Time-Resolved STED Microscopy

Diffraction does not allow light to be focused to a volume smaller than roughly one-half of the light wavelength along the lateral directions (x,y) and three times larger along the optical axis (z). Stimulated emission depletion (STED) microscopy (Hell et al., Opt. Letters, 19(11):780-782, 1994, Vicidomini et al., Nat. Methods, 8:571–573, 2018) overcomes this diffraction limit by reversibly silencing (depleting) fluorophores at predefined positions of the diffraction-limited excitation volumes. Only the non-silenced fluorophores in the complementary regions emit light, allowing features closer than the diffraction limit to be separated. In the most typical STED microscopy implementation, the fluorescent confinement is obtained by coaligning the Gaussian excitation beam of a scanning microscope with a second beam, called the STED beam, which (i) is tuned in wavelength to de-excite fluorophores via stimulated emission (SE) and (ii) is engineered to create a doughnut-shaped focal intensity distribution with a ‘zero’-intensity point in the center. Although the STED beam focal intensity distribution is diffraction limited, high intensities saturate the SE transition and keep virtually all the fluorophores in the ground state, except those located in a region around the ‘zero’-intensity point, whose size reaches sub-diffraction values and decreases with increasing STED beam intensity. Thus, scanning the coaligned beams together across a specimen leads to an image where the (sub-diffraction) spatial resolution is given by the size of the effective fluorescent volume around the ‘zero’.

Theoretically, STED microscopy resolution can reach the molecule’s size (the ultimate limit of a fluorescent microscope). In practice, it is limited by different factors, and in particular by the photo-damages effects. To efficiently deplete a fluorophore, SE has to win the competition with spontaneous emission, which typically occurs within a few nanoseconds after the excitation event (fluorophore’s excited-state lifetime). This short temporal window and the small cross-section of SE demand a high flux of stimulating photons. For example, to quench by half the fluorescence of a fluorophore with 4 ns excited-state fluorescence lifetime and 25 cm²/J stimulated emission cross-section requires 10 MW/cm² intensity (saturation intensity). Because a complete quenching of a fluorophore requires much higher intensities, and because the intensity reduces quadratically from the doughnut-crest to the ‘zero’-intensity point, effective resolution enhancements require > 1 GW/cm² intensity (at the doughnut-crest). At first, this request was achieved using expensive and complex mode-locked pulsed laser architectures (pulsed-STED microscopy), which, together with the photo-damage problem, initially slowed the growth and dissemination of STED microscopy.

A class of methods which mitigates both problems, i.e., which reduce the (peak) intensity -- to achieve a certain spatial resolution, and the system complexity, base on the analysis of the fluorescence dynamics (time-resolved STED microscopy). Since the SE process opens a new de-excitation pathway for an excited fluorophore, the fluorophores illuminated by the STED beam show a shorter effective excited-state fluorescence lifetime than the fluorophores not illuminated. In particular, the higher is the insensity of the STED beam the shorter is the effective lifetime. Because the doughnut-shaped distribution of the SE intensity, the fluorophores located at the doughnut-crest show the shortest lifetime, which increases up to the natural lifetime moving toward to the 'zero'-intensity point (Vicidomini et al., PLoS One, 8(1):e54421, 2013). This spatial lifetime signature of the fluorophore inside the excitation volume can be used to isolate the fluorescecne signal generated by the longer-lived fluorophores located at the 'zero'-intensity point from the signal generated from the fluorophores in the pheriphey. The final result is a detection volume, thus an (optical) spatial resolution, which reduces also without a complete depletion of the pheripheral fluorophores (incomplete depletion), and without increasing the STED beam intensity.

The first implementation using the time-resolved STED principle is the so-called gated-STED microscope (Vicidomini et al., Nat. Methods, 8:571–573, 2011, Moffit et al., Opt. Express, 19(5):4242-4254, 2011). The fluorescence signal of a STED microscope is registered in a time-correlated-single-photon-counting (TCSPC) mode and the image is formed solely with the photons registered after a certain time from the excitation events, i.e., time-gated detection. Thanks to this scheme, the photons contributing to the image are likely to be emitted by the long-lived fluorophores located in the inner part of the STED effective fluorescence volume. Initially, gated-STED microscopy used STED beams running in continuous-wave (CW), thus obtaining not only a sensitive reduction of the STED bams (peak) intensity to achieve effective sub-diffraction resolution, but also a reduction in complexity and costs. On the contrary, the time-gating benefits for the early architectures based on mode-locked pulsed lasers were negligible. However, while STED beams running in CW reduce the peak intensity, they also result in useless illumination: typically, excitation uses 80 MHz (or lower) pulsed beams and the fluorescence lifetime is only a few nanoseconds. Thus the duty-cycle is small. In light of this consideration and with the time-resolved STED principle in mind, the mode-locked lasers of the pulsed-STED implementations were replaced by (sub)nanosecond pulsed laser. In this gated pulsed-STED implementation the STED beam illumination is optimized and time-gating reduces the (peak) intensity (Castello et al., Microsc. Res. Tech., 79(9):785-791, 2016).

Notably, the analysis of the fluorescence dynamics to improve the resolution of conventional microscopy (Enderlein, Appl. Phys. Lett., 87:094105, 2005), the combination of TCSPC with STED microscopy (Auksorius et al., Opt. Lett., 33(2): 113-115, 2008), and the effects of SE on the fluorophore lifetime (Marsh et al., Chem. Phys. Lett., 366(3):398-405, 2002), were all topics already investigated. However, to the best of our knowledge, the combination of all these aspects to reduce the (peak) intensity in STED microscopy was never explored up to the introduction of gated-STED microscopy. Thereby, the gated-STED story can be well described by the famous sentence of Sir Isaac Newton: "If I have seen further, it is by standing upon the shoulders of giants."

The major limitation of gated-STED microscopy is that reducing the detection volume, i.e., the improvement of the (optical) resolution, is obtained at the cost of a signal reduction. Thus, for low fluorescent photon flux or high background, reducing the signal to noise/background ratio can cancel-out the effective resolution enhancement. Indeed, time-gated detection removes completely the fluorescent photons originating from the periphery of the effective fluorescent volume, but partially also the photons from the center. Two different approaches partially solve this limitation. The first approach uses image deconvolution (Castello et al., Appl. Phys. Lett., 105:234106, 2014): The TCSPC imaging modality provides a three-dimensional STED image (x,y,t, where t is the lag-time of the photon-arrival histogram), which is deconvolved with an effective temporal point-spread-function describing the fluorescence dynamics both spatially and temporally. The second method, called separation-by-lifetime-tuning (SPLIT), uses the phasor plot representation (Lanzanò et al., Nat. Commun., 6:6701, 2015): Each pixel of the TCSPC image is represented in the phasor domain. The phasor's linear property allows separating the photons emitted from the long-lived fluorophores in the center of the effective fluorescent volume from the photons emitted by the short-lived fluorophore at the periphery.

The aim of this project is to develop new algorithms which combine deconvolution and phasor plot representation to further ehnance the effective resolution of time-resolved STED microscopy. Within the same goal, the project will take also advantages from new photon detectors, lasers, optical strategies, and probes.

Slides and Video:

Team members:

Publications: