Single-Molecule Localization Super-Resolution
Microscopy: Deeper and Faster
Sébastien Herbert,1,2 Helena Soares,3 Christophe Zimmer,1,* and Ricardo Henriques1,*
1Institut Pasteur, Groupe Imagerie et Modélisation, CNRS URA 2582, 25 rue du Docteur Roux, 75015 Paris, France
2Frontiers in Life Sciences PhD Program, University Paris Diderot, 5 rue Thomas-Mann, 75013 Paris, France
3Institut Pasteur, Lymphocyte Cell Biology Unit, CNRS URA 1961, 28 rue du Docteur Roux, 75015 Paris, France
Abstract: For over a decade fluorescence microscopy has demonstrated the capacity to achieve single-molecule
localization accuracies of a few nanometers, well below the ;200 nm lateral and ;500 nm axial resolution limit
of conventional microscopy. Yet, only the recent development of new fluorescence labeling modalities, the
increase in sensitivity of imaging hardware, and the creation of novel image analysis tools allow for the
emergence of single-molecule-based super-resolution imaging techniques. Novel methods such as photoacti-
vated localization microscopy and stochastic optical reconstruction microscopy can typically reach a tenfold
increase in resolution compared to standard microscopy methods. Their implementation is relatively easy only
requiring minimal changes to a conventional wide-field or total internal reflection fluorescence microscope. The
recent translation of these two methods into commercial imaging systems has made them further accessible to
researchers in biology. However, these methods are still evolving rapidly toward imaging live samples with high
temporal resolution and depth. In this review, we recall the roots of single-molecule localization microscopy,
summarize major recent developments, and offer perspective on potential applications.
Key words: single molecule, super-resolution, fluorescence, microscopy, PALM, STORM
INTRODUCTION
Fluorescence microscopy is one of the main methods for
the study of cell biology. Through fluorescent tagging,
molecules such as DNA, RNA, and protein can be readily
differentiated within their cellular environment and ob-
served in a noninvasive manner. Notwithstanding, optical
diffraction in standard microscopes restricts resolution to
;200 nm laterally and ;500 nm axially ~Abbe, 1882!. In
the last decades, major efforts have been made to improve
both microscopy instrumentation and image analysis in
order to push the boundaries of imaging resolution, depth,
and speed ~Rino et al., 2009!. In recent years, these efforts
gave rise to the field of super-resolution microscopy: a novel
stream of approaches focused on achieving resolutions of
few nanometers ~typically 1–100 nm axially in two or three
dimensions!, while keeping most properties of classical
fluorescence microscopy ~Yildiz et al., 2003; Huang et al.,
2009!. Single-molecule localization microscopy ~SMLM!, a
family of super-resolution methods, relies on the capacity to
discern and pinpoint individual molecules even within
densely labeled environments ~see Figs. 1, 2!, achieving
molecular localization accuracies at the nanometer level in
experimental settings ~Yildiz et al., 2003!. This family of
methods edges toward the ,1 nm resolving power of
electron microscopy ~EM! and atomic force microscopy
~AFM! ~Giessibl, 1995; Erni et al., 2009!, while keeping the
molecular-specific labeling of fluorescence imaging.
Only recently did SMLM transition from a developmen-
tal phase to true biological applications. Several novel stud-
ies have demonstrated the capacity of SMLM to provide a
nanoscale view into ultrastructure not easily differentiated
with classical optical microscopes, such as lysosomes ~Betzig
et al., 2006!, the Golgi apparatus ~Betzig et al., 2006!,
microtubules ~Huang et al., 2008b!, clathrin-coated pits
~Huang et al., 2008b!, viral proteins and structures ~Manley
et al., 2008; Lelek et al., 2012!, bacterial structures ~Biteen
et al., 2008; Greenfield et al., 2009!, bacterial vacuole rup-
ture ~Ehsani et al., 2012!, nuclear pores ~Löschberger et al.,
2012!, and synaptic vesicles ~see Fig. 3!.
Here, we review the evolution of SMLM, from its
inception to its current state, as it progresses toward fast,
live-cell, and deep imaging.
THE RESOLUTION LIMIT
In microscopy, photons follow the optical path from the
light source into the sample and are finally collected by a
detector. Within this route they will interact with different
physical media such as lenses, filters, the sample itself, and
its embedding medium. Each of these factors will contrib-
ute to light scattering and visual blurring of the observed
sample. Additionally, the limited aperture of microscope
objectives will induce for each visible fluorophore a spatial
profile featuring a central spot with a series of concentric
Received June 2, 2012; accepted August 9, 2012
*Corresponding authors: E-mail: ricardoh@pasteur.fr, czimmer@pasteur.fr
Microsc. Microanal. 18, 1419–1429, 2012
doi:10.1017/S1431927612013347
Microscopy AND
Microanalysis
© MICROSCOPY SOCIETY OF AMERICA 2012
REVIEW ARTICLE

----!@#$NewPage!@#$----
rings, a silhouette known as the Airy diffraction pattern,
generally referred to as the point spread function ~PSF, see
Fig. 1!. These effects can be easily interpreted when observ-
ing spatial distinct fluorophores. In the 19th century, E.
Abbe described this effect analytically and defined the reso-
lution limit as the minimal distance allowing for subdiffrac-
tion points to be differentiated. This distance can be
calculated through Abbe’s formulas: l/~2NA! perpendicular
to the optical axis and ~2lh!/~2NA!2 along the optical axis,
where l is the wavelength of the light, h is the refraction
index of the medium, and NA the numerical aperture of the
objective lens in the imaging system ~Abbe, 1882!. In biolog-
ical systems, molecules of a few nanometers in size are
typically organized within molecular aggregates below Abbe’s
limit ~roughly 200 nm!. Standard microscopes are thus
incapable of accurately resolving most structures and inter-
actions at the molecular scale. Electron microscopy and
AFM are able to achieve resolutions under 1 nm ~Ando
et al., 2008; de Jonge et al., 2009!, but they lack the main
advantages of fluorescence optical microscopy such as live-
cell imaging, molecular specific multicolor labeling, and
relatively simple experimental protocols. The ideal micros-
copy technique would combine these features with the
subnanometer resolution of EM and AFM.
Figure 1. Point spread functions and the resolution limit: ~A!
representative PSF profile for a wide-field or TIRF microscope
viewed in the xy-plane ~scale bar � 200 nm! and xz-plane ~scale
bar � 600 nm!; ~B! two particles at resolvable and unresolvable
distances as seen in a microscope, red arrows demonstrate when
the centers of the particles are resolvable, green line demarks the
intensity profile shown in the plots, dotted red and blue lines
correspond to the profile of the upper and lower particles, black
line to the summed profile of both particles.
Figure 2. Acquisition and analysis of a biological structure through single-molecule localization methods. In this family
of methods, the microscope acquires a diffraction-limited sequence of images featuring subsets of temporarily active
and spatially distinct fluorophores, switched-on from a larger population of nonactive fluorophores. Generally, the first
frames show an unwanted base level of naturally active photoswitchable fluorophores, impeding the immediate
discrimination of single molecules ~Acquisition – t0!. The subsequent switch-off ~mainly by bleaching! of these
fluorophores leads to a reduced number of active fluorophores ~Acquisition – t1!. The constant illumination by an
excitation beam in combination with either a pulsed or continuous activation light ~not always required! allows a small
number of active fluorophores in each frame to be maintained ~Acquisition – t1, t2, . . .!. Analysis of these frames permits
accurate fluorophore detection and localization ~Analysis – t1, t2, . . .!. The superposition of the calculated particle
positions will generate a super-resolution reconstruction ~Analysis – Reconstruction!. This example makes use of a
synthetically generated structure for illustration purposes.
1420
Sébastien Herbert et al.

----!@#$NewPage!@#$----
PUSHING THE BOUNDARIES OF
FLUORESCENCE MICROSCOPY
Recent years have witnessed a boom in attempts to push the
current resolution boundaries of fluorescence microscopy,
focused on three main areas: imaging hardware, cell label-
ing, and image analysis. Many super-resolution methods
different from SMLM have been developed. Below, we give a
brief historical overview of their development before focus-
ing in SMLM approaches.
In one of the earliest attempts to achieve super-
resolution microscopy, Cremer and Cremer ~1978! pro-
posed the use of interference to confine the excitation beam
PSF into a single spot smaller than the diffraction limit.
Once combined with point scanning, this method allowed
for super-resolution imaging.
Later in 1995, the development of the 4Pi microscope
allowed for 140 nm axial and 210 nm lateral resolutions to
be achieved. In this setup, two opposing objectives lenses
are used to coherently illuminate and detect the same spot
~Hänninen et al., 1995!.
The first applications of total internal reflection fluores-
cence ~TIRF! appeared in the early 1980s. This method
generates an evanescent light wave that restricts sample
illumination to the first couple of hundred nanometers
directly above the coverslip ~Axelrod, 1981!. TIRF is ideal
for live cell imaging, providing images with a strong signal-
to-noise ratio by constraining the fluorescence excitation
depth in the z-axis. It has proved to be a powerful tool in
the study of cell membrane dynamics, cytoskeleton, or
cell-substrate contact regions ~Hu et al., 2007; Mattheyses
et al., 2010!. Initially TIRF microscopy required the use of
prisms, to induce an evanescence illumination field directly
over the coverslip, and an objective opposed to the prism, in
an upright-type microscope setup. The increase in the max-
imum NA ~1.4 to 1.65! of objectives has allowed for the
evanescence field to be generated directly by the objective
lens and thus facilitates the imaging of biological cells.
However, TIRF is restricted to a limited penetration depth
close to the surface of the samples ~depth , 200 nm!.
Structured illumination microscopy ~SIM! is a more
recent super-resolution approach. In this method, the illu-
mination beam is shaped by a diffraction grating to create a
sinusoidal-type illumination pattern directly in the sample.
A series of images is then acquired while varying the
position and rotation of this pattern. Computational recon-
struction methods then yield a super-resolution image by
exploring the additional frequency information extracted
from the acquired images. This method virtually doubles
the NA of the system, hence doubling the lateral resolution
to around 100 nm ~Gustafsson, 2000!. SIM relies only on
optics and computational processing; the sample prepara-
tion and labeling are exactly the same as in standard
fluorescence microscopy ~Kner et al., 2009!.
Another super-resolution method, stimulated emission
depletion ~STED!, uses two independent lasers to sharpen
and confine the effective excitation pattern to a subdiffrac-
tion area. To this end, a first focused laser excites the
fluorophores of a diffraction limited area in the sample, and
a second laser, engineered into a doughnut-shaped intensity
profile, forces the excited fluorophores at the periphery of
the diffraction limited spot back into the ground state. Only
the fluorophores at the center of the doughnut will be able
to emit light and be detected ~Hell & Wichmann, 1994; Klar
et al., 2000!. STED has been demonstrated to achieve 30–
40 nm in lateral and axial resolution within biological
samples ~Schmidt et al., 2008!.
The super-resolution optical fluctuation imaging ~SOFI!
method is able to analytically improve the optical resolution
of a dataset by analyzing the temporal uncorrelated fluctua-
tion in the emission of fluorophores ~Dertinger et al., 2009!.
Although SOFI in biological samples is generally limited to
100 nm lateral resolution, it has been demonstrated to work
with both conventional epifluorescence or TIRF microscopes.
High speed imaging with standard fluorophores has been
demonstrated in live cell imaging ~Dedecker et al., 2012!.
SMLM approaches have been developed in parallel to
these approaches. In many cases, SMLM can be used as an
alternative or a complement to the mentioned super-
resolution methods. For the remaining part of the review,
we will focus on SMLM.
BYPASSING THE RESOLUTION LIMIT WITH
SINGLE-PARTICLE LOCALIZATION
In fluorescence microscopy, an individual subdiffraction
particle can be localized with accuracy below the diffraction
Figure 3. 3D direct STORM ~dSTORM! image of fixed hippocam-
pal neurons stained with antibodies against synaptic markers:
~A! dSTORM image of an isolated axonal track ~gray: VGlut1, a
synaptic vesicle reporter; green: bassoon, a presynaptic site re-
porter!; ~B! super-resolution z-depth reconstruction of an individ-
ual synaptic vesicle from blue inset of image A; ~C! particle
detection profile from demarcated red line in images B, shows
;20 nm resolution. Cell preparation and labeling as described in
Henriques et al. ~2010!.
Single-Molecule Super-Resolution
1421

----!@#$NewPage!@#$----
limit if its spatial profile does not overlap with that of other
imaged objects ~Thompson et al., 2002! ~see Fig. 1!. In this
case, the center of the PSF—corresponding to the fluoro-
phore location—can be computed with an accuracy that is
limited mostly by the detected fluorophore signal relative to
the noise ~Gelles et al., 1988; Thompson et al., 2002; Ober
et al., 2004! ~see Fig. 1!.
Several analytical methods have been developed for
locating the center of a particle with subdiffraction accuracy
~Crocker & Grier, 1996; Thompson et al., 2002; Rogers et al.,
2007; Chao et al., 2009a!. Although the calculation of
two-dimensional ~2D! coordinates ~x, y! requires a single
image with the visible particle ~Ober et al., 2004!, the
determination of the z coordinate with a low localization
error demands an optical alteration of the PSF shape to
encode the position of the particle in the z-axis ~Kao &
Verkman, 1994! or multiple-focal planes to be acquired
~Juette et al., 2008; Chao et al., 2009b! ~detailed later in the
High-Resolution Single-Molecule Localization in 3D sec-
tion of this review!. Various different algorithms can be
employed for single-particle localization ~Cheezum et al.,
2001!, some of the most accurate 2D and three-dimensional
~3D! particle localization methods involve fitting the parti-
cle intensity profile to a Gaussian function approximation
of the PSF. High-precision fitting near the theoretical reso-
lution limit can be achieved through nonlinear least-squares
optimization ~Cheezum et al., 2001!, or even more accu-
rately through a maximum-likelihood estimator ~Abraham
et al., 2009!. Notwithstanding, fitting does entail a consid-
erable computational cost due to the multiple iterative
cycles needed for the position of the particles to be re-
trieved. Constrained center-of-mass ~Henriques et al., 2010!
and radial symmetry ~Ma et al., 2012; Parthasarathy, 2012!
have also been used for high-speed calculation of the
particle center with reduced computational burden when
compared to fitting methods, although at the cost of a
minor increase in the localization error.
High photon output of fluorophores is critical to
achieve localization accuracy bellow the diffraction limit
~Thompson et al., 2002!. Synthetic fluorescent dyes tend to
emit a considerably higher number of photons ~typically
tenfold more! than genetically encoded fluorophores ~Hen-
riques & Mhlanga, 2009!, allowing better localization accu-
racy. However, the use of large linkers such as antibodies to
tag the molecules of interest leads to an ;10 nm spacing
relative to the fluorophore ~Jares-Erijman & Jovin, 2003;
Ries et al., 2012!, thus incorporating an additional error to
localization of the molecule of interest. However, a solution
is provided by the use of small linkers to label proteins of
interest with dyes, such as nanobodies ~Ries et al., 2012!, the
fluorescein arsenical helix binder ~FLAsH! ~Lelek et al.,
2012!, or small peptide tags ~Klein et al., 2010; Wombacher
et al., 2010!.
Other factors influencing resolution include fluoro-
phore density, optical system stability, sample drift, and the
type of fluorophore localization. Due to these features, the
localization accuracies can vary considerably, typically rang-
ing between 1 to 40 nm ~Yildiz et al., 2003; Pertsinidis et al.,
2010; Schermelleh et al., 2010!.
The principle of subdiffraction accuracy single-particle
localization has been the foundation for seminal studies
including the work of Gelles et al. ~1988! in the movement
of kinesin-coated beads and of Yildiz et al. ~2003! on the
movement of Myosin V along actin filaments.
DETECTING INDIVIDUAL MOLECULES
WITHIN THE ENSEMBLE
To minimize the visible fluorophores spatial overlap re-
quired for high accuracy particle localization, the aforemen-
tioned methods are limited to single-molecule imaging of a
very low density of emitting fluorophores. In general, how-
ever, molecules to be imaged are present at unknown densi-
ties, typically too high to be resolved by standard microscopy.
Accurate single-molecule localization is impossible when
closely packed fluorophores emit in the same spectral range
simultaneously ~Fig. 1!; however, they could be individually
resolved if made to emit in different colors ~Betzig, 1995! or
at different points in time ~Lidke et al., 2005; Betzig et al.,
2006; Hess et al., 2006; Rust et al., 2006!. The concept of
high-accuracy localization of neighboring fluorophores with
distinct colors was experimentally demonstrated in 1998
~Van Oijen et al., 1998; Lacoste et al., 2000!. In 2004, the
principle of differentiating and localizing sequentially disap-
pearing fluorophores ~“bleached”! from crowded regions
was introduced by subtracting consecutive images in a
time-lapse sequence ~Gordon et al., 2004; Qu et al., 2004!.
Later in 2005, Lidke et al. used the stochastic blinking of
quantum dots as a method to transiently distinguish and
localize individual fluorophores in space and time ~Lidke
et al., 2005!. It was in 2006 that the independent work of
Betzig et al. ~2006! in photoactivated localization micros-
copy ~PALM!, Hess et al. ~2006! in fluorescence PALM
~FPALM!, and Rust et al. ~2006! in stochastic optical recon-
struction microscopy ~STORM! led to a breakthrough in
single-molecule super-resolution methodology. These groups
demonstrated that the use of photoswitchable fluorophores
allows control of the number of actively emitting fluoro-
phores in each acquired image; therefore, it is possible to
minimize the probability that in any given image two or
more fluorescent molecules spatially overlap. Photoswitch-
able fluorophores are able to reversibly or irreversibly switch
between two or more spectrally distinct states. The change
between states is generally induced by light and can be
partially controlled via the microscope sample illumination.
PALM, FPALM, and STORM require the majority of photo-
switchable fluorophores to stay in a nonvisible state. This is
generally achieved by a high-intensity excitation of the
visualized fluorescent state, forcing fluorophores to be revers-
ibly or irreversibly bleached. A small fraction of these fluo-
rophores can then be transiently switched to a visible state
by either naturally recovering from reversible photobleach-
ing or by photoactivation with a low-intensity light of
compatible wavelength. By acquiring a sequence of images
1422
Sébastien Herbert et al.

----!@#$NewPage!@#$----
while the photoactivation of subsets of fluorophores occur,
each frame will capture a small number of emitting individ-
ual fluorophores compatible with single-molecule localiza-
tion ~see Figs. 1, 2!. A super-resolution representation of the
dataset can then be recreated by plotting the positions of
the molecules localized in each frame of the sequence of
images ~see Fig. 2!. As each frame needs to have a suffi-
ciently small number of emitting fluorophores to keep the
overlapping probability low, hundreds to thousands of im-
ages commonly need to be acquired to accumulate enough
particle detections to accurately represent molecular and
cellular structures. The final resolution of the reconstruc-
tion depends on the localization accuracy of each individual
fluorophore, the density of localized molecules, and the
capacity to correct for imaging artifacts such as unwanted
motion ~drift! of the sample during the acquisition. Typi-
cally, SMLM approaches based on PALM and STORM will
lead to a 10–20-fold increase in resolution when compared
to the classical resolution limit ~;200 nm!, as seen in the
example of Figure 3.
PALM and STORM rely on photoswitchable properties
inherent or conferred to fluorophores. Photoswitchable
fluorophores can stochastically switch to an excitable state
when illuminated by activating light @e.g., near-ultraviolet
~UV! light applied to photoactivatable green fluorescent
protein ~PA-GFP!# and to disable it through high-powered
excitation light ~e.g., a strong irradiation of 488 nm laser
will switch-off the PA-GFP fluorescence, typically by irrevers-
ibly bleaching the molecule!. Different SMLM methods
rely on different types of photoswitchable fluorophores.
PALM-based approaches tend to use genetically encoded
photoswitchable fluorescent proteins ~e.g., PA-GFP, mEos2!,
while STORM-based methods tend to rely on the inducible
photoswitching properties of synthetic dyes when coupled
in certain dye pairs ~e.g., Cy3-Cy5!, where the excitation of
one dye ~Cy3! will induce the photoactivation of the
complementary fluorophore ~Cy5!. The direct-STORM
~dSTORM! approach further demonstrated the possibility
of inducible photoswitching properties in simple classical
fluorophores ~e.g., Cy5! without the requirement of a pair
of fluorophores ~Heilemann et al., 2009!. Several studies
have shown that special embedding media prevent the
irreversible photobleaching of fluorophores, granting photo-
switchable properties to fluorophores ~Folling et al., 2008;
Steinhauer et al., 2008; Vogelsang et al., 2008; Heilemann
et al., 2009!.
HIGH-RESOLUTION SINGLE-MOLECULE
LOCALIZATION IN 3D
Even though SMLM approaches allow localization of indi-
vidual molecules with subdiffraction lateral accuracy in 2D
images, the axial position of these particles needs to be
obtained through additional methods. Several approaches
have attempted to encode the axial position of fluorophores
into their observable 2D spatial profile ~see Fig. 4 and
Table 1!. These include astigmatism ~Huang et al., 2008a!,
BiPlane detection ~Juette et al., 2008!, double-helix PSF
formation ~Pavani et al., 2009!, and dual-objective interfer-
ometry ~Shtengel et al., 2009; Aquino et al., 2011!.
Astigmatism relies on the introduction of a cylindrical
lens in the emission optical path, as a result of which a
fluorescent particle will appear as a fluorescent spot elon-
gated along either the X or the Y direction depending on its
axial coordinate Z. The Z coordinate is thus encoded in the
ellipticity of each spot and can be computationally retrieved
~Huang et al., 2008a!. Huang et al. ~2008a! applied this
method to STORM microscopy, resolving the morphology
of mitochondria and mitochondria-microtubule contacts.
Combining astigmatism with multiple acquisitions at differ-
ent z-depths, they imaged entire cells in 3D ~typically 50 mm
wide and 3 mm thick! with a resolution up to 25 nm in xy
and 67 nm in z. A recent study ~Xu et al., 2012! combined
astigmatism with a dual-objective scheme. This system has
achieved resolutions up to ;10 nm in lateral direction and
Figure 4. Principles of 3D single-molecule super-resolution meth-
ods: ~A! retrieving the z-position of molecules from their observ-
able spatial-profile—astigmatism changes the fluorescent spots
ellipticity, DH-PSF converts the normal spatial profile into two
rotating lobes and BiPlane simultaneously acquires images from
different focal planes—z-coordinates for each spot can be esti-
mated through the analysis of their shape; ~B! relation between 3D
methods applied to single-molecule super-resolution, the axial
thickness of the optical slices, and the possible complementary
techniques for super-resolution z localization.
Single-Molecule Super-Resolution
1423

----!@#$NewPage!@#$----
;20 nm in axial direction and allowed resolution of indi-
vidual actin filaments. This boost in resolution was sup-
ported by the near doubling of photon collection through
the two opposed objectives, improved objective stabilization
and reduced noise sensitivity thanks to the redundancy of
the images obtained through the two objectives.
In contrast, in BiPlane FPALM ~BP FPALM! two detec-
tion planes are imaged at slightly different axial positions
simultaneously ~;700 nm!. Particle z coordinates are re-
trieved by analyzing the changes in their PSF profile within
the two focal planes. Authors reached 30 nm lateral and
75 nm axial resolutions while imaging 4 mm diameter beads
~Juette et al., 2008!. BiPlane has the benefit of keeping a
relatively stable axial resolution in Z, while in astigmatism
the resolution improvement diminishes with distance to
the focal plane. Both astigmatism and BiPlane are remark-
ably efficient at retrieving the axial subdiffraction posi-
tions of particles within a z-range of ;1 mm, without the
need of changing the focal position of the objective. Con-
sequently, no axial scanning is necessary to cover this
volume within a sample, thus allowing for fast 3D acquisi-
tion. Both methods also can be implemented with minor
modifications to either a TIRF or wide-field microscopy
apparatus.
Double-helix point spread function ~DH-PSF! stands
as a third and alternative method providing axial subdiffrac-
tion resolution. Using a spatial light modulator, this tech-
nique reshapes the PSF of the emitted light into a double
helix pattern, leading to two visible lobes in a 2D image.
These two lobes are rotated around their midpoint by an
angle that depends on the particle Z coordinate. Particles
can then be accurately localized axially by measuring this
angle, while the midpoint between the centroids of each
lobe provides the lateral position ~Pavani et al., 2009!. This
method has experimentally achieved 30 nm lateral and
50 nm axial resolution. While these resolutions are very
similar to the other methods, the z-range of DH-PSF is
slightly higher reaching ;2 mm.
Axial interference has been used to improve localiza-
tion in the z-axis by an approach featuring two opposing
objectives referred to as “interferometric PALM” ~iPALM!
~Shtengel et al., 2009!. This method achieves 3D super-
resolution localization of fluorophores in parallel to optical
sectioning. In iPALM, the fluorophore emission signal is
made to interference with itself, thereby encoding the
z-position of the emitter through an envelope of oscillations
of the resulting signal. In this way, authors correlate the
axial depth of the emitter and the relative intensity detected
by each camera to arrive at particle localization with a
resolution up to 10 nm axially and 23 nm laterally within a
250 nm thickness layer. However, this system suffers from
a limited working range problem and periodic artifacts ~Xu
et al., 2012!. Recently, the axial working range problem was
improved by combining 4pi with PALM microscopy ~Aquino
et al., 2011!. Here, a resolution of ;8–22 nm laterally and
;6 nm axially was achieved in an enlarged optical layer of
;650 nm around the focal plane.
Table 1.
Comparison of SMLM Methods.*
Method
Basis
Live Cell
XY Resolution
Z Resolution
Depth
Super-Resolution
Volume Acquisition
Time
Frame Rate
Multicolor
References
Astigmatism
STORM
Astigmatism z localization
No
;20–30 nm
;60–70 nm
;3 mm
ND
;20 Hz
Yes
Huang et al., Nat Methods ~2008b!
BP-FPALM
Biplane z localization
No
;30 nm
;75 nm
;9 mm
;1–10 min
;100 Hz
No
Juette et al., Nat Methods ~2008!
DH-PSF
Double-helix z localization
No
;10 nm
;20 nm
;2 mm
;1–10 min
;2 Hz
No
Pavani et al., PNAS ~2009!
iPALM
Interferometry of the emitted
signal
No
;8–22 nm
;6 nm
ND
;5–30 min
;100 Hz
Yes
Aquino et al., Nat Methods ~2011!
Dual-objective
STORM
Astigmatism z localization � dual
objective setup
No
;10 nm
;20 nm
;2 mm
;30–60 min
;60 Hz
No
Xu et al., Nat Methods ~2012!
IML-SPIM
SPIM optical sectioning � astigmatism
z localization
Yes
;63 nm
;140 nm
;100–150 mm
;2–3 min
;25–33 Hz
No
Zanacchi et al., Nat Methods ~2011!
3B
Bayesian localization algorithm
Yes
;50 nm
2D only
ND
;0.5–10 s
;50 Hz
No
Cox et al., Nat Methods ~2011!
CS-STORM
Compressed sensing algorithm
Yes
;60 nm
2D only
ND
;3 s
;60 Hz
No
Zhu et al., Nat Methods ~2012!
*Although the values shown are based on information contained in the corresponding publications, they should be perceived as rough estimations that can vary depending on the sample properties and the
imaging apparatus.
1424
Sébastien Herbert et al.

----!@#$NewPage!@#$----
CONSTRAINING ILLUMINATION IN
3D SPACE
The combination of SMLM with TIRF has been widely used
to constrain the z localization of fluorescing molecules
to an axial region of space below the diffraction limit
~,200 nm! ~Betzig et al., 2006; Rust et al., 2006!. Yet, TIRF is
restricted to imaging the sample surface directly over the
coverslip, preventing any deeper objects from being imaged.
In contrast, wide-field-based illumination of the sample per-
mits deeper imaging but at the cost of a considerable excita-
tion of fluorophores below and above the focal plane, thus
increasing unwanted fluorescence background and limiting
resolution.Alternative sample illumination methods compat-
ible with 3D SMLM have been recently developed. These
aim to excite only a small in-focus volume ~see Fig. 4 and
Table 1!, improving the signal-to-noise ratio, while both de-
creasing sample photodamage and phototoxicity in the case
of live cells ~see the Live-Cell Imaging section below!.
Highly inclined and laminated optical sheet ~HILO! is a
z-sectioning approach that can be performed with minimal
adaptations to a microscope capable of TIRF. In this method,
the illumination beam is moved to the lateral periphery of
the back focal plane of the objective lens just before induc-
ing a TIRF evanescent field. The resulting beam out of the
objective is at a subcritical angle with the coverslip. The
large refraction angle occurring at the surface of the cover-
slip shapes the excitation light into a highly inclined and
laminated thin optical sheet incident within the sample.
The thickness of the sheet can be changed with the diameter
of the illumination beam and has been shown to vary
between 5 and 10 mm ~Tokunaga et al., 2008!. Baddeley
et al. ~2011! applied with success HILO illumination for
super-resolution imaging.
Vaziri et al. ~2008! used two-photon temporal focusing
to confine activated fluorophores to a single optical slice
roughly 2 mm deep. This activation step is followed by
standard PALM imaging using wide-field illumination of
the entire sample. Only the photoactivated fluorophores in
the two-photon activation plane will emit fluorescence,
safeguarding nonactivated fluorophores in the out-of-focus
region from bleaching. This method has been successfully
applied and improved by York et al. ~2011! to image the
mitochondrial network within cells at a 50 nm lateral and
100 nm axial resolution, up to 8 mm deep in the sample
Similarly, a remarkable combination of selective-plane
illumination microscopy ~SPIM! and FPALM, named indi-
vidual molecule localization–selective plane illumination
Microscopy ~IML-SPIM!, has been demonstrated by Zanac-
chi et al. ~2011!. Using a first laser, a single thin layer of
photoactivatable fluorophores is activated ~ranging between
2–4 mm!; this region of space is then excited by a second
laser, allowing the emitted fluorescence to be acquired with
a second objective aligned perpendicularly to the excited
plane. This study demonstrates a lateral and axial resolution
of 63 and 140 nm, respectively, 150 mm deep into the
sample, without significant influence of scattering and opti-
cal aberrations. As stated by the authors, this method could
be further improved by using two-photon excitation for
selective plane illumination ~Truong et al., 2011!.
SPIM, two-photon activation, and HILO all provide
z-sectioning capabilities to SMLM approaches. Neverthe-
less, these methods alone cannot constrain the sample illu-
mination or photoactivation to subdiffraction regions, as
the induced optical slices are thicker than 1 mm ~Tokunaga
et al., 2008; York et al., 2011; Zanacchi et al., 2011!. Thus,
other methods such as those mentioned in the previous
section must be combined with these approaches to obtain
super-resolution along the axial direction deep in the sample.
LIVE-CELL IMAGING
Live-cell imaging becomes difficult when dealing with meth-
ods such as PALM and STORM, where a high number of
fluorophores need to be collected in a sufficiently fast man-
ner to correctly represent the structures of interest within a
time interval ~Henriques et al., 2011!.
Over the last years, dynamic live-cell imaging has been
one of the main points of focus for developments in single
molecule super-resolution microscopy. One of the key diffi-
culties is motion blur, which can arise if the molecules move
over distances larger than the spatial resolution during the
time taken to acquire sufficient frames for a meaningful
super-resolution reconstruction ~Shroff et al., 2008!.
Shroff et al. ~2008! argued that according to the Nyquist
criterion the final reconstruction resolution will be limited
to 2/j 1/D, where j is the spatial density of localized mol-
ecules and D the number of spatial imaging dimension
~typically D � 2 or 3!. To achieve a significant resolution,
the density needs to be high, which requires the acquisition
of a large number of frames. In contrast, the amount of
frames needs to be sufficiently low to represent a single time
point with minimal cell motion artifacts. This problem can
be partially solved by streaming images at high frequencies,
but this possibly leads to smaller signal-to-noise ratios of
the detected particles, unless higher laser illumination power
is used, which can however affect cell viability.
To attain high molecular localization densities within
short time periods, three key factors come into play: ~1! sen-
sitivity of the imaging hardware’s and its capacity to reduce
unwanted background, such as through 3D sectioning ~see
the previous section!; ~2! the algorithmic capacity to local-
ize and distinguish a large number of particles per frame
~see the next section!; ~3! fluorophore photon output and
contrast ratios ~note that some photoswitchable fluoro-
phores maintain residual fluorescence when in an off-
state! ~Henriques & Mhlanga, 2009!. The choice of
high-performance fluorophores ~point 3! becomes critical
when planning live-cell experiments. Although most fluo-
rescent photoswitchable proteins have high contrast ratios,
they supply relatively small photon outputs when compared
to fluorescent dyes. For example, a single Cy5 dye emits
;6,000 fluorescence photons compared to ;500 photons
for mEos2 ~Lukyanov et al., 2005; Bates et al., 2007;
Single-Molecule Super-Resolution
1425

----!@#$NewPage!@#$----
Henriques & Mhlanga, 2009!. For this reason, synthetic dyes
are attractive candidates as super-resolution labels. How-
ever, antibody labeling of live cells is restricted to the
membrane surface due to cell impermeability. Early live-cell
single-molecule super-resolution studies mostly relied on
genetically encoded photoswitchable fluorophores, allowing
resolutions of 40 to 70 nm in 2D within 30–60 s time
frames ~Biteen et al., 2008; Shroff et al., 2008!. Currently, a
major effort is being focused on exploring live-cell friendly
dye-labeling approaches such as by directly linking dyes to
proteins ~Jones et al., 2011! and small peptides sequences
~Klein et al., 2010; Wombacher et al., 2010!. For example,
the small FLAsH dye has been used to image the integrase
enzyme of HIV with 30 nm resolution, resolving morpho-
logical traits of the virus and applied to live-cell super-
resolution imaging ~Lelek et al., 2012!. In another instance,
anti-GFP nanobodies ~small, high-affinity antibodies! were
used to tag GFP proteins expressed within live hippocampal
neurons, achieving high density particle tracking while
minimizing the localization error ~;10 nm! stemming from
the larger standard antibodies ~Ries et al., 2012!. Remark-
ably, Jones et al. demonstrated live-cell imaging of clathrin
coated pits and transferrin with 1–2 s time resolution in 3D
space featuring 30 nm lateral and 50 nm axial resolution, by
either direct coupling of fluorophores to proteins or via
SNAP-tags ~Gautier et al., 2008! coupled to photoswitchable
cyanide dyes ~Jones et al., 2011!. Manley et al. ~2008!
showed the possibility of combining particle tracking with
stochastic photoswitching. Their technique named single-
particle tracking PALM ~sptPALM! is based on the low
excitation of photoactivated fluorophores, allowing them to
fluoresce for a few consecutive frames until bleached while
keeping a low amount of active fluorophores in each frame.
It becomes possible then to analyze the short trajectories of
each particle during its activated state. Based on this ap-
proach Manley et al. ~2008! acquired more than a thousand
trajectories in a single cell at 20 frames per second allowing
to combine spatial and dynamic information of protein
clustering in the cell membrane.
Obviously, an important aspect in live-cell imaging is
cell survival. In live-cell SMLM experiments, two main fac-
tors can promote cell death or abnormal behavior: ~1! chem-
ical toxicity of buffers that enhance photoswitching and ~2!
photodamage caused by sample illumination. Genetically en-
coded photoswitchable proteins do not generally require such
buffers and can thus be directly used in physiological sample
embedding settings. One exception is the work of Matsuda
et al. ~2010!, where standard GFP was shown to photocon-
vert from green to red when stimulated by blue light in the
presence of reduced riboflavin. Although this buffer allowed
successful use of GFP for PALM on fixed cells, live-cell imag-
ing was not possible due to the impermeability of cell mem-
branes to riboflavin and the need for anaerobic conditions.
SMLM using synthetic dyes will generally demand special
imaging buffers to induce photoswitching of synthetic fluo-
rescent dyes. These can have different degrees of chemical
toxicity, as recently compared in a review from Henriques
et al. ~2011!. SMLM methods can also cause considerable
phototoxicity through extensive use of high intensity illumi-
nation,normally required to maintain fast fluorophore photo-
switching cycles. This becomes even more critical when
near-UV light is needed to induce fluorophore photoactiva-
tion, as for the majority of photoswitchable genetically en-
coded protein ~Henriques & Mhlanga, 2009!. Therefore,
techniques that constrain light illumination in 3D space such
as TIRF ~see the previous section! or diminish the require-
ment for strong illumination intensities—for example, the
3B method ~Cox et al., 2011! described in the next section—
are extremely advantageous.
SPEEDING UP RECONSTRUCTION AND
IMAGING
One important aspect of single-molecule localization micros-
copy is the computational burden of analyzing thousands to
millions of detectable fluorophores in a timely manner. One
initial focus in the field was to develop highly efficient
algorithms achieving real-time analysis of PALM or
STORM data concurrent to the acquisition itself ~Hedde
et al., 2009; Henriques et al., 2010; Smith et al., 2010; Wolter
et al., 2010!. Thanks to such software, researchers are able to
generate super-resolution reconstructions of the sample
being imaged while images are streamed from the camera.
This feature allows optimal acquisition settings based on the
reconstruction—for example, adjusting the illumination
intensity to maintain a constant number of molecule detec-
tions over time.
When imaging dynamic structures in live cells, the bal-
ance between the number of frames needed to generate a
super-resolution time point and the correspondent acquisi-
tion time is crucial to resolve cellular ultrastructure mini-
mally corrupted by sample movement. The initial requisite
that individual fluorescent spots do not overlap, which se-
verely restricts the number of visible particles in each frame,
imposes a crucial limiting factor for the imaging speed. A
series of new algorithms have challenged this requisite and
have enabled a steep reduction of the number of images
needed to generate a super-resolution reconstruction. These
new algorithms allow simultaneous detection of multiple
overlapping spots from nearby fluorophores, while distin-
guishing them from high and complex backgrounds. One
class of recently proposed methods includes fitting paramet-
ric models consisting of multiple, rather than a single, PSF
functions ~Holden et al., 2011; Huang et al., 2011! and
sparcity-based reconstruction ~CS-STORM! ~Zhu et al.,
2012!. Both methods enable up to a tenfold increase in
molecular densities compared to standard PALM and STORM
particle detection methods ~Henriques et al., 2010!. CS-
STORM is able to achieve slightly higher particle density for
each frame when compared to multiple PSF fitting but at the
expense of a minor loss in resolution ~;20%!. Yet, these ap-
proaches are still highly sensitive to the signal-to-noise ratio.
Another class of reconstruction methods includes
bleaching/blinking assisted localization microscopy ~BaLM!
1426
Sébastien Herbert et al.

----!@#$NewPage!@#$----
~Burnette et al., 2011!, generalized single molecule high-
resolution imaging with photobleaching”~gSHRImP! ~Simon-
son et al., 2011! and Bayesian localization microscopy ~3B!
~Cox et al., 2011!. Unlike the algorithms above, these meth-
ods use the temporal fluctuations of fluorescence emission
~caused by photobleaching or photoswitching! to pinpoint
each fluorophore ~in the case of BaLM and gSHRImP! or to
generate a probability density map of fluorophore positions
~in 3B!. These approaches have been shown to work with
classical fluorophores, thus removing the need for special-
ized photoswitchable fluorophores, and even in some cases
dispensing of laser activation. Notably, 3B has been recently
demonstrated to achieve fast ~4s per image! live-cell imag-
ing of mCherry labeled cells, with 50 nm resolution using a
standard microscope with a xenon arc lamp.
THE FUTURE OF SINGLE-MOLECULE
SUPER-RESOLUTION
SMLM faces today a large number of developments in
photoswitchable markers, imaging speed, analysis time, 3D
molecular localization, and imaging depth. These improve-
ments should soon be further expanded through integra-
tion with each other. For example, the 3B method could be
further enhanced through optical sectioning, allowing for
the analysis of thicker samples.
Due to the fast pace of improvements in SMLM, it is
challenging for nonexperts to keep up with the state-of-the-
art of the field. However, the recent release of commercial
system dedicated to SMLM by major microscopy compa-
nies such as Leica, Carl Zeiss, and Nikon already provide
biologists access to base tools for super-resolution micros-
copy. We also foresee these companies updating their sys-
tems to support major recent SMLM developments, although
with some delay.
Super-resolution microscopy is now ready to provide
researchers with a novel view of nanoscale structures not
visible through standard optical microscopy. This thriving
field unlocks the possibility to study cellular structures
below the diffraction limit; for example, viral particles,
membrane receptors, protein aggregate formation, protein-
protein interactions, and many others. EM has been their
default imaging method in the past and still provides a
resolution level not yet achievable through SMLM. How-
ever, SMLM caries many nice features not possible through
EM such as simple sample preparation, molecular specific
labeling, and live-cell and deep tissue imaging.
The further combination of SMLM methods and stan-
dard microscopy modalities ~such as wide-field or confocal!
can provide users with a method to observe cells in a fast
manner and then “zoom-in” to cellular structures of interest
through super-resolution.
ACKNOWLEDGMENTS
We thank E. Fabre and M. Lelek at Institut Pasteur - Paris
for advice in the preparation of the manuscript. We would
also like to thank E.F. Fornasiero ~San Raffaele Scientific
Institute and Vita-Salute University, Milan, Italy! for the
labeling and preparation of cells shown in Figure 3. This
work was funded by Institut Pasteur and the Fondation
pour la Recherche Médicale. H.S. additionally funded by an
EMBO long-term fellowship and by a “financement jeune
chercheur” from Sidaction.
REFERENCES
Abbe, E. ~1882!. The relations of aperture and power in the
microscope. J Roy Micro Soc 2, 300–309.
Abraham, A.V., Ram, S., Chao, J., Ward, E. & Ober, R.J. ~2009!.
Quantitative study of single molecule location estimation tech-
niques. Opt Express 17, 23352.
Ando, T., Uchihashi, T., Kodera, N., Yamamoto, D., Miyagi, A.,
Taniguchi, M. & Yamashita, H. ~2008!. High-speed AFM and
nano-visualization of biomolecular processes. Eur J Physiol
456, 211–225.
Aquino, D., Schönle, A., Geisler, C., Middendorff, C.V., Wurm,
C.A., Okamura, Y., Lang, T., Hell, S.W. & Egner, A. ~2011!.
Two-color nanoscopy of three-dimensional volumes by 4Pi
detection of stochastically switched fluorophores. Nat Methods
8, 353–359.
Axelrod, D. ~1981!. Cell-substrate contacts illuminated by total
internal reflection fluorescence. J Cell Biol 89, 141–145.
Baddeley, D., Crossman, D., Rossberger, S., Cheyne, J.E., Mont-
gomery, J.M., Jayasinghe, I.D., Cremer, C., Cannell, M.B. &
Soeller, C. ~2011!. 4D super-resolution microscopy with con-
ventional fluorophores and single wavelength excitation in
optically thick cells and tissues. PLoS One 6, e20645.
Bates, M., Huang, B., Dempsey, G.T. & Zhuang, X. ~2007!.
Multicolor super-resolution imaging with photo-switchable flu-
orescent probes. Science 317, 1749–1753.
Betzig, E. ~1995!. Proposed method for molecular optical imag-
ing. Opt Lett 20, 237–239.
Betzig, E., Patterson, G.H., Sougrat, R., Lindwasser, O.W.,
Olenych, S., Bonifacino, J.S., Davidson, M.W., Lippincott-
Schwartz, J. & Hess, H.F. ~2006!. Imaging intracellular fluo-
rescent proteins at nanometer resolution. Science 313, 1642–1645.
Biteen, J.S., Thompson, M.A., Tselentis, N.K., Bowman, G.R.,
Shapiro, L. & Moerner, W.E. ~2008!. Super-resolution imag-
ing in live Caulobacter crescentus cells using photoswitchable
EYFP. Nat Methods 5, 947–949.
Burnette, D.T., Sengupta, P., Dai, Y., Lippincott-Schwartz, J.
& Kachar, B. ~2011!. Bleaching/blinking assisted localization
microscopy for superresolution imaging using standard fluores-
cent molecules. Proc Natl Acad Sci USA 108, 21081–21086.
Chao, J., Ram, S., Abraham, A.V., Sally Ward, E. & Ober, R.J.
~2009a!. A resolution measure for three-dimensional micros-
copy. Optics Comm 282, 1751–1761.
Chao, J., Ram, S., Ward, E.S. & Ober, R.J. ~2009b!. A comparative
study of high resolution microscopy imaging modalities using
a three-dimensional resolution measure. Opt Express 17,
24377–24402.
Cheezum, M.K., Walker, W.F. & Guilford, W.H. ~2001!. Quanti-
tative comparison of algorithms for tracking single fluorescent
particles. Biophys J 81, 2378–2388.
Cox, S., Rosten, E., Monypenny, J., Jovanovic-Talisman, T.,
Burnette, D.T., Lippincott-Schwartz, J., Jones, G.E. &
Heintzmann, R. ~2011!. Bayesian localization microscopy re-
veals nanoscale podosome dynamics. Nat Methods 9, 195–200.
Single-Molecule Super-Resolution
1427

----!@#$NewPage!@#$----
Cremer, C. & Cremer, T. ~1978!. Considerations on a laser-
scanning-microscope with high-resolution and depth of field.
Microsc Acta 81, 31–44.
Crocker, J.C. & Grier, D.G. ~1996!. Methods of digital video
microscopy for colloidal studies. J Colloid Interf Sci 179, 298–310.
Dedecker, P., Mo, G.C.H., Dertinger, T. & Zhang, J. ~2012!.
Widely accessible method for superresolution fluorescence imag-
ing of living systems. Proc Natl Acad Sci USA 109, 10909–10914.
de Jonge, N., Peckys, D.B., Kremers, G.J. & Piston, D.W. ~2009!.
Electron microscopy of whole cells in liquid with nanometer
resolution. Proc Natl Acad Sci USA 106, 2159–2164.
Dertinger, T., Colyer, R., Iyer, G., Weiss, S. & Enderlein,
J. ~2009!. Fast, background-free, 3D super-resolution optical
fluctuation imaging ~SOFI!. Proc Natl Acad Sci USA 106,
22287–22292.
Ehsani, S., Santos, J.C., Rodrigues, C.D., Henriques, R., Au-
dry, L., Zimmer, C., Sansonetti, P., Tran Van Nhieu, G. &
Enninga, J. ~2012!. Hierarchies of host factor dynamics at the
entry site of Shigella flexneri during host cell invasion. Infect
Immun 80, 2548–2557.
Erni, R., Rossell, M.D., Kisielowski, C. & Dahmen, U. ~2009!.
Atomic-resolution imaging with a sub-50-pm electron probe.
Phys Rev Lett 102, 96101.
Folling, J., Bossi, M., Bock, H., Medda, R., Wurm, C.A., Hein,
B., Jakobs, S., Eggeling, C. & Hell, S.W. ~2008!. Fluorescence
nanoscopy by ground-state depletion and single-molecule re-
turn. Nat Methods 5, 943–945.
Gautier, A., Juillerat, A., Heinis, C., Correa, I.R., Jr., Kinder-
mann, M., Beaufils, F. & Johnsson, K. ~2008!. An engineered
protein tag for multiprotein labeling in living cells. Chem Biol
15, 128–136.
Gelles, J., Schnapp, B.J. & Sheetz, M.P. ~1988!. Tracking kinesin-
driven movements with nanometre-scale precision. Nature 331,
450–453.
Giessibl, F.J. ~1995!. Atomic resolution of the silicon ~111!-~7�7!
surface by atomic force microscopy. Science 267, 68–71.
Gordon, M.P., Ha, T. & Selvin, P.R. ~2004!. Single-molecule
high-resolution imaging with photobleaching. Proc Natl Acad
Sci USA 101, 6462.
Greenfield, D., McEvoy, A.L., Shroff, H., Crooks, G.E., Win-
green, N.S., Betzig, E. & Liphardt, J. ~2009!. Self-organization
of the Escherichia coli chemotaxis network imaged with super-
resolution light microscopy. PLoS Biol 7, e1000137.
Gustafsson, M.G.L. ~2000!. Surpassing the lateral resolution limit
by a factor of two using structured illumination microscopy. J
Microsc 198, 82–87.
Hänninen, P., Hell, S., Salo, J., Soini, E. & Cremer, C. ~1995!.
Two-photon excitation 4Pi confocal microscope: Enhanced ax-
ial resolution microscope for biological research. Appl Phys Lett
66, 1698.
Hedde, P.N., Fuchs, J., Oswald, F., Wiedenmann, J. & Nien-
haus, G.U. ~2009!. Online image analysis software for photo-
activation localization microscopy. Nat Methods 6, 689–690.
Heilemann, M., van de Linde, S., Mukherjee, A. & Sauer, M.
~2009!. Super-resolution imaging with small organic fluoro-
phores. Angew Chem Int Edit 48, 6903–6908.
Hell, S.W. & Wichmann, J. ~1994!. Breaking the diffraction
resolution limit by stimulated emission: Stimulated-emission-
depletion fluorescence microscopy. Opt Lett 19, 780–782.
Henriques, R., Griffiths, C., Hesper Rego, E. & Mhlanga,
M.M. ~2011!. PALM and STORM: Unlocking live-cell super-
resolution. Biopolymers 95, 322–331.
Henriques, R., Lelek, M., Fornasiero, E.F., Valtorta, F., Zim-
mer, C. & Mhlanga, M.M. ~2010!. QuickPALM: 3D real-time
photoactivation nanoscopy image processing in ImageJ. Nat
Methods 7, 339–340.
Henriques, R. & Mhlanga, M.M. ~2009!. PALM and STORM:
What hides beyond the Rayleigh limit? Biotechnol J 4, 846–857.
Hess, S.T., Girirajan, T.P. & Mason, M.D. ~2006!. Ultra-high
resolution imaging by fluorescence photoactivation localization
microscopy. Biophys J 91, 4258–4272.
Holden, S.J., Uphoff, S. & Kapanidis, A.N. ~2011!. DAOSTORM:
An algorithm for high-density super-resolution microscopy.
Nat Methods 8, 279–280.
Hu, K., Ji, L., Applegate, K.T., Danuser, G. & Waterman-
Storer, C.M. ~2007!. Differential transmission of actin motion
within focal adhesions. Science’s STKE 315, 111.
Huang, B., Bates, M. & Zhuang, X. ~2009!. Super-resolution
fluorescence microscopy. Annu Rev Biochem 78, 993–1016.
Huang, B., Jones, S.A., Brandenburg, B. & Zhuang, X. ~2008a!.
Whole-cell 3D STORM reveals interactions between cellular
structures with nanometer-scale resolution. Nat Methods 5,
1047–1052.
Huang, B., Wang, W., Bates, M. & Zhuang, X. ~2008b!. Three-
dimensional super-resolution imaging by stochastic optical re-
construction microscopy. Science 319, 810–813.
Huang, F., Schwartz, S.L., Byars, J.M. & Lidke, K.A. ~2011!.
Simultaneous multiple-emitter fitting for single molecule super-
resolution imaging. Biomed Opt Express 2, 1377–1393.
Jares-Erijman, E.A. & Jovin, T.M. ~2003!. FRET imaging. Nat
Biotechnol 21, 1387–1395.
Jones, S.A., Shim, S.H., He, J. & Zhuang, X. ~2011!. Fast, three-
dimensional super-resolution imaging of live cells. Nat Meth-
ods 8, 499–505.
Juette, M.F., Gould, T.J., Lessard, M.D., Mlodzianoski, M.J.,
Nagpure, B.S., Bennett, B.T., Hess, S.T. & Bewersdorf, J.
~2008!. Three-dimensional sub-100 nm resolution fluorescence
microscopy of thick samples. Nat Methods 5, 527–529.
Kao, H.P. & Verkman, A. ~1994!. Tracking of single fluorescent
particles in three dimensions: Use of cylindrical optics to
encode particle position. Biophys J 67, 1291–1300.
Klar, T.A., Jakobs, S., Dyba, M., Egner, A. & Hell, S.W. ~2000!.
Fluorescence microscopy with diffraction resolution barrier
broken by stimulated emission. Proc Natl Acad Sci USA 97,
8206.
Klein, T., Löschberger, A., Proppert, S., Wolter, S., van de
Linde, S. & Sauer, M. ~2010!. Live-cell dSTORM with SNAP-
tag fusion proteins. Nat Methods 8, 7–9.
Kner, P., Chhun, B.B., Griffis, E.R., Winoto, L. & Gustafsson,
M.G.L. ~2009!. Super-resolution video microscopy of live cells
by structured illumination. Nat Methods 6, 339–342.
Lacoste, T.D., Michalet, X., Pinaud, F., Chemla, D.S., Alivisa-
tos, A.P. & Weiss, S. ~2000!. Ultrahigh-resolution multicolor
colocalization of single fluorescent probes. Proc Natl Acad Sci
USA 97, 9461.
Lelek, M., Di Nunzio, F., Henriques, R., Charneau, P., Arhel,
N. & Zimmer, C. ~2012!. Superresolution imaging of HIV in
infected cells with FlAsH-PALM. Proc Natl Acad Sci USA 109,
8564–8569.
Lidke, K., Rieger, B., Jovin, T. & Heintzmann, R. ~2005!. Super-
resolution by localization of quantum dots using blinking
statistics. Opt Express 13, 7052–7062.
Löschberger, A., van de Linde, S., Dabauvalle, M.C., Rieger,
B., Heilemann, M., Krohne, G. & Sauer, M. ~2012!. Super-
1428
Sébastien Herbert et al.

----!@#$NewPage!@#$----
resolution imaging visualizes the eightfold symmetry of gp210
proteins around the nuclear pore complex and resolves the
central channel with nanometer resolution. J Cell Sci 125,
570–575.
Lukyanov, K.A., Chudakov, D.M., Lukyanov, S. & Verkhusha,
V.V. ~2005!. Innovation: Photoactivatable fluorescent proteins.
Nat Rev Mol Cell Biol 6, 885–891.
Ma, H., Long, F., Zeng, S. & Huang, Z.L. ~2012!. Fast and precise
algorithm based on maximum radial symmetry for single mol-
ecule localization. Opt Lett 37, 2481–2483.
Manley, S., Gillette, J.M., Patterson, G.H., Shroff, H., Hess,
H.F., Betzig, E. & Lippincott-Schwartz, J. ~2008!. High-
density mapping of single-molecule trajectories with photoacti-
vated localization microscopy. Nat Methods 5, 155–157.
Matsuda, A., Shao, L., Boulanger, J., Kervrann, C., Carlton,
P.M., Kner, P., Agard, D. & Sedat, J.W. ~2010!. Condensed
mitotic chromosome structure at nanometer resolution using
PALM and EGFP-histones. PLoS One 5, e12768.
Mattheyses, A.L., Simon, S.M. & Rappoport, J.Z. ~2010!. Imag-
ing with total internal reflection fluorescence microscopy for
the cell biologist. J Cell Sci 123, 3621–3628.
Ober, R.J., Ram, S. & Ward, E.S. ~2004!. Localization accuracy in
single-molecule microscopy. Biophys J 86, 1185–1200.
Parthasarathy, R. ~2012!. Rapid, accurate particle tracking by
calculation of radial symmetry centers. Nat Methods 9, 724–726.
Pavani, S.R., Thompson, M.A., Biteen, J.S., Lord, S.J., Liu, N.,
Twieg, R.J., Piestun, R. & Moerner, W.E. ~2009!. Three-
dimensional, single-molecule fluorescence imaging beyond the
diffraction limit by using a double-helix point spread function.
Proc Natl Acad Sci USA 106, 2995–2999.
Pertsinidis, A., Zhang, Y. & Chu, S. ~2010!. Subnanometre
single-molecule localization, registration and distance measure-
ments. Nature 466, 647–651.
Qu, X., Wu, D., Mets, L. & Scherer, N.F. ~2004!. Nanometer-
localized multiple single-molecule fluorescence microscopy. Proc
Natl Acad Sci USA 101, 11298.
Ries, J., Kaplan, C., Platonova, E., Eghlidi, H. & Ewers, H.
~2012!. A simple, versatile method for GFP-based super-
resolution microscopy via nanobodies. Nat Methods 9, 582–584.
Rino, J., Braga, J., Henriques, R. & Carmo-Fonseca, M. ~2009!.
Frontiers in fluorescence microscopy. Int J Dev Biol 53,
1569–1579.
Rogers, S.S., Waigh, T.A., Zhao, X. & Lu, J.R. ~2007!. Precise
particle tracking against a complicated background: Poly-
nomial fitting with Gaussian weight. Phys Biol 4, 220.
Rust, M.J., Bates, M. & Zhuang, X. ~2006!. Sub-diffraction-limit
imaging by stochastic optical reconstruction microscopy
~STORM!. Nat Methods 3, 793–795.
Schermelleh, L., Heintzmann, R. & Leonhardt, H. ~2010!. A
guide to super-resolution fluorescence microscopy. J Cell Biol
190, 165–175.
Schmidt, R., Wurm, C.A., Jakobs, S., Engelhardt, J., Egner, A.
& Hell, S.W. ~2008!. Spherical nanosized focal spot unravels
the interior of cells. Nat Methods 5, 539–544.
Shroff, H., Galbraith, C.G., Galbraith, J.A. & Betzig, E.
~2008!. Live-cell photoactivated localization microscopy of nano-
scale adhesion dynamics. Nat Methods 5, 417–423.
Shtengel, G., Galbraith, J.A., Galbraith, C.G., Lippincott-
Schwartz, J., Gillette, J.M., Manley, S., Sougrat, R., Wa-
terman, C.M., Kanchanawong, P. & Davidson, M.W. ~2009!.
Interferometric fluorescent super-resolution microscopy re-
solves 3D cellular ultrastructure. Proc Natl Acad Sci USA 106,
3125.
Simonson, P.D., Rothenberg, E. & Selvin, P.R. ~2011!. Single-
molecule-based super-resolution images in the presence of
multiple fluorophores. Nano Lett 11~11!, 5090–5096.
Smith, C.S., Joseph, N., Rieger, B. & Lidke, K.A. ~2010!. Fast,
single-molecule localization that achieves theoretically mini-
mum uncertainty. Nat Methods 7, 373–375.
Steinhauer, C., Forthmann, C., Vogelsang, J. & Tinnefeld, P.
~2008!. Superresolution microscopy on the basis of engineered
dark states. J Am Chem Soc 130, 16840–16841.
Thompson, R.E., Larson, D.R. & Webb, W.W. ~2002!. Precise
nanometer localization analysis for individual fluorescent probes.
Biophys J 82, 2775–2783.
Tokunaga, M., Imamoto, N. & Sakata-Sogawa, K. ~2008!. Highly
inclined thin illumination enables clear single-molecule imag-
ing in cells. Nat Methods 5, 159–161.
Truong, T.V., Supatto, W., Koos, D.S., Choi, J.M. & Fraser, S.E.
~2011!. Deep and fast live imaging with two-photon scanned
light-sheet microscopy. Nat Methods 8, 757–760.
Van Oijen, A., Köhler, J., Schmidt, J., Müller, M. & Braken-
hoff, G. ~1998!. 3-Dimensional super-resolution by spectrally
selective imaging. Chem Phys Lett 292, 183–187.
Vaziri, A., Tang, J., Shroff, H. & Shank, C.V. ~2008!. Multilayer
three-dimensional super resolution imaging of thick biological
samples. Proc Natl Acad Sci USA 105, 20221–20226.
Vogelsang, J., Kasper, R., Steinhauer, C., Person, B., Heile-
mann, M., Sauer, M. & Tinnefeld, P. ~2008!. A reducing and
oxidizing system minimizes photobleaching and blinking of
fluorescent dyes. Angew Chem Int Ed Engl 47, 5465–5469.
Wolter, S., Schüttpelz, M., Tscherepanow, M., Van de Linde,
S., Heilemann, M. & Sauer, M. ~2010!. Real-time computa-
tion of sub-diffraction-resolution fluorescence images. J Mi-
crosc 237, 12–22.
Wombacher, R., Heidbreder, M., van de Linde, S., Sheetz,
M.P., Heilemann, M., Cornish, V.W. & Sauer, M. ~2010!.
Live-cell super-resolution imaging with trimethoprim conju-
gates. Nat Methods 7, 717–719.
Xu, K., Babcock, H.P. & Zhuang, X. ~2012!. Dual-objective
STORM reveals three-dimensional filament organization in the
actin cytoskeleton. Nat Methods 9, 185–188.
Yildiz, A., Forkey, J.N., McKinney, S.A., Ha, T., Goldman, Y.E. &
Selvin, P.R. ~2003!. Myosin V walks hand-over-hand: Single
fluorophore imaging with 1.5-nm localization. Science 300,
2061–2065.
York, A.G., Ghitani, A., Vaziri, A., Davidson, M.W. & Shroff,
H. ~2011!. Confined activation and subdiffractive localization
enables whole-cell PALM with genetically expressed probes.
Nat Methods 8, 327–333.
Zanacchi, F.C., Lavagnino, Z., Donnorso, M.P., Del Bue, A.,
Furia, L., Faretta, M. & Diaspro, A. ~2011!. Live-cell 3D
super-resolution imaging in thick biological samples. Nat Meth-
ods 8, 1047–1049.
Zhu, L., Zhang, W., Elnatan, D. & Huang, B. ~2012!. Faster
STORM using compressed sensing. Nat Methods 9~7!, 721–723.
Single-Molecule Super-Resolution
1429