Abstract
The challenge of increasing the spatial resolution of an optical microscope beyond the diffraction limit can be reduced to a spectroscopy task by proper manipulation of the molecular states. The nanoscale spatial distribution of the molecules inside the detection volume of a scanning microscope can be encoded within the fluorescence dynamics and decoded by resolving the signal into its dynamics components. Here we present a robust and general method to decode this information using phasor analysis. As an example of the application of this method, we optically generate spatially controlled gradients in the fluorescence lifetime by stimulated emission. Spatial resolution can be increased indefinitely by increasing the number of resolved dynamics components up to a maximum determined by the amount of noise. We demonstrate that the proposed method provides nanoscale imaging of subcellular structures, opening new routes in superresolution microscopy based on the encoding/decoding of spatial information through manipulation of molecular dynamics.
Introduction
The visualization of macromolecular complexes inside cells is a fundamental need in modern cell biology. The fact that the size of interest is in the range of 10 to 100 nm generated a growing interest for fluorescence microscopy methods able to form images with spatial resolution well below the diffraction limit (~200 nm). Since the early 1990s, several physical concepts were identified for breaking the diffraction limit, which triggered the development of superresolved fluorescence microscopy^{1,2,3}. The spatial resolution of a microscope has usually been expressed in terms of its pointspreadfunction (PSF) (the image of a pointlike source), stating that fluorophores that are closer than the PSF size cannot be resolved. With few exceptions (for instance, methods based on nonclassicality of light^{4} and nearfield interaction^{5}) all the superresolution fluorescence methods provide subdiffraction resolution by transiently switching the fluorophores between optically distinguishable states and causing the fluorophores within the same diffractionlimited (DL) PSFsized region to signal sequentially. Accordingly, the different superresolution concepts can be broadly classified based on how the fluorophore states are manipulated^{6}. In methods that employ stochastic switching and readout, where the DLPSF represents the image of a fluorophore, only one fluorophore, at that time, is likely to fluoresce in a region covered by the DLPSF^{7,8,9}, so that its position can be determined with a precision much higher than the DLPSF size^{10}. In methods that employ targeted switching and readout, where the DLPSF represents the focal volume of the scanned excitation beam, the fluorophores are selectively silenced at the periphery of the DLPSF, resulting in an effective (E) PSF of smaller size^{11,12}. Notably, structured illumination superresolution techniques can be classified as targeted methods^{13,14,15}, since they can be regarded as a parallelization of the singlepoint scanning methods.
The need for a generic superresolution technique to preclude the simultaneous emission of neighbouring fluorophores comes from the fact that the fluorescence signal versus the spatial coordinate is the most prominent physical quantity used for the construction of an image. Since at each pixel a given number of photons N are detected originating from an undetermined number of fluorophores at unknown positions within the DLPSFsized region, one way to reduce this indetermination is to act on the fluorophores either by reducing their number to one (stochastic switching) or by restricting their possible positions within the DLPSFsized region (targeted switching). However, adding a temporal dimension to the measure of fluorescence can make this concept based on restricting the number/position of fluorophores superfluous^{16,17,18}. This can be addressed in terms of information capacity theory^{19}: subdiffraction resolution can be obtained with a method that encodes information from the saturated spatial channels of the microscope system into the temporal channel and decodes it after the transmission^{20,21}.
To the best of our knowledge, the first superresolution approach making explicit use of the temporal dynamics of fluorescence was proposed by Enderlein (ref. 16). This approach, called dynamic saturation optical microscopy (DSOM), relies on the fact that, on modulated illumination and in the presence of darkstate fluorophore transitions, a spatial distribution of the illumination intensity induces a spatialdependent temporal dynamics. A confocalbased implementation has been realized exploiting the singlet–tripletstate transition^{22}. Unfortunately, an efficient DSOM implementation needs hundreds of microseconds tripletstate lifetime fluorophores, which favours the occurrence of unwanted photobleaching processes.
A similar spatialdependent temporal dynamics can be obtained exploring the singlet states transitions and using the process of stimulated emission (SE). The (singlet) excitedstate lifetime τ of a fluorophore reduces as a function of the intensity of the light beam inducing the SE process, the socalled stimulated emission depletion (STED) beam. Hence the spatial distribution of the intensity of the STED beam, will determine differences in the fluorescence decay rates of fluorophores at different positions within the DLPSF. This principle has been used recently to implement an efficient and versatile STED microscope: the gated continuouswaveSTED (CWSTED) microscope^{23,24,25}. In a STED microscope, the STED beam is shaped like a doughnut and coaligned with the regular Gaussian excitation beam. If the STED beam operates in CW (the CWSTED implementation) the intensity is normally too low to effectively quench all the fluorophores in the periphery and thus to significantly reduce the size of the EPSF. However, the STED beam modifies the fluorescence lifetime at the periphery but not at the centre of the EPSF, so that by using a pulsed excitation beam and a timeresolved measurement the collection of fluorescent photons after a timedelay T_{g} from the excitation events (timegated detection) allows the removal of a significant fraction of photons emitted in the periphery. Theoretically, the spatial resolution of the gated CWSTED microscope improves with the timedelay T_{g}, but since ‘wanted’ photons, stemming from the EPSF centre, are also discarded the limiting factor becomes the signaltonoise ratio (SNR)^{26}. This can be made worse by the presence of uncorrelated background signal, originating for instance from direct excitation by the STED beam, which is not removed by simple time gating^{27,28,29}. The very same principle of gated CWSTED is the basis of another superresolution configuration, which uses a regular Gaussian STED beam^{30} . In this implementation, the time windows of the timeresolved measurements correspond to images recorded with their own EPSFs, which form a basis set from which an optimized EPSF is obtained and can be used for reconstructing an image with subdiffraction spatial resolution^{30}. Notwithstanding this, despite the relatively simple optical setup implementation, the optimization of the parameters of the reconstructed EPSF for a given degree of depletion is less straightforward, especially in the presence of noise and background.
Here we present a general method to achieve, in principle, arbitrary spatial resolution, which relaxes the condition of sequential fluorescence emission. We encode information about the fluorophores nanometer scale spatial distribution within the temporal dynamics of the fluorophore’s transition and decode it using a fast and robust phasor approach. Since the encoding procedure uses temporal modulation of the sample illumination, the method also isolates any uncorrelated background signal. For example, by generating controlled gradients in the fluorophore’s (singlet) excitedstate lifetime via SE, we demonstrate that this method provides backgroundfree nanometer scale imaging of the subcellular structures. This approach opens a new route for superresolution microscopy based on the encoding and decoding of spatiotemporal information through manipulation of the fluorophore dynamics.
Results
The SPLIT method
Modulating the sample illumination and measuring the temporal dynamics of the fluorescence can alleviate the stringent condition of silencing all the fluorophores located on a specific portion of the DLPSF. The N photons observed at each pixel can still originate from fluorophores located at any position within the DLPSF but they could be emitted with different temporal dynamics according to the position of the generating fluorophore in the DLPSF. The maximum achievable spatial resolution is ultimately determined by the ability to distinguish between different temporal dynamics. The key point here is that the issue of resolving spatial features is translated into the spectroscopy problem of resolving temporal dynamics components. The scheme of the method that we call SPLIT (Separation of Photons by LIfetime Tuning) is depicted in Fig. 1a. Suppose that within the DLPSF of the microscope, we can distinguish two spatial components 1 and 2 characterized by different temporal dynamics. We make use of the phasor analysis of lifetime data^{31,32,33} to represent the two different temporal dynamics as two vectors in the phasor plot. The total number of photons detected at one pixel is the sum of the photons originating in the two spatial components plus the uncorrelated background (BKGD) N=N_{1}+N_{2}+N_{BKGD}, where only N_{1} represents the ‘wanted’ part of all the photons. Following the rules of phasors, the vector P=(g,s) associated with the intensity decay at one pixel can be expressed as the linear combination of the vectors P_{1}=(g_{1},s_{1}) and P_{2}=(g_{2},s_{2}) associated with the two components, with weights f_{1} and f_{2} given by the corresponding fractions of detected photons: P=(N_{1}P_{1}+N_{2}P_{2})/N=f_{1}P_{1}+f_{2}P_{2}. This is a linear system of equations in the unknowns f_{1} and f_{2}. We can write this system in the form P=M f, where f=(f_{1},f_{2}) is the vector of the fractional components and M_{ij} is the matrix
which describes the two different temporal dynamics in the phasor domain. For a given n × n matrix M, the solution of this system is given by f=M^{−1}P. Once we find f the images N_{i}(x,y) (i=1,…,n) of the photons emitted in each of the n subdiffraction volumes and the image N_{BKGD}(x,y) of the background can be calculated as N_{i}(x,y)=f_{i}(x,y)N(x,y) and . As a result the original image N(x,y) has been split into n+1 images based on the assumption that we can observe and distinguish, within our observation volume, n different dynamics, associated with n linearly independent vectors in the ndimensional phasor space. These dynamics, using the RESOLFT concept^{6}, are seen as generalized reversible states of an ensemble of molecules, as they do not correspond necessarily to a specific state of the molecule but rather to a temporal fingerprint. The overall dynamics observed in the DLPSF is described here using the linear combination properties of phasors: if we assume that there are only two components, the phasor will fall on the line connecting the phasors from pure components (P_{1} and P_{2}). If we take into account the uncorrelated background as a third component, the phasor P will fall in a triangle where the vertices are the phasors P_{1} and P_{2} and the phasor P_{BKGD} of the uncorrelated background (Fig. 1a). To separate more than two dynamics components (n>2) and the uncorrelated background, we extend the analysis to phasors obtained at multiple harmonic frequencies^{34,35}. The image formation process in SPLIT is depicted schematically in Fig. 1b. Shortly, the temporal information of the signal at each pixel is used to generate the g and s images. These images are then processed to obtain the final SPLIT images.
The SPLIT method in timeresolved CWSTED
We focus now on the specific case of SEinduced lifetime variations and on the CWSTED microscopy architecture, that is, a Gaussian excitation beam and a doughnutshaped STED beam (Fig. 2a). However, the proposed approaches can be easily adapted to other configurations and other state transitions. The first ingredient is a model to describe the n dynamics components into which to split the measured intensity pixelbypixel, namely, we need the matrix M. For simplicity, we assume (i) a Gaussian profile of the conventional DLPSF h(x′,y′,z′)=exp(−2r^{2}/w^{2})exp(−2z′^{2}/w_{z}^{2}), with w and w_{z} being the beam waists along the radial and axial direction, respectively; and r^{2}=x′^{2}+y′^{2} the radial distance from the focal point (x=0, y=0) (ii) a parabolic approximation for the doughnutshaped STED beam I_{STED}(r)=I_{STED}(w)r^{2}/w^{2}, with I_{STED}(w) the STED beam intensity at position r=w; (iii) a single exponential decay rate for the unperturbed fluorophores γ_{0}=1/τ_{0}, where τ_{0} is the unperturbed excitedstate lifetime. Under these assumptions the spatial distribution of the decay rate is approximated by a parabolic function γ(r^{2})=γ_{0}+γ_{0}k_{S}r^{2}/w^{2}, where k_{S}=I_{STED}(w)/I_{SAT} is the ratio between I_{STED}(w) and the saturation value I_{SAT} for which the probability of decay due to SE and spontaneous emission are equal (see Supplementary Note 1). Importantly, the value of k_{S} determines the relative variation of decay rate values within the EPSF of the CWSTED microscope (Fig. 2a).
The timedependent fluorescence signal F(x,y,t) at each pixel can be obtained by integrating the contribution of all the fluorophores located in the EPSF centred in the pixel position (x,y) (see Supplementary Note 1)
where C(r^{2}) describes the concentration of the fluorophores in a concentric region of radius r around the pixel position and K is a constant that depends on the quantum yield of the fluorophore, the maximum of the excitation intensity and the detection efficiency. In contrast to deconvolution methods, the proposed method does not try to reassign photons to the original position, thereby the position at which the fluorophores are located within each concentric cylinder is not important. In other words, this approach does not need prior knowledge of the EPSF of the CWSTED system, which makes this approach suitable also for nonexpert users. The temporal dynamics of F(x,y,t) encodes nanoscale spatial information in the distribution of exponential decay components. We split the integral and calculate n dynamics components defined uniquely by the parameters γ_{0} and k_{S} (see Methods), from which the decoding matrix M is derived.
We tested the proposed method on synthetic timeresolved CWSTED images obtained with known γ_{0} and k_{S}. Figure 2b shows the ability of the SPLIT method in separating the photons coming from the inner subdiffraction volume from those of the periphery and the uncorrelated background, whereas time gating is affected by an increasing fraction of background (Fig. 2b,c). Notably, the spatial features appearing on the background image are due to the approximation of the continuous distribution of dynamics to only two components (see Supplementary Fig. 1). The spatial resolution of the SPLIT image can be further increased using a higher number n of components (Fig. 2d and Supplementary Fig. 2a). In gated CWSTED microscopy this is done by increasing the timedelay T_{g} (Supplementary Fig. 2b). For instance, using n=4 it is possible to get the same spatial resolution of CWSTED but at a STED beam intensity, which is 1 order of magnitude lower (Supplementary Fig. 2c). Note that the separation of dynamics obtained in the SPLIT method is conceptually different from time gating. The separation in SPLIT is based on the analysis of variations of the signal over the whole time range (Fig. 1b). For this reason, a SPLIT image could be obtained at increasing values of n even when T_{g} is limited by the period T (the reciprocal of the repetition rate, typically in the order of 10^{7} Hz) (Supplementary Fig. 2d).
The SPLIT image exploits the additional spatial information potentially encoded in the g(x,y) and s(x,y) images (see Supplementary Note 2). This additional amount of spatial information is available on a STED image but not on a confocal image (Supplementary Fig. 3). The improvement in spatial resolution in a SPLIT image at increasing values of n comes from the analysis of increasingly higher temporal frequencies in the signal (see Supplementary Note 2). Thus, it comes from a better sorting of photons as a function of dynamics/locations. However, when noise is taken into account, the larger the number n of components the higher will be the noise propagated to the final images, quantified as the condition number k_{cond} of the matrix M to invert (Fig. 2d and Supplementary Fig. 4; see Supplementary Note 3). For a given level of depletion and for a given level of noise, there is a finite number of values of n for which the noise in the final image is below a desired threshold.
Experimental determination of unknown decoding parameters
To decode the spatial information hidden in the gradients of dynamics induced by the STED beam, we need to know, according to our model, only the two parameters γ_{0} and k_{S.} The parameter τ_{0}=1/γ_{0} is usually known for a specific fluorophore or can be easily measured from the sample with the very same instrumentation by setting the STED beam power to zero. The parameter k_{S}=I_{STED}(w)/I_{SAT} is proportional to the STED beam power but its precise value depends on the optical configuration and on the properties of the sample. It is interesting that, using our analytical model of the SEinduced lifetime variations, we are able to estimate the value of k_{S} from the same image F(x,y,t) by considering the average timeresolved decay of all the pixels of an image (see Supplementary Note 1),
where B denotes the uncorrelated background. To validate the model, we imaged 40 nm fluorescent beads at several STED beam powers (Fig. 3a). The twodimensional (2D) histogram of the values g(x,y) and s(x,y) associated with each pixel is represented in the phasor plot (Fig. 3b). The phasor of the confocal image (zero STED power) is centred to the position corresponding to a single exponential decay with τ_{0}=4.5 ns. The same value τ_{0}=1/γ_{0} is found by fitting the average photonarrival time histogram to equation (3) with k_{S}=0 (Fig. 3c). With the increasing of the STED beam power the phasor becomes elongated as different dynamics are sampled in the image. The precise value of k_{S} at each STED power can be determined by fitting the average photonarrival time histogram to equation (3) with τ_{0} fixed (τ_{0}=4.5 ns). The good agreement with the model is confirmed by the linearity between k_{S} and the STED beam power. The phasor associated with the theoretical decay expressed by equation (3) for τ_{0}=4.5 ns and increasing the value of k_{S} describes the expected trajectory of the average phasor of the image as a function of the STED power. To assess the validity of the method for the imaging of nonpointlike structures, we also performed simulations using more convoluted structures similar to those found in cytoskeletal networks (see Supplementary Fig. 5). Also in this case, by using the values of k_{S} obtained by fitting the average timeresolved STED decay of the image, we were able to separate the images of the superresolved components and the background.
SPLIT imaging of subcellular structures
We finally applied the SPLIT method to the imaging of biological structures, namely microtubules on fixed HeLa cells, as reported in Fig. 4. We compare results obtained by labelling tubulin with two different dyes, Alexa Fluor 488 (Fig. 4a) and Oregon Green 488 (Fig. 4b). The parameters γ_{0} and k_{S} are found from the same experimental timeresolved data sets by fitting the average decay of all the pixels (Supplementary Fig. 6). According to the model and the simulations, the expected fullwidth at halfmaximum (FWHM) of the SPLIT image (n=2) is of the order of 100 nm, as confirmed by the experimental intensity profiles (Fig. 4 and Supplementary Fig. 7). The SPLIT image is compared with the confocal and the gated CWSTED image (gating time is set to T_{g}=1 ns), showing the improvement in spatial resolution with simultaneous efficient removal of background. Under the same experimental conditions (λ_{STED}=560 nm) the Oregon Green 488 fluorophores exhibit more uncorrelated background due to STED beaminduced excitation, thereby in the relative gated STED image, the improvement in resolution is totally masked by the strong background induced by direct STED beam excitation. Even though specific methods for background subtraction^{27,28,29} have been developed recently, it is remarkable how the uncorrelated background photons are automatically separated in the calculation of the SPLIT image (Fig. 4c). This is possible because we are operating a separation of the signal in the frequency domain (where the uncorrelated background is wellseparated from the other components) rather than in the time domain (where the uncorrelated photons are evenly distributed). Here in addition to the different temporal dynamics of the excitation and STED beams, sufficient to remove the uncorrelated background, we are also decoding the spatial frequencies hidden in the gradient of dynamics induced by the STED beam.
Discussion
We have shown that superresolution can be achieved by proper spatial tuning of the fluorophore signal dynamics as a function of their position within the detection volume of a scanning microscope. A key element of this method is the phasor representation of the signal dynamics, which provides a procedure to recover the hidden spatial information lost during the timeaveraging process. To the best of our knowledge, this is the first time that a tool like the phasor, originally developed for spectroscopy, has been applied for imaging to increase its spatial resolution beyond the diffraction limit.
In particular, we have described a robust, fitfree and userfriendly method for separating the different dynamics components of a timeresolved CWSTED measurement and isolating the high spatial resolution components, even in the presence of uncorrelated background. The method is ultimately limited by the noise in the measurement of the dynamics but, importantly, we can predict the maximum number of components that we can resolve into for a given level of noise. The spatially associated dynamics components are generated starting from only two parameters (γ_{0} and k_{S}) that can be easily assessed in the same image. Notably, these parameters alone are sufficient to decode the subdiffraction spatial information without prior calibration of the confocal PSF.
This approach is different from other superresolution approaches making explicit use of the temporal dynamics of fluorescence. For instance, in the DSOM implementation, which exploits a singlet–tripletstate transition, the contribution of the faster decaying fluorophores in the centre of the DLPSF is extracted by fitting the signal at each pixel with a multiexponential model. In timegated STED, the size of the EPSF is reduced by restricting detection to late photons, an approach that is of straightforward operation but not equivalent to resolving dynamics components. Indeed, in timegated STED the subtraction of background requires further analysis, and, worth noting, the information hidden in the early photons is lost. In the Gaussian STED beam implementation, instead of performing spectroscopy on each pixel of the image, an image is associated to each bin of the time decay and a superresolved image is obtained from a reconstruction using a suitable linear combination of these images. Importantly, the derivation of the optimal coefficients for the linear combination needs the knowledge of the EPSF of the Gaussian STED implementation.
When the EPSF of the imaging system is known, a powerful approach to improve the spatial resolution is image deconvolution^{36}. Generally speaking, deconvolution uses the fact that each pixel of an image encodes spatial information of structures contained in the neighbouring pixels. Given the EPSF of the system, a deconvolution algorithm tries to recover such information. In comparison, our SPLIT approach does not use the apriori spatial information provided by the EPSF, which is unfortunately not straightforward information. However, it is clear that the spatial information provided by the EPSF and the temporal information provided by the signal dynamics can work in synergy. In particular, one could develop a dedicated deconvolution algorithm which uses a spatiotemporal EPSF^{26}, or apply a conventional spatial deconvolution algorithm on the SPLIT image. Indeed, the SPLIT method is a linear and spaceinvariant system, thus fully characterized by an EPSF (Supplementary Note 2).
Compared with all these methods, the SPLIT approach appears as a simple and efficient way to extract the high spatial frequencies encoded in the variations of fluorescence and the only method to operate the decoding explicitly in the frequency domain in a pixelbypixel manner. SPLIT exploits the additional spatial information potentially encoded in the g(x,y) and s(x,y) images describing the evolution of the fluorescence signal at each pixel.
It is worth noticing that the generality of the method is not compromised by the specific assumptions used to simplify the analytical description. For instance, even though we have considered lifetime variations generated along only the radial direction but not along the axial, the model can be easily adapted to take into account decay rate gradients generated by 3Dstructured STED beams^{37}. We have also assumed that unperturbed temporal dynamics could be described by a single exponential decay, though for many fluorophores this is not true. However, one of the major advantages of phasor analysis is precisely that both exponential and nonexponential decays are equally described as vectors in the phasor space, and it is on these vectors that we perform our analysis.
In STED microscopy, lifetime tuning and photon separation can already be achieved at lowintensity illumination and fast acquisition speed, which are among the key requirements for noninvasive live cell imaging. The fact that photons are spatially sorted and not suppressed makes it quite intriguing to explore this method in singlemolecule techniques like fluorescence correlation spectroscopy^{38}, where a high SNR is generally required and where STED beaminduced background appears as an obstacle towards full 3D nanoscale fluctuation imaging^{39}. The method is not limited to SE but can be applied to other types of dynamics or spectra in general, for instance the transition dynamics of reversibly photoswitchable proteins^{40,41} or, similar to DSOM, the singlettotriplet states dynamics. As long as the observed dynamics occur on a faster time scale compared with the pixel dwell time, the method has no limitations in terms of speed. Its resolution performances will be ultimately determined by our ability in generating and discriminating different dynamics components against the noise. This suggests looking for those physical phenomena that maximize transient dynamics or spectral gradients within the DLPSF rather than decrease the total number of emitted photons. In this perspective, an interesting point for future developments will be the design of optical probes whose photophysical or spectral properties could be easily modulated without detrimental effects. We envisage that additional advantages could be gained by exploiting the more complex dynamics of photophysical systems involving a larger number of molecular states^{42}. With a rapidly evolving imaging technology helping us to improve sensitivity and thereby SNR, we expect the next generation of fast, backgroundfree optical superresolution microscopy to make extensive use of spatiotemporal information encoding strategies.
Methods
Calculation of phasors and their linear combination
The phasor coordinates at a harmonic h corresponding to the timeresolved intensity I(t) are defined as^{34}:
where T is the period of excitation or a smaller value for which the function I(t) has already decayed to the uncorrelated background value. If the intensity at one point is due to the sum of two components plus the uncorrelated background, I(t)=I_{1}(t)+I_{2}(t)+I_{BKGD}, then its phasor can be expressed as a linear combination of the phasors of the two components and the phasor of the background (that is, a null vector):
where the total number of photons N detected at one pixel is the sum of the photons originating in the two spatial components plus the uncorrelated background N=N_{1}+N_{2}+N_{BKGD}. It can be seen that since the uncorrelated background is independent of t, its phasor coordinates are (0,0). The addition of uncorrelated background does not affect the value of the phase φ=tan^{−1}(s/g) but decreases the value of modulation m=(g^{2}+s^{2})^{1/2} of a phasor.
Separation of the intensity into n components
The calculation of an arbitrary number of fractional components n was obtained by considering a matrixvector representation and extending the phasor analysis to higher harmonics. If the intensity I(t) is sampled in N_{bin} time windows, then the maximum number of harmonics we can use is N_{bin}/2. P=(g, s, g^{(2)}, s^{(2)},…) is the nelement vector formed by the phasor coordinates derived from the intensity decay at one pixel. The last element of the vector P is g^{((n+1)/2)} (if n is odd) or s^{(n/2)} (if n is even). Provided that we know the temporal dynamics of the n components I_{j}(t), we defined M_{ij} as the n × n matrix whose column j is the vector with the phasor coordinates of the j^{th} component up to g_{j}^{((n+1)/2)} (if n is odd) or s_{j}^{(n/2)} (if n is even):
Then, provided that det M≠0, the nelements vector of the fractional components f=(f_{1},…,f_{n}) was calculated by f=M^{−1}P.
Calculation of dynamics components in timeresolved CWSTED
In timeresolved CWSTED microscopy, the exact temporal dynamics of F(x,y,t) depends on the function C(r^{2}), which acts as a pixeldependent weight on the exponential decay components exp(−γ(r^{2})t). To approximate the continuous distribution of decays in a discrete number n of components, we split the integral into n parts
where I_{i}(t) describes the average dynamics of the ith component
The boundaries r_{i} of the subdiffraction volumes were chosen in such a way that, for C(r^{2})=constant, all the timecorrelated photons were split in equal number among the n components (see Supplementary Note 1).
Simulations and data analysis
Simulations of timeresolved STED microscopy images of pointlike particles were performed using custombuilt software in ImageJ. The value of the intensity at pixel (x,y) and time t originating from N_{p} pointlike particles was set as: where , γ_{0}=1/τ_{0} is the spontaneous decay rate and r_{i}^{2} is the square of the distance from the ith particle. Stacks consisted of 128 frames of 64 × 64 pixels. The parameter S is the maximum intensity signal from a particle expressed in counts detected at one pixel in one frame of the stack. The parameter B represents a uniform level of background expressed in counts detected at one pixel in one frame of the stack. The resulting ideal image is successively corrupted by Poisson noise. For all the simulations reported in Fig. 2b, Supplementary Fig. 1, Supplementary Fig. 2, Supplementary Fig. 3 and Supplementary Fig. 4 the confocal waist was set to the value w=167 nm (corresponding to a FWHM=200 nm), the pixel size to 5.2 nm and the time step to Δt=0.097 ns. The other parameters were varied as indicated in the figures. The confocal image was obtained by adding all the frames of the confocal stack (k_{S}=0). The STED image was obtained by adding all the frames of the STED stack (k_{S}>0). The timegated STED image was obtained by adding only those frames of the STED stack for which t≥T_{g}.
Simulations of convoluted structures similar to those found in cytoskeletal networks (Supplementary Fig. 5) were performed using MATLAB (MathWorks). The cytoskeletal phantom was composed of 75 filaments with diameter of 30 nm. To each filament, we associated a value between 0 and 1 to simulate differences in the brightness of the structures. The maximum total number of photons detected from a single pixel position in one frame of the stack was set to S=120. A uniform level of background was set to the value B=0.5 counts per each pixel and per each frame of the stack. The phantom was convolved with a theoretical 3D (x,y,t) EPSF of a CWSTED microscope^{26} and the obtained image was corrupted by Poisson noise. In the example reported in Supplementary Fig. 5, the stack consisted of 64 frames of 256 × 256 pixels, showing an area of 10 × 10 μm. The confocal FWHM was set to 235 nm. The total time period was set to T=12.5 ns so that Δt=12.5/64 ns. The unperturbed decay rate was set to τ_{0}=3.15 ns and relative variation of decay rate was set to the value of k_{S}=12.7. Again, in this case, the STED image was obtained as the sum of all the frames in the stack. For the simulations of cytoskeletal structures the value of γ_{0} was known, whereas the value of k_{S} was determined by fitting equation (3) to the average timeresolved STED decay of the image (Supplementary Fig. 5).
The parameters γ_{0} and k_{S} relative to the experimental biological images were extracted from the full fields of view reported in Supplementary Fig. 6. We extracted first the value of γ_{0} by fitting equation (3) to the average confocal decay by fixing k_{S}=0. Then we extracted the value of k_{S} by fitting equation (3) to the average STED decay. By fitting the confocal decay to a single exponential decay, we obtain the value of τ_{0} (τ_{0}=2.7 ns for the Alexa Fluor 488 sample; τ_{0}=1.8 ns for the Oregon Green sample). Then we fix this parameter and estimate k_{S} from the fitting of the STED decay (k_{S}=7.8 for the Alexa Fluor 488 sample; k_{S}=4.9 for the Oregon Green sample). To test if the parameter k_{S} varied across the sample, we performed the same analysis in smaller regionsofinterest of different size (Supplementary Fig. 6c). The values of k_{S} extracted from regionsofinterest of size down to about 32 pixels were quite consistent between different regions and consistent with the k_{S} value extracted from the whole image. We used the parameters γ_{0} and k_{S} to split the timeresolved STED image into the superresolved components (1 and 2) and the background (BKGD).
The SPLIT analysis was implemented writing a custom code in MATLAB. For each pixel (x,y) of the timeresolved STED image, the phasor coordinates g(x,y) and s(x,y) were calculated using a FFT algorithm. The phasor plots reported in Fig. 3 are the 2D histograms of these values and were obtained using Globals for Images (Laboratory for Fluorescence Dynamics). The values of γ_{0} and k_{S} were used to generate the expected theoretical decays of the n spatial components and the decoding matrix M. The condition number k_{cond} of the matrix M was calculated in MATLAB.
All the fitting procedures were performed in OriginPro (OriginLab) using an unweighted least squares procedure.
Experiments
All the timeresolved CWSTED experiments were performed on a homebuilt CWSTED microscope^{25,29}. The excitation beam was provided by a supercontinuum source and the STED beam was provided by a CW visible fibre laser (VFL) emitting at 560 nm (VFLP1000560, MPB Communication Inc.). We generated the supercontinuum source by pumping a photoncrystalfibre (femtoWHITE800, NKT Photonics) with a femtosecond modelocked Ti:Sapphire laser of 150 fs pulse width, 80 MHz repetition rate (Chameleon, Vision II, Coherent). To obtain a doughnutshaped diffraction pattern at the focus the STED beam passed through a polymeric mask imprinting 0–2π helical phaseramps (VPPA1, RPC Photonics). The STED and the excitation beams were collinearly aligned using two dichroic mirrors (zt488RDC and z560sprdc, AHF analysentechnik), then deflected by two galvanometric scanning mirrors (6215HM40B, CTICambridge) and directed towards the objective lens (HCX PL APO 100/1.40/0.70 Oil, Leica Microsystems) by the same set of scan and tube lenses as the ones used in a commercial scanning microscope (Leica TCS SP5, Leica Microsystems). The fluorescence light was collected by the same objective lens, descanned and passed through the dichroic mirrors as well as through a fluorescence band pass filter (ET Bandpass 525/50 nm, AHF analysentechnik) before being focused (focal length 60 mm, AC254060AML, Thorlabs) into a fibre pigtailed single photon avalanche diode (PDF Series, Micro Photon Devices). Photonarrival times were detected at each pixel by a timecorrelatedsinglephotoncountingcard (SPC830, Becker & Hickl). Synchronization was obtained from the reference signal provided by the Ti:sapphire laser. All imaging operations were automated and managed by the software Imspector (Max Planck Innovation). For both the STED and excitation light, the average power P was measured at the back aperture of the objective lens. Due to losses in the objective lens, the power at the sample is actually lower by 15% and 12% at 488 nm and 560 nm, respectively.
Samples of 40nm diameter yellow–green fluorescent spheres (Yellow–Green, Invitrogen) were prepared as follows. The spheres were diluted in water by 1:3,000 (v/v). We dropped the dilute solution of fluorescent beads onto a polyLlysine (Sigma) coated glass coverslip, waited 10 min, washed it with water and dried the coverslip by blowing nitrogen onto it. Finally we mounted the coverslip with a special medium (Mounting Medium, Invitrogen) and we observed with the STED microscope.
For immunofluorescence imaging, HeLa cells were cultured on glass coverslips (18mm diameter) in Dulbecco’s modified Eagles medium (Invitrogen) supplemented with 10% foetal bovine serum (Invitrogen), 100 IU ml^{−1} penicillin and 100 μg ml^{−1} streptomycin (Invitrogen) at 37 °C in a humidified atmosphere containing 5% CO_{2} for 24 h. Plated cells were rinsed with phosphatebuffered saline (PBS) (0.1 M, pH 7.4) and fixed by incubation in 4% formaldehyde in PBS for 15 min. Fixed cells were washed with PBS and permeabilized for 30 min at room temperature with 3% normal bovine serum albumin and 0.1% Triton X100 in PBS. The cells were then incubated with the monoclonal mouse antiαtubulin antiserum (Sigma Aldrich) diluted in 3% bovine serum albumin 0.1% Triton/PBS (1:1,000) for 1 h at room temperature. Antiαtubulin antibody was revealed using Alexa Fluor 488 goat antimouse IgG (1:500, Molecular Probes) or Oregon Green 488 goat antimouse IgG (1:500, Molecular Probes). The coverslips were rinsed in PBS and then placed in an openbath imaging chamber containing PBS and observed with the STED microscope.
Additional information
How to cite this article: Lanzanò, L. et al. Encoding and decoding spatiotemporal information for superresolution microscopy. Nat. Commun. 6:6701 doi: 10.1038/ncomms7701 (2015).
References
 1
Hell, S. W. Toward fluorescence nanoscopy. Nat. Biotechnol. 21, 1347–1355 (2003) .
 2
Hell, S. W. Microscopy and its focal switch. Nat. Methods 6, 24–32 (2009) .
 3
Huang, B., Babcock, H. & Zhuang, X. Breaking the diffraction barrier: superresolution imaging of cells. Cell 143, 1047–1058 (2010) .
 4
Schwartz, O. et al. Superresolution microscopy with quantum emitters. Nano Lett. 13, 5832–5836 (2013) .
 5
Betzig, E. & Trautman, J. K. Nearfield optics: microscopy, spectroscopy, and surface modification beyond the diffraction limit. Science 257, 189–195 (1992) .
 6
Hell, S. W. Farfield optical nanoscopy. Science 316, 1153–1158 (2007) .
 7
Hess, S. T., Girirajan, T. P. K. & Mason, M. D. Ultrahigh resolution imaging by fluorescence photoactivation localization microscopy. Biophys. J. 91, 4258–4272 (2006) .
 8
Rust, M. J., Bates, M. & Zhuang, X. Subdiffractionlimit imaging by stochastic optical reconstruction microscopy (STORM). Nat. Methods 3, 793–796 (2006) .
 9
Betzig, E. et al. Imaging intracellular fluorescent proteins at nanometer resolution. Science 313, 1642–1645 (2006) .
 10
Deschout, H. et al. Precisely and accurately localizing single emitters in fluorescence microscopy. Nat. Methods 11, 253–266 (2014) .
 11
Hell, S. W. & Wichmann, J. Breaking the diffraction resolution limit by stimulated emission: stimulatedemissiondepletion fluorescence microscopy. Opt. Lett. 19, 780–782 (1994) .
 12
Hofmann, M., Eggeling, C., Jakobs, S. & Hell, S. W. Breaking the diffraction barrier in fluorescence microscopy at low light intensities by using reversibly photoswitchable proteins. Proc. Natl Acad. Sci. USA 102, 17565–17569 (2005) .
 13
Heintzmann, R., Jovin, T. M. & Cremer, C. Saturated patterned excitation microscopy a concept for optical resolution improvement. J. Opt. Soc. Am. A 19, 1599–1609 (2002) .
 14
Gustafsson, M. G. L. Nonlinear structuredillumination microscopy: Widefield fluorescence imaging with theoretically unlimited resolution. Proc. Natl Acad. Sci. USA 102, 13081–13086 (2005) .
 15
Chmyrov, A. et al. Nanoscopy with more than 100,000 'doughnuts'. Nat. Methods 10, 737–740 (2013) .
 16
Enderlein, J. Breaking the diffraction limit with dynamic saturation optical microscopy. Appl. Phys. Lett. 87, 094105 (2005) .
 17
Dertinger, T., Colyer, R., Iyer, G., Weiss, S. & Enderlein, J. Fast, backgroundfree, 3D superresolution optical fluctuation imaging (SOFI). Proc. Natl Acad. Sci. USA 106, 22287–22292 (2009) .
 18
Cox, S. et al. Bayesian localization microscopy reveals nanoscale podosome dynamics. Nat. Methods 9, 195–200 (2011) .
 19
di Francia, G. T. Resolving power and information. J. Opt. Soc. Am. 45, 497–499 (1955) .
 20
Lukosz, W. Optical systems with resolving powers exceeding the classical limit. II. J. Opt. Soc. Am. 57, 932–939 (1967) .
 21
Cox, I. J. & Sheppard, C. J. R. Information capacity and resolution in an optical system. J. Opt. Soc. Am. A 3, 1152–1158 (1986) .
 22
Humpolickova, J., Benda, A., Machan, R., Enderlein, J. & Hof, M. Dynamic saturation optical microscopy: employing darkstate formation kinetics for resolution enhancement. Phys. Chem. Chem. Phys. 12, 12457–12465 (2010) .
 23
Moffitt, J. R., Osseforth, C. & Michaelis, J. Timegating improves the spatial resolution of STED microscopy. Opt. Express 19, 4242–4254 (2011) .
 24
Vicidomini, G. et al. Sharper lowpower STED nanoscopy by time gating. Nat. Methods 8, 571–573 (2011) .
 25
Vicidomini, G. et al. Gated CWSTED microscopy: A versatile tool for biological nanometer scale investigation. Methods 66, 124–130 (2014) .
 26
Vicidomini, G. et al. STED nanoscopy with timegated detection: theoretical and experimental aspects. PLoS ONE 8, e54421–e54421 (2013) .
 27
Vicidomini, G., Moneron, G., Eggeling, C., Rittweger, E. & Hell, S. W. STED with wavelengths closer to the emission maximum. Opt. Express 20, 5225–5236 (2012) .
 28
Ronzitti, E., Harke, B. & Diaspro, A. Frequency dependent detection in a STED microscope using modulated excitation light. Opt. Express 21, 210–219 (2013) .
 29
Coto Hernàndez, I. et al. A new filtering technique for removing antiStokes emission background in gated CWSTED microscopy. J. Biophotonics 7, 376–380 (2014) .
 30
Marsh, R. J., Culley, S. & Bain, A. J. Low power super resolution fluorescence microscopy by lifetime modification and image reconstruction. Opt. Express 22, 12327–12338 (2014) .
 31
Clayton, A. H., Hanley, Q. S. & Verveer, P. J. Graphical representation and multicomponent analysis of singlefrequency fluorescence lifetime imaging microscopy data. J. Microsc. 213, 1–5 (2004) .
 32
Redford, G. I. & Clegg, R. M. Polar plot representation for frequencydomain analysis of fluorescence lifetimes. J. Fluoresc. 15, 805–815 (2005) .
 33
Digman, M. A., Caiolfa, V. R., Zamai, M. & Gratton, E. The phasor approach to fluorescence lifetime imaging analysis. Biophys. J. 94, L14–L16 (2008) .
 34
Weber, G. Resolution of the fluorescence lifetimes in a heterogeneous system by phase and modulation measurements. J. Phys. Chem. 85, 949–953 (1981) .
 35
Gratton, E., Jameson, D. M. & Hall, R. D. Multifrequency phase and modulation fluorometry. Annu. Rev. Biophys. Bioeng. 13, 105–124 (1984) .
 36
Bertero, M., Boccacci, P., Desiderá, G. & Vicidomini, G. Image deblurring with Poisson data: from cells to galaxies. Inverse Probl. 25, 123006 (2009) .
 37
Harke, B. et al. Resolution scaling in STED microscopy. Opt. Express 16, 4154–4162 (2008) .
 38
Elson, E. L. Fluorescence correlation spectroscopy: past, present, future. Biophys. J. 101, 2855–2870 (2011) .
 39
Ringemann, C. et al. Exploring singlemolecule dynamics with fluorescence nanoscopy. N. J. Phys. 11, 103054 (2009) .
 40
Grotjohann, T. et al. Diffractionunlimited alloptical imaging and writing with a photochromic GFP. Nature 478, 204–208 (2011) .
 41
Testa, I., D’Este, E., Urban, N. T., Balzarotti, F. & Hell, S. W. Dual channel RESOLFT nanoscopy by using fluorescent state kinetics. Nano Lett. 15, 103–106 (2015) .
 42
Schoenle, A. & Hell, S. W. Farfield fluorescence microscopy with repetitive excitation. Eur. Phys. J. D 6, 283–290 (1999) .
Acknowledgements
This work was supported in part by grants NIH P41GM103540 (E.G.), NIH P50GM076516 (E.G.), PRIN N. 2008S22MJC 005 (A.D. and G.V.). We thank Jenu Varghese Chacko for fruitful discussions, Marta d’Amora for providing experimental samples and Eileen Sheppard for English proofreading.
Author information
Affiliations
Contributions
L.L. and G.V. conceived the idea and analysed the results. E.G. and A.D. assisted in the project and analysed the results. L.L., I.C.H. and G.V. performed experiments. L.L and M.C. performed simulations. L.L. developed the model and wrote the software. L.L. and G.V. wrote the manuscript. A.D. and E.G. supported manuscript writing.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Supplementary information
Supplementary Information
Supplementary Figures 17, Supplementary Notes 13 and Supplementary References (PDF 1295 kb)
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
Lanzanò, L., Coto Hernández, I., Castello, M. et al. Encoding and decoding spatiotemporal information for superresolution microscopy. Nat Commun 6, 6701 (2015). https://doi.org/10.1038/ncomms7701
Received:
Accepted:
Published:
Further reading

Advances in fluorescence microscopy techniques to study kidney function
Nature Reviews Nephrology (2021)

Naryl pyrido cyanine derivatives are nuclear and organelle DNA markers for twophoton and superresolution imaging
Nature Communications (2021)

Preprocessing visualization of hyperspectral fluorescent data with Spectrally Encoded Enhanced Representations
Nature Communications (2020)

Exploiting the tunability of stimulated emission depletion microscopy for superresolution imaging of nuclear structures
Nature Communications (2018)

Local raster image correlation spectroscopy generates highresolution intracellular diffusion maps
Communications Biology (2018)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.