9140
Langmuir 2007, 23, 9140-9142
Intercalation Interactions between dsDNA and Acridine Studied by Single Molecule Force Spectroscopy Chuanjun Liu,† Zhenhua Jiang,† Yiheng Zhang,† Zhiqiang Wang,† Xi Zhang,*,† Fude Feng,‡ and Shu Wang‡ Key Lab of Optoelectronics and Molecular Engineering, Department of Chemistry, Tsinghua UniVersity, Beijing 100084, PR China, and Key Lab of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100080, PR China ReceiVed May 13, 2007. In Final Form: July 17, 2007 In this letter, we report on the direct measurement of the intercalation interactions between acridine and doublestranded DNA (dsDNA) using single molecule force spectroscopy. The interaction between acridine and dsDNA is broken by force of 36 pN at a loading rate of 5.0 nN/s. The most probable rupture force between acridine and dsDNA is dependent on the loading rate, indicating that the binding of acridine and dsDNA is a dynamic process. The combination of SMFS experimental data with the theoretical model clearly suggests the presence of two energy barriers along with an unbinding trajectory of acridine-dsDNA.
In this letter, we report on the direct measurement of intercalation interactions between acridine and double-stranded DNA (dsDNA) using atomic force microscopy (AFM)-based single molecule force spectroscopy (SMFS). In general, guest molecules may associate with dsDNA by groove binding or intercalation.1 Intercalators are the most important group of compounds that interact reversibly with dsDNA and are of great interest to researchers in different fields, including molecular recognition, chemotherapy, and nanomedicine.2,3 Intercalation occurs by the insertion of planar ligands into neighboring base pairs and results from collective interactions of different forces, such as electrostatic interactions, π-stacking, hydrogen bonding, van der Waals, and hydrophobic interactions, depending on the structure of intercalators.1,2 Acridine and its derivatives are typical intercalation agents for dsDNA and are active in antitumor treatment and chemotherapy.3 The investigation of the interaction between intercalators and dsDNA can provide an inherent basis for defining the molecular forces that drive a particular binding reaction and offer a means for understanding in quantitative detail the contributions of the ligands to the binding event.4 Classically, the binding of small molecules to dsDNA is investigated with various thermodynamic and biochemical techniques,5 which rely on ensemble measurements. The rapidly developing AFM-based SMFS technique allows for the measurement of minute force precisely at a single molecule level.6,7 Many intra- and intermolecular interactions have been detected using SMFS, such as antigen-antibody,8 host-guest,9 hydrogen bonding,10 coor* To whom correspondence should be addressed. E-mail: xi@ mail.tsinghua.edu.cn. Tel: +86-10-62796283. Fax: +86-10-62771149. † Tsinghua University. ‡ Chinese Academy of Sciences. (1) (a) Lerman, L. J. Mol. Biol. 1961, 3, 18. (b) Ihmels, E.; Otto, D. Top. Curr. Chem. 2005, 258, 161. (2) (a) Erkkila, K.; Odom, D.; Barton, J. Chem. ReV. 1999, 99, 2777. (b) Hannon, M. Chem. Soc. ReV. 2007, 36, 280. (3) (a) Brana, M.; Cacho, M.; Gradillas, A.; de Pascual-Teresa, B.; Ramos, A. Curr. Pharm. Des. 2001, 7, 1745. (b) Hutchins, R.; Crenshaw, J.; Graves, D.; Denny, W. Biochemistry 2003, 42, 13754. (4) (a) Chaires, J.; Satyanarayana, S.; Suh, D.; Fokt, I.; Przewloka, T.; Priebe, W. Biochemistry 1996, 35, 2047. (b) Chaires, J. Curr. Opin. Struct. Biol. 1998, 8, 314. (5) (a) Breslauer, K.; Remeta, D.; Chou, W.; Ferrante, R.; Curry, J.; Zaunckowski, D.; Snyder, J.; Marky, L. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 8922. (b) Chaires, J. Biopolymers 1997, 44, 201. (6) Krautbauer, R.; Pope, L.; Schrader, T.; Allen, S.; Gaub, H. E. FEBS Lett. 2002, 510, 154.
dination bonding,11 hydrophobic interactions,12 and π-π interactions.13 This letter represents the first attempt to employ SMFS to investigate the unbinding process between acridine and dsDNA and aims to understand the intercalation interactions and the nature of the driving force for the intercalation binding process. To implement the single molecule experiment, we immobilized dsDNA bearing a thiol group at the 5′ terminus onto a gold substrate through a thiol-gold bond and attached acridine to an AFM tip through a flexible poly(ethylene oxide) (PEO) chain (Mn, 6660 g/mol; polydispersity index, 1.15), as shown schematically in Figure 1a. The introduction of PEO as a spacer can provide a means for differentiating the force signals based on the extension length as well as for avoiding the disturbance of the nonspecific interaction between the AFM tip and the substrate.7f,11b,c For the tip modification, the tip was silanized by 3-aminopropyldimethylethoxysilane (APDES) and then was immersed in an N-(1-(9-acridinyl))-1,4-butanediamine-PEOPhCOOH (in brief, acridine-PEO-PhCOOH) solution in the presence of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and N-hydroxysuccinimide (NHS) (Supporting Information). In this way, it allows for acridine-PEO-PhCOOH to be covalently attached to the silanized tip surface. The force measurements were carried out in PBS solution at room temperature. The tip modified with acridine was brought into (7) (a) Janshoff, A.; Neitzert, M.; Oberdo¨rfer, Y.; Fuchs, H. Angew. Chem., Int. Ed. 2000, 39, 3212. (b) Hugel, T.; Seitz, M. Macromol. Rapid Commun. 2001, 22, 989. (c) Zhang, W.; Zhang, X. Prog. Polym. Sci. 2003, 28, 1271. (d) Butt, H.; Cappella, B.; Kappl, M. Surf. Sci. Rep. 2005, 59, 1. (e) Liu, C.; Shi, W.; Cui, S.; Wang, Z.; Zhang, X. Curr. Opin. Solid State Mater. Sci. 2005, 9, 140. (f) Kienberger, F.; Ebner, A.; Gruber, H.; Hinterdorfer, P. Acc. Chem. Res. 2006, 39, 29. (8) Florin, E.; Moy, V.; Gaub, H. E. Science 1994, 264, 415. (9) (a) Scho¨nherr, H.; Beulen, M.; Bu¨gler, J.; Huskens, J.; van Veggel, F.; Reinhoudt, D.; Vancso, G. J. J. Am. Chem. Soc. 2000, 122, 4963. (b) Kado, S.; Kimura, K. J. Am. Chem. Soc. 2003, 125, 4560. (c) Eckel, R.; Ros, R.; Decker, B.; Mattay, J.; Anselmetti, D. Angew. Chem., Int. Ed. 2005, 44, 484. (10) Zou, S.; Scho¨nherr, H.; Vancso, G. J. J. Am. Chem. Soc. 2005, 127, 11230. (11) (a) Conti, M.; Falini, G.; Samorı`, B. Angew. Chem., Int. Ed. 2000, 39, 215. (b) Kudera, M.; Eschbaumer, C.; Gaub, H. E.; Schubert, U. AdV. Funct. Mater. 2003, 13, 615. (c) Kersey, F.; Yount, W.; Craig, S. J. Am. Chem. Soc. 2006, 128, 3886. (12) (a) Cui, S.; Liu, C.; Zhang, W.; Zhang, X.; Wu, C. Macromolecules 2003, 36, 3779. (b) Meadows, P.; Bemis, J.; Walker, G. J. Am. Chem. Soc. 2005, 127, 4136. (c) Ray, C.; Brown, J.; Akhremitchev, B. J. Phys. Chem. B 2006, 110, 17578. (13) Zhang, Y. H.; Liu, C. J.; Shi, W. Q.; Wang, Z. Q.; Dai, L. M.; Zhang, X. Langmuir 2007, 23, 7911.
10.1021/la7013804 CCC: $37.00 © 2007 American Chemical Society Published on Web 08/03/2007
Letters
Langmuir, Vol. 23, No. 18, 2007 9141
Figure 1. (a) Schematic setup for measuring the intercalation interactions between acridine and dsDNA (basic sequence 5′-HS-(CH2)6-ATC TTG ACT ATG TGG GTG CTA ACT C-3′ and it complementary sequence 5′-GAG TTA GCA CCC ACA TAG TCA AGA T-3′) using SMFS. (b) Force curves measured between the AFM tip modified with acridine and the Au substrate immobilized with dsDNA in PBS (phosphate-buffered saline, pH 7.4) at room temperature (∼25 °C).
Figure 2. One force curve measured between the AFM tip modified with acridine and a Au substrate immobilized with dsDNA in PBS. The M-FJC model fitting is shown as a solid line.
contact with the substrate immobilized with dsDNA, making a molecular bridge between the AFM tip and the substrate. Upon retracting the piezo tube, the AFM tip and the substrate were separated. There should appear a force signal of rupture, which may correspond to the rupture of the molecular bridge between the acridine on the tip and the dsDNA on the substrate. We have directly measured the rupture force between acridine and dsDNA on the basis of the above procedures using SMFS. Figure 1b shows some typical force-extension curves (in brief, force curves), which were acquired under a retraction velocity of 500 nm/s. The force rises with the increasing extension of the polymer chain in each curve and then drops to zero suddenly after the rupture point is reached. Only one force signal is observed in the stretching event, and this force signal is completely separated from the nonspecific interaction because of the introduction of a PEO spacer. We have used the modified freely jointed chain (M-FJC) model to describe the elasticity of the polymer chain semiquantitatively,7b,c,14 although it is only a simple approximation (Supporting Information). The Kuhn length of PEO is 0.8 ( 0.1 nm, and the segment elasticity is 150 ( 12 N/m, as shown in Figure 2. These fitting parameters agree well with the reported data for single PEO chain stretching,15 thus indicating that the observed stretching events in the force-extension curve are attributed to the stretching of single PEO spacers. Therefore, the force signal originates from single chain stretching, and the rupture (14) Li, H.; Liu, B.; Zhang, X.; Gao, C.; Shen, J.; Zou, G. Langmuir 1999, 15, 2120. (15) Oesterhelt, F.; Rief, M.; Gaub, H. E. New J. Phys. 1999, 1, 6.1.
Figure 3. (a) Histogram of the rupture length of force curves at a loading rate of 5.0 ((1.5) nN/s. (b) Histogram of rupture force of the interaction between acridine and dsDNA at a loading rate of 5.0 ((1.5) nN/s. The most probable rupture force is 36 pN (with a distribution width of 12 pN).
force signal should be related to the unbinding of acridine and dsDNA. To find out how the force signals are related to the rupture event, we have statistically analyzed the rupture length in all force curves at a loading rate of 5.0 nN/s, as shown in Figure 3a. The rupture length is the distance that the AFM tip is moved away from the substrate until the rupture event occurs.7b,11b The distribution of rupture length is centered at 45 nm from Gaussian fitting. The contour length of the PEO spacer bearing an acridine attached to the tip can be calculated to be 54 nm, assuming that each monomer of PEO spacer molecules has a length of 0.358 nm.15 The rupture length is shifted to a shorter length than the contour length of PEO because acridine-dsDNA is usually ruptured before the PEO molecule is completely extended. There
9142 Langmuir, Vol. 23, No. 18, 2007
Figure 4. Plot of the most probable rupture force between acridine and dsDNA as a function of the corresponding loading rate on a logarithmical scale.
is also the possibility that the reduced rupture length results from the fact that the alignment of the spacer molecules is sometimes not completely parallel to the piezo axis. The broader distribution can be attributed to the polydispersity of the utilized PEO spacer (polydispersity index is 1.15) and the different locations of the modified molecules on the AFM tip with a diameter of 20-50 nm. The distribution of rupture length is centered at 45 nm from the Gaussian fitting, which is almost consistent with the contour length of the PEO spacer of about 54 nm. The consistency of the rupture length with the PEO spacer is crucial evidence of the rupture between acridine and dsDNA. Therefore, we have obtained the most probable rupture force of ∼36 pN by Gaussian fitting as shown in Figure 3b, and the force is ascribed to the rupture between acridine and dsDNA. To provide further evidence that the rupture events result from the unbinding of the acridine-dsDNA complex in the force curves, two control experiments have been performed. One experiment is to occupy the binding sites of dsDNA with excess acridine of 0.1 mg/mL (noted as saturated dsDNA) and then to measure the rupture force between the acridine-modified AFM tip and the saturated dsDNA. The force signals obtained in these control experiments are the same as the rupture between dsDNA and acridine. The only difference lies in the probability of achieving a rupture event, which has been significantly reduced from 4 to 0.8%, suggesting that the rupture events originate from the specific interaction between acridine and dsDNA. It is reasonable that the probability has not dropped to zero, considering the fact that the binding between dsDNA and acridine is a dynamic process.1,3 For another control experiment, we have synthesized PEO-PhCOOH and have utilized the same procedure to modify the AFM tip without acridine. SMFS experiments under the same condition have shown force curves without such rupture events except for nonspecific adhesion peaks. These facts further prove that the rupture events observed for the acridine-modified tips correspond to the specific interaction between acridine and dsDNA. According to the thermally driven unbinding theory of the activated decay of a metastable bound state, the measured rupture forces are not constant, depending on the temporal force evolution on the molecular complex, which is referred to as the loading rate (dF/dt, rf).16 The loading rate can be calculated from the slope of the force-time curve near the rupture point.10 We are wondering whether the most probable rupture force between acridine and dsDNA is dependent on the loading rate. For this (16) (a) Bell, G. Science 1978, 200, 618. (b) Evans, E.; Ritchie, K.; Biophys. J. 1997, 72, 1541. (c) Evans, E. Annu. ReV. Biophys. Biomol. Struct. 2001, 30, 105.
Letters
purpose, several hundred rupture events are measured at different loading rates, and the most probable rupture force at each loading rate is determined from Gaussian fitting (Supporting Information). The most probable rupture force is plotted against loading rate on a logarithmic scale, as shown in Figure 4. There indeed exists a loading rate that is dependent on the intercalation interactions, indicating that the binding of acridine and dsDNA is a dynamic process. Interestingly, there are two force regions with different slopes, as shown in Figure 4. To understand what the different slopes mean, we have employed a model of forced unbinding of receptor-ligand pairs proposed by Bell and Evans to analyze the energy landscape. The logarithmic plots of most probable rupture force (F) versus loading rate (rf) are fitted with the BellEvans model16 as shown in eq 1, where kΒT is the thermal energy (4.11 pN‚nm at 298 K), xβ is the effective distance between the bound and transition states projected along the force vector, rf is the force-dependent loading rate, and koff is the thermal offrate at zero force.
F)
( )
kB T kBT xβ ln(rf) + ln xβ xβ koffkBT
(1)
The slope of the line fit correlates with the potential barrier along the reaction coordinate of the system, yielding xβ. By extrapolation of the fit to zero external force, koff, which is the natural thermal off-rate for the system, can be obtained. The two linear regions correspond to two barriers in the energy landscape of the acridine-dsDNA complex. With this dynamic force spectroscopy (force-loading rate plot), details about the kinetics of the binding and information concerning the length scale of the interaction can be extracted.16 The line fit for low loading rate (Figure 4, dashed line) reveals a dissociation rate at zero force of koff1 ) 1.2 s-1 and a potential width of xβ1 ) 1.18 nm. At high loading rate (Figure 4, solid line) koff2 ) 21.7 s-1 and a potential width of xβ2 ) 0.28 nm are deduced. These two regimes suggest the presence of two energy barriers along with unbinding trajectory of acridine-dsDNA. The origin of the barriers is not fully understood at present. Considering the driving forces for the intercalation binding between acridine and dsDNA,1,3 we have reason to predict that the energy barrier located at 1.18 nm might correspond to long-range interaction, such as hydrophobic interaction in the system, and the energy barrier located at 0.28 nm is related to short-range interactions such as π stacking and hydrogen bonding. In summary, we have directly measured the strength of the intercalation interactions between acridine and dsDNA using SMFS and have employed the Bell-Evans model to analyze its dynamic force spectroscopy. The combination of SMFS experimental data with the theoretical model clearly suggests the presence of two energy barriers along with the unbinding trajectory of acridine-dsDNA. It is anticipated that such a method of chemical modification can be used to study the intercalation interactions of different ligands with dsDNA and could be extended to study not only monovalent but also polyvalent binding. Acknowledgment. We thank the Natural Science Foundation of China (20474035, 20574073) and the National Basic Research Program (2007CB808000) for financial support. We also thank Dr. Huaping Xu for helpful discussions. Supporting Information Available: Synthesis procedures, experimental details, and histograms of rupture forces. This material is available free of charge via the Internet at http://pubs.acs.org. LA7013804
