Chinese Chemical Letters  2015, Vol.26 Issue (12): 1490-1495   PDF    
Base pair distance analysis in single DNA molecule by direct stochastic optical reconstruction microscopy
Suresh Kumar Chakkarapania, Guenyoung Parka, Seong Ho Kanga,b     
a Department of Chemistry, Graduate School, Kyung Hee University, Yongin-si, Gyeonggi-do 17104, Republic of Korea;
b Department of Applied Chemistry and Institute of Natural Sciences, Kyung Hee University, Yongin-si, Gyeonggi-do 17104, Republic of Korea
Abstract: Precise fluorescence imaging of single λ-DNA molecules for base pair distance analysis requires a superresolution technique, as these distances are on the order of diffraction limit. Individual λ-DNA molecules intercalated with the fluorescent dye YOYO-1 were investigated at subdiffraction spatial resolution by direct stochastic optical reconstructionmicroscopy (dSTORM). Various dye-to-DNA base pair ratios were imaged by photoswitching YOYO-1 between the fluorescent state and the dark state using two laser sources. The acquired images were reconstructed into a super-resolution image by applying Gaussian fitting to the centroid of the point spread function. By measuring the distances between localized fluorophores, the base pair distances in single DNA molecules for dye-to-DNA base pair ratios of 1:50, 1:100, and 1:500 were calculated to be 17.1±0.8 nm, 34.3±2.2 nm, and 170.3±8.1 nm, respectively, which were in agreement with theoretical values. These results demonstrate that intercalating dye in a single DNA molecule can be photoswitched without the use of an activator fluorophore, and that super-localization precision at a spatial resolution of ~17 nm was experimentally achieved.
Key words: DNA     Single-molecule study     Subdiffraction resolution     Super-localization    
1. Introduction

Single molecule detection with high sensitivity and molecular specificity is a valuable tool for imaging and analyzing biological species at nanoscale resolution [1, 2]. Single molecule DNA studies have provided insight for genomic studies and analysis of DNAprotein binding sites [3]. DNA base pair studies have also provided valuable information regarding the determination of mRNA sequences,which determine the genetic code of living organisms [4]. In fluorescence detection,intercalating fluorescent dyes have mainly been used to label individual DNA molecules [5] with different dye-to-DNA base pair ratios. Until now,however,the exact location of the intercalating dye in a single DNA molecule has not been experimentally observed.

By localizing the position of the intercalating dye between base pairs of a single DNA molecule,the distance between adjacent base pairs could be measured. As the distance between adjacent base pairs in λ-DNA is approximately 0.34 nm,a super-resolution technique [6, 7, 8] is required to localize the position of the intercalating dye. Recently,super-resolution techniques such as photoactivated localization microscopy [9],fluorescence photoactivated localization microscopy [10],stochastic optical reconstruction microscopy (STORM) [11, 12],and direct STORM (dSTORM) [13, 14] have been reported to achieve a subdiffraction- limit image resolution to visualize fluorescence images at sub-nanometer resolution. In dSTORM,fluorophores are photoswitched between the fluorescent on and off states by an excitation laser and an activation laser in the absence of an activator fluorophore. Once the activated fluorophores are detected,they get into dark state and another set of fluorophores is activated (Fig. 1a) [14]. Transition between the fluorescent on and off states can be driven by a photoswitching buffer consisting of additives such as thiols [15, 16] or redox agents [13, 17]. Only a subset of fluorophores is activated at a given time point and their point spread function is detected. The number of fluorophores that are active at a given point in time can be controlled by adjusting the power of the excitation and activation lasers. The cycle between the fluorescent on and off states is repeated for separate fluorophore emissions at given time intervals. Fluorophore position in each frame can be localized by fitting the centroid of the point spread function (PSF) [18]. By reconstructing the localized fluorophores from individual stacks into a single frame,a super-resolution image can be obtained (Fig. 1b). Flors et al. achieved super-resolution imaging of λ-DNA labeled with intercalating dye YOYO-1 For information regarding intercalation dye YOYO-1 see Fig. S1 in Supporting information [19] and reported differences between 1:20 and 1:600 dye-to-DNA pair ratios [20, 21]. However,the exact distance between the localized fluorophores intercalated between base pairs has not yet been determined.

Download:
Fig. 1.(a) A subset of YOYO-1 intercalated between base pairs of λ-DNA was activated and detected. Once the fluorophore subset was deactivated, another subset was activated and the process was repeated to obtain several frames of images. (b) Data processing was done by applying Gaussian fitting to super-localize the centroid of the PSF. The process was repeated for several frames and reconstructed into a single frame super-resolution image. (c) λ-DNA molecules were loaded onto a bare coverslip (22 × 22), which was inverted and placed on the PLL-coated coverslip (24 × 60) followed by stretching of an individual λ-DNA molecule via laminar flow.

In this study,the distances between adjacent fluorophores (i.e., YOYO-1) localized between base pairs in single DNA molecule were measured at different dye-to-DNA base pair ratios using the super-resolution technique,dSTORM. To separate the fluorescence emission from the nearby fluorophores intercalated between the adjacent base pair,the fluorophores were photoswitched stochastically. Fluorophores positions were determined by finding the centroid of the activated fluorophores. Using super-localization, the distance between the adjacent fluorophores was measured, thereby showing the distance between base pairs of a single DNA molecule. In order to confirm the base pair distance,the measurements were made with different dye-to-DNA pair ratios ranging from 1:1 to 1:500. Experimental measurements of all ratios were in agreement with the theoretical values.

2. Experimental 2.1. dSTORM switching buffer and glass coverslips

100mmol/L switching buffer was used for dSTORM image acquisition. 10mmol/L phosphate buffered saline (PBS) was converted into a switching buffer by the addition of 0.5 mg/mL glucose oxidase (Sigma-Aldrich Inc.,St. Louis,MO,USA),50mmol/L β-mercaptoethylamine (MEA,Sigma-Aldrich),40 μg/mL catalase (Sigma-Aldrich),10% (w/v) glucose (Duksan Pure Chemical Co. Ltd., Ansan,Korea),and4%(v/v)2-mercaptoethanol (Sigma-Aldrich). The final pH was adjusted to 8.2 using 1mmol/L KOH solution (Sigma- Aldrich). Next,22mm× 22mm (Paul Marienfeld GmbH & Co. KG, Lauda-Konigshofen,Germany) and 24mm× 60mm (Thermo Fisher Scientific Inc.,Waltham,MA,USA) No. 1 Corning glass coverslips were cleaned by sonication in spectroscopic grade methanol (Duksan) for 30 min,and 24mm× 60mm coverslips were coated with 0.01% (w/v) poly-L-lysine (PLL,Sigma-Aldrich).

2.2. DNA sample preparation

λ-DNA (48 502 bp) was obtained from Promega (Madison,WI, USA) and labeled with YOYO-1 (Molecular Probes,Eugene,OR, USA) at various dye-to-DNA base pair ratios in 100 mmol/L switching buffer for single molecule imaging. DNA samples at various dye-to-DNA base pair ratios were incubated in the dark at room temperature for about 1 h and diluted to 1 pmol/L using the same switching buffer prior to dSTORM imaging.

2.3. Sample loading onto the coverslip for dSTORM

The stretching method for individual λ-DNA molecules on the cover glass surface was modified from a previously reported technique [22]. Briefly,4 μL of 1 pmol/L λ-DNA stained with YOYO-1 was loaded at the corner of a 22 mm × 22 mm coverslip (Fig. 1c),0.4 cm along the x-axis and 0.4 cm along the y-axis. The coverslip was turned and attached to the PLL-coated 24 × 60 mm coverslip to create a laminar flow stream between the two coverslips. The λ-DNA molecules were then stretched through laminar flow and immobilized on the PLL-coated coverslip.

2.4. dSTORM imaging system

dSTORM measurements were performed on a home-built upright Olympus BX53 microscope (Olympus Optical Co.,Ltd, Tokyo,Japan),which applied a dove prism-type total internal reflection fluorescence (TIRF) configuration (Fig. 2). The microscopic system was equipped with a 488 nm Ar+laser (35-LAP-431- 220,Melles Griot,Carlsbad,CA,USA) and a 405 nm diode-pumped solid-state laser (COMPACT-30G-405,World Star Tech.,Toronto, ON,Canada). The TIRF images were taken immediately before dSTORM imaging by irradiation with the 488 nm laser,whereas dSTORM images were acquired by simultaneous irradiation with the 488 nm laser for excitation and a 405 nm laser for activation. For simultaneous irradiation,two mechanical shutters (Uniblitz Electronics,Rochester,NY,USA) were used. The resulting emission was collected by an objective lens with numerical aperture 1.3 (×100 UPlanFI,Olympus) and recorded using an electronmultiplying cooled charge-coupled device camera (EMCCD, 512 × 512 pixel imaging array,iXon Ultra,Andor,Belfast,Ireland). Image acquisition,shutter speed,and laser exposure time were controlled using MetaMorph 7.8.6 software (Molecular Devices, Sunnyvale,CA,USA).

Download:
Fig. 2.(a) Schematic diagram and (b) photograph of the physical layout of the TIRF and dSTORM detection system. L, laser; M, mirror; DCM, dichroic mirror; MSH, mechanical shutter; OL, objective lens; F, filter; EM-CCD, electron-multiplying cooled charge-coupled device; P, prism.
2.5. dSTORM image analysis

One thousand image frames were recorded at a rate of 17 Hz per second and reconstructed into a single frame super-resolution image (Fig. S2 in Supporting information) using QuickPALM plugin [23] ImageJ software. For super-localization,the centroid of the fluorescence spot was fitted to a two-dimensional (2D) Gaussian function [24]:

where sx and sy are the standard deviations along x and y,xo and yo are the coordinates of the centers,zo is the constant from background noise,and A is the amplitude. The localization precision (σ) [25] was defined as:

where N is the number of collected photons,a is the pixel size of the detector,s is the standard deviation of the point spread function, and b is the background noise of the detector. During reconstruction, 300-500 photons were detected at 50 ms per molecule. For reconstruction of data using the QuickPALM plug-in (Fig. S3 in Supporting information),the signal to noise ratio value was 4-8, and full width half maxima value was calculated to be 2 using the following equation [26]:

where λ is the wavelength of excitation light,the combined term 2η sin a is the objective numerical aperture,Imax is the peak intensity of the depletion laser,and Is is the saturation intensity of the fluorophore. During the analysis of various ratios from 1:1 to 1:500,all ImageJ parameters were constant except for the value of the signal to noise ratio,as if the intensity of the signal decreases from 1:1 to 1:500. The final reconstructed image was rendered at a pixel size of 10 nm. The intensities of the fluorophores obtained from individual frames were averaged and compared for various ratio ranges.

3. Results and discussion

Fluorophores possessing large molar absorptivity and high fluorescence quantum yields are the preferred reagents to label λ- DNA for fluorescence imaging. The covalent linkage of YOYO-1 increases the binding affinity of the dye as it forms a quadruple cationic bi-chromophore complex [21]. The bis-intercalating property of YOYO-1 allows DNA to elongate and unwind from its super-coiled structure [27]. It has previously been reported that YOYO-1 stains DNA molecules homogeneously when incubated for 2 h at 50 ℃ in the dark [28]. Hence,bifunctional cyanine dyes like YOYO-1 are commonly used for labeling λ-DNA due to their high binding affinity and large fluorescence enhancement upon binding to λ-DNA base pairs [5, 20, 21]. As shown in earlier studies that used direct photoswitching of cyanine dyes [29] such as Cy5 and Alexa 647,fluorophores can be reversibly switched between the fluorescent and dark states with high efficiency without the use of an activator fluorophore [30]. In this work,YOYO-1 was photoswitched between the fluorescent and dark states without the use of an activator fluorophore. However,the mechanism for photoswitching YOYO-1 could be different from that of Cy5 and Alexa 647. As Cy5 and Alexa 647 tend to photoswitch with the reversible addition of a thiol radical via the formation of a polymethin-bridge [31],photoswitching of YOYO-1 occurs due to an electron transfer reaction between YOYO-1 and the reducing agent MEA [20].

λ-DNA was stained with YOYO-1 and incubated at room temperature in the dark for 1 h to achieve complete homogeneity. By varying the concentrations of λ-DNA and YOYO-1,samples with different dye-to-DNA base pair ratios ranging from 1:1 to 1:500 were prepared. As previously described,intercalating YOYO-1 dye in λ-DNA is homogenous,and on a PLL-coated coverslip at pH 8.2, λ-DNA can be stretched linearly from its coiled structure due to the combination of hydrophobic and electrostatic interactions with an adsorbing surface [22]. DNA samples were applied to the surface of a PLL-coated coverslip at a pH of 8.2 using a switching buffer and were simultaneously irradiated by two different lasers at wavelengths of 488 and 405 nm for dSTORM imaging and a wavelength of 488 nm for TIRF imaging. TIRF images for the various dye-to-DNA base pair ratios showed a considerable decrease (10876-2206) in fluorescence intensities from 1:1 to 1:500 (Fig. 3a). Drift correction was not done during TIRF imaging. Decreases in relative fluorescence intensity and the number of fluorophores detected with respect to dye-to-DNA base pair ratios from 1:1 to 1:500 (Fig. 3b and c) showed homogenous staining with YOYO-1 of λ-DNA in those ratios.

Download:
Fig. 3.(a) TIRF images of λ-DNA at various dye-to-DNA base pair ratios from 1:1 to 1:500. (b) Plot of relative fluorescence intensity and (c) number of photons detected from individual DNA molecules. The fluorescence intensity decreased as the concentration of dye decreased from 1:1 to 1:500 (n = 5).

To verify that an individual YOYO-1 molecule was reversibly switched between the fluorescent and dark states,YOYO-1 labeled λ-DNA at a 1:100 dye-to-DNA base pair ratio was adsorbed on a PLL-coated coverslip and was imaged using a PBS switching buffer. Upon laser irradiation at 488 nm,the fluorophore YOYO-1 emitted a constant number of photons and then switched back to the dark state. The fluorophore was activated by irradiation with the 405 nm wavelength laser. Thus,the switching ability of YOYO-1 was confirmed. In order to determine the photoswitching kinetics of YOYO-1,the fluorophores were irradiated at a wavelength of 488 nm at different laser powers. Once the fluorophores attained a dark state,they were irradiated with the 405 nm activation laser with different laser powers. The plot (Fig. 4a and b) of the rate constant of fluorescence emission with different laser powers shows a linear increase in the fluorescence emission during both 405 nm and 488 nm laser irradiation. During photoactivation,the 488 nm wavelength continuously irradiated the sample at 30mW laser power. Thus,our results show that the number of photons that are activated at a given time completely depends on the power of both the activation laser and the excitation laser. This switching ability of the intercalating dye upon irradiation with two laser wavelengths while embedded on a switching buffer and without the use of an activator fluorophore make it one of the best reagents for imaging λ-DNA under dSTORM conditions.

Download:
Fig. 4.Photoswitching kinetics for the (a) fluorescence off state by irradiation at a wavelength of 488 nmand the (b) fluorescence on state by irradiation of YOYO-1 in λ-DNA at a wavelength of 405 nm as a function of laser power (n = 5).

To find the optimum incubation time for homogenous staining of DNA with YOYO-1,the 1:100 dye-to-base pair ratio sample was incubated in dark at room temperature and data was acquired at different time interval (Fig. S4 in Supporting information). Image acquired at once the DNA was stained with YOYO-1,showed a poor localization of fluorophores (Fig. S4a),the data acquired after 30 min showed a better localization but the width of the peaks were not even hence the staining remains non-homogenous (Fig. S4b). The data acquired after 1 h incubation showed a precise localization of fluorophore with even width of the peaks confirms the homogenous staining of the fluorophores. Therefore,all the dye-to-DNA base pair ratios from 1:1 to 1:500 were incubated in dark at room temperature to achieve homogenous staining. The intercalated YOYO-1 dye was photoswitched without the use of an activator fluorophore to generate dSTORM images,and the laser powers were adjusted in such a manner to allow only a subset of the fluorophores to be active at any given time in the field of view. For excitation,the 488 nm laser with a power of 90mW was used, and for activation,the 405 nm laser with a power of 2 mW was used for dSTORM acquisition. To measure the distance between the base pairs of λ-DNA molecules,various dye-to-DNA ratios from 1:1 to 1:500 (Fig. 5a) were imaged and analyzed using the dSTORM technique. Intensities of the fluorophores obtained from individual frames were quantified and compared for various ratios ranging from 1:1 to 1:500 (Fig.[38TD$DIF] 5b),and the results showed that the difference in intensity decreases from 1:1 to 1:500.

Download:
Fig. 5.(a) Intercalation of YOYO-1 dye in λ-DNA and the distance between the fluorophore for different ratios ranging from 1:1 to 1:500. (b) Plot of total fluorescence intensity obtained during dSTORM acquisition of an individual DNA molecule vs. dye-to-DNA base pair ratio shows a decrease in intensity as the concentration of dye decreased from 1:1 to 1:500 (n = 5).

During dSTORM acquisition,the stage of the microscope tends to during reconstruction of the super-resolution image (Fig. 6c). To compare the differences in resolution between the TIRF and dSTORM images,parts of the images were magnified (Fig. 6a and c). The width of the λ-DNA in the TIRF image was calculated to be 43 ± 1 nm(Fig. 6d),whereas the width in the super-resolution image before drift correction shows two peaks (Fig. 6e). Once the drift correction has been made on both the x and y-axes (Fig. S5 in Supporting information),the width of the single λ-DNA was calculated to be 21 ± 1 nm (Fig. 6f).

Download:
Fig. 6.(a) TIRF image of λ-DNA stained with YOYO-1 at a 1:100 dye-to-DNA base pair ratio. (b) dSTORM image before drift correction and (c) after drift correction. Widths of the individual λ-DNA in a (d) TIRF image, and a (e) dSTORM image before drift correction and (f) after drift correction.

λ-DNA base pair ratios of 1:1,1:5,and 1:10 possessed theoretical distances between base pairs of 0.34,1.7,and 3.4 nm respectively,and these values were too small to be resolved. Therefore,a significant result was not achieved in measuring the distances between the localized fluorophores in those ratios (Fig. S6 in Supporting information). The dye-to-DNA base pair ratios of 1:50,1:100,and 1:500 possessed theoretical distances between localized fluorophores of 17 nm,34 nm,and 170 nm respectively,and these values were well within the diffraction limit. With respect to dye-to-DNA base ratio,fluorophore distribution was well observed (Fig. S7 in Supporting information). TIRF and dSTORM (Fig. 7a and b) images of 1:50,1:100,and 1:500 dye-to-DNA base pair ratios show the differences in resolution. A part of the image was magnified (Fig. 7c) to show the distance between the localized fluorophores. To obtain the distance between the adjacent fluorophore intercalated in the base pair of DNA molecule,a line selection option was used in ImageJ software. Fluorophores were selected with the line selection option and a plot profile plugin was used to obtain the plot of the relative fluorescence unit with the distance of the fluorophores in nanometers. 2D surface plots for the ratios 1:50,1:100,and 1:500 showed distances between peaks of 17.1 ± 0.8 nm (n = 957), 34.3 ± 2.2 nm (n = 473),and 170.3 ± 8.1 nm (n = 95),respectively (Fig. 7d). The errors in measurement could be ascribed to YOYO-1 that binds to the DNA molecule by bis-intercalation and a weak external binder to the phosphate group of DNA molecule. The narrow base of the peak for 1:50,1:100,and 1:500 suggests that individual fluorophores were well localized and far apart from one another, such that the distance between peaks represents the distance between localized intercalated fluorophores. The 2D surface plots for the ratios 1:50,1:100,and 1:500 are shown with the same scale bar (40 nm) on the x-axis (Fig. S8 in Supporting information). The peaks became narrow in the following order: 1:500,1:100,and 1:50. This shows that the number of particles localized in the 1:500 ratio is larger than that of 1:100 and 1:50. Thus,the homogenous nature of YOYO-1 in λ-DNA was confirmed and the distance between the peak edges reflects the distance between the base pairs of λ-DNA. The value obtained from the graph is in agreement with the theoretical value of the base pair distance of λ-DNA,which indicated a spatial resolution of ~17 nm was achieved.

Download:
Fig. 7.(a) TIRF and (b) reconstructed dSTORM images for ratios of 1:50, 1:100, and 1:500. (c) Magnified images. (d) 2D surface plots of relative fluorescence intensity (RFI) vs. distance in nm for ratios of 1:50, 1:100, and 1:500 showing distances of 17 nm, 34 nm, and 170 nm respectively.
4. Conclusion

In summary,upon binding with λ-DNA,the cyanine fluorophore YOYO-1 was photoswitched between the fluorescent and dark states with exposure to an activation laser and an excitation laser. Upon simultaneous exposure to these two lasers,the fluorophores were switched and several cycles of images were obtained. These images were then reconstructed into a single frame and were analyzed by applying a Gaussian distribution and PSF to generate a super-localized image with a subdiffractionlimit. The distances between two localized YOYO-1 molecules at various dye-to-DNA base pair ratios were measured to determine the distance between base pairs. The results showed that the distances between base pairs for dye-to-base pair ratios of 1:50, 1:100,and 1:500 agreed with the theoretical values of 17 nm, 34 nm,and 170 nm respectively. Super-resolution imaging of l- DNA by intercalating dye with a subdiffraction-limit of ~17 nm was experimentally achieved for the first time. Therefore,these results suggest that the dSTORM is a valuable tool that can be used in single DNA molecule studies with high sensitivity and specificity and nanoscale resolution.

Acknowledgment

This research was supported by a grant from Kyung Hee University in 2015 (No. KHU-20150618).

References
[1] M.D. Barnes, W.B. Whitten, J.M. Ramsey, Detecting single molecules in liquids, Anal. Chem. 67 (1995) 418A-423A.
[2] S.H. Kang, S. Lee, E.S. Yeung, Detection of single enzyme molecules inside nanopores on the basis of chemiluminescence, Angew. Chem. Int. Ed. 49 (2010) 2603-2606.
[3] S. Howorka, J. Hesse, Microarrays and single molecules: an exciting combination, Soft Matter 10 (2014) 931-941.
[4] D.O. Clary, D.R. Wolstenholme, The mitochondrial DNA molecule of Drosophila yakuba: nucleotide sequence, gene organization, and genetic code, J. Mol. Evol. 22 (1985) 252-271.
[5] S.H. Kang, S. Lee, E.S. Yeung, Digestion of individual DNA molecules by lexonuclease at liquid-solid interface, Analyst 135 (2010) 1759-1764.
[6] B.H. Chen, L.A. Gilbert, B.A. Cimini, et al., Dynamic imaging of genomic loci in living human cells by an optimized CRISPR/Cas system, Cell 155 (2013) 1479-1491.
[7] A.S. Stender, K. Marchuk, C. Liu, et al., Single cell optical imaging and spectroscopy, Chem. Rev. 113 (2013) 2469-2527.
[8] Z.Y. Tian, A.D.Q. Li, Photoswitching-enabled novel optical imaging: innovative solutions for real-world challenges in fluorescence detections, Acc. Chem. Res. 46 (2013) 269-279.
[9] E. Betzig, G.H. Patterson, R. Sougrat, et al., Imaging intracellular fluorescent proteins at nanometer resolution, Science 313 (2006) 1642-1645.
[10] S.T. Hess, T.P.K. Girirajan, M.D. Mason, Ultra-high resolution imaging by fluorescence photoactivation localization microscopy, Biophys. J. 91 (2006) 4258-4272.
[11] M. Bates, B. Huang, G.T. Dempsey, X.W. Zhuang, Multicolor super-resolution imaging with photo-switchable fluorescent probes, Science 317 (2007) 1749-1753.
[12] M.J. Rust, M. Bates, X.W. Zhuang, Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM), Nat. Methods 3 (2006) 793-796.
[13] M. Heilemann, S. van de Linde, M. Schü ttpelz, et al., Subdiffraction-resolution fluorescence imaging with conventional fluorescent probes, Angew. Chem. Int. Ed. 47 (2008) 6172-6176.
[14] M. Heilemann, S. van de Linde, A. Mukherjee, M. Sauer, Super-resolution imaging with small organic fluorophores, Angew. Chem. Int. Ed. 48 (2009) 6903-6908.
[15] N.R. Conley, J.S. Biteen, W.E. Moerner, Cy3-Cy5 covalent heterodimers for singlemolecule photoswitching, J. Phys. Chem. B 112 (2008) 11878-11880.
[16] G.T. Dempsey, M. Bates, W.E. Kowtoniuk, et al., Photoswitching mechanism of cyanine dyes, J. Am. Chem. Soc. 131 (2009) 18192-18193.
[17] G.T. Dempsey, J.C. Vaughan, K.H. Chen, M. Bates, X.W. Zhuang, Evaluation of fluorophores for optimal performance in localization-based super-resolution imaging, Nat. Methods 8 (2011) 1027-1036.
[18] A. Yildiz, J.N. Forkey, S.A. McKinney, et al., Myosin V walks hand-over-hand: single fluorophore imaging with 1.5-nm localization, Science 300 (2003) 2061-2065.
[19] LifeTechnologies, https://www.lifetechnologies.com/order/catalog/product/ Y3601, (accessed February 2015).
[20] C. Flors, C.N.J. Ravarani, D.T.F. Dryden, Super-resolution imaging of DNA labelled with intercalating dyes, ChemPhysChem 10 (2009) 2201-2204.
[21] C. Flors, Photoswitching of monomeric and dimeric DNA-intercalating cyanine dyes for super-resolution microscopy applications, Photochem. Photobiol. Sci. 9 (2010) 643-648.
[22] S.H. Kang, M.R. Shortreed, E.S. Yeung, Real-time dynamics of single-DNA molecules undergoing adsorption and desorption at liquid-solid interfaces, Anal. Chem. 73 (2001) 1091-1099.
[23] QuickPALM, http://rsbweb.nih.gov/ij/, (accessed February 2015).
[24] L.S. Lerman, Structural considerations in the interaction of DNA and acridines, J. Mol. Biol. 3 (1961) 18-30.
[25] R.E. Thompson, D.R. Larson, W.W. Webb, Precise nanometer localization analysis for individual fluorescent probes, Biophys. J. 82 (2002) 2775-2783.
[26] C. Höppener, L. Novotny, Exploiting the light-metal interaction for biomolecular sensing and imaging, Q. Rev. Biophys. 45 (2012) 209-255.
[27] K. Gü nther, M. Mertig, R. Seidel, Mechanical and structural properties of YOYO-1 complexed DNA, Nucl. Acids Res. 38 (2010) 6526-6532.
[28] C. Carlsson, M. Jonsson, B. Akerman, Double bands in DNA gel electrophoresis caused by bis-intercalating dyes, Nucl. Acids Res. 23 (1995) 2413-2520.
[29] S. van de Linde, A. Lö schberger, T. Klein, et al., Direct stochastic optical reconstruction microscopy with standard fluorescent probes, Nat. Protocols 6 (2011) 991-1009.
[30] D. Bar-On, S. Wolter, S. van de Linde, et al., Super-resolution imaging reveals the internal architecture of nano-sized syntaxin clusters, J. Biol. Chem. 287 (2012) 27158-27167.
[31] S. van de Linde, M. Sauer, M. Heilemann, Subdiffraction-resolution fluorescence imaging of proteins in the mitochondrial inner membrane with photoswitchable fluorophores, J. Struct. Biol. 164 (2008) 250-254.