Orthogonal Probing of Single Molecule Heterogeneity by Correlative

Subscriber access provided by Gothenburg University Library

Article

Orthogonal Probing of Single Molecule Heterogeneity by Correlative Fluorescence and Force Microscopy Wout Frederickx, Susana Rocha, Yasuhiko Fujita, Koen Kennes, Herlinde De Keersmaecker, Steven De Feyter, Hiroshi Uji-i, and Willem Vanderlinden ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.7b05405 • Publication Date (Web): 19 Dec 2017 Downloaded from http://pubs.acs.org on December 20, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Nano is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Orthogonal Probing of Single Molecule Heterogeneity by Correlative Fluorescence and Force Microscopy Wout Frederickx1, Susana Rocha1, Yasuhiko Fujita1, Koen Kennes1, Herlinde De Keersmaecker1, Steven De Feyter1,*, Hiroshi Uji-i1,2,*, Willem Vanderlinden1,3,* 1

KU Leuven-University of Leuven, Department of Chemistry, Division of Molecular Imaging

and Photonics, Celestijnenlaan 200F, B-3001, Leuven, Belgium 2

Hokkaido University, Research Institute for Electronic Science, Nanomaterials and Nanoscopy,

Kita 10 Nishi 20, North Ward, Sapporo, Japan

3

Department of Physics, Nanosystems Initiative Munich, and Center for NanoScience, LMU

Munich, Amalienstrasse 54, 80799 Munich, Germany

[email protected] [email protected] [email protected]

1 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 40

Abstract Correlative imaging by fluorescence and force microscopy is an emerging technology to acquire orthogonal information at the nanoscale. Whereas atomic force microscopy (AFM) excels at resolving the envelope structure of nanoscale specimen, fluorescence microscopy can detect specific molecular labels, which enables the unambiguous recognition of molecules in a complex assembly. Whereas correlative imaging at the micrometer-scale has been established, it remains challenging to push the technology to the single-molecule level. Here, we used an integrated setup to systematically evaluate the factors that influence the quality of correlative fluorescence and force microscopy. Optimized data processing to ensure accurate drift correction and high localization precision, results in image registration accuracies of ~ 25 nm on organic fluorophores, which represents a two-fold improvement over the state of the art in correlative fluorescence and force microscopy. Furthermore, we could extend the Atto532 fluorophore bleaching time ~ 2-fold, by chemical modification of the supporting mica surface. In turn, this enables to probe the composition of macromolecular complexes by stepwise photo-bleaching with high confidence. We demonstrate the performance of our method by resolving the stoichiometry of molecular sub-populations in a heterogeneous EcoRV-DNA nucleoprotein ensemble. Keywords: correlative imaging, atomic force microscopy (AFM), single molecule localization microscopy (SMLM), protein-DNA interactions, stepwise photobleaching


Page 3 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Self-assembly of bio-macromolecules into multipart complexes underlies functionality and regulation in virtually all cellular processes. Structural heterogeneity is inherent, and often critically important, to these systems.1-3 Functional and mechanistic insights can be obtained using single molecule imaging techniques with nanometer scale resolution, that capture heterogeneity in supramolecular architecture directly. In this context, the last two decades have seen the development of optical nanoscopy techniques that can visualize mesoscale structures with a resolution ranging down to 20-50 nm, provided that the molecules can be densely labelled with fluorophores.4-5 An alternative technique that can sample conformational heterogeneity beyond this length scale, and in a label-free manner is atomic force microscopy (AFM).6-15 Conventionally, AFM samples only the envelope of surface features, and is insensitive to the chemical nature of the constituents. Therefore, detailed information of the composition in multipart assemblies remains obscure. To get around this limitation, several approaches to detect specific molecules in a heterogeneous ensemble have been proposed. A first approach, AFMbased recognition imaging, relies on the (bio-)chemical functionalization of the probe-tip. Transient binding between the modified tip, and a specific protein on the scanned surface, is reflected in the dynamic behavior of the cantilever.16-19 AFM-based recognition does not necessarily require modification of the molecule of interest, and works particularly well on molecules which are well-oriented on the supporting surface, e.g. membrane proteins. However, when the molecule has no well-defined orientation on the surface, or when it is buried within a larger complex, alternative approaches are needed.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 40

An alternative route towards orthogonal probing of macromolecular structure, is based on the combination of AFM with optical nanoscopy techniques.20-21 Scanning probe techniques such as near field scanning optical microscopy (NSOM)22 and tip-enhanced Raman spectroscopy (TERS)23 can provide chemical insight at the nanoscale, but their sensitivity is often insufficient for biological applications.24

In this context, a more promising approach is correlative force and fluorescence microscopy. Correlated imaging at the micrometer length-scale, such as AFM-based topographic or mechanical property mapping of fluorescently stained cells, is well-established.25-27 In contrast, the visualization of purified or reconstituted subcellular structures below the diffraction limit remains challenging. Nevertheless, the potential of integrated setups to correlate topographic and optical information at the nanoscale, or at the level of single molecules has been demonstrated in a handful of publications over the last few years.28-34 On the one hand, researchers have developed integrated tools for correlated AFM and superresolution fluorescence microscopy. For instance, a combination of AFM and stimulated emission depletion microscopy (STED) could resolve a correlation of local cell elasticity with cytoskeleton architecture.28 In addition, AFM correlated with localization-based super-resolution methods, such as photo-activated localization (PALM) and stochastic optical reconstruction microscopy (STORM), have been used to map heterogeneous sample labeling.29-30 Recently, the morphological dynamics of the cell surface and focal adhesion complexes in mammalian cells was imaged by correlated AFM and PALM.31 These examples demonstrate that densely labelled biomolecular assemblies can be directly compared in topographic and super-resolved fluorescence images.


Page 5 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

On the other hand, when the dimensions of the specimen are smaller than the optical resolution, a direct structural comparison is no longer achievable. In this case, the images must be registered using fiducial markers that can be visualized by both techniques. In a first approach, quantumdot (QD) labeled protein-DNA complexes have been visualized together with non-bound QDs as fiducial markers, with a registration accuracy of ~ 10 nm.32 However, the size of these QDs are in the same order or even larger than the protein size, possibly interfering with protein functionality. Moreover, QDs heavily blink, making them unreliable as internal references. Other groups have therefore used fluorescent styrene beads as fiducial markers to detect specific components in heterogeneous nucleoprotein complexes.33-34 In these reports, registration of overview topographic and fluorescence images of single organic dyes, achieved an accuracy ~ 50 nm. To further advance atomic force – single molecule localization microscopy (FM-SMLM) technology, several issues should be addressed. First, identification and localization of specific molecules within a multipart complex demands high image registration accuracy, ideally approaching AFM resolution. Second, structural interrogation at the nanometer scale requires improvements of topographic resolution in the registered image, without sacrificing registration accuracy. Third, to exploit the fluorescence signal of single molecules beyond localization, detection should be highly sensitive, and the signal should be continuous over a maximal time.

In this contribution, we have quantitatively evaluated, and optimized, the parameters that influence the quality of correlated AFM-SMLM. We use an excitation scheme and analysis routine that, together, improve the sensitivity, precision, and accuracy of AFM-SMLM correlative imaging beyond the current state of the art. Further, we demonstrate that interactions


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 40

with the substrate reduce the sensitivity and survival time of organic fluorophores. Surfacemodification with poly-L-lysine significantly diminishes this effect, which in turn enables highconfidence quantized photo-bleaching analysis. We validate our experimental design by orthogonal probing of nucleoprotein complexes, and demonstrate the detection of structural and compositional sub-populations in the ensemble.

Results

Experimental design for correlative AFM-SMLM To probe heterogeneity in complex nanoscale assemblies, we employ a correlative AFM-SMLM setup that comprises a commercial AFM mounted on a home built inverted fluorescence microscope (Fig. 1A) specifically designed to minimize mechanical instabilities. Sample preparation involves drop casting a mixture of biomolecular complexes site-specifically labelled with single fluorophores, and fluorescent fiducial markers on a transparent substrate. The fiducial markers serve as readily distinguishable features in both fluorescence and AFM images, and enable high-accuracy image-overlay (Fig. 1B). In the fluorescence channel, movies are recorded for a few minutes, and at a framerate of ~ 9 s-1 (corresponding to an integration time of 110 ms per frame, a total of 1500 frames are acquired). Least-square 2D Gaussian fitting to individual point spread functions (PSF) yields the coordinates of the centroids of the fiducial markers and biomolecule-bound fluorophores.35 To improve the signal to noise ratio /, the accumulation time can be increased by sliding window averaging of frames. Each accumulated image generates one localization map (Fig. 1C). Accordingly, acquisition of frames, yields − + 1 localization maps. These maps are corrected for drift, and then super-imposed. The final


Page 7 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

localization calculates the center of mass of drift-corrected PSF centroid clusters. In addition, the background corrected intensities can be plotted as fluorescence intensity time-traces (Fig. 1C).

Figure 1. Experimental design of correlative scanning force and fluorescence microscopy. A. Experimental setup. Sample and fiducial markers are adsorbed on a transparent substrate and mounted on an inverted microscope equipped with a tip-scanning AFM. B. Principle of image correlation. The position of both fiducial markers and single dyes is estimated by 2D Gaussian fitting. The error on localization of the fiducial markers is slightly smaller than for the single dyes. The calculated positions of the beads in the fluorescence channel are superimposed on their localization in the AFM channel by an affine transformation. C. Fluorescence data analysis. (i) Fluorescence is acquired for frames at a framerate = 9 s-1. To increase the signal to noise ratio (/), the raw data is averaged using sliding 7 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 40

window averaging, with window size . (ii) Next, the location of the fluorescence particles is calculated in each frame by least-square 2D Gaussian fitting of the point spread function and (iii) the different x,y positions are plotted in one frame. (iv) Drift is thereafter corrected based on the coordinates of the fiducial markers centroids in subsequent frames. (iv) Last, the final localization and associated uncertainty are calculated as the average and width of the distribution of all centroid coordinates. For each fluorescent particle, the fluorescence intensity time trace can be calculated.

Optimizing AFM-SMLM data quality is a multi-dimensional problem To extract information at the single molecule level, it is imperative to understand, and optimize the factors that affect the quality of correlative single molecule localization microscopy (SMLM) and atomic force microscopy (AFM). Image quality is a function of the resolution in the AFM topography channel, the sensitivity / and localization precision in the fluorescence channel, and the image registration accuracy (target registration error, TRE; localization registration error, LRE) (Fig. 2). In this section, we will introduce the factors that render these different parameters mutually dependent, and that make the optimization process a multi-dimensional problem.

Lateral resolution in AFM topographs is defined as the minimum distance between two features for which the sample dip exceeds the vertical noise.36 High resolution imaging of objects in the single nanometer range (e.g. nucleoprotein complexes), is therefore highly dependent on the bandwidth of the AFM hardware and control system, and requires appropriate isolation from external vibrations. In addition, substrate roughness should be minimized. This is typically achieved by sample deposition on an atomically flat surface, like muscovite mica.8


Page 9 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 2. Defining the parameters that quantify the quality of correlative AFM-SMLM, using the example of a labelled protein-DNA complex. The AFM topography image is represented in brown scale while the high contrast colour scale represents the super resolution fluorescence data. The lateral resolution of SFM topography data is given by . The localization precisions of the fiducial markers and the single fluorophore are given by and respectively. The image registration accuracy is quantified either by the pair-wise distance between features detected in both channels (fiducial markers; TRE: target registration error), or the pair-wise distance between the fluorophore positions (LRE: localization registration error) in both channels. Note that the fluorophore itself is not visualized by SFM, but its position can be estimated from the position of the protein.

However, mica exhibits a broad absorption spectrum that reduces the observed brightness of the fluorophores (Fig. 3a). In addition, the birefringent properties of mica impart substantial 9 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 40

aberration that distort the PSF.37 Thus, mica intrinsically reduces the signal-to-noise / in the fluorescence channel. To minimize these effects, while providing sufficient mechanical stability and an appropriate optical working distance, thin slices (< 10µm) of mica are glued to glass coverslips.32-33, 38-39 Further, the proximity between fluorophore and mica potentially affects the fluorescence signal by photochemical and/or photophysical interactions.40-41 It is however unclear whether, and to what extent, such interactions affect the excited state of fluorophores.

Accurate registration is key to correlated imaging, as it enables to assign distinct fluorescent signals to specific molecular architectures observed within the AFM channel. Image registration accuracy can be quantified by the target registration error (TRE), which is defined as the difference between corresponding points visible in both channels, after registration.42 However, for single molecule imaging, it is pivotal to know how accurate a single fluorophore, invisible in AFM, can be localized in the registered image. We therefore adopt the definition of the localization registration error (LRE) of Cohen and Ober, which refers to the uncertainty in localizing a single fluorophore in the registered image (Fig. 2).43-44 We note that the TRE defines the lower limit for the LRE.

Thus, the optical properties of fluorophores and fiducial markers, the supporting surface, the bandwidth, and stage drift of the microscope, as well as the software routines for data processing, are factors that need all to be considered when aiming for high quality AFM-SMLM correlative imaging.

Excitation scheme


Page 11 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

We employ a scheme wherein emission of single fluorophores and fiducial markers, after simultaneous excitation, is detected on the same region of the CCD camera.45-48 This has the advantage that drift of both signals is concurrently corrected for. Furthermore, by proper consideration of the optical properties of both the chromophores and fiducials, their /, and hence their localization precision σ, can be matched.49-50 In turn, this should result in a similar registration accuracy (TRE ~ LRE) in the correlated image. Desirable properties for the bio-compatible fluorophores include a high quantum yield, minimal blinking, and favorable photo-stability. For biomolecular site-specific labelling, we therefore selected the commercial fluorophore Atto532. In terms of fiducial markers, different candidates have been reported. We discarded quantum dots as a fiducial marker: because of their long-lived dark-states, they require long (~ 5-10 s) integration times to enable drift correction, which in turn limits the time-resolution of the fluorescence readout. In contrast, we identified the FluoSpheres® carboxylate-modified microspheres to exhibit a stable emission (Fig. SI 1A). Furthermore, they are relatively small (~20 nm), water-soluble, and are easily immobilized on the substrate.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 40

Figure 3. An excitation scheme that yields comparable sensitivity for fluorosphere® fiducial markers, and individual Atto532 fluorophores deposited on muscovite mica. A. absorption and emission spectra of Atto532, fluospheres® and absorption spectrum of muscovite mica. Atto532 and fluospheres® are excited by a 532 nm laser wavelength (black dotted line). The emission was separated from the excitation light by a 545nm longpass filter. B. Fluorescence (left) and SFM (right) image of Atto532 labeled DNA molecules and fluosphere® (white circles) are detected in both channels. The field of view is 12x12µm. C. Sensitivity / of fluorescence detection for single Atto532 dye and internal reference, as a function of the number of accumulated frames . D. Localization precisions σ for Atto532 and internal references as a function of the number of accumulated frames n. Note that the variance of decreases for increasing . E. Target Registration Error (TRE) as a function of the total number of frames for 1 and 10. 12 ACS Paragon Plus Environment

Page 13 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Solid square = Atto532, open circles = internal references. Errors are the standard deviation of the distributions.

Critical for our excitation scheme, is the spectral separation of the absorption of the fiducial markers from the absorption of Atto532 (Fig.3A). To achieve a similar / for both single fluorophores and fiducial markers, samples are irradiated at 532 nm. This guarantees optimal excitation efficiency for the Atto532 fluorophores, whereas the fluorophores encapsulated in the fiducial markers are sub-optimally excited. This approach avoids CCD camera saturation by fiducial markers at laser powers needed to detect single dyes. However, a sufficiently high / for high quality 2D Gaussian fitting cannot be guaranteed given the absorption spectrum of mica (Fig. 3A). This can be overcome by increasing the accumulation time by sliding window averaging (Fig. 1C). In the following paragraphs, we evaluate the impact of on the localization precision , and on the accuracy and distribution of the registration errors and (Fig. 2). We present a rationale for optimal data processing.

Quantitative assessment of the factors that control sensitivity, precision and accuracy In this section, we quantify how increasing the fluorescence sensitivity affects the localization precision of both single dyes and internal references simultaneously which ultimately results in a better image registration accuracy. To this end, linear DNA molecules end-labelled with a single Atto532 dye, are co-deposited with fiducial markers. The DNA construct is generated by polymerase chain reaction amplification, whereby one primer is modified at its 5’ end with


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 40

Atto532. This approach guarantees a single fluorophore per molecule of DNA, at a well-defined position.33

The sensitivity of fluorescence detection is quantified in terms of the signal to noise ratio ⁄. Increased ⁄ empowers applications such as stepwise photo-bleaching, and improves the quality of PSF fitting. Fig. 3B depicts that single Atto532 fluorophores and fiducial markers are simultaneously detected, according to the design of our excitation scheme. The / of both fluorescent species was determined for different accumulation times , for the duration of the entire acquisition time (approximately 150 s). For 1, individual Atto532 fluorophores and fiducial markers are detected with a similar, / of ~1.5. The / value improves rapidly with the number of accumulated frames (Fig. 3C). Accordingly, increasing improves the quality of the 2D Gaussian fit to individual PSFs (Fig. SI 2A) and enables proper drift correction. Drift correction of the fluorescence data is performed using the average fiducial correction wherein the change in position, averaged over all fiducial markers, is used to correct each frame (Fig. SI 2B).51

Next, we assess the quality of the localization precision σ of Atto532 and internal references obtained at different accumulation times . Both for Atto532 and fiducial markers, σ decreases with and levels off at 10, reflecting the accuracy of the 2D Gaussian fit (Fig. 3D and SI 2A). Note that, besides the average value of σ, also its variability decreases with increasing . We find that an appropriate localization precision σ ≤ 5 nm is reached for most (~70%) particles for an accumulation time of ≥ 1 (/ ≥ 15 ; Fig. SI 2C).


Page 15 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Finally, the Target Registration Error (TRE) was determined for different and . The is calculated by the leave-one-out method:32 each fiducial marker is alternatingly left out of the image registration calculation and the distance between its localized position in the registered image and its position in the AFM image is determined. When is small, both the average TRE and its variability are critically dependent on , reflecting the effect of σ. For instance, when m = 3, we find a TRE of 25 ± 11 nm and 14 ± 6 nm for an accumulation time of 0.11 s ( 1) respectively 1.1 s ( 10). In contrast, for ≥ 500, the TRE is not significantly affected by the accumulation time, with a TRE of ~ 11 nm for both an accumulation time of 0.11 s ( 1) and 1.1 s ( 10) (Fig. 3E). Moreover, the TRE for an accumulation times of 1.1 s ( 10) is not strongly dependent on the number of frames and is close to the Cramer-Rao Lower Bound (CRLB) limit for image registration between two fluorescent channels of different color (SI Fig. 2D).42, 44 This indicates that drift in AFM channel is insignificant, which is further confirmed by measuring an AFM calibration grid (SI Fig. 3).

It should be noted that the TRE only describes the accuracy with which the fiducial markers can be registered. As they fluoresce throughout the entire movie, the TRE approaches the CRLB at sufficient high , despite low . In contrast, single organic fluorophores will bleach after a certain time. Nevertheless, as single fluorophores and fiducial markers have a comparable / and σ (Fig. 3C and D), a similar effect of n and m on their registration accuracy is expected. High registration accuracies on single fluorophores can therefore only be guaranteed, in case they survive sufficiently long. More specifically, the fluorescence signal should be accumulated for at least 13 frames to obtain 10 and ≥ 3, which requires a dye survival time of at least 1.4 s to obtain TRE < 15 nm.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 40

We conclude that, in the framework of image registration, ≥ 10 allows for accurate drift correction and high accuracy image registration, while keeping in mind that should not be too large, to maintain good time resolution. Furthermore, single dyes should survive sufficiently long (minimally 1.4 s), to ensure a high ⁄ and registration accuracies.

Modifying the surface chemistry of mica increases Atto532 longevity Tracing single fluorophore intensity over time enables several opportunities. One application is direct fluorophore counting via stepwise photo-bleaching. For unbiased detection of bleaching events, we use an algorithm based on Kerssemakers et al.52 The algorithm detects individual steps, depending on /, and on the duration of the intensity plateau. A single photo-bleaching step can typically be detected with high confidence, e.g. dyes with a survival time of only 1.9 s have a 75% probability to report on monomers (Fig. SI 4A). However, to confirm the oligomerization state of a hexameric complex with the same confidence, a minimal survival time of 11.5 s is required. Single dyes should thus be sufficiently photostable for the high confidence detection of higher multimers. To better understand the process of photo-bleaching, we evaluated the spectroscopic properties of Atto532, covalently attached to DNA after adsorption on mica (Fig. 4A). We made a quantitative comparison of the survival time, for two popular approaches that enable DNA adsorption onto mica: the DNA was either deposited on Poly-L-Lysine (PLL) functionalized mica or on bare mica from a buffer containing Mg2+ ions.53-54 We find that the lifetime τ for survival of Atto532 on PLL-mica (τ = 14 s) is approximately two-fold higher as compared to the survival lifetime on Mg-mica (τ = 7 s) (Fig. 4 B,C).


Page 17 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

To understand this effect, confocal single-molecule, single-photon timing spectroscopy was used to analyze the excited state lifetimes and blinking dynamics of Atto532 on both surfaces (Fig. SI 4B, C). The excited state lifetime of the dye on both modified substrates is lower than on glass (~ 2 ns versus 3.3 ns, which is close to the reported lifetime of 3.8 ns in solution), indicative of a stronger interaction with these surfaces compared to glass. Since the excited state lifetimes are similar on PLL-mica and Mg-mica, the deactivation of the fluorescence must occur on similar time scales but does not directly explain the observed differences in survival lifetimes. Analysis of the fluorescence blinking behaviour, however, shows that the scaling exponents of the offtimes are clearly distinct, with a more negative exponent on PLL-mica as compared to Mg-mica (Fig. SI 4D). The negative exponent is indicative for a trapping mechanism.55-57 Our results are in line with a mechanism whereby Atto532 undergoes an excited state electron (or hole) transfer to the substrate, where it performs a random walk or is trapped. After some time, the charges can recombine to undergo the normal photophysical deactivation process until it gets re-excited. A smaller off-exponent indicates that the back-charge transfer is hindered more. Therefore, the remaining radical Atto532 has increased probability for photo-bleaching when deposited on bare mica. Accordingly, in the AFM-SMLM setup, a fluorescent signal could be assigned to ~70 % of the Atto532-labelled DNA molecules on PLL-mica, whereas only ~ 50 % of the DNA molecules on Mg-mica exhibited detectable fluorescence intensities. This could, at least in part, be explained by fast photo-bleaching of single fluorophores.

Optimal detection of single fluorophores and rationale for high quality correlative AFMSMLM


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 40

The end-labelled DNA construct additionally serves as an ideal tool to directly quantify the . Whereas Atto532 cannot be directly resolved in AFM topography images, we know by design that the fluorophores are in the proximity (~ 1nm) of the DNA end. Following Sanchez et al., we determined the by measuring the distances between the nearest end of the DNA in the AFM image, and the coordinates of the mean position of the centroid cluster determined in the fluorescence channel (Fig. 4A).33 For = 10, we find = 27 ± 15 nm (Fig. 4D). This value represents a ~ two-fold improvement as compared to the current state of the art.33 The difference between the LRE (~27 nm) and TRE (~11 nm) can be attributed to chromatic aberration effects caused by the different emission wavelengths of the Atto532 and the fiducial markers (Fig. 3A). In addition, it was found that the overlay of a higher resolution AFM topograph (3x3 µm2, ~ 1.5 nm pixel size) with the overview image (12x12 µm2, ~ 6 nm pixel size), does not significantly affect the (Fig. SI 5A-D). Therefore, AFM image quality can be improved to the level achievable by a conventional stand-alone AFM, with the added value brought by correlated imaging. Our systematic evaluation of the parameters that quantify correlative AFM-SMLM data quality, enables presenting a rationale for best practice. First, simultaneous detection of fiducial markers and single fluorophores enables concurrent drift correction. Excitation at the absorption maximum of single fluorophores maximizes their brightness. To prevent saturation of fiducial fluorescence, their absorption spectrum needs to be sufficiently shifted. Second, time averaging over consecutive frames using a sliding window increases the /. In our case, for 10, we achieve an average / ≥ 15 for both single Atto532 and internal references, which enables


Page 19 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

accurate drift correction and consequently high localization precision for all particles. In turn, this allows image registration accuracies close to the CRLB. Third, high confidence stepwise photo-bleaching analysis is dependent on the fluorophore survival time. In that context, photochemical interaction with the substrate need to be considered. We found that on PLL-mica and for ⁄ 15.7 ± 6.5 ! = 10), Atto532 can be localized with a precision of = 4.2 ± 2.0 nm, and registered with the topography image with = 27 ±15 nm. Last, we demonstrate that high resolution topographs of labelled biomolecules can be overlaid by cross correlation with the overview topography image, without sacrificing the registration accuracy.

Figure 4. Orthogonal imaging of Atto532-labelled DNA. A. AFM topography image of the Atto532-labelled DNA construct, depicts the measurement scheme for determination of LRE, and highlights the lateral resolution that can be obtained B. Intensity time-trace of Atto532 covalently attached at a 5’ end of DNA, depicting stepwise photo-bleaching. C. Survival times of Atto532 deposited either on PLL functionalized mica or on bare mica in a Mg2+ containing buffer. The blue and red lines represent a fit to an exponential decay function (" #$/% & for PLL


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 40

and Mg2+, respectively. The Atto532 survival lifetime can be twofold extended when deposited on PLL functionalized mica. (N=36) D. The distribution of LRE values (N = 61) is centered at ~27 ± 15 nm. The width of the bars is 10 nm, taking into account a minimum localization precision of 10 nm for single Atto532 (see Fig. 3C). Errors represent the standard deviations of the distributions.

Resolving geometrical and compositional heterogeneity in EcoRV-DNA nucleoprotein complexes We now demonstrate the potential of correlated AFM-SMLM to resolve geometry, composition, and stoichiometry of multipart nucleoprotein complexes. As a proof-of-concept, we evaluate the DNA-binding properties of the EcoRV enzyme, a prominent tool in biotechnology, and a paradigm in the field of protein-DNA target search. EcoRV is a dimeric restriction enzyme that recognizes its target sequence with high specificity. After target localization, and prior to dsDNA cleaving, EcoRV bends the DNA by ~50°.58-60

First, we investigated the interaction of Atto532-labeled EcoRV(C21S/K58C) with a 500 basepair DNA fragment comprising the EcoRV target sequence. The fluorescent label was installed site-specifically on the genetically engineered cysteine residue of each monomer, via maleimide chemistry, with a labelling efficiency of ~ 70 %. Mixing of the protein with the DNA fragment in the presence of calcium ions, allows nucleoprotein formation while restriction catalysis is impeded. On adsorption of the sample onto PLL-mica, we find that EcoRV is bound with high selectivity at its cognate DNA sequence (Fig. SI 6A,B). Owing to the high lateral resolution of the topography images, we could quantify the bend angle distribution at a length


Page 21 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

scale of 7.5 nm (Fig. 5A,B). The bend angle distribution was fit using the sum of two folded Gaussians, where the first population (64% of all complexes) represents the state wherein EcoRV does not impose a protein-induced bend angle, and the second population (centered at 43 ± 8 deg; 36% of all complexes) corresponds to the bend state (Fig. 5B). In addition, we estimated the stoichiometry of the nucleoprotein complex based on the apparent volume. To account for variability of the sharpness of AFM probes and changing environmental conditions, the apparent volumes were normalized with respect to free DNA volume per unit length.61 Based on a calibration curve constructed using other proteins of known molecular weight, the dimeric nature of EcoRV was confirmed (Fig. 5C and Fig. SI 7). In addition, we performed an independent analysis to evaluate the stoichiometry of nucleoprotein complexes, based on stepwise photo-bleaching (Fig. 5D). Provided a labelling efficiency of 70 %, a dimeric stoichiometry describes our experimental distribution of the number of photo-bleaching steps best (Fig. SI 9), in agreement with our volumetric analysis, X-ray crystallography,58 and biochemical data.62

Volumetric analysis of nucleoprotein complexes to determine their stoichiometry becomes problematic when the size or complexity of the DNA substrate increases (Fig. SI 8).61 To highlight the strength of stepwise photo-bleaching, we next evaluated the binding of fluorescently labelled EcoRV with supercoiled plasmid DNA. We opted for the pBR322 plasmid, which contains a unique EcoRV restriction site within its 4361 bp sequence. We found only ~ 35 % of nucleoprotein complexes bound at a single dsDNA segment (Fig. SI 9A,B). These complexes exhibited step distributions in line with a dimer stoichiometry. Surprisingly, the


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 40

remaining ~ 65 % of complexes appeared at DNA crossovers in the supercoiled molecule, implying simultaneous binding to two segments of dsDNA (Fig. 5E,F). This was further confirmed for wild type, non-labelled EcoRV, excluding DNA-dye interaction (Fig. SI 9C). There are two plausible explanations for this observation. At the level of the EcoRV dimer, only a single DNA can fit in the catalytic cleft formed at the dimer interface. Still, it should not be excluded that a second binding site exists elsewhere in the protein. Alternatively, the crossoverbound complexes constitute a higher order EcoRV stoichiometry. Stepwise photo-bleaching analysis provides evidence for the latter scenario: the distributions of the number of steps are in excellent agreement with predictions for a tetrameric stoichiometry (Fig. 5F and Fig. SI 9E). Thus, the orthogonal information obtained via high quality correlative AFM-SMLM evidences a heterogeneous nucleoprotein stoichiometry that depends on DNA topology.


Page 23 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

Figure 5. Orthogonal probing of heterogeneity in EcoRV-DNA nucleoprotein complexes. A. EcoRV binding to its target sequence. The white arrow indicates the EcoRV protein. B. Bend angle distribution of EcoRV, measured at a length scale of 15 nm. The bend angle is defined as the deviation from linearity, as shown in the inset. The distribution was fit to a double folded Gaussian. The black line represents the population of EcoRV fraction that does not impose DNA bending (mean and standard deviation of 0° and 25°, respectively). The blue line represents the population of EcoRV that imposes DNA bending (mean and standard deviation of 42° and 8°, respectively). ( 93& C. Normalized volume distribution of DNA-bound EcoRV ( 65). The average volume corresponds to the expected volume of an EcoRV dimer (inset). The unit of normalized volume is a direct consequence of the normalization method. D. Fluorescence intensity time trace of the complex shown in A. The dimeric stoichiometry is confirmed by two photobleaching steps E. EcoRV complexed to supercoiled plasmid DNA. The white arrow shows DNA bridging by EcoRV. F. Fluorescence intensity time trace of the complex shown in E. Four photobleaching events are detected, in line with a tetrameric EcoRV stoichiometry.

Conclusion

Spatial and temporal heterogeneity at the nanoscale underlie the properties and mechanisms of biological samples and materials. To further our understanding of these systems, technological advances are a prerequisite. Correlative imaging allows to probe the sample in an orthogonal fashion, which in turn can increase detail of our understanding. Based on a systematic assessment of the parameters that influence data quality, we have presented a rationale to correlated AFM and fluorescence imaging methodology. The


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 40

simultaneous detection of single dyes and internal markers with comparable sensitivity is key towards a high localization precision and registration accuracy. Additional improvements in LRE can be achieved by implementing active stabilization of the imaging optics in our setup63 or the development of alternative fiducial markers with similar / and emission wavelengths of the single dyes. Furthermore, extending the survival time of organic fluorophores enables highconfidence stepwise photo-bleaching analysis. The proximity of the muscovite mica support, used to optimize lateral resolution of topographic images, was found detrimental for the fluorophore longevity. Spectroscopic analysis indicates that on mica, the primary route towards photo-bleaching involves charge transfer from the excited fluorophore to the surface. By modifying the mica surface with poly-L-lysine, we could extend the fluorophore survival time more than twofold, which in turn results in higher quality data. Further research towards the optimization of correlative force-fluorescence microscopy would therefore benefit from the use of dyes with “self-healing” properties,64 and surface modifications that decouple surface-dye interactions.

We validated our experimental design by studying the interaction of site-specifically labelled EcoRV with DNA. High-resolution topographic imaging and stepwise photo-bleaching uncovered a hitherto unresolved nucleoprotein complex, comprised of an EcoRV tetramer bridging two segments of dsDNA. Previous reports demonstrated that coiling of DNA enhances the target search of EcoRV, but the structural basis remains thus far unknown.65 We speculate that the tetrameric stoichiometry of EcoRV bound to supercoil DNA crossings might constitute an intermediate that enhances target search via intersegmental transfer. While the potential functional role of this newly uncovered EcoRV-DNA complex requires further investigations, its


Page 25 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

resolution by AFM-SMLM clearly demonstrates the power to sample structural and compositional heterogeneity at the single molecule level. We anticipate that the progresses described herein, will enable to provide insights into other multipart protein-protein, or proteinnucleic acid architectures. Beyond the current state of the art, the systematic evaluation in this report, provides a critical baseline for further technological improvements.

Last, the technical and methodological advances for increasing fluorescence sensitivity to the level of single organic fluorophores as presented herein, will aid the development of instruments capable of reaching out beyond static nanoscale architecture and composition, e.g. by implementation of single pair fluorescence resonance energy transfer,66 tip-enhanced fluorescence67 or single-turnover mapping of catalysts.68

Materials and methods

DNA substrates and labeling Fluorescent DNA fragments were generated by amplifying a 1000bp DNA fragment of the bacterial

pBR322

plasmid

with

Atto532-AATGCGCTCATCGTCATCC

and

CTGCCAAGGGTTGGTTTG. EcoRV recognition site containing DNA fragments were generated

by

amplifying

a

500bp

fragment

from

pBR322

with

primers

AATGCGCTCATCGTCATCC and CGACGCTCTCCCTTATGC. All primers were purchased from Integrated DNA Technologies.

EcoRV purification and labeling


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 40

EcoRV was purified as previously described.69-70 Briefly, EcoRV expressing E. coli cells were suspended in lysis buffer (30 mM KH2PO4-KOH pH 7.2, 800 mM NaCl, 0.5 mM EDTA, 1 mM DTT) and lysed by sonication. After centrifugation, the supernatant was subjected to affinity chromatography using a HisTrap HP 5ml column (GE Healthcare) equilibrated with lysis buffer. EcoRV was then eluted with lysis buffer supplemented with 250mM Imidazole. Finally, the proteins were loaded onto a HiTrap Heparin HP column (GE Healthcare) and eluted by increasing the concentration of the high salt buffer.

EcoRV was labeled by incubating 50µM of the protein with a 10-fold excess of Atto532maleimide for three hours at room temperature. Unbound dye was removed using Amicon©ultra0.5 centrifugal filters (10KDa MWCO). The degree of labeling was calculated by () *+,- #./0123 !*-45678-45 9:+,- &∗>1+,-

with ?@ABCD =51800 M-1cm-1, ?E$$BFGH =11500 M-1 cm-1 and IJHKL =0.11

which corrects for the fact that Atto532 also absorbs at 280nm.

Sample preparation To account for its absorption and birefringent properties, mica was cleaved until it was thin enough for single molecule fluorescence imaging. For mechanical stability, mica sheets were glued by poly(dimethylsiloxane) (PDMS) to a glass coverslip and cured at 80°C. Prior to sample deposition, the mica was cleaved further until only a minimal thickness remained. Next, 20µl of 0.01% poly-l-lysine (PLL) solution was deposited on the mica and incubated for 30 seconds, before rinsing with 50mL of milliQ water. Dropcasting of 5’ Atto532 labeled DNA (10µl; 0.125ng/µl) mixed with FluoSpheres® carboxylate-modified microscopheres (0.02µm, crimson fluorescent, ThermoFisher) in 10mM 2-amino-2-(hydroxymethyl-propane-1,3-diol 26 ACS Paragon Plus Environment

Page 27 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

(Tris) buffer containing 200mM Na+, was deposited on the PLL functionalized mica and incubated for 30 seconds, prior to rinsing with 20mL of MQ water. 1nM of Atto532 labeled EcoRV was incubated with either a 0.125ng/µl EcoRV recognition site containing DNA fragment or 0.125ng/µl pBR322 plasmid DNA in a 10mM Tris buffer containing 10mM Ca2+, 100mM Na+ and 1mM dithiothreitol (DTT) and incubated for 15 minutes at room temperature. Just before sample deposition, the solution was mixed with fluorescent beads. 10µl of the solution was deposited on PLL functionalized mica, incubated for 30” before being washed with 20mL of MQ water.

Single molecule fluorescence spectroscopy The experiments were carried out on a home-built scanning fluorescence confocal microscopy system based on an Olympus IX71 inverted microscope. A piezo-driven scanning stage (Physik Intrumente P5173CL), which is controlled by a home-written software program, allows for imaging the sample point by point in a raster scanning fashion. Upon laser illumination, the fluorescence emission signal from the sample was collected by a 100× oil immersion objective (Olympus UPLFLN 1.3NA) and collected via an avalanche photodiode (APD, PerkinElmer CD3226). The excitation wavelength of 488 nm was obtained using the second harmonic of a Ti:sapphire laser (Tsunami, Spectra Physics). The Ti:sapphire laser was pumped by an intracavity frequency doubled Millenia laser (Spectra Physics). Appropriate filters were used to suppress the excitation wavelength.

Correlated fluorescence and force microscopy


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 40

Measurements were performed with a CombiScopeTM 1000 (AIST-NT) equipped on a home built transmission-type optical microscope. Excitation light from a diode laser (532 nm) was reflected by a dichroic mirror (Chroma, Z532RDC) and then illuminated onto the sample via wide-field configuration by an oil-immersion objective lens (Nikon, CFI S Fluor, 100x, N.A.1.3). Fluorescence was collected with the same objective and was captured by an electron multiplying charge coupled device (EMCCD) camera (Andor, iXon 897) operated at -65 ˚C, through the dichroic mirror and a long pass filter (Linos, 454 LP). However, in contrast to conventional multicolor fluorescence imaging, emission from both fluorophores and fiducial markers pass through the same optical elements on the same region of the CCD camera.42, 71-74 To minimize vibrations and instabilities, the microscope was installed on an anti-vibration table (Accurion, Halcyonics-micro), the microscope objective was withdrawn from the surface and the laser and EMCCD cooling were switched off. During fluorescence imaging, each frame was recorded with an acquisition time of 110 ms and at least 1500 frames were captured. AFM images were acquired by AC240TS probes (Olympus) in amplitude modulation mode. To determine the dependence of registration accuracy on the number of reference points, a sample with ~ 20 internal references in the topography FOV was measured. For each number of reference points, the registration was repeated 40 times with randomly selected internal references.

Analysis of correlated imaging Unless stated otherwise, fluorescent frames were accumulated per 10 by a moving average to increase the signal to noise ratio (S/N). In house written routines were used to fit each fluorescent signal in each frame by a 2D Gaussian by the least-square method after which each


Page 29 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

frame was drift corrected. Beads were recognized as the fluorescent signals still present in the last 15% of frames. In the AFM images, beads were detected using the triangle method. To accurately determine the bead positions and to correct for parachuting effects, the bead topography was fitted by a 2D Gaussian. The highest pixel closest to the Gaussian center was selected as the bead position. Then, the localized positions of the beads in the fluorescence channel were superpositioned with the positions in the AFM image by an affine transformation and consecutively the whole fluorescence image is correlated with the AFM image. High resolution AFM images were registered with the overview correlated image by an image similarity metric.

AFM analysis AFM images were analyzed by the Scanning Probe Image Processor (SPIP, Image Metrology). Images were flattened by correcting either each line with a polynomial fit (overview AFM images) or by elevating the individual x-profiles so that their height distribution obtains the best match (histogram alignment, high resolution AFM images). Analysis of length, volume and heights was carried out by the “particle and pore analysis module”. To determine protein volumes, the volumes of the entire nucleoprotein complex and naked DNA molecules located in the vicinity of the complex were determined. Next, the average naked DNA volume (< OPQ* >) was subtracted from the nucleoprotein volume (OST ) to obtain the protein volume. Finally, the protein volume (nm3) was normalized relative to the average volume (nm3) per nanometer (nm) length of DNA (OPQ*,VSW ), to account for tip shape, air


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 40

humidity and other factors influencing the apparent volume of proteins (OSB W !OST −< OPQ* >&/OPQ*,VSW ). The resulting unit for the normalized volumes is thus nm. Origin 8.0 was used to statistically process the data, plot figures and fit distributions. The distribution of individual LRE values was fitted to a folded Gaussian function X Y" Y"

6!9[Z2/& -0-

6!96Z2/&-0-

+

.

Prior to fitting the EcoRV DNA bend angles, the bend angle distribution of naked DNA was determined and fitted to a folded Gaussian X Y"

6!96\] &-0-

+ Y"

6!9[\] & -0-

where a, α1 and c

corresponds to the height, bending angle (0°) and standard deviation (25°) of free DNA, respectively. EcoRV bending angles were subsequently fitted to a double folded Gaussian X Y"

6!96\] &-0-

+ Y"

6!9[\] & -0-

+ ^"

6!96\- &-_-

+ ^"

6!9[\- & -_-

where a, α1 and c corresponds to the

height, bending angle (0°) and standard deviation (25°) of free DNA determined earlier, respectively. The values earlier determined for α1 and c were kept constant. d, α2 and f corresponds to the height, bending angle and standard deviation of the EcoRV induced bending angle, respectively, and are optimized using global fitting over both datasets.

ACKNOWLEDGEMENTS. We thank W. Wende for kindly providing the EcoRV(C21S/K58C) expression plasmid, J. Hofkens and M. van der Auweraer for the use of the single molecule fluorescence spectroscopy setup, J. Lipfert and P. Walker for thoughtful discussions and critical reading of the manuscript, J. Demeulemeester and K. Čermáková for help with protein purification, and other members of the Debyser lab for constructive discussions. We acknowledge funding from KU Leuven through the IDO program for financial support; WF, SR and WV like to thank Fonds Wetenschappelijk Onderzoek (FWO) for personal fellowships.


Page 31 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

SUPPORTING INFORMATION Available: Supporting Figures S1-9. This material is available free of charge via the Internet at http://pubs.acs.org.

TABLE OF CONTENT GRAPHIC

References

1.

Solomatin, S. V.; Greenfeld, M.; Herschlag, D., Implications of Molecular Heterogeneity

for the Cooperativity of Biological Macromolecules. Nat. Struct. Mol. Biol. 2011, 18, 732-734. 2.

van Oijen, A. M.; Blainey, P. C.; Crampton, D. J.; Richardson, C. C.; Ellenberger, T.;

Xie, X. S., Single-Molecule Kinetics of Lambda Exonuclease Reveal Base Dependence and Dynamic Disorder. Science 2003, 301, 1235-1238. 3.

Engelkamp, H.; Hatzakis, N. S.; Hofkens, J.; De Schryver, F. C.; Nolte, R. J.; Rowan, A.

E., Do Enzymes Sleep and Work? Chem. Commun. (Camb.) 2006, 935-940. 4.

Sahl, S. J.; Moerner, W. E., Super-Resolution Fluorescence Imaging with Single

Molecules. Curr. Opin. Struct. Biol. 2013, 23, 778-787. 5.

Xiao, J.; Dufrene, Y. F., Optical and Force Nanoscopy in Microbiology. Nat. Microbiol.

2016, 1, 1-13. 31 ACS Paragon Plus Environment

ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

6.

Page 32 of 40

Binnig, G.; Quate, C. F.; Gerber, C., Atomic Force Microscope. Phys. Rev. Lett. 1986, 56,

930-933. 7.

Lyubchenko, Y. L., Preparation of DNA and Nucleoprotein Samples for Afm Imaging.

Micron 2011, 42, 196-206. 8.

Lyubchenko, Y. L.; Shlyakhtenko, L. S., Imaging of DNA and Protein-DNA Complexes

with Atomic Force Microscopy. Crit. Rev. Eukaryot. Gene Expr. 2016, 26, 63-96. 9.

Henderson, E.; Haydon, P. G.; Sakaguchi, D. S., Actin Filament Dynamics in Living

Glial Cells Imaged by Atomic Force Microscopy. Science 1992, 257, 1944-1946. 10.

Hoh, J. H.; Schoenenberger, C. A., Surface Morphology and Mechanical Properties of

Mdck Monolayers by Atomic Force Microscopy. J. Cell Sci. 1994, 107 1105-1114. 11.

Mou, J.; Yang, J.; Shao, Z., Atomic Force Microscopy of Cholera Toxin B-Oligomers

Bound to Bilayers of Biologically Relevant Lipids. J. Mol. Biol. 1995, 248, 507-512. 12.

Hansma, H. G.; Vesenka, J.; Siegerist, C.; Kelderman, G.; Morrett, H.; Sinsheimer, R. L.;

Elings, V.; Bustamante, C.; Hansma, P. K., Reproducible Imaging and Dissection of Plasmid DNA under Liquid with the Atomic Force Microscope. Science 1992, 256, 1180-1184. 13.

Vanderlinden, W.; Blunt, M.; David, C. C.; Moucheron, C.; Kirsch-De Mesmaeker, A.;

De Feyter, S., Mesoscale DNA Structural Changes on Binding and Photoreaction with Ru[(Tap)2phehat]2+. J. Am. Chem. Soc. 2012, 134, 10214-10221. 14.

Vanderlinden, W.; Lipfert, J.; Demeulemeester, J.; Debyser, Z.; De Feyter, S., Structure,

Mechanics, and Binding Mode Heterogeneity of Ledgf/P75-DNA Nucleoprotein Complexes Revealed by Scanning Force Microscopy. Nanoscale 2014, 6, 4611-4619. 15.

Friedrichs, J.; Taubenberger, A.; Franz, C. M.; Muller, D. J., Cellular Remodelling of

Individual Collagen Fibrils Visualized by Time-Lapse Afm. J. Mol. Biol. 2007, 372, 594-607.


Page 33 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

16.

Hinterdorfer, P.; Baumgartner, W.; Gruber, H. J.; Schilcher, K.; Schindler, H., Detection

and Localization of Individual Antibody-Antigen Recognition Events by Atomic Force Microscopy. Proc. Natl. Acad. Sci. 1996, 93, 3477-3481. 17.

Dupres, V.; Menozzi, F. D.; Locht, C.; Clare, B. H.; Abbott, N. L.; Cuenot, S.; Bompard,

C.; Raze, D.; Dufrene, Y. F., Nanoscale Mapping and Functional Analysis of Individual Adhesins on Living Bacteria. Nat. Methods 2005, 2, 515-520. 18.

Stroh, C.; Wang, H.; Bash, R.; Ashcroft, B.; Nelson, J.; Gruber, H.; Lohr, D.; Lindsay, S.

M.; Hinterdorfer, P., Single-Molecule Recognition Imaging Microscopy. Proc. Natl. Acad. Sci. 2004, 101, 12503-12507. 19.

Pfreundschuh, M.; Alsteens, D.; Hilbert, M.; Steinmetz, M. O.; Muller, D. J., Localizing

Chemical Groups While Imaging Single Native Proteins by High-Resolution Atomic Force Microscopy. Nano Lett. 2014, 14, 2957-2964. 20.

Betzig, E.; Lewis, A.; Harootunian, A.; Isaacson, M.; Kratschmer, E., Near Field

Scanning Optical Microscopy (Nsom): Development and Biophysical Applications. J. Biophys. 1986, 49, 269-279. 21.

Stöckle, R. M.; Suh, Y. D.; Deckert, V.; Zenobi, R., Nanoscale Chemical Analysis by

Tip-Enhanced Raman Spectroscopy. Chem. Phys. Lett. 2000, 318, 131-136. 22.

Lewis, A.; Taha, H.; Strinkovski, A.; Manevitch, A.; Khatchatouriants, A.; Dekhter, R.;

Ammann, E., Near-Field Optics: From Subwavelength Illumination to Nanometric Shadowing. Nat. Biotechnol. 2003, 21, 1378-1386. 23.

Zhang, Z.; Sheng, S.; Wang, R.; Sun, M., Tip-Enhanced Raman Spectroscopy. Anal.

Chem. 2016, 88, 9328-9346.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

24.

Page 34 of 40

Verma, P., Tip-Enhanced Raman Spectroscopy: Technique and Recent Advances. Chem.

Rev. 2017, 117, 6447-6466. 25.

Hecht, E.; Thompson, K.; Frick, M.; Wittekindt, O. H.; Dietl, P.; Mizaikoff, B.; Kranz,

C., Combined Atomic Force Microscopy–Fluorescence Microscopy: Analyzing Exocytosis in Alveolar Type Ii Cells. Anal. Chem. 2012, 84, 5716-5722. 26.

Xu, X.; Li, Z. Y.; Cai, L. Y.; Calve, S.; Neu, C. P., Mapping the Nonreciprocal

Micromechanics of Individual Cells and the Surrounding Matrix within Living Tissues. Sci. Rep. 2016, 6, 1-9. 27.

Colom, A.; Casuso, I.; Rico, F.; Scheuring, S., A Hybrid High-Speed Atomic Force-

Optical Microscope for Visualizing Single Membrane Proteins on Eukaryotic Cells. Nat. Commun. 2013, 4, 1-8. 28.

Harke, B.; Chacko, J. V.; Haschke, H.; Canale, C.; Diaspro, A., A Novel Nanoscopic

Tool by Combining Afm with Sted Microscopy. Opt. Nanoscopy 2012, 1, 1-6. 29.

Monserrate, A.; Casado, S.; Flors, C., Correlative Atomic Force Microscopy and

Localization-Based

Super-Resolution

Microscopy:

Revealing

Labelling

and

Image

Reconstruction Artefacts. ChemPhysChem 2014, 15, 647-650. 30.

Bondia, P.; Jurado, R.; Casado, S.; Dominguez-Vera, J. M.; Galvez, N.; Flors, C., Hybrid

Nanoscopy of Hybrid Nanomaterials. Small 2017, 13, 1-7. 31.

Odermatt, P. D.; Shivanandan, A.; Deschout, H.; Jankele, R.; Nievergelt, A. P.; Feletti,

L.; Davidson, M. W.; Radenovic, A.; Fantner, G. E., High-Resolution Correlative Microscopy: Bridging the Gap between Single Molecule Localization Microscopy and Atomic Force Microscopy. Nano Lett. 2015, 15, 4896-4904.


Page 35 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

32.

Fronczek, D. N.; Quammen, C.; Wang, H.; Kisker, C.; Superfine, R.; Taylor, R.; Erie, D.

A.; Tessmer, I., High Accuracy Fiona-Afm Hybrid Imaging. Ultramicroscopy 2011, 111, 350355. 33.

Sanchez, H.; Kertokalio, A.; van Rossum-Fikkert, S.; Kanaar, R.; Wyman, C., Combined

Optical and Topographic Imaging Reveals Different Arrangements of Human Rad54 with Presynaptic and Postsynaptic Rad51–DNA Filaments. Proc. Natl. Acad. Sci. 2013, 110, 1138511390. 34.

Sanchez, H.; Paul, M. W.; Grosbart, M.; van Rossum-Fikkert, S. E.; Lebbink, J. H. G.;

Kanaar, R.; Houtsmuller, A. B.; Wyman, C., Architectural Plasticity of Human Brca2-Rad51 Complexes in DNA Break Repair. Nucleic Acids Res. 2017, 45, 4507-4518. 35.

Yildiz, A.; Forkey, J. N.; McKinney, S. A.; Ha, T.; Goldman, Y. E.; Selvin, P. R., Myosin

V Walks Hand-over-Hand: Single Fluorophore Imaging with 1.5-Nm Localization. Science 2003, 300, 2061-2065. 36.

Bustamante, C.; Rivetti, C., Visualizing Protein-Nucleic Acid Interactions on a Large

Scale with the Scanning Force Microscope. Ann. Rev. Biophys. Biomol. Struct. 1996, 25, 395429. 37.

El-Bahrawi, M. S.; Nagib, N. N.; Khodier, S. A.; Sidki, H. M., Birefringence of

Muscovite Mica. Opt. Laser Technol. 1998, 30, 411-415. 38.

Rocha, S.; De Keersmaecker, H.; Hutchison, J. A.; Vanhoorelbeke, K.; Martens, J. A.;

Hofkens, J.; Uji-i, H., Membrane Remodeling Processes Induced by Phospholipase Action. Langmuir 2014, 30, 4743-4751.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

39.

Page 36 of 40

Rocha, S.; Hutchison, J. A.; Peneva, K.; Herrmann, A.; Mullen, K.; Skjot, M.; Jorgensen,

C. I.; Svendsen, A.; De Schryver, F. C.; Hofkens, J.; Uji-i, H., Linking Phospholipase Mobility to Activity by Single-Molecule Wide-Field Microscopy. ChemPhysChem 2009, 10, 151-161. 40.

Balme, S.; Janot, J.-M.; Déjardin, P.; Vasina, E. N.; Seta, P., Potentialities of Confocal

Fluorescence for Investigating Protein Adsorption on Mica and in Ultrafiltration Membranes. J. Membr. Sci. 2006, 284, 198-204. 41.

Schulz, O.; Zhao, Z.; Ward, A.; Koenig, M.; Koberling, F.; Liu, Y.; Enderlein, J.; Yan,

H.; Ros, R., Tip Induced Fluorescence Quenching for Nanometer Optical and Topographical Resolution. Opt. Nanoscopy 2013, 2, 1-8. 42.

Churchman, L. S.; Okten, Z.; Rock, R. S.; Dawson, J. F.; Spudich, J. A., Single Molecule

High-Resolution Colocalization of Cy3 and Cy5 Attached to Macromolecules Measures Intramolecular Distances through Time. Proc. Natl. Acad. Sci. 2005, 102, 1419-1423. 43.

Cohen, E. A. K.; Ober, R. J., Analysis of Point Based Image Registration Errors with

Applications in Single Molecule Microscopy. IEEE Trans. Signal Process. 2013, 61, 6291-6306. 44.

Cohen, E. A. K.; Kim, D.; Ober, R. J., Cramer-Rao Lower Bound for Point Based Image

Registration with Heteroscedastic Error Model for Application in Single Molecule Microscopy. IEEE Trans. Med. Imaging 2015, 34, 2632-2644. 45.

Rust, M. J.; Bates, M.; Zhuang, X., Sub-Diffraction-Limit Imaging by Stochastic Optical

Reconstruction Microscopy (Storm). Nat. Methods 2006, 3, 793-795. 46.

Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J.

S.; Davidson, M. W.; Lippincott-Schwartz, J.; Hess, H. F., Imaging Intracellular Fluorescent Proteins at Nanometer Resolution. Science 2006, 313, 1642-1645.


Page 37 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

47.

Bates, M.; Dempsey, G. T.; Chen, K. H.; Zhuang, X., Multicolor Super-Resolution

Fluorescence Imaging Via Multi-Parameter Fluorophore Detection. ChemPhysChem 2012, 13, 99-107. 48.

Shechtman, Y.; Weiss, L. E.; Backer, A. S.; Lee, M. Y.; Moerner, W. E., Multicolour

Localization Microscopy by Point-Spread-Function Engineering. Nat. Photonics 2016, 10, 590594. 49.

Thompson, R. E.; Larson, D. R.; Webb, W. W., Precise Nanometer Localization Analysis

for Individual Fluorescent Probes. J. Biophys. 2002, 82, 2775-2783. 50.

Mortensen, K. I.; Churchman, L. S.; Spudich, J. A.; Flyvbjerg, H., Optimized

Localization Analysis for Single-Molecule Tracking and Super-Resolution Microscopy. Nat. Methods 2010, 7, 377-381. 51.

Yi, J.; Manna, A.; Barr, V. A.; Hong, J.; Neuman, K. C.; Samelson, L. E., Madstorm: A

Superresolution Technique for Large-Scale Multiplexing at Single-Molecule Accuracy. Mol. Biol. Cell 2016, 27, 3591-3600. 52.

Kerssemakers, J. W. J.; Laura Munteanu, E.; Laan, L.; Noetzel, T. L.; Janson, M. E.;

Dogterom, M., Assembly Dynamics of Microtubules at Molecular Resolution. Nature 2006, 442, 709-712. 53.

Bussiek, M.; Mücke, N.; Langowski, J., Polylysine‐Coated Mica Can Be Used to Observe

Systematic Changes in the Supercoiled DNA Conformation by Scanning Force Microscopy in Solution. Nucleic Acids Res. 2003, 31, e137-e137. 54.

Bustamante, C.; Vesenka, J.; Tang, C. L.; Rees, W.; Guthold, M.; Keller, R., Circular

DNA Molecules Imaged in Air by Scanning Force Microscopy. Biochemistry 1992, 31, 22-26.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

55.

Page 38 of 40

Zondervan, R.; Kulzer, F.; Kol'chenk, M. A.; Orrit, M., Photobleaching of Rhodamine 6g

in Poly(Vinyl Alcohol) at the Ensemble and Single-Molecule Levels. J. Phys. Chem. A 2004, 108, 1657-1665. 56.

Kennes, K.; Dedecker, P.; Hutchison, J. A.; Fron, E.; Uji-i, H.; Hofkens, J.; Van der

Auweraer, M., Field-Controlled Charge Separation in a Conductive Matrix at the SingleMolecule Level: Toward Controlling Single-Molecule Fluorescence Intermittency. ACS Omega 2016, 1, 1383-1392. 57.

Yeow, E. K.; Melnikov, S. M.; Bell, T. D.; De Schryver, F. C.; Hofkens, J.,

Characterizing the Fluorescence Intermittency and Photobleaching Kinetics of Dye Molecules Immobilized on a Glass Surface. J. Phys. Chem. A 2006, 110, 1726-1734. 58.

Winkler, F. K.; Banner, D. W.; Oefner, C.; Tsernoglou, D.; Brown, R. S.; Heathman, S.

P.; Bryan, R. K.; Martin, P. D.; Petratos, K.; Wilson, K. S., The Crystal Structure of Ecorv Endonuclease and of Its Complexes with Cognate and Non-Cognate DNA Fragments. EMBO J. 1993, 12, 1781-1795. 59.

Horton, N. C.; Perona, J. J., DNA Cleavage by Ecorv Endonuclease: Two Metal Ions in

Three Metal Ion Binding Sites. Biochemistry 2004, 43, 6841-6857. 60.

Zahran, M.; Daidone, I.; Smith, J. C.; Imhof, P., Mechanism of DNA Recognition by the

Restriction Enzyme Ecorv. J. Mol. Biol. 2010, 401, 415-432. 61.

Fuentes-Perez, M. E.; Gwynn, E. J.; Dillingham, M. S.; Moreno-Herrero, F., Using DNA

as a Fiducial Marker to Study Smc Complex Interactions with the Atomic Force Microscope. J. Biophys. 2012, 102, 839-848. 62.

Pingoud, A.; Jeltsch, A., Structure and Function of Type Ii Restriction Endonucleases.

Nucleic Acids Res. 2001, 29, 3705-3727.


Page 39 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Nano

63.

Pertsinidis, A.; Zhang, Y.; Chu, S., Subnanometre Single-Molecule Localization,

Registration and Distance Measurements. Nature 2010, 466, 647-651. 64.

Altman, R. B.; Terry, D. S.; Zhou, Z.; Zheng, Q.; Geggier, P.; Kolster, R. A.; Zhao, Y.;

Javitch, J. A.; Warren, J. D.; Blanchard, S. C., Cyanine Fluorophore Derivatives with Enhanced Photostability. Nat. Methods 2011, 9, 68-71. 65.

van den Broek, B.; Lomholt, M. A.; Kalisch, S.-M. J.; Metzler, R.; Wuite, G. J. L., How

DNA Coiling Enhances Target Localization by Proteins. Proc. Natl. Acad. Sci. 2008, 105, 15738-15742. 66.

Ha, T.; Ting, A. Y.; Liang, J.; Caldwell, W. B.; Deniz, A. A.; Chemla, D. S.; Schultz, P.

G.; Weiss, S., Single-Molecule Fluorescence Spectroscopy of Enzyme Conformational Dynamics and Cleavage Mechanism. Proc. Natl. Acad. Sci. 1999, 96, 893-898. 67.

Ma, Z.; Gerton, J. M.; Wade, L. A.; Quake, S. R., Fluorescence near-Field Microscopy of

DNA at Sub-10 Nm Resolution. Phys. Rev. Lett. 2006, 97, 1-4. 68.

Roeffaers, M. B. J.; Sels, B. F.; Uji-i, H.; De Schryver, F. C.; Jacobs, P. A.; De Vos, D.

E.; Hofkens, J., Spatially Resolved Observation of Crystal-Face-Dependent Catalysis by Single Turnover Counting. Nature 2006, 439, 572-575. 69.

Bonnet, I.; Biebricher, A.; Porte, P. L.; Loverdo, C.; Benichou, O.; Voituriez, R.; Escude,

C.; Wende, W.; Pingoud, A.; Desbiolles, P., Sliding and Jumping of Single Ecorv Restriction Enzymes on Non-Cognate DNA. Nucleic Acids Res. 2008, 36, 4118-4127. 70.

Schulze, C.; Jeltsch, A.; Franke, I.; Urbanke, C.; Pingoud, A., Crosslinking the Ecorv

Restriction Endonuclease across the DNA-Binding Site Reveals Transient Intermediates and Conformational Changes of the Enzyme During DNA Binding and Catalytic Turnover. EMBO J. 1998, 17, 6757-6766.


ACS Nano 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

71.

Page 40 of 40

Gahlmann, A.; Ptacin, J. L.; Grover, G.; Quirin, S.; von Diezmann, A. R.; Lee, M. K.;

Backlund, M. P.; Shapiro, L.; Piestun, R.; Moerner, W. E., Quantitative Multicolor Subdiffraction Imaging of Bacterial Protein Ultrastructures in Three Dimensions. Nano Lett. 2013, 13, 987-993. 72.

Pertsinidis, A.; Mukherjee, K.; Sharma, M.; Pang, Z. P.; Park, S. R.; Zhang, Y.; Brunger,

A. T.; Sudhof, T. C.; Chu, S., Ultrahigh-Resolution Imaging Reveals Formation of Neuronal Snare/Munc18 Complexes in Situ. Proc Natl Acad Sci U S A 2013, 110, E2812-2820. 73.

Erdelyi, M.; Rees, E.; Metcalf, D.; Schierle, G. S.; Dudas, L.; Sinko, J.; Knight, A. E.;

Kaminski, C. F., Correcting Chromatic Offset in Multicolor Super-Resolution Localization Microscopy. Opt. Express 2013, 21, 10978-10988. 74.

Lehmann, M.; Rocha, S.; Mangeat, B.; Blanchet, F.; Uji, I. H.; Hofkens, J.; Piguet, V.,

Quantitative Multicolor Super-Resolution Microscopy Reveals Tetherin Hiv-1 Interaction. PLoS Pathog. 2011, 7, e1002456.


Orthogonal Probing of Single Molecule Heterogeneity by Correlative

Recommend Documents