The kinetics of YOYO-1 intercalation into single molecules of double-stranded DNA

Biochemical and Biophysical Research Communications 403 (2010) 225–229

Contents lists available at ScienceDirect

Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc

The kinetics of YOYO-1 intercalation into single molecules of double-stranded DNA Marcel Reuter 1, David T.F. Dryden ⇑ School of Chemistry and COSMIC, The University of Edinburgh, West Mains Road, The King’s Buildings, Edinburgh EH9 3JJ, UK

a r t i c l e

i n f o

Article history: Received 19 October 2010 Available online 10 November 2010 Keywords: Lambda DNA YOYO-1 DNA intercalator Single molecule imaging TIRF microscopy Microfluidics

a b s t r a c t The cyanine dye, YOYO-1, has frequently been used in single DNA molecule imaging work to stain double-stranded DNA as it fluoresces strongly when bound. The binding of YOYO-1 lengthens the DNA due to bis-intercalation. We have investigated the kinetics of binding, via this increase in DNA length, for single, hydrodynamically-stretched molecules of lambda DNA observed via Total Internal Reflection Fluorescence (TIRF) microscopy. The rate and degree of lengthening in 40 mM NaHCO3 (pH 8.0) buffer depend upon the free dye concentration with the reaction taking several minutes to reach completion even in relatively high, 40 nM, concentrations of YOYO-1. In the absence of overstretching of the DNA molecule, we determine the second order rate constant to be 3.8 ± 0.7  105 s1 M1, the dissociation constant to be 12.1 ± 3.4 nM and the maximum DNA molecule extension to be 36 ± 4%. The intercalation time constant (inverse of the pseudo-first order rate constant), s, decreased from 309 to 62 s as YOYO-1 levels increased from 10 to 40 nM. The kinetics of binding help with interpretation of the behavior of DNA–YOYO-1 complexes when overstretched and establish defined conditions for the preparation of DNA–YOYO-1 complexes. Ó 2010 Elsevier Inc. All rights reserved.

1. Introduction In the last decade single-molecule experiments have revolutionised the fields of chemistry, biology and biophysics [1–4]. For example, by looking at individual molecular trajectories such important issues as protein–protein and protein–DNA interactions have been addressed while avoiding the averaging effects inherent to traditional ensemble methods. One important scaffold to study the interaction of proteins with double-stranded DNA (dsDNA) at the single molecule level is surface-tethered or microbeadtethered, stretched bacteriophage lambda DNA [5]. Commonly, DNA is stretched by the application of either optical tweezers, magnetic tweezers or hydrodynamic flow in microfluidic devices and visualized by epi-fluorescence or Total Internal Reflection Fluorescence (TIRF) microscopy [6–15] using the DNA bis-intercalating dye YOYO-1, a member of a diverse range of cyanine dyes, which fluoresces strongly only when bound to DNA [6,16–23]. While using the lambda DNA scaffold for single molecule studies of the prokaryotic homologous recombination enzyme AddAB from Bacteroides fragilis [15], we observed that YOYO-1 present in the flow buffer at concentrations larger than 10 nM was able

⇑ Corresponding author. E-mail address: [email protected] (D.T.F. Dryden). Present address: Erasmus Medical Centre, Department of Genetics, Dr. Molewaterplein 50, 3015GE Rotterdam, The Netherlands. 1

0006-291X/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2010.11.015

to intercalate into and lengthen dsDNA within minutes, but at concentrations between 0 and 5 nM YOYO-1 no DNA length change was detected. This lengthening effect originates from the fact that each YOYO-1 molecule separates two pairs of neighboring DNA base pairs by 0.4 nm [24,25]. Recently, the behavior of dsDNA in the overstretching region of DNA in combination with a high YOYO-1 concentration of 100 nM using a single-molecule optical tweezers instrument has been described [7,12,14,26]. It was found that stretched dsDNA–YOYO-1 would, if rapidly stretched further, take up further molecules of YOYO-1 to re-establish equilibrium conditions on a time scale of order 1 s. However, in general, experimental staining procedures have typically specified incubation times between DNA and YOYO-1 of 2–120 min [13,18,19,23] and no consensus is apparent in the literature. Given the frequent use of YOYO-1 in single-molecule experiments, it is curious that as far as we are aware, no analysis of the kinetics of YOYO-1 binding has been reported. Earlier kinetic studies on DNA interactions with other dyes and intercalators focussed mainly on typical bulk biochemical conditions [22,27–29]. The second order rate constants for binding were found to be much lower than expected for a diffusion-controlled reaction. The low values were attributed to binding sites on the DNA only rarely becoming available due to fluctuations in the structure of successive base pairs. In this paper, we characterize the kinetics of YOYO-1 intercalation into single molecules of hydrodynamically-stretched dsDNA as a function of the YOYO-1 concentration present in the flow buffer.

226

M. Reuter, D.T.F. Dryden / Biochemical and Biophysical Research Communications 403 (2010) 225–229

2. Materials and methods All experimental details of the microscope, microfludics, YOYO-1 (Invitrogen) and DNA preparation have been described in exhaustive detail elsewhere [15]. DNA from bacteriophage lambda (New England Biolabs) was tethered to the surface via a Neutravidin–biotin linkage. For imaging experiments a filter-sterile 40 mM NaHCO3 (pH 8.0) buffer containing 0.04 mg/ml catalase, 0.5 mg/ml glucose oxidase, 4 mg/ml b-mercaptoethanol and 12.5% w/v glucose was used. YOYO-1 intercalation was triggered by varied concentrations of YOYO-1 in the same buffer. Two MatLab (MathWorks, UK) routines were written for kymograph presentation and time-dependent DNA length data extraction. The latter routine calculated the mean pixel value for each timeline in a kymograph image and set all pixel values in this time line that are smaller than the mean to zero. The first and last two subsequent non-zero values in one time line were taken as molecule start and end, respectively. Data from dark frames, which are a result of occasional shutter and camera miscommunication, were deleted from the traces before fitting. Origin (Microcal, UK) was used for fitting and presenting the kinetic data. 3. Results Individual surface-tethered lambda DNA molecules, pre-stained at a minimal dye to base pair ratio of 1/100 so that they could be located in the field of view and then stretched, were observed to lengthen when different concentrations of YOYO-1 were introduced into the flow (Figs. 1 and 2). Comparing the average DNA length observed prior to the addition of further YOYO-1, which is around 15 lm at the shear rate employed, to the reported literature value of the contour length of around 16 lm [1], we estimate that we apply less than an average of 5 pN of force on the molecule by means of hydrodynamic flow in the used microfluidic devices. Hence the double-stranded structure of the DNA is not being overstretched and distorted. Kymograph images showing this lengthening were produced from TIRF videos acquired at these different YOYO-1 concentrations (Fig. 2). The initial stretching when the shear flow was started is complete in <2 s (Fig. 1). Each kymograph shows a selected single molecule as it appears in successive camera frames. In this way time-dependent changes become obvious to the eye. The YOYO-1 concentration is in great excess over the DNA concentration and, given the strong affinity between YOYO-1 and dsDNA (the association constant is believed to be >1010 M1) [16,18,19], the observed kinetics should be pseudo-first order. Between 10 and 20 individual lambda DNA molecules were analyzed for each YOYO-1 concentration and their lengths extracted over time using a MatLab routine (Fig. 2). From these

traces and the kymograph images, it was clear that YOYO-1 intercalation into lambda DNA followed a mono-exponential decay described by the equation l ¼ lend  ðlend  lstart Þ  expðt=sÞ. In this equation, lstart and lend are, respectively, the apparent DNA molecule length l before and after YOYO-1 intercalation (which does not proceed to saturation at lower YOYO-1 concentrations during the observation time); t is the time and s is the time constant. During the process of YOYO-1 intercalation, the number of free DNA sites gradually decreases thereby giving rise to an exponential behavior. Some lambda DNA molecules were shorter than expected which was attributed to either photo-cleavage or shear forces breaking the molecule. Nevertheless, the initial apparent DNA length had no influence on the time constant s of YOYO-1 intercalation. The distributions of the individual time constants, s, are shown in Fig. 3. At YOYO-1 concentrations at and below 5 nM, no additional dye intercalation into dsDNA was detected within 10 min. At higher YOYO-1 concentrations, DNA intercalation was observed within 10 min. The intercalation time constants, s, decreased from 309 to 62 s as YOYO-1 levels increased from 10 to 40 nM. A time constant of 35 s has been previously determined with 100 nM YOYO-1 in a glucose-containing buffer similar to that used in our experiments [7]. Our data upon extrapolation would agree with this value. While the distribution of s was asymmetric for the two YOYO-1 concentrations of 10 and 20 nM, it became symmetric at a concentration of 40 nM. For the first two conditions, the mean value of s is larger than the median of the data set. The rate law for our observations is rate = k  [YOYO-1]  [binding sites on DNA]. Calculating the pseudo-first order rate constants (1/s = kpseudo = k  [YOYO-1]) from the average s allows a second order rate constant, k, for the interaction of 3.8 ± 0.7  105 s1 M1 to be calculated. During the intercalation of YOYO-1 into lambda DNA, the relative DNA molecule length increased by 16 ± 4%, 24 ± 7% and 27 ± 7% for 10, 20 and 40 nM YOYO-1, respectively (Fig. 4). As each YOYO-1 molecule stretches DNA by 0.4 nm and covers 4 DNA base pairs [24,25], one would expect a maximum length change of 29%. A simple fit (Fig. 4) using a one site binding equation (fractional extension = {maximum extension}  [YOYO-1]/{Kd + [YOYO1]}) to our data indicates a maximum extension of 36 ± 4% corresponding to one YOYO-1 per 3 base pairs and a dissociation constant, Kd, of 12.1 ± 3.4 nM, a bit lower than expected from previous estimates of association constants (Ka = 1/Kd) of >1010 M1 [16,18,19]. This discrepancy can be attributed to using an overly simple binding equation rather than a complicated binding equation which accounts for overlapping of sites and dye distribution along the DNA as our data are not of sufficient resolution to justify its use [6,30]. At 40 nM YOYO-1 one reaches almost complete DNA saturation within roughly 100 s, a time scale for interaction also found in recent literature [12,14,26], although very high concentrations of intercalator, in which the mode of intercalation is poorly

Fig. 1. Colour-inverted TIRF image sequence showing the hydrodynamic stretching of a surface-tethered single molecule of lambda DNA stained with YOYO-1 subjected to a shear rate of 510 s1 commencing at 1.6 s.

M. Reuter, D.T.F. Dryden / Biochemical and Biophysical Research Communications 403 (2010) 225–229

227

Fig. 2. Kymograph images and DNA length plots over time (l vs. t) showing the lengthening of individual lambda DNA molecules as a function of YOYO-1 concentration in the flow buffer. A constant shear rate of 510 s1 was maintained during the image acquisition. The signal-to-noise ratio differs between individual kymographs, depending on whether they were located in the center or at the edge of the evanescent-excitation field. The horizontal banding visible in some kymographs is an image acquisition artefact and can be ignored. The time constant, s, for each fitted curve is given in the insets.

defined, were used in these studies. These timescales are much shorter than those suggested by the supplier [23]. 4. Discussion Our data show that the binding of YOYO-1 lengthens individual DNA molecules on the second to minute timescale with a

saturation of one YOYO-1 per three base pairs being reached if enough dye is present. This result and the determination of binding affinity are in reasonable agreement with previously published work. We have been unable to find a value for the rate constant for binding of YOYO-1 to DNA in the literature so we have determined the second order rate constant for binding. This rate constant is much lower than expected for a diffusion-controlled

228

M. Reuter, D.T.F. Dryden / Biochemical and Biophysical Research Communications 403 (2010) 225–229

Fig. 3. Box plot of the distribution of YOYO-1 intercalation time constants, s, depending on the YOYO-1 concentration. It shows the mean of the data (indicated by the small rectangle within the box), the median and the lower and upper quartiles of the distribution (indicated by the lines across the box), the whiskers (vertical lines) which represent the standard deviation, as well as the smallest and largest values of the data set (two crosses). The number of DNA molecules analyzed was 16, 10 and 15 for 10, 20 and 40 nM YOYO-1, respectively.

40 35

% Extension

30 25 20 15 10 5 0 0

5

10 15 20 25 30 35 40 45 50 [YOYO] / nM

Fig. 4. The percentage increase in length of the DNA molecules as a function of YOYO-1 concentration after 600 s incubation time. The data are fitted to a single site binding equation giving a maximum extension of 36 ± 4% and a dissociation constant of 12.1 ± 3.4 nM.

reaction but this is in agreement with older results on other intercalators obtained in ensemble experiments [22,27–29]. The low values were attributed to binding sites on the DNA only rarely becoming available due to fluctuations in the structure of successive base pairs. In other words, the binding process involved an initial diffusion-limited encounter complex between the DNA and the dye, followed by a slow process of intercalation reliant upon relatively rare distortions of the stacking of base pairs to allow ingress of the dye into the base pair stack. Our kinetic data can help with the interpretation of other single molecule measurements on complexes of YOYO-1 and DNA. It has frequently been noticed that the force extension behavior of DNA in single molecule stretching experiments changes when YOYO-1 is present [7,12,14,26]. These experiments have found that the persistence length of the DNA [31] becomes shortened as the concentration of YOYO-1 increases and the rate of stretching influences the behavior of the DNA. The change in persistence length at high forces, the ‘‘overstretching regime’’, as a function of [YOYO-1] has recently been explained. Direct microscopic observation of single DNA molecules stained with YOYO-1 has shown regions of single-stranded DNA (ssDNA) appearing upon overstretching [11,14]. These ssDNA regions do not fluoresce and appear as gaps in the image of the DNA. Thus the variation in the ‘‘average’’ persistence length as a

function of [YOYO-1] can be explained by the DNA containing both double-stranded and single-stranded regions, each with their expected persistence length. The proportion of each region changes as the force changes. However, ssDNA is not present at lower forces but the persistence length was still reported to decrease as YOYO-1 was added. More recently Günther et al. failed to observe any change in persistence length upon addition of YOYO-1 [13]. We suggest that this may be due to a difference in experimental conditions. Early reports using YOYO-1 have a single tethered DNA molecule either in a defined YOYO-1 concentration ([11,12,14,26], our data) or in buffer after removal of unbound YOYO-1 [11] whereas Günther et al. [13] bathe their tethered DNA molecule in a solution of not only defined YOYO1 concentration but also a high, constant concentration of free DNA. Thus, in some of their experiments, there will be no unbound YOYO-1 at all. Thus stretching will not be influenced by the kinetics of YOYO-1 binding to any freshly exposed sites on the DNA. In the earlier experiments it was suggested that the dependence on the stretching rate and the concentration of YOYO-1 in solution indicated that the system was out of equilibrium during the measurements [12,14,26]. In these stretching experiments, the single dsDNA molecule is bathed in a solution of defined YOYO-1 concentration. Changing the conformation of the DNA would therefore allow the release or uptake of YOYO-1 as it is in vast excess in terms of molecular numbers. During fast stretching (1 s in total), it was observed that the persistence length was apparently greater than when the stretching was slow (20 s in total). Furthermore when stretching stopped, the force on the DNA relaxed on the second timescale. Our second order rate constant is consistent with these observations. The YOYO-1 concentrations used (i.e. 100 nM) would have a time constant of 20 s from extrapolating our data for binding (Fig. 3). The fast stretching is complete before significant amounts of YOYO-1 can bind to sites on the DNA revealed during the stretching. During slow stretching, the YOYO-1 has time to bind and hence the amount of dye bound changes throughout the stretching process making analysis extremely complex. Thus our measurement of the second order rate constant for binding of YOYO-1 and DNA, a basic parameter long overdue for determination, contributes to the understanding of DNA stretching experiments and should be measured routinely in future single-molecule experiments performed under different solvent conditions. Acknowledgments We thank Andrew Garry (COSMIC) for technical assistance, Drs. Jochen Arlt (COSMIC) and Cristina Flors (Chemistry) for optics advice and Prof. Eric Greene (Columbia University) for advice on DNA stretching. M.R. was supported by the Marie-Curie Network ‘‘From FLIM to FLIN’’ (MRTN-CT-2005-019481) and the RASOR grant from the BBSRC (BB/C511599/1). M.R. acknowledges an equipment grant from the University of Edinburgh Development Trust (ID 3072). D.T.F.D. gratefully acknowledges The Master and Fellows of Emmanuel College, Cambridge for the award of a Derek Brewer Visiting Fellowship during which time this manuscript was prepared. References [1] J.F. Allemand, D. Bensimon, V. Croquette, Stretching DNA and RNA to probe their interactions with proteins, Curr. Opin. Struct. Biol. 13 (2003) 266–274. [2] C. Joo, H. Balci, Y. Ishitsuka, C. Buranachai, T. Ha, Advances in single-molecule fluorescence methods for molecular biology, Annu. Rev. Biochem. 77 (2008) 1– 26. [3] J. van Mameren, E.J. Peterman, G.J. Wuite, See me, feel me: methods to concurrently visualize and manipulate single DNA molecules and associated proteins, Nucleic Acids Res. 36 (2008) 4381–4389. [4] M.L. Visnapuu, D. Duzdevich, E.C. Greene, The importance of surfaces in single molecule bioscience, Mol. Biosyst. 4 (2008) 394–403.

M. Reuter, D.T.F. Dryden / Biochemical and Biophysical Research Communications 403 (2010) 225–229 [5] N.E. Murray, A. Gann, What has phage lambda ever done for us? Curr. Biol. 17 (2007) R305–R312. [6] S. Gurrieri, K.S. Wells, I.D. Johnson, C. Bustamante, Direct visualization of individual DNA molecules by fluorescence microscopy: characterization of the factors affecting signal/background and optimization of imaging conditions using YOYO, Anal. Biochem. 249 (1997) 44–53. [7] M.L. Bennink, O.D. Schärer, R. Kanaar, K. Sakata-Sogawa, J.M. Schins, J.S. Kanger, J. Greve, B.G. de Grooth, Single-molecule manipulation of doublestranded DNA using optical tweezers: interaction studies of DNA with RecA and YOYO-1, Cytometry 36 (1999) 200–208. [8] P.R. Bianco, L.R. Brewer, M. Corzett, R. Balhorn, Y. Yeh, S.C. Kowalczykowski, R.J. Baskin, Processive translocation and DNA unwinding by individual RecBCD enzyme molecules, Nature 409 (2001) 374–378. [9] E. Theofanidou, T.J. Su, J. Arlt, D.T.F. Dryden, W.J. Hossack, J. Crain, W.C.K. Poon, Condensation of hydrodynamically stretched using single molecule fluorescence imaging and optical tweezers, Proc. SPIE 5514 (2004) 225–233. [10] A. Graneli, C.C. Yeykal, T.K. Prasad, E.C. Greene, Organized arrays of individual DNA molecules tethered to supported lipid bilayers, Langmuir 22 (2006) 292– 299. [11] J. van Mameren, P. Gross, G. Farge, P. Hooijman, M. Modesti, M. Falkenberg, G.J.L. Wuite, E.J.G. Peterman, Unraveling the structure of DNA during overstretching by using multicolour single-molecule fluorescence imaging, Proc. Natl. Acad. Sci. USA 106 (2009) 18231–18236. [12] C.U. Murade, V. Subramaniam, C. Otto, M.L. Bennink, Interaction of oxazole yellow dyes with DNA studied with hybrid optical tweezers and fluorescence microscopy, Biophys. J. 97 (2009) 835–843. [13] K. Günther, M. Mertig, R. Seidel, Mechanical and structural properties of YOYO-1 complexed DNA, Nucleic Acids Res. (2010), doi:10.1093/nar/gkq434. [14] C.U. Murade, V. Subramaniam, C. Otto, M.L. Bennink, Force spectroscopy and fluorescence microscopy of dsDNA–YOYO-1 complexes: implications for the structure of dsDNA in the overstretching region, Nucleic Acids Res. 38 (2010) 3423–3431. [15] M. Reuter, F. Parry, D.T.F. Dryden, G.W. Blakely, Single-molecule imaging of Bacteroides fragilis AddAB reveals the highly processive translocation of a single motor helicase, Nucleic Acids Res. 38 (2010) 3721–3731. [16] A.N. Glazer, H.S. Rye, Stable dye–DNA intercalation complexes as reagents for high-sensitivity fluorescence detection, Nature 359 (1992) 859–861. [17] A. Larsson, C. Carlsson, M. Jonsson, B. Albinsson, Characterization of the binding of the fluorescent dyes YO and YOYO to DNA by polarized light spectroscopy, J. Am. Chem. Soc. 116 (1994) 8459–8465.

229

[18] C. Carlsson, M. Jonsson, B. Akerman, Double bands in DNA gel electrophoresis caused by bis-intercalating dyes, Nucleic Acids Res. 23 (1995) 2413–2420. [19] H.S. Rye, A.N. Glazer, Interaction of dimeric intercalating dyes with single stranded DNA, Nucleic Acids Res. 23 (1995) 1215–1222. [20] G. Cosa, K.S. Focsaneanu, J.R.N. McLean, J.P. McNamee, J.C. Scaiano, Photophysical properties of fluorescent DNA–dyes bound to single- and double-stranded DNA in aqueous buffered solution, Photochem. Photobiol. 73 (2001) 585–599. [21] A. Dietrich, V. Buschmann, C. Müller, M. Sauer, Fluorescence resonance energy transfer (FRET) and competing processes in donor–acceptor substituted DNA strands: a comparative study of ensemble and single-molecule data, Rev. Mol. Biotechnol. 82 (2002) 211–231. [22] T. Biver, A. de Biasi, F. Secco, M. Venturini, S. Yarmoluk, Cyanine dyes as intercalating agents: kinetic and thermodynamic studies on the DNA/Cyan40 and DNA/CCyan2 systems, Biophys. J. 89 (2005) 374– 383. [23] Invitrogen website, , Product Information: Dimeric Cyanine Nucleic Acid Stains. [24] H.P. Spielmann, D.E. Wemmer, J.P. Jacobsen, Solution structure of a DNA complex with the fluorescent bis-intercalator TOTO determined by NMR spectroscopy, Biochemistry 34 (1995) 8542–8553. [25] F. Johansen, J.P. Jacobsen, 1H NMR studies of the bis-intercalation of a homodimeric oxazole yellow dye in DNA oligonucleotides, J. Biomol. Struct. Dyn. 16 (1998) 205–222. [26] A. Sischka, K. Toensing, R. Eckel, S.D. Wilking, N. Sewald, R. Ros, D. Anselmetti, Molecular mechanisms and kinetics between DNA and DNA binding ligands, Biophys. J. 88 (2005) 404–411. [27] R. Bittman, L. Blau, Stopped-flow kinetic studies of actinomycin binding to DNAs, Biochemistry 14 (1975) 2138–2145. [28] K.R. Fox, M.J. Waring, Stopped-flow kinetic studies on the interaction between echinomycin and DNA, Biochemistry 23 (1984) 2627–2633. [29] R.B. Macgregor Jr., R.M. Clegg, T.M. Jovin, Viscosity dependence of ethidium– DNA intercalation kinetics, Biochemistry 26 (1987) 4008–4016. [30] H. Flyvbjerg, S.A. Keatch, D.T.F. Dryden, Strong physical constraints on sequence-specific target location by proteins on DNA molecules, Nucleic Acids Res. 34 (2006) 2550–2557. [31] G.S. Manning, The persistence length of DNA is reached from the persistence length of its null isomer through an internal electrostatic stretching force, Biophys. J. 91 (2006) 3607–3616.