Axial Colocalization of Single Molecules with Nanometer Accuracy

Subscriber access provided by Queen Mary, University of London

Axial co-localization of single molecules with nanometer accuracy using MIET Sebastian Isbaner, Narain Karedla, Izabela Kaminska, Daja Ruhlandt, Mario Raab, Johann Bohlen, Alexey I. Chizhik, Ingo Gregor, Philip Tinnefeld, Joerg Enderlein, and Roman Tsukanov Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b00425 • Publication Date (Web): 21 Mar 2018 Downloaded from http://pubs.acs.org on March 22, 2018

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 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 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.

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 15 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

Nano Letters

Axial co-localization of single molecules with nanometer accuracy using MIET

Sebastian Isbaner1, Narain Karedla1,2, Izabela Kaminska3,4,6, Daja Ruhlandt1, Mario Raab3,5, Johann Bohlen3,5, Alexey Chizhik1, Ingo Gregor1, Philip Tinnefeld3,5, Jörg Enderlein*,1,2 , Roman Tsukanov*,1

1 2

Third Institute of Physics – Biophysics, Georg August University, Göttingen, Germany

DFG Research Center “Nanoscale Microscopy and Molecular Physiology of the Brain” (CNMPB), Göttingen, Germany

3

Institute for Physical and Theoretical Chemistry, Braunschweig Integrated Centre of Systems Biology (BRICS), and Laboratory for Emerging Nanometrology (LENA), Braunschweig University of Technology, Braunschweig, Germany 4

Institute of Physics, Faculty of Physics, Astronomy, and Informatics, Nicolaus Copernicus University, Torun, Poland 5

6

Department of Chemistry and Center for NanoScience, Ludwig-Maximilians-Universität München, München, Germany

Current address: Institute of Physical Chemistry of the Polish Academy of Sciences, Warsaw, Poland

Keywords Single molecule fluorescence, super-resolution microscopy, plasmonics, DNA origami, MIET microscopy, fluorescence lifetime Abstract Single-molecule localization based super-resolution microscopy has revolutionized optical microscopy and routinely allows for resolving structural details down to a few nanometers. However, there exists a rather large discrepancy between lateral and axial localization accuracy, the latter typically three to five times worse than the former. Here, we use single-molecule Metal Induced Energy Transfer (smMIET) to localize single molecules along the optical axis, and to measure their axial distance with an accuracy of 5 nm. smMIET relies only on fluorescence lifetime measurements and does not require additional complex optical setups.

ACS Paragon Plus Environment

Nano Letters 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

Introduction The last two and a half decades have seen a tremendous development of new optical superresolution techniques in microscopy and their applications to medicine, biology and biophysics. Among these techniques are STED1, PALM2, STORM3, dSTORM4, SOFI5, DNA PAINT6 and many others (for a comprehensive recent review see Ref 7). These techniques are capable of laterally resolving structural details with a resolution down to a few nanometers, more than two orders of magnitude better than the classical diffraction limit (~200 nm). Several ingenious concepts have been developed to achieve a comparably high resolution also along the optical axis. Among these methods are the use of special phase plates for achieving high STED confinement along the optical axis8, or biplane imaging9, astigmatic imaging10, and helical wavefront shaping11 for axial super-resolution localization of single molecules in PALM and STORM. Nevertheless, all these methods provide an axial resolution that is by a factor of three to five worse than the lateral resolution, similar to conventional microscopy. As a result, several specialized methods have been invented for achieving maximal spatial resolution along the optical axis. Among them are supercritical angle fluorescence (SAF) imaging12, variable angle total internal reflection fluorescence microscopy (vaTIRF)13, 4pi-STED microscopy14, and interference PALM (iPALM)15. Whereas the first two methods can deliver/provide an axial resolution of a few nanometers, their lateral resolution is still diffraction-limited. On the other hand, the last two methods deliver indeed a nearly isotropic nanometer optical resolution, but for the cost of exceptionally complex and expensive technologies that are not easy to implement or to maintain. When it comes to the measurement of distances on the nanometer scale, another classical method comes to mind: Förster Resonance Energy Transfer (FRET)16. In FRET, one labels a structure of interest with two fluorescing molecules, a donor and an acceptor, with the fluorescence emission spectrum of the donor overlapping the absorption spectrum of the acceptor. Upon excitation of the donor, its excited state energy can electromagnetically be transferred to the acceptor, if the acceptor is a few nanometers away from the donor. Because the efficiency of this energy transfer is highly sensitive to the distance between the donor and the acceptor, FRET can be used to measure inter- and intramolecular distances between two positions with Ångström accuracy on a length scale from ~2 nm to 10 nm. However, labeling a protein or a molecular complex with two fluorescent dyes can be challenging. Also, the yield of double-labelling is low while the cost of commercial synthesis is high. Another issue in FRET is that it is often challenging to convert a measured energy transfer efficiency into the correct distance value. This is due to technical complications such as possible direct acceptor excitation by the donor excitation laser, bleed-through of donor emission into the acceptor channel, or uncertainties in the calibration of the so-called γ-factor (ratio between emitted acceptor photons to transferred donor excitation energy quanta). Even more, an unresolved problem in FRET is the exact determination of the so-called κ-factor describing the relative orientation of the donor’s emission dipole and the acceptor’s absorption dipole17-18. In order to extract a correct donor-acceptor distance, each particular FRET pair has to be accurately calibrated, e.g. by using double-stranded DNA19 or DNA origami20 in the same buffer and experimental conditions as used in the actual system of interest. Here, we present an alternative approach for high-precision (nanometer accuracy) singlemolecule co-localization measurements along the optical axis. Our approach employs Metal-


Page 2 of 15

Page 3 of 15 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

Nano Letters

Induced Energy Transfer (MIET), which exploits the fact that an excited fluorescent molecule can efficiently transfer its excited state energy to surface plasmons21 in a metal film, similar to the resonant energy transfer from a donor to an acceptor molecule in FRET. However, in contrast to FRET, the interaction range between a fluorescent molecule and a metal film is roughly one order of magnitude (up to ~100 nm), the optical absorption spectrum of a metal film is very broad (thus any fluorescent dye within the visual spectral range can be used), and the energy transfer efficiency depends only on one orientation angle (the angle between the emission dipole axis and the surface normal of the metal film). Moreover, MIET requires the specific labeling of only one site of a biomolecule, which is much easier than the double-labeling required by FRET. We introduced the concept of MIET in Chizhik A.I. et al22, where we used it to map the topography of the basal membrane of fluorescently labeled live cells with ~3 nm axial resolution. In a subsequent paper, we demonstrated that MIET can be applied even for localizing single molecules along the optical axis with a similar localization accuracy23. In all these publications, MIET was always used to measure the absolute distance of an emitter (or ensembles of emitters) from a surface. Thus, we introduced dual-color MIET for measuring the distance between two intracellular lipid membranes (inner and outer nuclear envelope) in fixed cells, by first measuring the distance of the two differently labeled structures from the surface, and thereafter calculating the distance between them, see Chizhik A.M. et al24. However, similar to the single-molecule localization techniques such as STORM and PALM, it would be desirable to be able to measure intra- or intermolecular distances by using only one kind of label and then to co-localize several individual labels on the same structure (i.e. in a diffraction-limited spot). This is the topic of the current paper. Here, we perform smMIET-based co-localization of single fluorescent molecules with ten nanometer accuracy along the optical axis, and we verify the accuracy of the method by using three-dimensional DNA origami structures25 with well-known geometry as reference. The only experimental requirement for smMIET is the availability of a single-molecule sensitive confocal microscope with the capability of measuring fluorescence lifetimes.


Nano Letters 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

Results and Discussion Single-molecule MIET experiment In order to exemplify and verify our method of axial co-localization of single molecules, we employ rigid DNA origami pillar structures with well-defined geometry and orientation. Due to the freedom of design and structural reproducibility of DNA origami, together with the possibility to attach dyes on the origami with high site specificity, fluorophores can be precisely positioned in space at any desired position with nanometer accuracy. Also, DNA origami rulers have been recently used as traceable distance measurement standard for fluorescence microscopy26. We have chosen the dye Atto647N as the fluorescent label because of its exceptional photo-stability and high quantum yield. Orientation-specific surface immobilization of the DNA pillars was achieved in two different ways. In the first approach, we used thiol groups at the bottom base of the pillars for binding them directly to the gold covered surface of a coverslip, see Figure 1, (A) and (E). In the second approach, we used covalently attached biotins for binding them via neutravidin to a thin SiO2 spacer layer on top of the metallized coverslip surface, as shown in Figure 1, (B) to (D) (see Methods section for more details).

Figure 1. The library of DNA origami structures and surfaces used in this work: (A-C) Pillar with one fluorophore at 82 nm height, on top of gold or silver metal layers. For these, thiol-gold and biotin-neutravidin specific surface immobilization strategies were employed. (D) Pillar with one fluorophore at 82 nm height on top of glass coverslip, for reference measurement. (E) Pillar with two fluorophores at 26 nm and 58 nm height, respectively. Here, a gold layer without SiO2 spacer was used.

Eight anchoring sites with thiol groups (or biotins) on the bottom of the structure were used to achieve a vertical orientation of the surface-immobilized pillars. Anti-oxidant (Trolox) was added to the buffer solution to suppress dye blinking27, so that bleaching steps of single emitters could be easily identified within recorded fluorescence intensity time traces28. Photons between bleaching steps (or between the measurement start and the first bleaching step) were used to calculate individual TCSPC histograms, which correspond to a fixed number of fluorescent molecules (see Figure 2 for analysis scheme, and Figure 3 for raw data examples). For samples with two emitters on one DNA origami pillar, two bleaching steps were clearly detectable in more than 80% of the observed spots. In a few cases, the first bleaching step occurred too fast and the resulting numbers of collected photons were too small for reliable lifetime fitting. Such trajectories were excluded from the analysis.


Page 4 of 15

Page 5 of 15 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

Nano Letters

Figure 2. Metal-Induced Energy Transfer: experiment and analysis flow. (A) Experiment geometry and MIET coverslip composition. (B) Exemplary data of intensity time trajectory and TCSPC histogram. A single emitter shows a monoexponential decay curve (blue) which is used to fit the fluorescence lifetime. After the emitter bleaches, the TCSPC histogram shows gold luminescence and background signal only (green). (C) Lifetime histogram for 157 single molecules. (D) The MIET lifetime-distance dependence is used to convert the fluorescence lifetime into height. (E) Final height histogram. All data was taken from the measurement of a single-emitter labeled pillar immobilized with biotinneutravidin on a gold coated coverslip with 10 nm SiO2 spacer (P1-B-Gold).

Figure 3. smMIET experiment: Intensity time traces and TCSPC histograms. (A) Pillar with one fluorophore on top of a gold layer; surface immobilization via thiol-gold binding. (B) Pillar with one fluorescent dye on top of silicon dioxide layer; surface immobilization via biotin-neutravidin binding. (C) Pillar with one fluorescent dye on top of glass coverslip (reference measurement); surface immobilization via biotin-neutravidin binding. (D) Pillar with two fluorescent dyes on top of a gold layer; thiol-gold surface immobilization. The TCSPC histograms correspond to the fluorescence signal before (blue) and after photobleaching (green). The latter consists of the short decay of the gold luminescence and a constant background. The red histogram in (D) corresponds to the time after the photobleaching of the first emitter until the bleaching of the second emitter.


Nano Letters 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

Single dye axial localization To characterize the ability of smMIET to localize the position of a single molecule along the optical axis, we used DNA origami pillars with a single dye attached at a designed height of 82 nm from the bottom of the structure. We used gold and silver metal layers for smMIET, which show the strongest fluorescence-surface-plasmon interaction in the visible spectral region. As already mentioned in the previous section, we used thiol-gold and biotin-neutravidin binding for oriented specific immobilization of the fluorescently labeled DNA origami pillars. The measured lifetime and height histograms are displayed in Figure 4. For the thiol-immobilized DNA origami pillars on a gold surface we found a height of (76.2 ± 3.8) nm above the surface, see Figure 4 (C1). For the biotin-immobilized DNA origami pillar on top of a SiO2 spacer and gold, the peak in the histogram appeared at (73.9 ± 4.5) nm, Figure 4 (C2), and for biotin-immobilized DNA origami pillar on top of a SiO2 spacer and silver, the peak appeared at (75.3 ± 5.1) nm, Figure 4 (C3). For all samples, our analysis yielded axial positions of the dye that are about 10% smaller than the design value of 82 nm. This is particularly surprising because we have previously shown that smMIET can localize molecules that are immobilized directly on a SiO2 spacer layer above the metal with nanometer accuracy. Moreover, we find lower heights for both gold and silver smMIET measurements. To verify the structural dimensions of the DNA origami, we performed 3D DNA Point Accumulation for Imaging in Nanoscale Topography (DNA PAINT)29-32. For this we used DNA origami pillars with the two docking sites positioned 90 nm apart. Also for this case, we find that the measured distance between the dyes is about 10% smaller than the design value, see Figure S2 in Supporting Information. The most probable explanation for these consistently smaller values is a deviation of the design height from the actual height. The design height is calculated using a simplified model of DNA obviously not correctly describing the intermark distance that is influenced by interhelix distances, bending, twist, and possible tilt of the structure. Another reason for this deviation could be due to electrostatic screening of the negatively charged DNA backbone in the used buffers with high concentrations of Mg++ ions (12.5 mM MgCl2)30. Measured distances deviating from the design distances have been also observed previously for DNA origami structures under similar experimental conditions29. Another observation is that we measure nearly the same vertical position of the dye whether the DNA origami pillar is directly bound to the surface via thiols, or via a biotin-neutravidin chemistry, which should theoretically add another ~10 nm. This can be explained by assuming that the biotin-neutravidin immobilization introduces a considerable tilt in the orientations of the pillars, which reduces the observed height value. This is supported by the 3D DNA PAINT experiment mentioned above, showing a broad distribution of orientation angles with an average tilt angle of 35° with respect to the vertical axis (Figure S2), which translates into a 1 ‒ cos(35°) ~18% height reduction.


Page 6 of 15

Page 7 of 15 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

Nano Letters

Figure 4. smMIET measurements of single-dye vertical position on DNA origami pillars. Measurements were performed on different substrates with different metal coating (gold and silver) and using different immobilization techniques (thiol-gold and biotin-neutravidin). Shown are: (A1-4) different experimental designs, (B1-4) lifetime histograms (N denotes the number of molecules in the histogram), and (C1-3) height histograms. The last panel (A4, B4) represents a control measurement on glass and is shown here for reference. In all lifetime histograms, the free-space lifetime is indicated by a red line. For each histogram, measurements on 150 – 400 individual origami structures were accumulated.


Nano Letters 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

Axial co-localization of two dyes For co-localization experiments, we used thiol-immobilized DNA origami pillars with two fluorophores attached at heights of 26 nm and 58 nm, Figure 5 (A). Using step-wise photobleaching, the fluorescence lifetime of each emitter can be determined and converted into a height value, see Figure 5 (B1-2), and Methods section for details. In short, we detect the bleaching steps (Figure 5 (B1)) and extract the corresponding TCSPC histograms (Figure 5 (B2)). Then, we fit the lifetime of the single emitter (red curve) and fix this lifetime in the biexponential fit of the two emitters (blue curve). The lifetimes are then converted into heights and histogrammed for all molecules, see Figure 5 (C1). Two peaks in the histogram appear at (23 ± 3) nm and (51 ± 3) nm, which is consistent with the deviation of ~10% from the design value discussed in the previous section. By subtracting the first value from the latter, the difference of these two values is (28 ± 5) nm and gives an estimate of the height difference between the two emitters. Alternatively, the distance can be calculated from each individual origami structure and then histogrammed. In that case, the average distance between the emitters was found to be (32 ± 11) nm, as shown in panel (C2) in Figure 5. The values of both approaches, (28 ± 5) nm and (32 ± 11) nm, are in excellent agreement with the design value of 32 nm. Although the first value seems to be more precise, the clear advantage of the second approach is the fact that it provides actual co-localization of two non-blinking emitters on the same structure.

Figure 5. Co-localization of two emitters on the same structure using smMIET. (A) Experiment design. (B1) Representative single-molecule intensity time traces with automatically detected bleaching steps: two emitters (blue), one emitter (red), and background (green). (B2) Normalized TCSPC histograms of the three sections of B1 in corresponding color. (C1) The fitted lifetimes were converted into height values and histogrammed. Alternatively, the height difference between the two emitters on each single structure was calculated and histogrammed (C2).


Page 8 of 15

Page 9 of 15 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

Nano Letters

Conclusions We reported the application of smMIET for co-localizing two single fluorescent emitters along the optical axis with nanometer accuracy. For this purpose, we used stepwise photobleaching to find the fluorescence lifetime values of each emitter on one structure, which allowed us to determine their individual heights from the surface and thus their mutual axial distance. The determined distance of (32 ± 11) nm is in excellent agreement with the design value of 32 nm. Our work demonstrates the potential of smMIET for intramolecular distance measurements on the nanometer scale, with a dynamic range of up to ~100 nm, which makes our method suitable for structural studies of large biomolecular complexes beyond the conventional FRET range of ~10 nm. Furthermore, one can tune the lifetime-to-height dependency to the required range by choosing a different metal or dielectric coating, for example indium tin oxide33. In principle, it is feasible to extend smMIET for axial co-localization of multiple emitters. However, with increasing number of emitters, using stepwise photobleaching for extracting individual lifetimes becomes increasingly inaccurate. In that case, stochastic blinking approaches such as dSTORM, PALM or alternatively DNA PAINT will be a much better choice, but will require dyes that show both suitable photo-switching and stable, environment-independent fluorescence lifetimes. Unfortunately, the dye Atto647N used in the present study (chosen for its exceptional photo-stability and resulting high photon yield) cannot be easily used for dSTORM, and we are currently screening for better dye alternatives. Finally, we hope that our present study is a first step towards nanometer-accuracy threedimensional localization/co-localization of fluorescent emitters, by combining the superior axial localization accuracy of smMIET with the lateral localization accuracy of dSTORM/PALM/DNA PAINT in a combined wide-field/fluorescence-lifetime imaging setup.

Methods DNA origami DNA origami structures were prepared following the procedures described previously25. The yield and completeness of DNA origami structures were validated using AFM imaging (Figure S7, S8 in Supporting Information). For our particular structures (one or two labels, biotin or thiol surface immobilization), simple staples were replaced with labeled staples (special modified sequences). The final sample concentration was 1-2 nM. Before immobilization, stock samples were diluted 30-50 times in imaging buffer (Tris 10 mM, EDTA 1 mM, MgCl2 12.5 mM, Trolox 2 mM) to achieve sufficiently low final surface concentration (50-100 structures in region of interest with dimensions of 20 µm × 20 µm). Additionally, the sample was gently washed three times with imaging buffer before sealing to remove structures that did not attach to the surface. Experimental setup Measurements were performed using a commercial single-molecule fluorescence microscope (MicroTime 200, PicoQuant): A 640 nm Laser (LDH-D-C-640, PicoQuant), equipped with a clean-


Nano Letters 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

up filter (Z640/10, Chroma Technology), was used to excite the fluorophores (laser power 510 µW before objective). An achromatic λ/4 wave plate was used to generate circularly polarized light, which was then coupled into a polarization-maintaining single mode optical fiber. At the fiber output, the light was collimated and reflected by a dichroic mirror (FITC/TRITC Chroma Technology) into the side port of an inverted microscope (IX71, Olympus). The laser light was focused by an objective (UAPON 100x oil, 1.49 NA, Olympus) which also collected the fluorescence (epi-fluorescence setup). The collected fluorescence light was focused through a 150 μm pinhole and then onto a single-photon counting module (SPCM-CD 3516 H, Excelitas Technologies). A long pass filter (F76-649, AHF) and an emission bandpass filter (BrightLine HC 692/40, Semrock) were used to block scattered light and gold/silver photoluminescence. Laser pulsing with 20 MHz repetition rate was done electrically using a dedicated laser driver (PDL 828 “Sepia II”, PicoQuant). Photon detection times were recorded by a TCSPC electronics (HydraHarp 400, PicoQuant) with a resolution of 32 ps. Samples were placed on a piezo stage (P-562.3CD with controller E-710.3CD, Physik Instrumente), and images were acquired using the SymphoTime software (PicoQuant). An area of 20 µm × 20 µm was scanned, and individual emitters were localized in this area. Then, the identified emitters were addressed individually, and fluorescence lifetime measurements were performed until photobleaching. Between these individual measurements, focus quality was checked by a back reflection camera and readjusted if necessary. Surface immobilization protocol For immobilizing DNA origami pillars on the surface, we used thiol-gold chemistry. The bottom part of the origami structure had eight strands modified with thiols (see Supplementary Information for exact thiol positions) that form strong bonds with a gold-coated surface and ensure vertical orientation of the structure. Sealed samples were incubated with imaging buffer containing low concentrations of origami and 100 mM MgCl2 for 30 minutes, and then were gently rinsed with imaging buffer and sealed with a coverslip. Samples were placed on the measurement setup for at least two hours prior to measurement to ensure thermal equilibration and to decrease sample drift. For biotin-neutravidin immobilization, gold-covered coverslips were coated with an additional layer of 10 nm SiO2 (Figure 2A). Prior to the experiment, the coverslips were rinsed with isopropanol, and dried with nitrogen flow. To avoid sample evaporation during the experiment, a silicon sheet (GBL664384, Sigma-Aldrich) with a round hole was applied to the coverslips and, after finishing sample preparation, sealed with a coverslip on top. In a next step, the coverslips were incubated with 80 µL of BSA-biotin (0.5 mg/mL, A7641, Sigma-Aldrich) for 60 min, washed at least three times with 100 µL of buffer A (Tris 10 mM, EDTA 1 mM, NaCl 100 mM), then with 80 µL of neutravidin (0.5 mg/mL, 31000, ThermoFisher) for 5 minutes, and finally washed three times with buffer A. Then, TE buffer with 12.5 mM MgCl2 including origami structures diluted 3050 fold were introduced and incubated for 60 min. Finally, the sample was rinsed three times with imaging buffer (Tris 10 mM, EDTA 1 mM, MgCl2 12.5 mM, Trolox 2 mM) and sealed. Metal-coating of coverslips


Page 10 of 15

Page 11 of 15 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

Nano Letters

A semitransparent metallic film, consisting of 10 nm gold or silver on top of a 2 nm titanium film for glass adhesion, was deposited on a cleaned glass coverslip (24 mm × 24 mm, thickness 170 µm). Metal films were generated by chemical vapor deposition using an electron beam source (Univex 350, Leybold) under high-vacuum conditions (~10-6 mbar). During vapor deposition, film thickness was continuously monitored with an oscillating quartz unit. Coverslips with an additional SiO2 spacer were prepared with an additional layer of 1 nm titanium and then depositing a 10 nm SiO2 layer.

Data evaluation Fluorescence lifetime measurements were analyzed by a custom written MATLAB program. Firstly, intensity time traces with 20 ms bin width were calculated, and bleaching/blinking steps were identified. For this purpose, we use the “Step Transition and State Identification” algorithm from Shuang and coworkers34. Intensity time traces with too many or too few on-states (bleaching steps) were eliminated from further analysis. In a next step, TCSPC histograms were calculated for each individual state (constant intensity level in time trace), and histograms corresponding to the same intensity level were added (eliminating the effect of blinking events). The histogram corresponding to the lowest intensity level (above background) was fitted with a mono-exponential decay function. We used either tail-fitting with a cutoff of 0.5 ns after the maximum of the TCSPC histogram, or applied full deconvolution using an instrumental response function (IRF) that was extracted from background-only signal. In the latter case, a fast decay component was added to account for luminescence from the metal layer. We used a NelderMead Simplex algorithm for minimizing the negative log-likelihood function, with the fluorescence decay lifetimes, their amplitudes, and an IRF color shift as the fit parameters. For the next higher intensity level, two exponential decays were fitted, where one lifetime was fixed to the value obtained from the previous lifetime fit. If the χ2-value and/or the fitted lifetime was similar to the previously fixed lifetime, the histograms were added and fitted together with only one lifetime. This procedure helped to exclude additional levels due to fluctuations in the intensity. Additionally, we excluded all lifetimes from fits with a χ2-value larger than 2. In total, roughly 70% of the points were excluded from the final lifetime histograms. For converting lifetime into distance values, one calculates ab initio the distance-versus-lifetime function using classical Maxwell’s electrodynamics of dipole emission in front of a layered substrate. For doing this, one needs to know three properties of the fluorophore: Its free-space lifetime (far away from any surface), its quantum yield, and its emission wavelength. We measured the fluorescence lifetime of Atto647N on the origami pillars in solution and determined a free-space lifetime value of 4.5 ns. Assuming that coupling of the dye to DNA does only change the non-radiative rate of the excited-to-ground state transition, we estimated the quantum yield of the dye on the origami to be 80%. In the calculations, we set the emission wavelength of the dye to 688 nm, which is the mean emission wavelength of Atto647N within our emission filter window (692/40 nm bandpass). With these parameters, we obtained a value of 4.37 ns for the lifetime of the dye at the design height of 82 nm in the absence of metals, which is in good agreement with the measured value of (4.2 ± 0.2) ns for P1B on glass, see Figure 4 (A4). As mentioned previously, the shorter experimental lifetime could be due to non-


Nano Letters 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

vertical pillars, which contribute shorter lifetimes, thus shifting the average towards shorter values. Such pillars are dimmer in the smMIET experiment due to considerable dye quenching, as compared to reference measurements on glass. Values for the complex-valued refractive indices of gold, silver, and titanium were taken from literature35, and with these values and the exact knowledge of the sample geometry, distance-versus-lifetime curves (MIET curves) were calculated (see Figure S1 in Supplementary Information). For these calculations, we assumed a random orientation of the dye’s emission dipole. These MIET curves are then used to convert measured lifetime values into height values. Finally, we fitted one (Fig. 4) or two (Fig. 5) Gaussian functions with a constant offset to the height histogram (panel C in Fig. 4 and 5) to obtain the mean and standard deviation. The constant offset results probably from erroneous fits and non-vertical (“fallen”) pillars. ASSOCIATED CONTENT Supporting Information Theoretical MIET curves, 3D DNA PAINT measurement, simulations for data analysis verification, DNA origami pillar design and sequences, AFM images. AUTHOR INFORMATION Corresponding Authors *R.T.: E-mail: [email protected], Phone: +49 551 39 22297 *J.E. E-mail: [email protected], Phone: +49 551 39 13833 Notes The authors declare no competing financial interest. Acknowledgment This work was supported by the DFG Cluster of Excellence 'Center for Nanoscale Microscopy and Molecular Physiology of the Brain (CNMPB)'. SI, DR, and NK are grateful to the DFG for financial support via project A14 and A05 of the SFB 937, and project A06 of the SFB 860, respectively. RT was supported by a postdoctoral fellowship of the Minerva Foundation in 2016/2017 for a 19 month research stay at Georg August University, Göttingen. IK is grateful for the support by the Mobility Plus grant 1269/MOB/IV/2015/0 from the Polish Ministry of Science and Higher Education (MNiSW). PT acknowledges the excellence cluster NIM (Nanosystems Initiative Munich), and the DFG (TI 329/9- 1, AC 279/2-1, INST 188/401-1 FUGG). MR was funded by the Braunschweig International Graduate School of Metrology B-IGSM and the DFG Research Training Group GrK1952/1 “Metrology for Complex Nanosystems”. JB was supported by the Cluster of Excellence Nanosystems Initiative Munich (NIM) of LMU Munich. The authors are grateful to Guillermo Acuna and Arindam Ghosh for fruitful discussions. Amna Abdalla Mohammed Khalid performed AFM imaging for surface roughness characterization. References (1) Hell, S. W.; Wichmann, J., Optics letters 1994, 19 (11), 780-782. (2) Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.; Hess, H. F., Science 2006, 313 (5793), 1642-1645. (3) Huang, B.; Bates, M.; Zhuang, X., Annual review of biochemistry 2009, 78, 993-1016.


Page 12 of 15

Page 13 of 15 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

Nano Letters

(4) Van De Linde, S.; Löschberger, A.; Klein, T.; Heidbreder, M.; Wolter, S.; Heilemann, M.; Sauer, M., Nature protocols 2011, 6 (7), 991-1009. (5) Dertinger, T.; Colyer, R.; Iyer, G.; Weiss, S.; Enderlein, J., Proceedings of the National Academy of Sciences 2009, 106 (52), 22287-22292. (6) Jungmann, R.; Steinhauer, C.; Scheible, M.; Kuzyk, A.; Tinnefeld, P.; Simmel, F. C., Nano Letters 2010, 10 (11), 4756-4761. (7) Sahl, S. J.; Hell, S. W.; Jakobs, S., Nature Reviews Molecular Cell Biology 2017. (8) Willig, K.; Harke, B.; Medda, R.; Hell, S. W., Nature methods 2007, 4 (11), 915-918. (9) Juette, M. F.; Gould, T. J.; Lessard, M. D.; Mlodzianoski, M. J.; Nagpure, B. S.; Bennett, B. T.; Hess, S. T.; Bewersdorf, J., Nature methods 2008, 5 (6), 527-529. (10) Huang, B.; Wang, W.; Bates, M.; Zhuang, X., Science 2008, 319 (5864), 810-813. (11) Backlund, M. P.; Lew, M. D.; Backer, A. S.; Sahl, S. J.; Grover, G.; Agrawal, A.; Piestun, R.; Moerner, W., Proceedings of the National Academy of Sciences 2012, 109 (47), 19087-19092. (12) Ruckstuhl, T.; Verdes, D., Optics express 2004, 12 (18), 4246-4254. (13) Olveczky, B. P.; Periasamy, N.; Verkman, A., Biophysical Journal 1997, 73 (5), 2836-2847. (14) Hell, S. W.; Schmidt, R.; Egner, A., Nature Photonics 2009, 3 (7), 381-387. (15) Shtengel, G.; Galbraith, J. A.; Galbraith, C. G.; Lippincott-Schwartz, J.; Gillette, J. M.; Manley, S.; Sougrat, R.; Waterman, C. M.; Kanchanawong, P.; Davidson, M. W.; Fetter, R. D.; Hess, H. F., Proceedings of the National Academy of Sciences of the United States of America 2009, 106 (9), 3125-30. (16) Clegg, R. M., Laboratory techniques in biochemistry and molecular biology 2009, 33, 1-57. (17) Lee, N. K.; Kapanidis, A. N.; Wang, Y.; Michalet, X.; Mukhopadhyay, J.; Ebright, R. H.; Weiss, S., Biophysical Journal 2005, 88 (4), 2939-2953. (18) Sisamakis, E.; Valeri, A.; Kalinin, S.; Rothwell, P. J.; Seidel, C. A. M., Walter, N. G., Ed. Academic Press: 2010; Vol. 475, pp 455-514. (19) Di Fiori, N.; Meller, A., Biophysical Journal 2010, 98 (10), 2265-2272. (20) Steinhauer, C.; Jungmann, R.; Sobey, T. L.; Simmel, F. C.; Tinnefeld, P., Angewandte Chemie International Edition 2009, 48 (47), 8870-8873. (21) Karedla, N., Springer: 2017. (22) Chizhik, A. I.; Rother, J.; Gregor, I.; Janshoff, A.; Enderlein, J., Nat Photon 2014, 8 (2), 124127. (23) Karedla, N.; Chizhik, A. I.; Gregor, I.; Chizhik, A. M.; Schulz, O.; Enderlein, J., ChemPhysChem 2014, 15 (4), 705-711. (24) Chizhik, A. M.; Ruhlandt, D.; Pfaff, J.; Karedla, N.; Chizhik, A. I.; Gregor, I.; Kehlenbach, R. H.; Enderlein, J., ACS Nano 2017. (25) Puchkova, A.; Vietz, C.; Pibiri, E.; Wünsch, B.; Sanz Paz, M.; Acuna, G. P.; Tinnefeld, P., Nano Letters 2015, 15 (12), 8354-8359. (26) Raab, M.; Jusuk, I.; Molle, J.; Buhr, E.; Bodermann, B.; Bergmann, D.; Bosse, H.; Tinnefeld, P., Scientific reports 2018, 8 (1), 1780. (27) Ha, T.; Tinnefeld, P., Annual Review of Physical Chemistry 2012, 63 (1), 595-617. (28) Gordon, M. P.; Ha, T.; Selvin, P. R., Proceedings of the National Academy of Sciences of the United States of America 2004, 101 (17), 6462. (29) Schmied, J. J.; Forthmann, C.; Pibiri, E.; Lalkens, B.; Nickels, P.; Liedl, T.; Tinnefeld, P., Nano Letters 2013, 13 (2), 781-785. (30) Schmied, J. J.; Raab, M.; Forthmann, C.; Pibiri, E.; Wünsch, B.; Dammeyer, T.; Tinnefeld, P., Nature protocols 2014, 9, 1367. (31) Iinuma, R.; Ke, Y.; Jungmann, R.; Schlichthaerle, T.; Woehrstein, J. B.; Yin, P., Science 2014, 344 (6179), 65.


Nano Letters 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

(32) Jungmann, R.; Avendaño, M. S.; Woehrstein, J. B.; Dai, M.; Shih, W. M.; Yin, P., Nature methods 2014, 11, 313. (33) Moerland, R. J.; Hoogenboom, J. P., Optica 2016, 3 (2), 112-117. (34) Shuang, B.; Cooper, D.; Taylor, J. N.; Kisley, L.; Chen, J.; Wang, W.; Li, C. B.; Komatsuzaki, T.; Landes, C. F., The journal of physical chemistry letters 2014, 5 (18), 3157-3161. (35) Rakić, A. D.; Djurišić, A. B.; Elazar, J. M.; Majewski, M. L., Appl. Opt. 1998, 37 (22), 52715283.


Page 14 of 15

Page 15 of 15 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

Nano Letters

TOC Figure


Axial Colocalization of Single Molecules with Nanometer Accuracy

Recommend Documents