gradients for water suppression, 2.0 s relaxation delay, 64 K data points, 16 ppm spectrum width, 128 scans. All NMR experiments were carried out at 298 K.

1.6 Docking experiments

Calculations were carried out using DockingServer (http://www.dockingserver.com). Gasteiger partial charges were added to the ligand atoms. Non-polar hydrogen atoms were merged, and rotatable bonds were defined. Docking calculations were carried out on untitled protein model. Essential hydrogen atoms, Kollman united atom type charges, and solvation parameters were added with the aid of AutoDock tools[10]. The crystal structure of the telomeric G-quadruplex (PDB ID 2JSM) HTG23 was used as an initial model to study the interaction between the liliflorin A and telomeric DNA. Ligand structures were constructed in Chemdraw.

2 Results and discussion

2.1 Liliflorin A stable human telomeric G-quadruplex HTG21: dGGG(TTAGGG)3in Na+solution

Circular dichroism, CD, is a useful technique to gain information about G-quadruplex DNA. It is also used to monitor the thermal melting and the kinetics of the formation of G-quadruplex[11]. The temperature, at which the G-quadruplex folded structure decomposed into the DNA unfold strand is called the melting temperature (Tm) that can be used to judge the stability of G-quadruplex structure. If ligands bind to and strengthen G-quadruplex structure, theTmvalue of G-quadruplex will be enhanced. By analyzing the melting curves shifts at the sensitive wavelength in the CD spectrum, G-quadruplexTmcan be calculated and used to estimate the stability of complex of ligand binding to G-quadruplex.

HTG21 is reported to form different topological structures in different monovalent cation buffers. In Na+solution, a basket-type structure is formed[12], whereas a mixture of hybrid-1 and hybrid-2 type structures are formed in K+solution[13]. Due to its simplicity, a basket-ball structure was firstly chosen by us to observe if liliflorin A could stabilize HTG21 by CD spectrum. Berberine and Quercetin[14]were used as positive controls. The change of the absorption at 295 nm in the CD spectrum, a typical signal corresponding to HTG21 in Na+was recorded. The concentration of all the compounds was changed gradually from 0.0 to 200.0 μmol·L-1respectively. The data showed that theTmvalue of HTG21 was enhanced in accordance to the incensement of Liliflorin A concentration. The highestTmwas observed at 72.53 ℃ under the concentration of 75.0 μmol·L-1liliflorin A. Compared with theTmvalue of HTG21 only in Na+solution, the ΔTmwas 1.94 ℃. In addition, liliflorin A showed comparative ability on enhancing HTG21Tm, compared to berberine and quercetin at the same concentration (Table 1).

Table 1 The melting temperatures of treated HTG21 (5.0 μmol·L-1strand concentration) after reacting with compounds in a series of concentrations in 10.0 mmol·L-1Tris-HCl buffer and 100 mmol·L-1NaCl at 25.0 ℃

concentrations/(μmol·L-1)TmvalueofHTG21/℃aLiliflorinABerberineQuercetin070 5970 5970 5925 069 8071 0770 3650 070 1071 8671 0575 072 5372 4772 17100 072 3772 5272 79200 072 5371 7073 90

a： All results are expressed as mean ±SE for all groups (n=3)

Due to the higher potassium concentration within the cell, G-quadruplex structures in the presence of K+is more relevant biologically than those topological structures in Na+[15]. Thus, K+solution was used in all the later experiments.

2.2 Liliflorin A selectively stabilize human telomeric G-quadruplex in FRET-melting

Because of its sensitivity and flexibility, fluorescence resonance energy transfer (FRET) is widely used to investigate conformational changes of G-quadruplexes, and also becomes very popular to study the interaction between ligands and G-quadruplex[16]. The melting curve could be described through the normalized fluorescent quenching vs temperature plotting though FRET-melting experiment, due to a large difference between the fluorescence properties of the folded and unfolded doubly labeled oligonucleotides. By the analysis of the fluctuation of the melting curve under the heating process, theTmcan be given to evaluate the stability of complex of ligand binding to G-quadruplex. In the experiment, HTG21 is labeled with a FAM (fluorescent donor) on the 5’ end and a TAMRA (fluorescent acceptor) on the 3’ end. This doubly labeled oligomer was called F21T.

2.2.1 Concentration-dependent experiment of interaction between Liliflorin A and F21T

The melting temperature of F21T in Tris-HCl buffer containing of 60.0 mmol·L-1K+was deeply studied under a series of Liliflorin A concentration. Under the concentration ranging from 1.0 to 4.0 μmol·L-1, the melting curve of F21T indicated a high temperature shift gradually [Fig.1(a)], and the calculated ΔTmwas 0.26, 1.07, 1.61 and 3.22 ℃ respectively. The enhancement ofTmstopped as Liliflorin A concentration increased to 5.0 and 6.0 μmol·L-1[Fig.1(b)]. The data showed that Liliflorin A interacted with F21T, and stabilized the G-quadruplex in a concentration-dependent manner. The highestTm69.93 ℃ appeared at 4.0 μmol·L-1.

Fig.1 FRET experiment was carried in 10.0 mmol·L-1Tris-HCl buffer and 60.0 mmol·L-1KCl

(a): Melting curves of F21T (0.2 μmol·L-1) in the presence of liliflorin A in various concentrations. Curves with normalized FAM fluorescence to a 0-1 range； (b): ΔTmof F21T in the presence of liliflorin A in 1.0, 2.0, 3.0, 4.0, 5.0 and 6.0 μmol·L-1respectively

2.2.2 Competitive FRET-melting experiment

To further prove the binding ability of liliflorin A to F21T, a competitive FRET-melting experiment was studied, in which the excess of unlabeled 26-bp duplex-DNA (ds26) was added to the mixture system of 0.2 μmol·L-1F21T and 4.0 μmol·L-1liliflorin A. As a result, little change of the ΔTmcould be recorded (Fig.2), even when the concentration of ds26 reached 10.0 μmol·L-1, indicating that the duplex-DNA ds26 had no influence on the interaction between F21T and liliflorin A.

Fig.2 Melting curves of the mixture of 0.2 μmol·L-1F21T and 4.0 μmol·L-1liliflorin A in 3.0 and 10.0 μmol·L-1ds26 respectively in 10.0 mmol·L-1Tris-HCl buffer and 60.0 mmol·L-1KCl

2.2.3 Interaction between liliflorin A and F10T by FRET-melting experiment

F10T, was another folded structure of nucleic acids which differed from G-quadruplex in having the hairpin loop structure. In FRET-melting experiment, although liliflorin A with concentration at 1.0, 2.0, 3.0, 4.0, 5.0 and 6.0 μmol·L-1was added into the mixture of F10T respectively, ΔTmof F10T remained unchanged, indicating that no interaction between liliflorin A and the hairpin loop F10T [Fig.3(a)].

2.2.4 Interaction of liliflorin A between F-myc-T and F-kit1 G-quadruplex by FRET-melting experiment

The other two G-quadruplexes found in the promoter regions ofmycandkitgenes were also studied by FRET-melting experiment[17]. F-myc-T and F-kit1, corresponding to the doubly labeled sequences were used respectively. ΔTmof both of the G-quadruplex were almost 0 in all the experiments, showing that the thermal stabilization of the G-quadruplex formed by c-kitsequence and c-mycsequence were not influenced, due to little interaction between liliflorin A and G-quadruplex [Fig.3(b) and (c)].

2.3 Interactions between liliflorin A and HTG21, c-myc2345 and c-kitG-quadruplex in the NMR experiments

Nuclear magnetic resonance spectroscopy (NMR) is an essential tool in the study of G-quadruplex nucleic acids. Imino protons atδ10～12 ppm in1H-NMR spectrum corresponding to guanine imino protons in G-tetrad formation[18]were characteristic signals for G-quadrplex. Changes in the chemical shift values of the relevant imino protons could be observed upon interaction between ligand and G-quadruplex. Depending on the types of the changes, the binding mode and the strength of the binding between the ligand and G-quadruplex can be proposed[5, 19].

Fig.3 Melting curves of 0.2 μmol·L-1(a) F10T, (b) F-myc-T, (c) F-kit1 in the presence of liliflorin A at 1.0, 2.0, 3.0, 4.0, 5.0 and 6.0 μmol·L-1respectively in 10.0 mmol·L-1Tris-HCl buffer and 60.0 mmol·L-1KCl

To verify the selective interaction between liliflorin A and HTG21 observed in FRET,1H-NMR experiments were performed. There were more than 12 guanine imino protons signals atδ10～12 ppm in the1H-NMR of HTG21 G-quadruplex, indicating a mixture of conformations in K+solution[13]. After the addition of liliflorin A into HTG21 solution, the1H-NMR spectra of the mixture was recorded every one hour. It can be observed that five peaks became broad and shifted upfield [Fig.4(a) and Fig.5] gradually from 1 to 3 hours, whereas, no more changes appeared after 3 hours, showing that liliflorin A bound to HTG21 in 3 hours. Moreover, there were not any changes in1H-NMR spectra of the mixture of liliflorin A and c-myc2345 or c-kitG-quadruplex [Fig.4(b) and (c)], suggesting no interaction between liliflorin A and c-myc2345 or c-kitG-quadruplex.

Fig.41H-NMR spectra of 1.0 mmol·L-1(a) HTG21, (b) c-myc2345, (c) c-kitG-quadruplex after reacting with 1 mmol·L-1liliflorin A ina: 0 h,b: 1 h,c: 2 h,d: 3 h, ande: 4 h in 80% phosphate buffer (20.0 mmol·L-1KH2PO4/K2HPO4, 70.0 mmol·L-1KCl, 90%H2O/10% D2O, pH 7.4) and 20% DMSO at 298 K

2.4 Molecular Docking

There has been an increasing interest in using docking method to carry out efficient and robust docking calculations of promising drug candidates[20]. Docking Server is a website that handles all aspects of molecular docking from ligand and bio-macromolecules set-up, provides full control on the setting of specific parameters of ligand and bio-macromolecules set up and docking calculations. Here, human telomere sequence dTAGGG(TTAGGG)3Tel23 was chosen as the G-quadruplex model[21](2JSM in PDB). The two more bases than Tel21 reinforce the G-quadruplex of Tel23, thus its structure model of HTG23 can be found in PDB.

Fig.5 The chemical shifts’ changes of the five guanine imino protons in HTG21 after reacting with liliflorin A for 3 hours

According to the results from Docking, liliflorin A binds to HTG23 in 1∶1 binding stoichiometry (Fig.6). The inhibi-

Fig.6 Hypothetical molecular models showing the interactions of liliflorin A with human telemoric G-quadruplex Tel23 (PDB ID: 2JSM). The loop of G-quadruplex is shown in cartoon, and G-tetrad is shown in sticks, while liliflorin A was represented with the red sticks

tion constant (Ki) is 387.40 μmol·L-1and the free energy of binding is -4.65 kcal·mol-1(Table 2). Moreover, the docking results indicated that liliflorin A binds to HTG23 at G9, G10, G16 and G17 (Fig.6). The calculation results confirmed the interaction between liliflorin A and HTG21 observed by spectroscopic experiments.

Table 2 Binding energies obtained from docking of liliflorin A to HTG23 in the rank of five calculating

3 Conclusions

Here, an efficient screening of natural products which can interact with G-quadruplex by spectroscopic methods was reported. The results of CD, FRET and NMR spectra, as well as the results of Molecular Docking were summarized, and the following conclusions can be drawn:

(1) In CD experiments, compared with berberine and quecertin, the melting temperature of HTG21 in sodium salt buffers was enhanced, indicating liliflorin A interacted with HTG21 and stabilized the G-quadruplex.

(2) In FRET experiments, HTG21Tmpresented variation with the same trend of liliflorin A concentration, the highest ΔTmwas 3.2 ℃ at 4.0 μmol·L-1. In the competitive circumstance, double strings DNA ds26 had no influence on the ΔTmof HTG21. Furthermore, theTmof F10T and G-quadruplex formed by c-mycand c-kitsequence presented little fluctuation regardless of liliflorin A’s existence, indicating the selectivity of liliflorin A toward HTG21.

(3) In NMR experiments, the spectra of HTG21 showed vivid alteration after reacting with liliflorin A in 3 hours. No changes were observed in those spectra of c-myc2345 and c-kitsequence in 3 hours. This phenomenon verified the better selectivity of liliflorin A toward HTG21.

(4) In Molecular Docking, the results suggested liliflorin A binds to HTG23 at G9, G10, G16 and G17，which were located on wide groove at the first and second G-tetrad planes.

Collectively, liliflorin A, isolated fromM.lilifloraby us, was reported here as the first lignan derivative, which presented a new type of ligand of G-quadruplex, and may generate a new candidate for antitumor drug design targeting on human telomeric G-quadruplex.

[1] Maizels N. Nat. Struct. Mol. Bio., 2006, 13: 1055.

[2] Shay J W, Bacchetti S. Eur. J. Cancer, 1997, 33: 787.

[3] Healy K C. Oncol. Res., 1995, 7: 121.

[4] Cosconati S, Marinelli L, Trotta R, et al. J. Am. Chem. Soc., 2009, 131: 16336.

[5] Haudecoeur R, Stefan L, Denat F, et al. J. Am. Chem. Soc., 2013, 135: 550.

[6] Wang W S, Lan X C, Wu H B, et al. Planta. Med., 2012, 78: 141.

[7] Kawai K, Fujitsuka M, Majima T. Chem. Commun., 2005, 1476.

[8] Li Z, Tan J H, He J H, et al. Eur. J. Med. Chem., 2012, 47: 299.

[9] Ma Y, Ou T M, Hou J Q, et al. Bioorg. Med. Chem., 2008, 16: 7582.

[10] Morris G M, Goodsell D S, Halliday R S, et al. J. Comput. Chem., 1998, 19: 1639.

[11] Paramasivan S, Rujan I, Bolton P H. Methods, 2007, 43(4): 324.

[12] Ambrus A, Chen D, Dai J X, et al. Nucleic Acids Res., 2006, 34(9): 2723.

[13] Dai J X, Carver M, Punchihewa C, et al. Nucleic Acids Res., 2007, 35(15): 4927.

[14] Bhadra K, Kumar G S. Biochim. Biophys. Acta, 2011, 1810: 485.

[15] Redon S, Bombard S, Elizondo-Riojas M A, et al. Nucleic. Acids. Res., 2003, 31: 1605.

[17] Dash J, Shirude P S, Hsu S D, et al. J. Am. Chem. Soc., 2008, 130(47): 15950.

[18] Da Silva M W. Methods, 2007, 43: 264.

[19] Mita H, Ohyama T, Tanaka Y, et al. Biochem., 2006, 45(22): 67652.

[20] Sponer J, Spackova N. Methods, 2007, 43: 278.

[21] Phan A T, Kuryavyi V, Luu K N, et al. Nucleic. Acids. Res., 2007, 35(19): 6517.

O657.3

A

紫玉兰素A与人端粒G-四链体相互作用的光谱学研究，第一个木脂素类衍生物与G-四链体相互作用

刘婷婷1， 周 爽1， 贾千澜1， 王文蜀1,2*， 闫晓倩1， 张文浩3， 王帅旗1， 焦玉国1

1. 中央民族大学生命与环境科学学院， 北京 100081

2. 中央民族大学， 北京市食品环境与健康工程研究中心， 北京 100081

3. 清华大学， 生物医学测试中心， 北京 100084

人端粒G-四链体结构是指端粒末端富含鸟嘌呤(G)的DNA 序列在一价阳离子(如K+和Na+)诱导下通过G碱基间Hoogsteen氢键连接形成的DNA二级结构。 能够稳定端粒G-四链体的配体通常为端粒酶抑制剂， 其可能成为抗肿瘤药物。 应用CD， FRET， NMR光谱方法第一次较全面地研究了一种木脂素衍生物， 紫玉兰素A (liliflorin A)与人端粒序列dGGG(TTAGGG)3G-四链体HTG21的相互作用， 采用分子对接技术进一步研究紫玉兰素A与人端粒序列dTAGGG(TTAGGG)3G-四链体HTG23的结合位点。 CD实验数据表明紫玉兰素A提高HTG21解链温度， FRET实验测得4.0 μmol·L-1紫玉兰素A可以将HTG21稳定温度提高3.2 ℃。 NMR实验结果表明， 加入紫玉兰素A三小时后HTG21核磁谱图出现明显变化。 分子对接结果表明紫玉兰素A结合到HTG23较宽沟槽上， 结合位点为G9， G10， G16和G17。 紫玉兰素A是第一个能够选择性稳定人端粒G-四链体HTG21的木脂素类衍生物配体。 实验结果为以人端粒G-四链体为靶点的抗肿瘤药物设计提供了新型候选化合物。

紫玉兰素A； G-四链体； 人端粒； 光谱分析； 相互作用

2015-05-08，

2015-11-02)

2015-05-08; accepted： 2015-11-02

The National Natural Science Foundation of China (31200260), The First-class University and the First-rate Discipline Construction Projects of Minzu University of China (YLDX01013, 2015MDTD08C), together with 111 Project (B08044), The National College Students’ innovation and entrepreneurship training program (GCCX 2014110017, GCCX 2015110012)

10.3964/j.issn.1000-0593(2016)03-0896-07

Biography： LIU Ting-ting, (1987—), Doctoral Candidate of College Life and Environmental Sciences, Minzu University of China e-mail: liutingting1204@163.com *Corresponding author e-mail: wangws@muc.edu.cn

*通讯联系人