Seminar 1

Stimulated emission depletion microscopy

Author: Neža Golmajer ZimaMentor: prof. dr., Janez Štrancar

Ljubljana, march 2018

Abstract

In this paper the stimulated emission depletion microscopy or STED is explained. The STED is asuper-resolution microscopy based on fluorescence with an additional doughnut-shaped laser whit which

fluorescence is repressed. The resolution of the STED is up to 30 nm and it is used for biological samples bothin in vitro and in vivo microscopy.

Contents1 Introduction 3

2 Diffraction barrier and fluorescence 32.1 Diffraction barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2 Fluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

3 STED 6

4 Conclusion 9

1 IntroductionThroughout the history, seeing small objects was a big desire. With a naked eye we see objects up to a fewhundred micrometres. Although the first magnifying lenses were known from the ancient Greeks, first microscopewas build in 17th century [1] and people were able to see objects in size up to one micrometre. But until theend of the 20th century the limit of visible light microscopy was set at approximately 200 nm. This limit wasovercome by electron microscope years before due to much shorter wavelength of the electrons. The limit ofelectron microscope is approximately 1 nm, but sample needs to be fixated, so it is not possible to observedynamic processes. Other problems are also that volumes must be small and multi-target labelling (that isusing different types of fluorophores at ones to study different interactions or processes at once) is a challenge[2]. In 1994 first idea of super resolution light imaging was presented by Stefan Hell who proposed a stimulatedemission depletion microscopy or shorter STED [3].

2 Diffraction barrier and fluorescenceTo understand STED microscopy, we first need to understand basics of fluorescence and characteristics ofobjectives.

2.1 Diffraction barrierEach microscope consists of an objective (or objective lenses) which is a part of a microscope that focuses a lightfrom a sample. Different parts of the sample have different characteristics thus scattered or fluorescent lightis also different for each part; but if two parts are too close together we cannot differentiate them apart. Theminimum distance of two point sources of light is determined by a diffraction barrier which can be calculatedfrom the Fraunhofer diffraction of a point source of light by a circular aperture. Intensity of a diffracted lightis given as:

I(x) = I0(2J1(x)x

), (1)

where J1 is a first Bessel function, x = 2ΠrNAλ , r is a distance from optical axis in focal plain, NA = nsinΘ is

a numerical aperture, n is a refractive index, Θ is the largest possible angle of incoming light in an objective[5]. Numerical apertures are between 0,25 for air objectives to 1,4 for very good (and expensive) immersionobjectives [6]. Because the objective is a lot wider than a spot in a sample we are observing light comes in theobjective under an angle. Larger angle means more light from the sample and a better objective (Figure 1).Zero of the first Bessel function is at 3.8317, so we can calculate minimal distance of two dots perpendicular tooptical axis:

dmin = λ

2NA, (2)

and parallel to the optical axis:z = 2nλ

(NA)2 . (3)

The equation 2 is also called the Abbe diffraction limit. A calculation of dmin and z for NA of quality objectives(NA = 1.4) at shortest wavelength of visible light (λ = 380nm) gives approximate values: dmin ≈ 150nm andz ≈ 580nm.

2.2 FluorescenceFluorescence is a natural phenomenon that was first described in 17th century and was observed in infusioncalled lignum nephriticum derived from the wood of Pterocarpus indicus and Eysenhardtia polystachya. When in

Figure 1: On the left: Graphic representation of numerical aperture. In the picture d = 2r. Adapted from Ref.[7]. On the right: Different distances of two dots on the screen and under every image the intensity of light isshown Adapted from Ref. [8].

19th century G. G. Stokes described the ability of fluorite (mineral form of calcium fluoride) to change invisiblespectrum of light into a blue light, he named it fluorescence [1].

Fluorescence is based on excitation and relaxation of electrons in atoms or in molecules called fluorophores.Before we illuminate the flourophore with an excitation laser, its electrons are in a ground state. If photon’senergy is the same as an energy gap between the ground state and one of the vibrant state in the first excitedstate, then the fluorophore’s electrons will be promoted into one of the vibrant states of its first excited state.Because the first excited state is less energetically favourable than the ground state, relaxation occurs. Atfirst electron relaxes into zero vibrant state and then into the ground state when photon is emitted (Figure 3).Energy of the emitted photon is lower than the absorbed energy of the excited electron thus a spectrum of theemitted light is red shifted (Figure 3). Excitation takes about a femtosecond, relaxation from a higher to alower vibrant state a few picoseconds, and the emission few nanoseconds.Fluorescence has two major problems: light intensity and so-called bleaching effect. An intensity of light isequal to:

I = IexσQe−tτ , (4)

where σ – a cross section for the fluorescence, Q - a quantum yield, τ - a decay time.High intensities of the excitation light can damage a biological samples, whether too low intensities cancause problems with an observation since the sample does not emit enough light. At long exposure times,a (photo)bleaching effect occurs. The photo bleaching is decrease in the fluorescence intensity, caused by theexcitation light which alters the fluorophores so they do not emit any more. Therefore the intensity of theemitted light decreases and the sample begins to fade. Although fluorescence has two major problems (lightintensity and bleaching effect) it is widespread due to its advantages. In a fluorescent microscopy we knowwhere exactly a fluorescent light comes from so we know an exact position of a fluorophore. Fluorophore canbe bind to a special place in the sample (such as a mitochondrion, nucleus, etc.) so exact position of observedpart of the sample is known. Due to the wavelength of the fluorescence only one photon is needed to observeit. The wavelength of the fluorescence is in a visible spectrum whereas surrounding is in IR spectrum. Thelast but not least od the advantages is a contrast. In a fluorescent microscopy contrast is much better than thecontrast of a bright field microscopy where a scattered light is observed.The fluorescene is observed with a microscopes that have a few special components- excitation filter, a dichroicmirror and a emission filter - which let through only narrow part of the spectrum. (Figure 2) In microscopes

where fluorescence is not used, whole spectrum is observed and chromatic aberrations can occur. In fluorescencemicroscopy chromatic aberrations are reduced due to the filters and the dichroic mirror.

Figure 2: Schematics of a fluorescence microscope. Light from a light source, which can be a laser, a LEDdiode or a xenon lamp, goes through an excitation filter, a dichroic mirror and an objective on the sample.Fluorescence light than goes back through the objective, the dichroic mirror and a emission filter on an ocularor a camera. Adapted from Ref. [9].

Figure 3: On the left: The Jablonski diagram for fluorescence and stimulated emission. Adapted from Ref. [10].S0 and S1 are the ground state and the first exited state, the lines represent the vibrant states. With yellowfluorescence is represented and with red the stimulated emission. On the right: The intensity of the excitationand the emission of a fluorophore in dependece of a wavelength. Adapted from Ref. [11].

3 STEDSTED microscopy or stimulated emission depletion microscopy is new super resolution microscopy techniquewhere we use two lasers: one for the excitation and one for the depletion of fluorophores. It is also callednanoscopy since it detects particles on a nanoscale. The technique was developed by Stefan W. Hell in 1994 andin 2014 he won the Nobel Prize in Chemistry for his discovery [12]. The fluorescence is the main process in theSTED microscopy. With excitation laser the fluorophores are excited into the first excited state so that theystart to emit photons and with a second laser faster emission is stimulated thus electrons return into the groundstate in picoseconds to femtoseconds instead in nanoseconds. Fast relaxation leads to impoverished fluorescenceand very low intensity of it.

Let’s look at it more detailed. Like I explained in the previous paragraph, the excitation laser exciteselectrons from the ground state to one of their vibrant states in the first excited state. Afterwards the highintensity laser (named the STED laser) with the energy of an energy gap between the ground and the firstexited state is turned on. When the intensity of the STED laser is high enough the fluorophores are returnedinto the ground state and they do not fluoresce any more. The intensity of the STED laser and the depletionof fluorophores is related (Figure 3). Resolution is defined with the intensity and the shape of the STED laserand can be expressed as:

∆r = A√ISTED/Isat + 1

, (5)

where ISTED – intensity of the STED laser, Isat – saturated intensity, A - minimal distance of two points givenby equations 2 and 3 [10]. To achieve the super resolution of the STED laser is a doughnut-shaped.(Figure 5)The intensity of the STED laser in a rim must be higher than the saturated intensity for the fluorescence to bestopped, but in the centre of the STED laser the intensity has to be lower than the saturated intensity so thatit does not stop the fluorescence. The saturated intensity is the intensity at which the STED laser can stopthe fluorescence. The diameter of a centre part of a doughnut-shaped STED is defined by the intensity of theSTED laser and is usually 30 nm to 40 nm, but in one case it was shown that it can be even as low as 5.8 nm[14]. When image is taken only the centre, lit part, is seen on camera (Figure 4). This is an example to howone image is taken. For the image of the whole sample, the STED laser is moved around. It has to be takeninto account that the resolution is in size of the STED laser’s centre thus we cannot differentiate between theimage of two or more particles smaller than the centre of doughnut-shaped STED laser and the image of oneparticle in the size of STED’s centre.The STED microscopy is most commonly used in a bological samples both in in vivo an in vitro microscopy. Inin vitro microscopy the resolution is really at 30 nm (Figure 7). It is also used in in vivo microscopy but theresolution is not as good as 30 nm where tissue changes can be observed in real time. (Figure 6)Although the STED is a really good method for taking high resolution images of biological samples it hasa few disadvantages. The intensity of the STED laser in order of megawatts or even gigawatts on squarecentimeters (MW/cm2 to GW/cm2) is needed due to fast transition between states. Intensity that high candamage biological samples. The second problem is life span of a the fluorophores which is up to 108 excitationcycles. In STED for every image at least 200 nm of sample is excited thus one part of a sample is excited (anddepleted) more than once. For example: if we have a square of 200 nm and want to take an image of it, everypart of this square is excited a hundred times. That is why we cannot take more than a few images of the samesample.

Due to these problems new nanoscopy techniques are being developed. For different problems of STED,other microscopies are developing. A reversible saturable optical linear fluorescence transition (RESOLFT) isSTED-like microscopy that uses much lower intensity of light due to a different process of fluorescence. Metastates of fluorophores called ON and OFF are used instead of stimulated emission. Said meta states are morestable than the first excited state of electron, therefore the laser intensity is lower. Meta states transitionsbetween the ON state and the OFF state are in time domain of a milliseconds. In the ON state, fluorophorescan emit light, but not in the OFF state. The only difference between the RESOLFT and the STED is that

Figure 4: On top: Intensity of STED laser and of fluorescence in dependence of position of exitation. Onbottom: Cartoon shows how STED works. In a first row basic fluorescence is shown and in second row is theSTED principle Adapted from Ref. [13].

Figure 5: On the left: Doughnut-shaped STED laser beam in xy-plane (on top) an in xz-plane (on bottom).In the centre: Connection between intensity of fluorescence and intensity of STED laser. Both are in arbitraryunits. Adapted from Ref. [13]. On the right: Simplified Jablonski diagram for stimulated emission. Adaptedfrom Ref. [10].

the intensity of the depletion laser is much lower - in kilowatts per square centimetre (kW/cm2) instead of inmegawatts per square centimetre (MW/cm2). Consequently, the images can be taken for a longer period oftime without damaging the sample. One of disadvantages of the RESOLFT is that fluorescence is not stoppedimmediately but kinetics occurs. A photo-activated localization microscopy (PALM) and a stochastic opticalreconstruction microscopy (STORM) have a solution for the bleaching due to many cycles of the fluorescence.Both work at the same principle of the random fluorescence of only one molecule with a uniform illumination.A disadvantage of the PALM or the STORM is that the resolution or a location of a molecule is proportionalto 1√

numberofphotonsthus a lot of photons are needed to determine a location of the molecule in the sample [15].

Figure 6: With STED microscopy we can take up to 28 frames per second therefore we can image dynamicprocesses in biological samples. Image above shows an example of in-vivo imaging of mouse’s brain. Seventeenimages were taken in the time frame of 55 minutes. Brain dynamics of mouse is filmed while mouse is underanaesthesia. Adapted from Ref. [16].

Figure 7: In this image we can see the real capability of STED. We see cell membranes in green and nanotubesin red. On the upper left is high definition high resolution image of the part of the sample, on the lower leftis a single Ti02 nanotube which is wrapped with lipids. It is very impressive that we can see 30 nm particle.Yellow colour means that both green and red colours were detected.

4 ConclusionThe STED is a super-resolution microscopy based on the fluorescence. In comparison to a confocal or a wide-field microscopy with a resolution around 200 nm, the STED microscopy has a resolution of 30 nm. Such aresolution is achieved with an additional so-called STED laser having a doughnut-shape and a high intensity.The fluorescence is impoverished everywhere except in the center of the STED laser. The STED microscopy isused in biological samples due to the high resolution and the possibility of fast imaging which allows to trackchanges in the samples in a in vivo microscopy. Because the STED is not an optimal microscopy method for thein vivo microscopy due to damaging the tissue, other methods like the RESOLFT, the PALM and the STORMare being developed.

References[1] https://en.wikipedia.org/wiki/Fluorescence [17. 2. 2018]

[2] R. Chéreau, J. Tønnesen, U. V. Nägerl, STED microscopy for nanoscale imaging in living brain slices,Methods, Vol. 88, (2015) 57-66

[3] Hell S, Wichmann J., Breaking the diffraction resolution limit by stimulated emission: stimulated-emission-depletion fluorescence microscopy, Optics Letters. Vol. 19, No. 11 (1994) 780-782.

[4] http://slideplayer.com/slide/9071429/ [17. 2. 2018]

[5] http://mafija.fmf.uni-lj.si/seminar/files/2014_2015/In-vivo_mikroskopija_bioloskih_vzorcev.pdf [17. 2.2018]

[6] https://www.microscopyu.com/microscopy-basics/numerical-aperture [17. 2. 2018]

[7] http://openspim.org/images/5/53/Lens_na.png [17. 2. 2018]

[8] http://www.writeopinions.com/optical-resolution [17. 2. 2018]

[9] https://en.wikipedia.org/wiki/Fluorescence_microscope [17. 2. 2018]

[10] B. Huang, M. Bates, X. Zhuang, Super resolution fluorescence microscopy, Annual review of biochemistry,Vol. 78, (2009) 993-1016

[11] https://www.thermofisher.com/si/en/home/life-science/protein-biology/protein-biology-learning-center/protein-biology-resource-library/pierce-protein-methods/fluorescent-probes.html [17. 2. 2018]

[12] https://www.nobelprize.org/nobel_prizes/chemistry/laureates/2014/press.html [17. 2. 2018]

[13] https://svi.nl/STEDMicroscopy [17. 2. 2018]

[14] E. Rittweger, K. Y. Han, S. E. Irvine, C. Eggeling, S. W. Hell, STED microscopy reveals crystal colourcentres with nanometric resolution, Nature Photonics, Vol. 3, No. 3, (2009) 144-147

[15] https://www.youtube.com/watch?v=YyBGiZZSslY [17. 2. 2018]

[16] W. Wegner, P. Ilgen, C. Gregor, J. van Dort, A. C. Mott, H. Steffens, K. I. Willig, In vivo mouse andlive cell STED microscopy of neuronal actin plasticity using far-red emitting fluorescent proteins,ScientificReports, Vol. 7, (2017)