Hawkes P.W., Spence J.C.H. (Eds.) Science of Microscopy. V.1 and 2

Подождите немного. Документ загружается.

818 S.W. Hell and A. Schönle

()

()=













−

(

)

∑

(32)

transformation

kgggggg

()

⊗

()

++ ⊗+ ⊗⊗+

()

δξξ ξ



… (33)

At low intensities, ζ and therefore ξ is so small that only the linear term

is relevant and the convolution extends the support to 6k as discussed

above. The larger the maximum intensity, the more important higher

orders of the Taylor series will become. These involve multiple auto-

convolutions of the function g extending the support further and

further.

While a quantitative treatment in frequency space is more compli-

cated and less intuitive than the one introduced in the previous section,

the following analysis gives a feel for the effect of the saturation factor

and also illustrates the possible vast expansion of the OTF support. For

the sake of simplicity we assume a Gaussian form of the light distribu-

tion function

f(x) = 1 − exp(−x

/2a

) (34)

The properly normalized m-fold auto-convolution of g is then given by

⊗= −

gk a m ak m

() / ( /( ))22

π exp

(35)

Now let us assume that the useful support ends at a frequency where

the OTF is attenuated to a small fraction ε of its value at small frequen-

cies. For large saturation factors the inﬂ uence of the convolution with

the confocal OTF on the cut-off frequency can be neglected and we

have to calculated the sum in brackets in (33). Substituting (35) into

equation (33) and approximating the sum by an integral we get for the

term in brackets

ok i iak,/lnexplnςπξ ξ

(

)

≅− −

()

(36)

For large saturation factors we can write lnξ = lnζ − ln(ζ + 1) ≅ −1/ζ

and obtain

ok ak,exp(/)ςπζ ς

(

)

≅−2

(37)

This means that the attenuation of the modulus of the OTF at large

frequencies is anti-proportional to the square-root of the saturation

factor. This is equivalent to saying that the resolution increases with

the square root of the saturation factor just as we expected from our

previous analysis.

4.1 STED Microscopy

STED microscopy produces subdiffraction resolution and subdiffrac-

tion-sized ﬂ uorescence volumes in exactly the manner described above

by the depletion of the ﬂ uorescent state of the dye. Depletion inherently

implies saturation of the depleting transition. At present, it is realized

in a (partially confocalized) spot-scanning system due to a number of

technical advantages, but it has been conceptually clear from the outset

With g(r) = 1 − f(r) and = ζ/( ζ + 1 ) we obtain the OTF after Fourier

Chapter 12 Nanoscale Resolution in Far-Field Fluorescence Microscopy 819

490nm

244nm

STED

source

EXCT.

source

DET

97nm

104nm

d) e)

fluorescent

non-fluorescent

450 650

λ[nm]

Excitation

STED

Detection

Figure 12–7. Stimulated emission depletion (STED) was the ﬁ rst implementation of the RESOLFT

principle. (a) Dye molecules are excited into the S

(state A) by an excitation laser pulse. (b) Fluorescence

is detected over most of the emission spectrum. How ever, molecules can be quenched back into the

ground state S

(state B) using stimulated emission before they ﬂ uoresce by irradiating them with a light

pulse at the edge of the emission spectrum shortly after the excitation pulse and before they are able to

emit a ﬂ uorescence photon. Saturation is realized by increasing the intensity of the depletion pulse and

consequently inhibiting ﬂ uorescence everywhere except at the “zero points” of the focal distribution of

the depletion light. (c) Schematic of a point-scanning STED microscope. Excitation and depletion beams

are combined using appropriate dichroic mirrors (DC). The excitation beam forms a diffraction-limited

excitation spot in the sample (inset in d) while the depletion beam is manipulated using a phase-plate

(PP) or any other device to tailor the wavefront in such a way that it forms an intensity distribution with

a nodal point in the excitation maximum (left inset in e). The third inlay shows the resulting quenching

probability when saturating the depletion process. (d) and (e) show an experimental comparison

between the confocal PSF and the effective PSF after switching on the depleting beam. Note the doubled

lateral and ﬁ ve-fold improved axial resolution. The reduction in dimensions (x, y, z) yields ultrasmall

volumes of subdiffraction size, here 0.67 al (Klar et al., 2000), corresponding to an 18-fold reduction

compared to its confocal counterpart. The spot size is not limited on principle grounds but by practical

circumstances such as the quality of the zero and the saturation factor of depletion. (See color plate.)

that nonconfocalized detection is viable as well (Hell and Wichmann,

1994). The principal idea, a schematic setup and an exemplary mea-

surement of the resolution increase, is shown in Figure 12–7. The ﬂ uo-

rophore in the ﬂ uorescent state S

(state A) is stimulated to the ground

820 S.W. Hell and A. Schönle

state S

(state B) with a doughnut-shaped beam. The saturated deple-

tion of S

conﬁ nes ﬂ uorescence to the central intensity zero. With typical

saturation intensities ranging from 1 to 100 MW/cm

, saturation factors

of up to 120 have been reported (Klar et al., 2000, 2001). This should

yield a 10-fold resolution improvement over the diffraction barrier, but

imperfections in the doughnut have limited the improvement to 5 to

7-fold in experiments (Klar et al., 2001).

As already stated, light microscopy resolution can be described either

in real space or in terms of spatial frequencies. In real space, the resolu-

tion is assessed by the FWHM of the focal spot. The measurements

depicted in Figure 12–8 were carried out with an excitation wavelength

of λ = 635 nm, an oil immersion lens with a numerical aperture of 1.4,

and with the smallest possible probe: a single ﬂ uorescent molecule

(Westphal and Hell, 2005; Westphal et al., 2003). Figure 12–8a shows

the measured proﬁ le of the PSF in the focal plane (x) for a conventional

0.01

0.1

03040

Convent.

STED

STED deconv.

PSFa)

x [nm]

-200 -100 0 100 200

Convent.

222 nm

STED

40 nm

OTF

I [a.u.]

0.0

1.0

-1

[1/µm]

-100 0 100

62 nm

Mol.

no.1

Mol.

no.2

33 nm

Figure 12–8. (a) Comparison of the effective PSF’s lateral intensity proﬁ le for

confocal and STED microscopy indicating an ∼5.5-fold resolution increase in

the latter. (b) Lateral cuts through the effective OTFs giving the bandwidth of

the lateral spatial frequencies passed to the image. The data plotted in (a) and

(b) are gained by probing the ﬂ uorescent spot of a scanning microscope with

a single molecule of the ﬂ uorophore JA 26 using a numerical aperture of 1.4

(oil) objective lens and at wavelengths of 635 nm (excitation), 650–720 nm (ﬂ uo-

rescence collection), and 790 nm (STED). The inset demonstrates subdiffrac-

tion resolution with STED microscopy. Two identical molecules located in the

focal plane that are only 62 nm apart can be entirely separated by their inten-

sity proﬁ le in the image. A similar clear separation by conventional micro-

scopy would require the molecules to be at least 300 nm apart. (Date adapted

from Westphal et al.)

Chapter 12 Nanoscale Resolution in Far-Field Fluorescence Microscopy 821

ﬂ uorescence microscope along with its sharper subdiffraction STED

ﬂ uorescence counterpart. STED leads to improvement in resolution by

a factor of approximately 5.5.

Figure 12–8b shows the OTF of a conventional microscope along with

the enlarged OTF of the STED ﬂ uorescence microscope. As expected,

the effective OTF’s support in the confocal case ends at approximately

(2/635 nm + 2/720 nm) = 5.91/µm. For STED, we included the OTF after

successful linear deconvolution, which restores higher spatial frequen-

cies that are not swamped by noise. The region of usable OTF support

is approximately marked by the region where frequencies are enhanced

by the deconvolution process without producing artifacts and is ∼5.5

times larger than for the confocal case. This marks a fundamental

breaking of Abbe’s diffraction barrier in the focal plane. The inlay dem-

onstrates the resulting subdiffraction resolution exempliﬁ ed by the

linearly deconvolved STED image of two molecules at a distance of

62 nm. They are distinguished in full by two narrow peaks (Westphal

et al., 2003). As a result of deconvolution, the individual peaks are

sharper (33 nm FWHM) than the initial peak of 40 nm FWHM.

Very recently, utilizing STED wavelengths of λ = 750–800 nm, a lateral

FWHM of down to 16 nm has been achieved in experiments with single

JA 26 molecules spin-coated on a glass slide.

Measuring the resolution

as a function of the applied STED intensity conﬁ rmed the predicted

increase of the resolving power with the square root of the saturation

factor (see Figure 12–9). Of course the cutoffs presented are based on a

somewhat arbitrary deﬁ nition of what can be considered “usable fre-

quencies” at a certain signal-to-noise ratio. However low this threshold

is set, the confocal support cannot extend beyond ∼6/µm while the

STED-OTF’s support is theoretically unlimited.

The one-dimensional (1 D) phase-plate yielding 16 nm FWHM is opti-

mized for maximum resolution improvement in the lateral direction

perpendicular to the polarization of the depleting light and leads to an

intensity distribution with two strong peaks at either side of the excita-

tion maximum (Keller et al., in preparation). Because the depleting light

is polarized, the resolution gain depends on the orientation of the mole-

cules. However, a considerable increase in resolution is still possible for

the second phase-plate, which yields a doughnut-shaped intensity distri-

bution and thus an almost isotropic resolution increase in the lateral

directions when using circularly polarized light.

To “squeeze” the ﬂ uorescence spot in both lateral directions two

STED beams aberrated with 1 D phase-plates oriented at 90° to each

other can be combined. Together with circularly polarized excitation,

almost uniform resolution in the focal plane is achieved as shown in

Figure 12–10. A series of xy images acquired with different STED beam

powers demonstrate the resolution increase and concomitant widening

of the OTF when the applied saturation factor increases (Schönle et al.,

in preparation). This combination of two incoherent beams causes the

resolution to depend on the orientation of the transition dipole and

results in spikes along the x and y direction of the OTF when imaging

randomly oriented ﬂ uorophores (see Figure 12–10). New phase-plates

have been proposed to avoid such effects and to improve the effective

822 S.W. Hell and A. Schönle

saturation factors at a given total STED power. Incoherent combination

can then be used to improve the resolution in all three spatial dimen-

sions. The resulting PSFs exhibit very weak dependence on dipole ori-

entation (Keller et al., in preparation) and allow application of STED to

the imaging of biological specimen and reliable subsequent linear

deconvolution (Willing et al., in preparation).

STED microscopy has also been successfully applied to the imaging

of biological samples. Subdiffraction images with three-fold enhanced

axial and doubled lateral resolution have been obtained with mem-

brane-labeled bacteria and live budding yeast cells (Klar et al., 2000).

While there is some evidence for increased nonlinear photobleaching

of some dyes when increasing the depletion intensity (Dyba and Hell,

2003), there is no reason to believe that the intensities currently applied

would be detrimental to live cells. This is not surprising since the

intensities are two to three orders of magnitude lower than those used

in multiphoton microscopy (Denk et al., 1990). Moreover, STED has

proven to be single molecule sensitive, despite the proximity of the

Conf.

0.5

0.05

4.4 21 34

1/∆x [µm

-1

]

0 10 30

1+NA

max

[MW/cm

]

0 200

200

235nm

131nm

72nm

40nm

31nm

26nm

800 1000

δ [nm]

0.45

Conf. STED

x [nm]

-100-200 100 200

0.5

16nm

Figure 12–9. STED microscopy reduces the ﬂ uo-

rescence focal spot size to a degree far below the

diffraction limit: (a) spot of a confocal micro-

scope (left) compared with that in an STED

microscope (right) utilizing a y-oriented inten-

sity valley for STED (upper right inset, not to

scale) squeezing the spot in the x direction to

16 n m widt h. (b) As also observed in Figure 12–8,

the bandwidth in STED is fundamentally

increased over confocal microscopy. The graph

shows the normalized magnitude of optical

transfer function (OTF) as a function of inverse

distance. For the “1D” depletion scheme, the

usable support of the OTF is increased by almost

a factor of 8. When using a doughnut-shaped

depletion beam with a “wider” intensity zero, the

OTF support is still extended almost ﬁ ve-fold. (c)

The average focal spot size decreases with the

STED intensity following a square-root law, in

agreement with Eq. (27). Because the resolution

depends on molecule orientation, the spot sizes

were measured for several tens of single mole-

cules. The curves follow the mean values

(squares) and the inset discloses the histogram

of the measured spot sizes at 1100 (MW/cm

)

with the minimum FWHM at 16 nm and a 26 nm

average.

Chapter 12 Nanoscale Resolution in Far-Field Fluorescence Microscopy 823

1µm

0.001

0.2

0.1

0.01

-32

1/∆x [µm

-1

]

0.001

0.2

0.1

0.01

0.001

0.2

0.1

0.01

0.001

0.2

0.1

0.01

0 a.u. 1

0.001

0.2

0.1

0.01

-32

1/∆x [µm

-1

]

I [MW/cm

]

0.2

0.75

0.25

F/F

0.3

800

200

b) [84 MW/cm

]

c) [187 MW/cm

]

d) [290 MW/cm

]

0.001 0.2

Figure 12–10. Images of a wetted Al

matrix featuring z-oriented holes (Whatman plc, Brentford,

UK) with a spin cast of a dyed (JA 26) polymethyl methacrylate solution. The rings formed in this way

are ∼250 nm in diameter and are barely resolved in confocal mode. (a–d) The confocal image (a) and

STED images with two depleting beams perpendicularly polarized and aberrated by “1D” phase-plates

(b–d). The excitation PSF (g) and the STED PSF for y polarization (h) and x polarization (i) are shown

on the right. The STED intensity was chosen at the spots marked in the saturation curve (f). The smaller

effective spot size also results in an extended OTF as seen in the second column. Here, the insets show

the 2D Fourier transformation of the images in the left and the graphs show a proﬁ le along the x direc-

tion. Note the logarithmic scales. The Fourier transform of the image is given by the product of OTF

and the Fourier transform of the object [Eq. (2)]. For such regular structures, an estimate for the

modulus of the OTF can therefore be gained by estimating the latter and solving for the OTF. The

dashed line shows the Fourier transform of a ring with a diameter of 275 nm and a width of 50 nm and

the estimated OTF is presented in (e). (f) The suppression of ﬂ uorescence resulting from stimulated

emission. The phase-plates were removed and the ratio of ﬂ uorescence without STED light (F

) and

with the STED beams switched on (F) was recorded. The intensities are pulse intensities per beam at

the global maximum. (See color plate.)

824 S.W. Hell and A. Schönle

STED wavelength to the emission peak. In fact, individual molecules

have been switched on and off by STED upon command (Kastrup and

Hell, 2004; Westphal et al., 2003).

The power of STED and 4Pi microscopy has been synergistically com-

bined to demonstrate for the ﬁ rst time an axial resolution of 30–40 nm in

focusing light microscopy (Dyba and Hell, 2002). The intensity distribu-

tion of the depleting light is formed by a 4Pi setting with destructive

interference at the geometric focus leading to a zero intensity there and

two neighboring maxima at a distance of approximately λ/4. This

results in superior xz images, and the technique has initially been suc-

cessfully applied to membrane-labeled bacteria (Dyba and Hell, 2002).

More recently, STED-4Pi microscopy has been extended to i mmunoﬂ uo-

rescence imaging (Figure 12–11). A spatial resolution of ∼50 nm has been

demonstrated in the imaging of the microtubular meshwork of a mam-

malian cell.

These results indicate that the basic physical obstacles to

attaining a 3D resolution of the order of a few tens of nanometers have

been overcome. Since the samples were mounted in an aqueous buffer

(Dyba and Hell, 2002; Dyba et al., 2003), the results indicate that the

optical conditions for obtaining subdiffraction resolution are met under

the physical conditions encountered in live cell imaging.

It is to be expected that ultrasmall detection volumes created by

STED will also be useful in a number of sensitive bioanalytical tech-

niques. Fluorescence correlation spectroscopy (FCS) (Magde et al.,

1972) relies on small focal volumes to detect rare mole cular species or

interactions in concentrated solutions (Eigen and Rigler, 2001; Elson

and Rigler, 2001). While volume reduction can be obtained by nano-

fabricated structures (Levene et al., 2003), STED may prove instrumen-

tal in attaining ultrasmall spherical volumes at the nanoscale inside

samples that do not allow for mechanical conﬁ nement. The latter fact

is particularly important to avoid an alteration of the measured ﬂ uctua-

tions by the nanofrabricated sur face walls.

In fact, the viability of STED FCS has recently been shown in an

experiment (Kastrup et al., 2005). In a particular implementation STED

FCS has witnessed a reduction of the focal volume by a factor of ﬁ ve

along the optic axis and a concomitant reduction of the axial diffusion

time. The initial experiments showed that for particular dyewavelength

combinations the evaluation of the STED FCS data might be complicated

by a seemingly uncorrelated background at the outer wings of the ﬂ uo-

rescence spot where STED may not completely suppress the signal.

Further investigations will show whether this challenge is easily over-

come in the near future. In any case, published results suggest a further

decrease of the volume by another order of magnitude (Westphal et al.,

2003; Irie et al., 2002). An inherent disadvantage of STED is the necessity

of an additional pulsed light train that is tuned to the red edge of the

emission spectrum of the dye. Nevertheless STED is to date the only

known method for “squeezing” a ﬂ uorescence volume to the zeptoliter

scale without making mechanical contact. Thus, the creation of ultrasmall

volumes, tens of nanometers in diameter, by STED may be a pathway to

improving the sensitivity of ﬂ uorescence-based bioanalytical tech-

niques (Weiss, 2000; Laurence and Weiss, 2003).

Chapter 12 Nanoscale Resolution in Far-Field Fluorescence Microscopy 825

400

800

1200

1600

01234

Z / µm

Counts

Z / µm

FWHM

53nm

01234

Confocal

STED-4Pi

25nm

60nm

Immunolabeled

microtubule

Overview

Monolayer

Microtubules

Monolayer

Microtubules

Figure 12–11. Subdiffraction immunoﬂ uorescence imaging with STED-4Pi microscopy. (a) Overview

image (xy) of the microtubular network of an HEK cell. (b) Sketch of typical dimensions of a labeled

microtubule ﬂ uorescently decorated via a secondary antibody. (c) and (d) Standard confocal and STED-

4Pi xz image recorded at the same site of the cell; the straight line close to the cell stems from a mono-

molecular ﬂ uorescent layer attached to the adjacent coverslip. In both images, the pixel size was 95 ×

9.8 n m in the x and z direction, respectively; the dwell time per pixel was 2 ms. Note the fundamentally

improved clarity in (d). The STED-4Pi microscope’s PSF features two low side lobes caused by the sec-

ondary minima STED intensity distribution. These lobes are <25% and were removed in the STED-4Pi

image using linear ﬁ ltering as outlined in the text [see Eq. (16ff)]. (e) and (f) Corresponding proﬁ les of

the image data along the dashed lines in (b) and (c) quantify the improved axial resolution of the STED-

4Pi microscopy mode (f) over the confocal benchmark. Peaks 1, 2, and 3 due to microtubules are broader

than the response to the monolayer. Note the ability of the STED-4Pi microscope to distinguish adjacent

features. (See color plate.)

An important step toward far-reaching applicability of STED micro-

scopy was the demonstration of the suitability of laser diodes both for

excitation and for depletion (Westphal et al., 2003). However, several

issues remain to be addressed. Due to the considerably smaller detec-

tion volumes, the signal per pixel is reduced and the amount of pixels

to be recorded increases. Therefore, it will be important to incorporate

STED into fast, ideally parallelized scanning systems.

826 S.W. Hell and A. Schönle

While the transition to shorter wavelengths will further increase

resolution by a factor of ∼1.5, it is most likely that the ultimate resolu-

tion limit in STED will be set by the stability of the marker used. The

photostability of current markers was considerably improved by

stretching the depleting pulse to >300 ps (Dyba and Hell, 2003), but it

might not be easily possible to attain saturation factors ζ > 200 in the

near future. Nevertheless, according to Eq. (27) ζ = 200 should already

yield an improvement by one order of magnitude, provided that the

actual intensity value at the “intensity zero” is indeed negligible at this

saturation level.

As explained above (Eq. 30), the actual “depth” of the zero codeter-

mines the attainable resolution, because for relatively high saturation

factors the saturable transition also becomes effective at the zero point

or points. So far, typical depths were in the range of γ = 1–2.5% of the

global maximum of the depleting intensity I(r). The zero could be a

single point, as in a single beam scanning system, but in the case of a

parallelized system, it may also be a line or an array of points or lines.

The actual depth of the zeros will certainly depend on the particular

setup and the quality of optical components and proper alignment.

Independently of implementation details, active optical elements such

as wavefront phase modulators will be a valuable tool to further “deepen”

the zeros, which in turn will allow the full potential of the attained satu-

ration level to be exploited for improvement in resolution.

4.2 Variations of RESOLFT Microscopy and Producing Large

Saturation Factors at Low Power

At this point, we reiterate that RESOLFT is not restricted to the process

of stimulated emission, but can exploit any reversible (linear) transition

driven by light; the attainable resolution is determined by the ratio of

driving intensity and the competing transition rate k

. If the applicable

intensity is limited by the onset of photodamage to the marker or even

to the sample, marker constructs must be found where high saturation

levels are attained at lower intensities. This is certainly the case if the

rate competing with the transition to be saturated is lower.

One such example is the GSD mentioned earlier. In this version of the

RESOLFT concept the ground state (now state A) is depleted by target-

ing an excited state (B) with a comparatively long lifetime (Hell and

Kroug, 1995; Hell, 1997), such as the meta-stable triplet state T

. In many

ﬂ uorophores T

can be reached through the S

with a quantum efﬁ ciency

of 1–10% (Lakowicz, 1983). Being a forbidden transition, the relaxation

of the T

is 10

–10

times slower than that of the S

, thus yielding I

= 0.1–

100 kW/cm

. The signal to be measured (from the intensity zero) is the

ﬂ uorescence of the molecules that remained in the singlet system; this

measurement can be accomplished through a synchronized further

excitation (Hell and Kroug, 1995). For many ﬂ uorophores, this approach

is not straightforward, because T

is involved in the process of photo-

bleaching, but there are potential alternatives such as the meta-stable

states of rare earth metal ions that are fed through chelates.

Also proposed has been depleting the ground state S

by populating

the S

(now B) (Heintzmann et al., 2002). This is the technically simplest

Chapter 12 Nanoscale Resolution in Far-Field Fluorescence Microscopy 827

realization of saturated depletion, since it requires only excitation

wavelength matching. However, as the ﬂ uorescence emission maps the

spatially extended “majority population” in state B, the super resolved

images (represented by state A) are negative images hidden under a

bright signal from B. Hence photon noise from the large signal might

swamp the ﬂ uorescence minima that occur when intensity zeros, where

no ﬂ uorescence is excited, are colocalized with ﬂ uorophores. The sub-

sequent computational extraction of the positive image is therefore

very dependent on an excellent signal-to-noise ratio. The saturation

intensity is of the same order as in STED microscopy, because the satu-

ration of ﬂ uorescence also competes with the spontaneous decay of S

This probably results in photostability issues similar to the case of

STED. In fact, the photobleaching should be exacerbated, since the satu-

rated transition is effected with higher energy photons that are gener-

ally more prone to facilitating photochemical reactions. Pumping the

dye to a higher state rather than into the ground state also favors pho-

tolability. Moreover, the fact that a large number of dye molecules

constantly undergo excitation–emission cycles to image a compara-

tively small spot adds to the problem. Finally, saturation of the S

will

be possible only if the long-lived triplet state is not allowed to build up

during repeated excitation. As most dyes feature a triple relaxation rate

of >1 µs (that strongly depends on the environment), effective triplet

relaxation requires a pulse repetition rate <500 kHz. Nevertheless, due

to the simplicity of raw data acquisition it may remain an attractive

method for the imaging of very bright and photostable samples.

One possible solution to the quest for large saturation factors at low

intensities should be compounds with two (semi)stable states (Hell

et al., 2003; Dyba and Hell, 2002). If the rate k

(and the spontaneous

rate k

) almost vanish, large saturation factors are attained at very low

intensities. The lowest useful intensity is then determined by the

slowest acceptable imaging speed, which is ultimately determined by

the switching rate. A favorable aspect is that in most bistable com-

pounds the speed of the actual switching mechanism, i.e., of the con-

formational change, is less than a few nanoseconds, which is much

faster than the typical pixel dwell time in scanning. In the ideal case,

the marker indeed is a bistable ﬂ uorescent compound that can be pho-

toswitched at separate wavelengths, from a ﬂ uorescent state A to a dark

state B, and vice versa, where spontaneous rates will not inﬂ uence this

compromise.

Recently, a photoswitchable coupled molecular system, based on a

photochromic diary lethene derivative and a ﬂ uorophore, has been

reported (Irie et al., 2002). Using the kinetic parameters reported, Eq.

(27) predicts that focusing of less than 100 µW of deep-blue “switchoff

light” to an area of 10

−8

for 50 µs should yield better than 5 nm spatial

resolution. Future targeted optimization of photochromic or other com-

pounds to fatigue-free switching and visible light operation could there-

fore open up radically new avenues in microscopy and data storage (Hell

et al., 2003).

For live cell imaging, ﬂ uorescent proteins have many advantages

over synthetic dyes. Many of them feature dark states with light-driven

transitions (Hell, 1997; Hell et al., 2003). If the spontaneous lifetimes of