pubs.acs.org/NanoLett
Intrinsically Resolution Enhancing Probes for Confocal Microscopy Jan Vogelsang, Thorben Cordes, Carsten Forthmann, Christian Steinhauer, and Philip Tinnefeld* Angewandte Physik-Biophysik, and Center for NanoScience, Ludwig-Maximilians-Universita¨t, Amalienstrasse 54, 80799 Munich, Germany ABSTRACT In recent years different implementations of superresolution microscopy based on targeted switching (STED, GSD, and SSIM) have been demonstrated. The key elements to break the diffraction barrier are two distinct molecular states that generate a saturable nonlinear fluorescence response with respect to the excitation intensity. In this paper, we demonstrate that a nonlinearity can even be encoded in fluorescent probes, which then increase the resolution of a standard confocal microscope. This nonlinearity is achieved by an intensity dependent blocking of the resonance energy transfer between a donor and one or more acceptor fluorophores, utilizing radical anion states of the acceptor. In proof-of-principle experiments, we demonstrate a significant resolution increase using probes with different numbers of acceptor fluorophores. Quantitative description by a theoretical model paves the way for the development of fluorescent probes that can more than double the resolution of essentially any confocal microscope in all three dimensions. KEYWORDS Superresolution microscopy, FRET, single-molecule spectroscopy, excited-state saturation, nonlinear fluorescence
T
he vision to overcome the diffraction limit with farfield optics has become reality.1-4 It has been recognized that in all superresolution fluorescence microscopy techniques two distinct states of a fluorophore, a bright and a dark state, are the key to superresolution.5 Based thereon, two different experimental approaches have been realized, one using “stochastic switching” of the molecules between the bright and the dark state and the other one using a “targeted switching” scheme. Whereas subsequent localization of fluorophores represents an exciting “stochastic readout” approach that requires the possibility to prepare the largest fraction of fluorophores in the field of view in a dark state, the “targeted readout” scheme utilizes the two distinct states to generate molecules in the bright and dark state at predefined positions. A preferential category for the targeted switching is described by the RESOLFT concept (reversible saturable optical fluorescence transitions) where fluorophores are optically driven between two states to spatially select whether fluorescence is allowed or prohibited.6-8 The saturable transitions yield an exponential law enabling highly nonlinear fluorescence response with respect to the excitation intensity.7 An nth order and “positively bent” nonlinearity (second deviation >0, from now on referred to as superlinearity) can, however, already be achieved with adequate fluorescent probes in a conventional confocal microscope and also leads to superresolution.9,10 In these probes, the fluorescence increases steeper than linear at increasing excitation energy.
Generally, superresolution methods either require elaborate modification of the excitation scheme (as in case of, e.g., STED, SSIM, SPEM, and GSD)11-14 and/or computational post processing to reconstruct images (PALM, STORM, FPALM, dSTORM, GSDIM, and Blink-Microscopy).15-20 It is also noteworthy that increasing resolution is commonly accompanied by increasing demands on the fluorescent probes, which have to be improved with respect to functionality (e.g., switchability, labeling procedures, physical size, and so forth) and photostability.21,22 Fluorescent dyes and probes are thus moving into the center of fluorescence microscopy development. In this paper, we present fluorescent probes that show high photostability compared to standard fluorophores,22 and intrinsically exhibit a superlinear fluorescence response that directly translates into resolution enhancement in confocal microscopy in all three dimensions. These probes are called “energy transfer blockade probes” (ETBPs) and consist of a donor fluorophore surrounded by one to n acceptors (with n ) 0, 1, 2, ...), which are not necessarily fluorophores. Upon excitation, the donor efficiently transfers its excited state energy to the acceptors via fluorescence resonance energy transfer (FRET). To achieve a superlinear fluorescence response of the donor, the acceptor has to show a photoinduced transition to a state that does not act as an acceptor for FRET. The resulting superlinearity is related to the fact that the observed donor emission depends on two different processes: the excitation rate of the donor increases linearly with higher excitation intensity. Additionally, the probability that the donor is not quenched by the acceptor(s) also increases due to decreasing ON-times of the acceptor(s) at higher excitation intensity. In other words, the saturation
* To whom correspondence should be addressed. Tel.: +49 892180 1438. Fax: +49 892180 2050. E-Mail:
[email protected]. Received for review: 11/15/2009 Published on Web: 01/08/2010 © 2010 American Chemical Society
672
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
of the acceptor(s) is converted into superlinear fluorescence of the donor. For the realization of superlinear fluorescent probes, it has been argued that saturation of the singlet state could block the energy transfer to the acceptor.9,23 Since one photon is required for excitation of the acceptor (e.g., through FRET), and a second photon for excitation of the donor, which then fluoresces because it cannot transfer its energy to the acceptor anymore, the fluorescence intensity of the donor depends on the square of the excitation energy for one acceptor. The dependence is similar to the one in multiphoton and multiexciton excitation with the difference of another physical concept and that the excitation wavelength remains the wavelength of one-photon excitation.24,25 The nonlinearity then scales with the number of acceptors, that is, an (nth + 1) order nonlinearity is obtained.9,23 A true “exciton blockade” or “frustrated energy transfer” has been theoretically investigated by the group of Stefan Hell,6 but to the best of our knowledge has not been successfully realized despite several attempts.26,27 This is mainly due to the fact that FRET occurs not only to acceptors in the ground state but also to acceptors in excited states.28-30 Since the spin of the acceptor is not changing, energy transfer to all these states is Fo¨rster allowed and is referred to as singletsinglet annihilation and singlet-triplet annihilation, respectively. These annihilation processes have, for example, been studied on the ensemble and single-molecule level for perylene imide derivatives that exhibit energy hopping, singlet-singlet annihilation as well as singlet-triplet annihilation.28,29 As expected from transient singlet-singlet absorption and triplet-triplet absorption spectra of many common fluorophores such as rhodamine- and cyaninederivatives that exhibit significant extinction in the visible spectral region, annihilation processes also occur for combinations of fluorophores commonly used in confocal microscopy and cannot completely be avoided.31,32 On the other hand, it is known that in contrast to perylene imides,29 radicalanionstatesofmanyfluorophoressuchasrhodamines, oxazines, and cyanines do not exhibit extinction above ∼500 nm (see, for example, refs 33-35). In our new experimental approach, we realize an energy transfer blockade using radical anion states as saturable dark states that are not in resonance with donor emission in ETBPs. We use single ETBPs to demonstrate subdiffraction resolution in a confocal microscope. A quantitative theoretical description finally highlights the parameters that have to be optimized to fully exploit the potential of the approach. Experimental Realization of the ETBP. An ETBP was constructed from dsDNA as rigid scaffold labeled with ATTO647N as donor and zero, one or two ATTO680 as acceptor(s) at a distance of 12 basepairs for a 0-, 1- or 2-acceptor ETBP, respectively (see Figure 1). The 12 basepair distance was used to prevent non-FRET interchromophore interactions. © 2010 American Chemical Society
FIGURE 1. Scheme of the energy transfer blockade probes (ETBPs). 0-acceptor ETBP (a), 1-acceptor ETBP (b), and 2-acceptor ETBP (c).
In the ETBPs, the donor should emit stably and the acceptor(s) has to exhibit excitation dependent blinking to radical anion states. The strategy for stable donor emission with a blinking acceptor is based on the recently achieved understanding and control of single-molecule photophysics using a reducing and oxidizing system (ROXS).22 In the presence of both reductant and oxidant, potentially longlived intermediate states such as triplet and radical ion states are rapidly depopulated via electron transfer reactions yielding very stable emission without observable blinking for most dyes such as Cy3B and ATTO647N (see Figure 2a). This means that these dyes are continuously in an ON-state despite frequent but very brief transitions to triplet and radical ion states (Figure 2a).22 Depending on the redox potentials of fluorophore and redox agents, however, radical ion states can have significantly longer lifetimes even in the presence of reductant and oxidant. The radical anion state of oxazines, for example, is so stable that it is not efficiently depopulated by the oxidants methylviologen (MV) or oxygen. Consequently, transitions to the radical anion state yield an oxidant-concentration dependent OFF-time of 1-106 ms.36 This OFF-state is directly visualized in single-molecule transients of oxazines such as ATTO680 in the presence of ascorbic acid (AA) and MV (Figure 2b). To check whether the acceptor in the OFF-state still acts as FRET acceptor or whether the FRET is blocked we used the 1-acceptor ETBP (see Figure 1b). DNAs were immobilized via a biotin/streptavidin interaction to BSA/BSA-biotin coated cover slides (see Materials and Methods in Supporting Information for details on sample preparation and experimental setup). Donor and acceptor emission were separated using an appropriate beamsplitter and the fluorescence was detected by two avalanche photodiodes in a confocal singlemolecule setup.36 Fluorescence transients of single FRET pairs were recorded exciting at 640 nm in the presence of AA (250 µM) and MV (100 µM). Oxygen was removed 673
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
FIGURE 2. Scheme of the photophysics in a 1-acceptor ETBP. (a,b) Typical fluorescence transients of immobilized dsDNA labeled with ATTO647N (a) or ATTO680 (b) using AA (250 µM) and MV (100 µM) at an excitation intensity of 1.7 kW/cm2 for ATTO647N and 3.5 kW/cm2 for ATTO680 at λExc ) 640 nm. (c) Representative fluorescence transient of an immobilized 1-acceptor ETBP bearing ATTO647N and ATTO680 using the same buffer and excitation conditions as described for ATTO647N (a). The fluorescence was split by a dichroic beamsplitter (BS685) so that the green transient primarily represents ATTO647N emission and the red transient ATTO680 emission. The black transient in (c) corresponds to the proximity ratio (PR ) FDA/(FDD + FDA)) demonstrating the fluctuating FRET. The red/green bar indicates switching between the two states of the acceptor. In the ON-state the acceptor can accept the energy from the donor (d), while in the OFF-state, FRET is blocked and the donor fluoresces (e). The lifetime of the acceptor ON-state is controlled by the reductant concentration and excitation intensity while its OFF-state is controlled by the oxidant concentration (d,e).
sentially identical evidencing 100% donor dequenching when the acceptor is OFF.
enzymatically via a glucose oxidase/catalase system (see Supporting Information for details).37 When excited via FRET, the acceptor shows similar blinking as with direct excitation (compare Figure 2b, red transient, and Figure 2c, red transient). When the acceptor is ON, FRET occurs and the donor is efficiently quenched (Figure 2d). Every time the acceptor enters an OFF-state the fluorescence of the donor recovers indicating that in contrast to other approaches and especially in contrast to blinking due to triplet states,32 the donor emission is not in resonance with the acceptor in the OFF-state. The anticorrelation of the green and red transient is also evident from the transient of the proximity ratio (PR) (Figure 2c, black transient), which is the ratio of acceptor fluorescence to the sum of donor and acceptor fluorescence. Obviously, in the radical anion state of ATTO680, the FRET is blocked and ATTO647N emission occurs (Figure 2e). For the quality of an ETBP, it is important that quenching by the acceptor radical anion is absolutely negligible. We therefore inspected transients, in which the acceptor bleached before the donor and compared the intensity of the donor after acceptor photobleaching with the intensity of the donor when the acceptor is in its radical anion state. As shown in Supporting Information Figure S2 these intensities are es© 2010 American Chemical Society
The next step toward resolution enhancement with an ETBP is to validate the superlinear fluorescence response of the donor. Therefore, the process leading to the acceptor radical anion has to be photoinduced resulting in a superlinear fluorescence dependence of the donor with increasing excitation intensity. We recorded single-molecule PR transients and decreased the excitation intensity every 10 s during the acquisition (from 1.9 to 0.34 kW/cm2). Subsequently, 10 s periods were analyzed with respect to their PR. A fraction of a transient recorded at an average excitation intensity of 1.7 kW/cm2 is shown in Figure 3a together with the corresponding PR histogram. The ratio of the number of ON- and OFF-state bins in the transient, #on to #off, obtained from fitting the histogram by a two-peak Gaussian function, directly reports on the excitation dependence of the ON-time since τOFF is known to be intensity independent.20,36 Additional control experiments were performed to exclude any changes of the #on/#off ratio over time. Therefore the #on/#off ratio from two intensity ramps on the same molecule were compared, yielding unchanged blinking kinetics (see Supporting Information Figure S3). The 674
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
This proportionality neglects parameters such as detection efficiency, quantum yield and absorption cross section that are not necessary for the quantification of resolution. Here, Ci corresponds to the contrast of the donor fluorescence. The contrast describes the ratio of each intensity level FD,i of the donor with i acceptors being ON with respect to the intensity level FD,0 of the donor with 0 acceptors being ON (Ci ) (FD,0 - FD,i)/FD,0). The contrast is influenced by the FRET efficiency, the background intensity in the donor channel and possible crosstalk from acceptor fluorescence into the donor channel. The value Ri describes the ratio of ON- to OFF- times at a certain excitation intensity and is defined as:
αi )
I0τon,i(I0) τoff
(2)
FIGURE 3. Demonstration of the superlinear fluorescence behavior of a 1-acceptor ETBP. (a) Proximity ratio transient of an ETBP (250 µM AA, 1 mM MV, and oxygen removal) with 10 ms binning excited at λExc ) 640 nm with 1.7 kW/cm2. The right panel shows a histogram of the PR-values. The histogram is additionally fitted with a two peak Gaussian function. (b) Plot of the ratio #on/#off versus excitation intensity (black dots) of one ETBP fitted with an allometric function f(x) ) a·xb (black curve) and plot of the superlinear donor fluorescence versus excitation intensity (gray dots) of the same molecule fitted using eq 1 (see below). The fit is based on a 1-acceptor ETBP plus an offset with the experimentally obtained values C1 ) 0.71 and R1 ) 1.8 kW/cm2 (gray, see below for definitions of C1 and R1). For each data point the average intensity over 10 s was plotted against the excitation intensity.
In this equation τon,i(I0) corresponds to the average ONtime the acceptors spend in the ON-state at an excitation intensity I0 with i acceptors actually being in the ON-state. τoff denotes the average OFF-time of the acceptors. Ci and Ri can be written in terms of a 1-acceptor ETBP, that is, C1 and R1 as follows
procedure was repeated for different excitation intensities yielding the excitation power dependence depicted in Figure 3b (black dots). Averaging over twenty molecules and fitting by an allometric function f(x) ) a·xb (Figure 3b, black curve), yielded a value of b ) -0.99 ( 0.38 and proved the anti proportional dependence of ON-time on excitation intensity. This decrease of the acceptor ON-times is accompanied by a superlinear dependence of the donor fluorescence on excitation intensity plotted in Figure 3b (gray dots) that is fitted using eq 1 (see below). Theoretical Treatment of the Superlinear Fluorescence. In the following the superlinear fluorescence dependence of the donor on excitation intensity is quantified and the consequences for the optical resolution are discussed. With the assumption, that the energy transfer rate is the same to all acceptors within an n-acceptor ETBP, the donor fluorescence FD is given by (see Supporting Information for a detailed derivation)
and
∑ (1 - Ci)(ni ) 1 + αI i n
FD(I) ∝ I
i)0
(
© 2010 American Chemical Society
)( -i
1+
αi+1 I
)
(
Ci ) 1 - 1 + i
αi )
(
C1 1 - C1
)
-1
C1 + dxAD α1 -1 1 - C1 A +i + dxD C1
)
(3)
(4)
where dxDA denotes the direct excitation of the acceptor at the donor excitation wavelength, which has to be taken into account since it alters the blinking kinetics of the acceptors. For the FRET-pair ATTO647N and ATTO680 excited at 640 nm direct excitation was determined to dxDA ) 0.44. C1 and R1 are determined from transients such as in Figure 3a and Supporting Information Figure S2a. A histogram of the donor intensity exhibits two populations resulting from the blinking of the acceptor (Supporting Information Figure S2b, black bars). Fitting the histogram with a twopeak Gaussian function yields the mean values for FD,0 and FD,1 that allow straightforward calculation of C1 (see Supporting Information eq S2). From 40 ETBP molecules we found C1 ) 0.71 ( 0.05. Since the ratio #on/#off is equal to the ratio τon,1(I0)/τoff, eq 2 can be rewritten to
i-n
(1)
675
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
τon,1(I0) α1 #on ) ) #off τoff I0
(5)
To determine R1, the ratio #on/#off was analyzed for a 1-acceptor ETBP at different excitation intensities I0. Fitting the plot of #on/#off versus I0 with eq 5 yields R1 (see Figure 3b, black dots). Since R1 is controllable by the concentrations of reductant and oxidant, we chose 250 µM AA and 1 mM MV for comparable fast blinking kinetics (τon,1(I0 ) 1.5 kW/ cm2) ) 16 ( 5 ms, τoff ) 10 ( 2 ms) and high photostability.36 Under these conditions we determined R1 averaged over 20 molecules to R1 ) 1.8 ( 0.8 kW/cm2. To visualize the expected superlinear fluorescence behavior of the donor eq 1 was plotted for a 1-acceptor and 2-acceptor ETBP, respectively, for different values of C1 (see Supporting Information Figure S4). Theoretical Evaluation of Achievable Resolution with ETBPs. The point-spread function (PSF) of a confocal microscope is a measure for its resolution and is given by the product of the excitation volume and the detection volume. In our experiments, the size of the detection volume is chosen to be several fold larger compared to the excitation volume so that the signal-to-noise ratio is maximized and the theoretical treatment is simplified since only the effective width of the excitation volume WExc,Eff has to be taken into account. Approximating the excitation intensity profile with a Gaussian function allows estimating the width of the excitation volume to WExc ∼ 230 nm at λExc ) 640 nm, a numerical aperture of the objective of NA ) 1.40 and a perfect illumination of the back aperture. Because of the superlinear fluorescence of the ETBP, the detected PSF is altered with respect to the excitation volume. To calculate the effective width of the excitation volume WExc,Eff, the Gaussian intensity profile with a width of 230 nm is inserted into eq 1 and the resulting fluorescence profile is fitted by a Gaussian function. This procedure allows estimating WExc,Eff of the ETBP as well as the influence of the excitation intensity. Figure 4a shows the dependence of WExc,Eff for a 1-acceptor ETBP on the excitation intensity and varying values of R1 and C1. The calculations were performed for different realistic values of R1 and C1, recalling that R1 was determined by the ROXS controlled blinking kinetics and C1 is strongly related to the FRET efficiency. Interestingly, a minimum of the PSF exists for each condition at a specific excitation intensity. For a constant value of R1, for example, the minimum WExc,Eff is strongly decreasing with increasing C1 (Figure 4a, light gray) indicating that the FRET efficiency is a crucial parameter for resolution enhancement with ETBPs. In contrast, R1 has no influence on the obtainable resolution but determines the optimal excitation intensity (Figure 4a, dark gray). Accordingly, direct excitation of the acceptor dxDA has no consequence on the obtainable resolution improvement, since dxDA only influences Ri and therefore the optimal excitation intensity (see eq 4). This shows © 2010 American Chemical Society
FIGURE 4. Effective width of the excitation volume WExc,Eff dependence with respect to the excitation intensity is shown. (a) WExc,Eff shown for a 1-acceptor ETBP at different values for C1 and R1 and direct excitation dxDA ) 0.44. (b) WExc,Eff shown for different numbers of acceptor fluorophores in an ETBP at constant - experimentally obtained - values for C1 and R1 (C1 ) 0.71; R1 ) 1.8 kW/cm2). (c) Best obtainable WExc,Eff with respect to the number of acceptors at two different values for C1 (gray dots, C1 ) 0.71; black dots C1 ) 0.999).
that the ETBP can be adjusted to reasonable values of excitation intensity and also of acquisition time by adapting ROXS (or by using a different acceptor with different redox potential).22,36,38 Another important factor for the resolution is the number of acceptors whose influence on WExc,Eff is plotted in Figure 4b for the C1 and R1 value of the ETBP used (C1 ) 0.71 and R1 ) 1.8 kW/cm2). More acceptors increase the superlinearity of the donor fluorescence and thus increase the achievable resolution to below 180 nm for n > 3. Figure 4b also demonstrates that the minimum WExc,Eff is achieved at higher excitation intensity for increasing n. Since the FRET efficiency strongly influences the achievable resolution and because FRET efficiencies close to unity have already been demonstrated on the level of single molecules (see, e.g., ref 676
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
39), we evaluate the realistic potential of ETBPs by plotting the best achievable WExc,Eff versus the number of acceptors for the contrast of C1 ) 0.71 used in this work (gray dots in Figure 4c) and a maximum C1 of C1 ) 0.999 (black dots in Figure 4c). Accordingly, a WExc,Eff below 100 nm can be achieved with probes bearing 10 acceptors. From signal-tonoise considerations it is important to note that the fluorescence of the donor in a 10-acceptor ETBP (C1 ) 0.71) is reduced by 73% compared to a donor without acceptors at the optimal excitation intensity. Experimental Demonstration of Intrinsic Subdiffraction Resolution. For an experimental verification of the intrinsic resolution enhancement by ETBPs we used constructs with zero, one, and two acceptor(s) (Figure 1). On the basis of the theoretical treatment (see Figure 4a), we adapted R1 to an optimal excitation intensity of ∼1.7 kW/ cm2 for a 1-acceptor ETBP and ∼2.4 kW/cm2 for a 2-acceptor ETBP using ROXS (250 µM AA, 1 mM MV). Confocal scans of the surface revealed the positions of individual ETBPs such as the one in the inset of Figure 5a. The false color image indicates the higher FRET value at the rim and the lower FRET value in the center that is due to the described saturation of the acceptor and superlinearity of the donor. To obtain statistics on PSFs we scanned the centers of each spot several times over a range of 1.5 µm with a step size of 50 nm, and a time resolution of 1 s per step (as indicated in the spot of Figure 5a). This yields transients such as the one shown in Figure 5a (red ) acceptor emission, green ) donor emission). To determine the PSF-width of each scan, the average donorfluorescence for each scanning step was plotted against the position (Figure 5a, donor average, gray) and fitted with a Gaussian function (Figure 5a, Gaussian fit, black). Repeating the procedure for >200 single ETBPs yields histograms of PSF-widths displayed in Figure 5b. Gaussian fits to the PSF distributions show that experimentally ∼260 nm wide PSFs are obtained for the donor only molecules (Figure 5b, gray), ∼236 nm for the 1-acceptor ETBP (Figure 5b, blue), and ∼224 nm for the 2-acceptor ETBP (Figure 5b, red). On the basis of the 260 nm measured for the donor-only sample, we calculated the expected resolution with the ETBPs using eqs 1-4 (see Supporting Information for details). The calculated values of 232 nm for the 1-acceptor ETBP and 219 nm for the 2-acceptor ETBP deviate from the experimental values only slightly, which we assign to minor inhomogeneities of photophysical parameters from ETBP to ETBP (the excitation intensity was optimized for the average molecule and slightly varies from molecule to molecule) and the not fully achieved photodynamic equilibrium in the outer parts of the spots. As pointed out, a key parameter for the achievable resolution is the contrast value C1 (see Figure 4c). To show the potential of ETBPs for resolution enhancement in confocal microscopes we emulate 100% FRET efficiency, that is, a C1-value of 1. This is possible because we studied single © 2010 American Chemical Society
FIGURE 5. Determination of the PSF-width for different ETBPs in a confocal microscope. (a) One-dimensional scan of a single immobilized 1-acceptor ETBP. The fluorescence of the donor (green) and acceptor (red) channel was recorded for 1 s per 50 nm step and displayed at 5 ms time binning. The black graph corresponds to a Gaussian fit over the average donor fluorescence intensity for each step. The upper left inset shows a complete two-dimensional scan of a 1-acceptor ETBP (50 nm/pixel). The white line illustrates the one-dimensional scan. (b) Normalized PSF-histograms obtained from Gaussian fits of the donor fluorescence for a 0- (gray), 1- (blue), and 2- (red) acceptor ETBP. (c) Corresponding normalized PSF-Histograms for an idealized C1-value of 1 for a 0- (gray), 1- (blue) and 2(red) acceptor ETBP. The values were obtained by omitting donor photons emitted when one of the acceptors was ON.
ETBPs whose blinking was adapted to the millisecond range. By omitting the donor photons emitted when the acceptor was ON, C1 ) 1 is emulated. Therefore, the fluorescence transients measured to determine the PSFs were binned in 5 ms steps and analyzed with respect to the PR. For bins with PR > 0.6, the photons of the donor were omitted yielding new PSF-widths histogrammed in Figure 5c with a mean of ∼221 nm for the 1-acceptor ETBP and ∼199 nm for the 2-acceptor ETBP (see Figure 5c). These values deviate 677
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
slightly more from the theoretical values of 210 and 185 nm because the binning and thresholding are not perfectly separating the states. The almost doubling of the resolution improvement emulating C1 ) 1, however, indicates the importance of a high FRET efficiency for the further development of intrinsically resolution enhancing probes for confocal microscopy. Overcoming the diffraction limit is a major focus in microscopy development with an enormous impact on biological research. Here we present the first realization of probes termed ETBPs that intrinsically yield improved resolution in a confocal microscope by exhibiting a superlinear fluorescence dependence on excitation intensity. We have demonstrated in proof-of-principle experiments a significant resolution improvement with values which are in good agreement with the theoretical description of the superlinear fluorescence of ETBPs. The presented theoretical description reveals the parameters to improve ETBPs on the way to resolution enhancing probes that could be used in any confocal microscope without technical alterations of the setup or computational post-treatment of the data. Thus, especially the FRET-efficiency has to be maximized and the number of acceptors has to be increased opening up the fascinating potential of sub-100 nm resolution just by developing fluorescent probes and not the microscope setup. The currently long acquisition times can be reduced by accelerating the blinking kinetics of the acceptor using a different redox system or adapting the redox properties of the fluorophores. Further resolution improvements can be achieved by synergistically combining ETBPs with other illumination schemes such as 4Pi or two-photon microscopy.24,40 The optimal ETBP will be constructed of a central donor fluorophore surrounded by a number of acceptor fluorophores. A dendritic structure could, for example serve as a scaffold for such a construct.41 While we only treated the case of parallel FRET pathways in this work, hierarchical FRET systems with additional cascade-FRET from the acceptor to another acceptor could also lead to a further improvement not being limited by steric requirements of one donor surrounded by many identical acceptors.42,43 Finally, it appears likely that such probes might also function in living cells since ATTO647N does not exhibit blinking in tracking experiments,44 while other dyes such as oxazines exhibit blinking in living cells (e.g., ATTO655, see Supporting Information Figure S5).19 Alternatively to superresolution, ETBPs could also aid in increasing signal-to-noise e.g. through modulated excitation and an optical lock-in detection scheme.45
Supporting Information Available. This material is available free of charge via the Internet at http://pubs.acs.org. REFERENCES AND NOTES (1) (2)
(3) (4)
(5) (6) (7) (8)
(9)
(10) (11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
Acknowledgment. The authors are grateful to Markus Sauer and Jo¨rg Enderlein for stimulating discussion. The project was supported by the Biophotonics III Programm of the BMBF/VDI (Grant 13N9234) and the Nanosystems Initiative Munich (NIM). C.S. is grateful to the Elite Network of Bavaria (IDK-NBT) for a doctoral fellowship. © 2010 American Chemical Society
(21)
(22)
678
Hell, S. W. Far-field optical nanoscopy. Science 2007, 316 (5828), 1153–1158. Fernandez-Suarez, M.; Ting, A. Y. Fluorescent probes for superresolution imaging in living cells. Nat. Rev. Mol. Cell Biol 2008, 9 (12), 929–43. Huang, B.; Bates, M.; Zhuang, X. Super-resolution fluorescence microscopy. Annu. Rev. Biochem. 2009, 78, 993–1016. Ji, N.; Shroff, H.; Zhong, H.; Betzig, E. Advances in the speed and resolution of light microscopy. Curr. Opin. Neurobiol. 2008, 18 (6), 605–16. Hell, S. W. Microscopy and its focal switch. Nat. Methods 2009, 6 (1), 24–32. Hell, S. W. Strategy for far-field optical imaging and writing without diffraction limit. Phys. Lett. A 2004, 326 (1-2), 140–145. Hell, S. W. Toward fluorescence nanoscopy. Nat. Biotechnol. 2003, 21 (11), 1347–1355. Hell, S. W.; Dyba, M.; Jakobs, S. Concepts for nanoscale resolution in fluorescence microscopy. Curr. Opin. Neurobiol. 2004, 14 (5), 599–609. Schonle, A.; Hanninen, P. E.; Hell, S. W. Nonlinear fluorescence through intermolecular energy transfer and resolution increase in fluorescence microscopy. Ann. Phys. 1999, 8 (2), 115–133. Heintzmann, R.; Gustafsson, M. G. L. Subdiffraction resolution in continuous samples. Nat. Photon. 2009, 3 (7), 362–364. Klar, T. A.; Jakobs, S.; Dyba, M.; Egner, A.; Hell, S. W. Fluorescence microscopy with diffraction resolution barrier broken by stimulated emission. Proc. Natl. Acad. Sci. U.S.A. 2000, 97 (15), 8206–8210. Gustafsson, M. G. L. Nonlinear structured-illumination microscopy: Wide-field fluorescence imaging with theoretically unlimited resolution. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (37), 13081– 13086. Heintzmann, R.; Jovin, T. M.; Cremer, C. Saturated patterned excitation microscopy - a concept for optical resolution improvement. J. Opt. Soc. Am. A 2002, 19 (8), 1599–1609. Bretschneider, S.; Eggeling, C.; Hell, S. W. Breaking the diffraction barrier in fluorescence microscopy by optical shelving. Phys. Rev. Lett. 2007, 98 (21), 218103. Betzig, E.; Patterson, G. H.; Sougrat, R.; Lindwasser, O. W.; Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; LippincottSchwartz, J.; Hess, H. F. Imaging intracellular fluorescent proteins at nanometer resolution. Science 2006, 313 (5793), 1642–1645. Rust, M. J.; Bates, M.; Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat Methods 2006, 3 (10), 793–5. Hess, S. T.; Girirajan, T. P.; Mason, M. D. Ultra-high resolution imaging by fluorescence photoactivation localization microscopy. Biophys. J. 2006, 91 (11), 4258–72. Heilemann, M.; van de Linde, S.; Schuttpelz, M.; Kasper, R.; Seefeldt, B.; Mukherjee, A.; Tinnefeld, P.; Sauer, M. SubdiffractionResolution Fluorescence Imaging with Conventional Fluorescent Probes. Angew. Chem., Int. Ed. 2008, 47 (33), 6172–6176. Folling, J.; Bossi, M.; Bock, H.; Medda, R.; Wurm, C. A.; Hein, B.; Jakobs, S.; Eggeling, C.; Hell, S. W. Fluorescence nanoscopy by ground-state depletion and single-molecule return. Nat. Methods 2008, 5 (11), 943–5. Steinhauer, C.; Forthmann, C.; Vogelsang, J.; Tinnefeld, P. Superresolution microscopy on the basis of engineered dark states. J. Am. Chem. Soc. 2008, 130 (50), 16840–16841. Rittweger, E.; Han, K. Y.; Irvine, S. E.; Eggeling, C.; Hell, S. W. STED microscopy reveals crystal colour centres with nanometric resolution. Nat. Photon. 2009, 3 (3), 144–147. Vogelsang, J.; Kasper, R.; Steinhauer, C.; Person, B.; Heilemann, M.; Sauer, M.; Tinnefeld, P. A reducing and oxidizing system minimizes photobleaching and blinking of fluorescent dyes. Angew. Chem., Int. Ed. 2008, 47, 5465–5469. DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
(23) Hanninen, P. E.; Lehtela, L.; Hell, S. W. Two- and multiphoton excitation of conjugate-dyes using a continuous wave laser. Opt. Commun. 1996, 130 (1-3), 29–33. (24) Denk, W.; Strickler, J. H.; Webb, W. W. Two-photon laser scanning fluorescence microscopy. Science (Washington, DC, U. S.) 1990, 248 (4951), 73–6. (25) Hennig, S.; van de Linde, S.; Heilemann, M.; Sauer, M. Quantum Dot Triexciton Imaging with Three-Dimensional Subdiffraction Resolution. Nano Lett. 2009, 9 (6), 2466–2470. (26) Berglund, A. J.; Doherty, A. C.; Mabuchi, H. Photon statistics and dynamics of fluorescence resonance energy transfer. Phys. Rev. Lett. 2002, 89 (6), 068101-1–068101-4. (27) Melnikov, S. M.; Yeow, E. K. L.; Uji-i, H.; Cotlet, M.; Mullen, K.; De Schryver, F. C.; Enderlein, J.; Hofkens, J. Origin of simultaneous donor-acceptor emission in single molecules of peryleneimide-terrylenediimide labeled polyphenylene dendrimers. J. Phys. Chem. B 2007, 111 (4), 708–719. (28) Hofkens, J.; Cotlet, M.; Vosch, T.; Tinnefeld, P.; Weston, K. D.; Ego, C.; Grimsdale, A.; Muellen, K.; Beljonne, D.; Bredas, J. L.; Jordens, S.; Schweitzer, G.; Sauer, M.; De Schryver, F. Revealing competitive Foerster-type resonance energy-transfer pathways in single bichromophoric molecules. Proc. Natl. Acad. Sci. U.S.A. 2003, 100 (23), 13146–13151. (29) Vosch, T.; Cotlet, M.; Hofkens, J.; Van Biest, K.; Lor, M.; Weston, K.; Tinnefeld, P.; Sauer, M.; Latterini, L.; Muellen, K.; De Schryver, F. C. Probing Foerster Type Energy Pathways in a First Generation Rigid Dendrimer Bearing Two Perylene Imide Chromophores. J. Phys. Chem. A 2003, 107 (36), 6920–6931. (30) Gopich, I.; Szabo, A. Theory of photon statistics in single-molecule Forster resonance energy transfer. J. Chem. Phys. 2005, 122 (1), No. 0147071-01470718. (31) Tinnefeld, P.; Buschmann, V.; Weston, K. D.; Biebricher, A.; Herten, D.-P.; Piestert, O.; Heinlein, T.; Heilemann, M.; Sauer, M. How single molecule photophysical studies complement ensemble methods for a better understanding of chromophores and chromophore interactions. Recent Res. Dev. Phys. Chem. 2004, 7 (Pt. 1), 95–125. (32) Tinnefeld, P.; Buschmann, V.; Weston, K. D.; Sauer, M. Direct observation of collective blinking and energy transfer in a bichromophoric system. J. Phys. Chem. A 2003, 107 (3), 323–327. (33) Kandori, H.; Kemnitz, K.; Yoshihara, K. Subpicosecond Transient Absorption Study of Intermolecular Electron-Transfer between Solute and Electron-Donating Solvents. J. Phys. Chem. 1992, 96 (20), 8042–8048. (34) Chibisov, A. K. Triplet-States of Cyanine Dyes and Reactions of Electron-Transfer with Their Participation. J. Photochem. 1977, 6 (3), 199–214.
© 2010 American Chemical Society
(35) Korobov, V. E.; Chibisov, A. K. Primary Processes in Photochemistry of Rhodamine Dyes. J. Photochem. 1978, 9 (5), 411–424. (36) Vogelsang, J.; Cordes, T.; Forthmann, C.; Steinhauer, C.; Tinnefeld, P. Controlling the fluorescence of ordinary oxazine dyes for single-molecule switching and superresolution microscopy. Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (20), 8107–8112. (37) Harada, Y.; Sakurada, K.; Aoki, T.; Thomas, D. D.; Yanagida, T. Mechanochemical Coupling in Actomyosin Energy Transduction Studied By Invitro Movement Assay. J. Mol. Biol. 1990, 216 (1), 49–68. (38) Cordes, T.; Vogelsang, J.; Tinnefeld, P. On the mechanism of trolox as antiblinking and antibleaching reagent. J. Am. Chem. Soc. 2009, 131 (14), 5018–5019. (39) Metivier, R.; Nolde, F.; Müllen, K.; Basche, T. Electronic Excitation Energy Transfer between Two Single Molecules Embedded in a Polymer Host. Phys. Rev. Lett. 2007, 98 (4), 47802. (40) Bewersdorf, J.; Schmidt, R.; Hell, S. W. Comparison of (IM)-M-5 and 4Pi-microscopy. J. Microsc. (Oxford, U.K.) 2006, 222, 105– 117. (41) Cotlet, M.; Gronheid, R.; Habuchi, S.; Stefan, A.; Barbafina, A.; Muellen, K.; Hofkens, J.; De Schryver, F. C. Intramolecular Directional Foerster Resonance Energy Transfer at the SingleMolecule Level in a Dendritic System. J. Am. Chem. Soc. 2003, 125 (44), 13609–13617. (42) Cotlet, M.; Vosch, T.; Habuchi, S.; Weil, T.; Muellen, K.; Hofkens, J.; De Schryver, F. Probing Intramolecular Foerster Resonance Energy Transfer in a Naphthaleneimide-Peryleneimide-Terrylenediimide-Based Dendrimer by Ensemble and Single-Molecule Fluorescence Spectroscopy. J. Am. Chem. Soc. 2005, 127 (27), 9760–9768. (43) Heilemann, M.; Kasper, R.; Tinnefeld, P.; Sauer, M. Dissecting and reducing the heterogeneity of excited-state energy transport in DNA-based photonic wires. J. Am. Chem. Soc. 2006, 128 (51), 16864–75. (44) Espenel, C.; Margeat, E.; Dosset, P.; Arduise, C.; Le Grimellec, C.; Royer, C. A.; Boucheix, C.; Rubinstein, E.; Milhiet, P. E. Singlemolecule analysis of CD9 dynamics and partitioning reveals multiple modes of interaction in the tetraspanin web. J. Cell Biol. 2008, 182 (4), 765–776. (45) Marriott, G.; Mao, S.; Sakata, T.; Ran, J.; Jackson, D. K.; Petchprayoon, C.; Gomez, T. J.; Warp, E.; Tulyathan, O.; Aaron, H. L.; Isacoff, E. Y.; Yan, Y. L. Optical lock-in detection imaging microscopy for contrast-enhanced imaging in living cells. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (46), 17789–17794.
679
DOI: 10.1021/nl903823s | Nano Lett. 2010, 10, 672-679
