Biology Reference
In-Depth Information
random fluorophores such that the final superresolution image is built up similarly as
described earlier for STORM.
Over the last few years, invaluable information about the nature of cell membrane
nano- and microdomains has been obtained using localization-based superresolution
techniques. For instance, the organization of the Lck protein into 120 nm membrane
clusters (
Fig. 6.2
) was shown to be dependent on lipid raft integrity and protein con-
formational state, both affecting cell signaling (
Owen et al., 2010; Rossy, Owen,
Williamson, Yang, & Gaus, 2013
). To obtain more quantitative data on cluster or
domain formation, the dataset of single-molecule localizations can be analyzed using
Ripley´s function (
Ripley, 1977
) or pair correlation function (
Sengupta et al., 2011
).
From these algorithms, relevant data such as cluster sizes and the receptor densities
inside those clusters can be obtained.
To reconstruct high-quality superresolution images, a single-molecule-sensitive
camera should detect all emitters labeling a specific structure with sufficient number
of photons to get position accuracies down to tens of nanometers. Moreover, the im-
aging conditions and the reconstruction algorithms are critical to obtain biologically
meaningful data and thus optimized imaging protocols for each subject under inves-
tigation are usually required. A precise quantitative assessment of protein clustering
is still challenging due to the possible multiple localization assignments of a single
molecule that is photoblinking on long timescales, which will therefore erroneously
appear as a cluster. Furthermore, since no true optical imaging is being generated, all
the optical contrast mechanisms associated with light, that is, intensity, polarization,
wavelength, and lifetime, are lost on the reconstructed image.
6.2.3.2
Far-field patterned illumination techniques (STED)
An alternative superresolution method that also relies on the photophysical proper-
ties of fluorophores is STED microscopy (
Klar & Hell, 1999
). Conceptually, the
method proposes to reduce the spatial extent of the excitation light by stimulating
the depletion of the emission on the outer regions of the diffraction-limited excitation
profile. The principle is based on the fact that an excited fluorophore can be stimu-
lated back to the ground state without photon emission. This can be achieved by il-
luminating the fluorophore with red-shifted light after excitation. With enough
power for the depletion beam, it is possible to reduce the probability of fluorescence
significantly. Then, by using a conventional diffraction-limited spot to excite the
fluorophores and overlay this with a donut-shaped deexcitation (or depletion) profile,
one can reshape the region where emission is allowed down to a few nanometers
(
Rittweger, Han, Irvine, Eggeling, & Hell, 2009
). To form a true nanoscopic image,
this effective nanosized spot is raster-scanned through the sample. For STEDmicros-
copy, the resolution,
r
STED
, can be tuned by the intensity of the STED beam,
I
,
through
r
STED
I
/
I
s
)
1/2
, where
r
d
is the size of the diffraction-limited spot
and
I
s
the intensity of the STED beam needed to deplete the fluorescence to 1/e.
However, because of its mere principle, STED requires accurate control of the po-
sition, phase, and amplitude of two laser beams (for single color fluorescence), and
r
d
*(1
þ
