close

Вход

Забыли?

вход по аккаунту

?

Branching Out of Single-Molecule Fluorescence Spectroscopy Challenges for Chemistry and Influence on Biology.

код для вставкиСкачать
Reviews
P. Tinnefeld and M. Sauer
Single-Molecule Detection
Branching Out of Single-Molecule Fluorescence
Spectroscopy: Challenges for Chemistry and Influence
on Biology
Philip Tinnefeld* and Markus Sauer*
Keywords:
conformational dynamics · electron
transfer · fluorescent labels · FRET
(fluorescence resonance
energy transfer) · singlemolecule spectroscopy
Dedicated to Professor Jrgen Wolfrum
on the occasion of his 65th birthday
Angewandte
Chemie
2642
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200300647
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
In the last decade emerging single-molecule fluorescencespectroscopy tools have been developed and adapted to analyze
individual molecules under various conditions. Single-moleculesensitive optical techniques are now well established and help to
increase our understanding of complex problems in different
disciplines ranging from materials science to cell biology. Previous
dreams, such as the monitoring of the motility and structural
changes of single motor proteins in living cells or the detection of
single-copy genes and the determination of their distance from
polymerase molecules in transcription factories in the nucleus of a
living cell, no longer constitute unsolvable problems. In this
Review we demonstrate that single-molecule fluorescence
spectroscopy has become an independent discipline capable of
solving problems in molecular biology. We outline the challenges
and future prospects for optical single-molecule techniques which
can be used in combination with smart labeling strategies to yield
quantitative three-dimensional information about the dynamic
organization of living cells.
From the Contents
1. Introduction
2643
2. Fundamentals and Technical
Developments in SMFS
2645
3. Multiparameter SMFS, Photon
Statistics and Analysis
2648
4. SMFS to Study Structure and
Dynamics
2651
5. Use of SMFS to Study Photophysics
in Multichromophoric Systems
2654
6. Molecular Photonic Wires
2655
7. SMFS in Biology
2657
8. Colocalization and Distance
Measurements Between Single
Molecules
2660
9. Summary and Outlook
2661
1. Introduction
Major technological breakthroughs and intensive efforts
in pushing the detection limits of microscopic techniques in
the last two decades have been the basis for fulfilling many
scientists dream: to detect, manipulate, and control matter on
an atomic and molecular scale. Successful approaches to
detect and study single atoms or molecules on surfaces have
used near-field interactions with tunneling electrons or forces
from sharp tips, in scanning tunneling microscopy (STM)[1] or
atomic force microscopy.[2] On the other hand, optical
methods enable the observation of single molecules from a
certain distance unperturbed by tips. Moreover, advances in
optical spectroscopy and microscopy have made it possible
not only to detect and identify freely diffusing or immobilized
fluorescent molecules, but also to realize spectroscopic
measurements and monitor dynamic processes. Much of the
fascination about single-molecule fluorescence spectroscopy
(SMFS) stems from the fact that it provides the basis for a new
and more direct scientific approach. That is, SMFS enables us
to directly compare models, usually drawn from considerations of individual molecular systems, with experimental
observations. In contrast to ensemble measurements which
yield information only on average properties, single-molecule
experiments provide information on individuals, such as
distributions and time trajectories of properties that would
otherwise be hidden. Whether each individual molecule has
slightly different but constant properties (static disorder) or
changes its properties over time (dynamic disorder), cannot
be answered correctly by averaging the observable. Since in
SMFS only one molecule is observed at a time, any
synchronization procedure to induce, for example, a conformational change of the molecular structure is avoided.
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Therefore, SMFS represents a versatile tool to measure
conformational dynamics in biopolymers, for example, protein folding or contact formation in peptides, on a wide range
of time scales, extending from nanoseconds to seconds, at the
single-molecule level.[3–5]
One of the most prominent and challenging examples that
requires SMFS is the sequencing of a single DNA strand. In
contrast to current DNA-sequencing schemes based on
Sangers enzymatic chain-termination method, fluorescencebased single-molecule DNA sequencing would allow a single
fragment of DNA, several tens of kilobases or more in length,
to be read at a theoretical rate of more than several hundred
bases per second. The concept is to “watch” the stepwise
incorporation of DNA bases by polymerase enzymes into a
single DNA strand (Figure 1 a).[6, 7] If each base can be
identified as it is incorporated into the DNA strand, the
sequence can be obtained. Analogous to this approach, the
reverse process could also be monitored, that is, an exonuclease could be used to degrade a single DNA strand base-bybase from one end while identifying each base after it is
cleaved (Figure 1 b).[8–11] Multiple DNA strands cannot be
[*] Dr. P. Tinnefeld, Prof. Dr. M. Sauer
Applied Laserphysics und Laserspectroscopy
Faculty of Physics
University of Bielefeld
Universittsstrasse 25
33615 Bielefeld (Germany)
Fax: (+ 49) 521-106-2958
E-mail: tinnefeld@physik.uni-bielefeld.de
sauer@physik.uni-bielefeld.de
DOI: 10.1002/anie.200300647
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2643
Reviews
P. Tinnefeld and M. Sauer
As simple as it sounds, the realization of singlemolecule DNA sequencing constitutes one of the
greatest challenges biologists, chemists, and physicists have confronted. From the biological point of
view, major problems arise through the use of
fluorescently labeled nucleotides. The bulky fluorescent dye attached to the nucleotide could hinder
the polymerase- or exonuclease-mediated incorporation or degradation of nucleotides. Besides these
problems, the well-defined selection of a single DNA
strand and the detection and identification of each
nucleotide through the spectroscopic characteristics
of the fluorescent label with high accuracy are very
demanding tasks. While the primary goal to
sequence a single DNA fragment has not been
achieved yet, each development step towards that
goal is productive for the generation and improvement of analytic research systems capable of following biochemical processes (e.g., enzymatic activities)
at the single-molecule level.
Current high-resolution methods, such as X-ray
Figure 1. Illustration of the proposed procedure for polymerase-, and exonuclease-mediated
crystallography and NMR spectroscopy, have prosequencing of a single DNA molecule. a) Polymerase or DNA is immobilized in zero-mode
vided a vast array of structural detail for biological
waveguide nanostructures. The rapid decay of illumination incident to the entrance of such
molecules, yet the output of these methods is limited
21
[6]
guides provides zeptoliter (10 L) effective observation volumes within the guides. Zeroby the static molecular description that they deliver
mode waveguides thereby enable single-molecule detection studies in the micromolar
and by ensemble averaging. Single-molecule methrange, which is the concentrations required for efficient enzymatic DNA synthesis.[302]
Owing to the small observation volume the typical residence time of freely diffusing single
ods provide an alternative set of approaches that will
fluorescently labeled nucleotide molecules is in the range of a few microseconds. On the
lead to a more direct view of the action of molecules
other hand, when recognized by the DNA polymerase and incorporated into the growing
without the need to infer functional processes from
DNA strand, the residence time increases to several milliseconds, sufficiently long for its
static structures. For example, many of the molecular
unequivocal fluorescence spectroscopic identification. b) In exonuclease-mediated sequencmachines that are essential for cell function and
ing, the DNA to be sequenced should be copied using a biotinylated primer, a DNA polysurvival, such as replication, transcription, transmerase, and the four nucleotide triphosphates (dNTPs), each base type contains a different
lation, and protein folding, are asynchronous and
fluorescent label with distinct laser-induced fluorescence. The single DNA fragment is
bound to a microsphere or other solid support and transferred into a flowing sample
have different kinetic pathways. Biomolecular
stream by mechanical micromanipulation or optical trapping. Upon addition of a 3’!5’
assemblies and machines can only be fully underexonuclease, fluorescent nucleotide monophosphate molecules (dNMPs) will be cleaved
stood if the structure and function of the components
and transported to the detection area down stream where they are identified based on their
of living systems is known. This discipline, known as
[9–11, 82]
In both schemes, the DNA sequence is directly
characteristic fluorescence properties.
structural biology, is in contrast to other fields of
retrieved from the detected signal sequence.
molecular biology an area of research where progress depends critically on sophisticated instruments and
used because the different enzyme rates on the different
labeling strategies.
DNA strands would result in rapid dephasing. This elegant
SMFS is now used in physics, chemistry, and biology and
alternative to common sequencing techniques requires the
has already been elaborated in excellent reviews.[12–29] The
minimum conceivable starting material to obtain a sequence.
Philip Tinnefeld, born 1974, studied chemistry in Mnster, Montpellier, and Heidelberg.
In 2002, he received his PhD at the University of Heidelberg under the supervision of
Prof. Jrgen Wolfrum. In 2001, he was recipient of a Schloessmann Award of the Max
Planck Society for his ideas on the development of a molecular photonic wire. He carried out postdoctoral research with Prof.
Shimon Weiss (UCLA) and with Prof.
Frans C. De Schryver (Leuven) on the development of new single-molecule fluorescence
techniques for the investigation of molecular
interactions and dynamics. Since 2003 he is group leader in the department of Applied Laserphysics and Laserspectroscopy at the University of
Bielefeld.
2644
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Markus Sauer was born 1965 in Pforzheim.
He studied chemistry in Karlsruhe, Saarbrcken, and Heidelberg, and finished his
PhD in physical chemistry at the University
of Heidelberg in 1995 under the guidance of
Prof. Jrgen Wolfrum. After a short-term
visit at LBNL, Berkeley in the group of Prof.
Shimon Weiss, he received the BioFuture
award in 1998 to perform independent
research on single-molecule handling, detection, and identification. In 2002 he finished
his habilitation in Heidelberg at the Institute
of Physical Chemistry. Since 2003 he is Professor for Applied Laserphysics and Laserspectroscopy at the University of Bielefeld.
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
ongoing success of SMFS has been enabled not only by the
development of optical single-molecule techniques but also
through new refined organic-synthesis methods and the large
repertoire of molecular-biology techniques. The ability to
specifically label many different sites of macromolecules, for
example, provides a great tool box for the application of
spectroscopic techniques, such as fluorescence resonance
energy transfer (FRET).[30] However, most biologically relevant SMFS studies have been performed in vitro, that is, the
molecules under investigation were isolated and often
immobilized to ease their observation and minimize background. The ultimate goal is to investigate these molecules in
their natural environment, under native conditions, that is, in
living cells. Only then can we observe them in their natural
conditions and study the effect of molecular crowding, for
example. Can SMFS successfully be utilized to study communication pathways, signaling cascades, or RNA polymerases
during transcription inside the nucleus of a live cell?
Herein, recent advances of single-molecule spectroscopy
and microscopy with emphasis on biological applications are
reviewed. First, we will briefly explain the fundamentals for
successful detection and spectroscopy of single fluorescent
molecules. In particular, we will discuss the merits of widefield and confocal microscopy. Then, we highlight various
energy-transfer methods, capable of monitoring distance
changes between single molecules, and demonstrate how
multiparameter SMFS techniques combined with smart data
analysis can be utilized to study and control complex multichromophoric systems and photonic wires and to quantitatively determine the number of interacting molecules and
their relative orientations. Furthermore, we review new
techniques, such as high-precision distance microscopy, to
measure distances between single molecules well below the
optical-resolution limit. Examples will be given how every
accessible spectroscopic parameter can contribute to a better
understanding of complex biological systems. Finally, we give
some perspectives of how SMFS, in combination with optical
tweezers, new labeling strategies, and other spectroscopic
techniques, for example, surface-enhanced Raman spectroscopy, might be successfully adapted to live-cell analysis.
2. Fundamentals and Technical Developments in
SMFS
Considering all the technical developments and scientific
efforts that were required for the first successful detections of
single fluorescent molecules in 1989[31] and 1990,[32] it is hard
to imagine that it only takes a human eye, properly adjusted
to the dark, and a microscope with high collection efficiency
to detect single, ordinary fluorophores, such as a rhodamine
molecule, by its fluorescence emission. To aid this detection, a
variety of techniques have been applied to SMFS, of these
wide-field microscopy using a state-of-the-art charge coupled
device (CCD) detector is the simplest.
When the central player, usually a fluorophore, is
illuminated by laser light it is excited from the ground-state,
S0, into high-lying vibrational levels of the first excited singletstate, S1. This state undergoes rapid, non-radiative, internal
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
conversion to the lowest S1 level. The optical saturation limit
is the maximum rate that a dye molecule can be cycled
between S0 and S1 and is dependent on the fluorescence
lifetime of the dye, tf (Figure 2 a). Several depopulation
Figure 2. a) Three-electronic-state model describing the general photophysical behavior of single fluorescent molecules under laser illumination. b) Fluorescence signals observed in consecutive time windows
(bins) of 1 ms from a 1011 m aqueous solution of Cy5 molecules
(count rate nc versus time t) using confocal fluorescence microscopy
and a pulsed diode laser emitting at l = 635 nm for excitation. When a
molecule passes the observation volume, up to 200 photon counts are
detected per millisecond. c) Fluorescence intensity trajectory of a
single Cy5 molecule immobilized on a dry cover slide. One commonly
observed source of single-molecule fluorescence-intensity fluctuations
involves discrete intensity jumps from “on” (high intensity) to “off”
(background) states arising from quantum jumps of the single molecule to non-emissive “dark” states, for example, triplet states. After
approximately 5.6 s irreversible photobleaching occurs, reflected in a
spontaneous drop of the fluorescence intensity to the background
level.
pathways, such as intersystem crossing into the triplet state,
compete with fluorescence emission, thus reducing the
number of emitted photons. Furthermore, the number of
photons that can be detected from a single fluorophore is
limited by irreversible photobleaching, which is an intrinsic
property of all conventional organic fluorophores. Common
dye molecules can emit up to several 106 photons before
irreversible photobleaching occurs. In general, the fluorescence signal is governed by the selection of appropriate
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2645
Reviews
P. Tinnefeld and M. Sauer
fluorophores which exhibit a high extinction coefficient e at
order of 50–100 nm is achieved by bringing a sample to within
the excitation wavelength l, a high fluorescence quantum
5–10 nm of a subwavelength sized optical aperture so that
yield Ffl, and a high photostability. Additionally, the ideal
photons from the aperture do not have enough distance to
experience diffraction (Figure 3 a). Although NSOM offers
fluorophore should have a fairly short fluorescence lifetime
advantages, such as the ability to correlate spectroscopic
(i.e. in the nanosecond range) so that it can go through the
information with topographic information, it suffers from
excitation and emission cycle at high frequency and should
not show fluorescence intermittencies
(off states) as a result of, for example,
triplet transitions (Figure 2).
The most important parameter in
SMFS is the signal-to-noise ratio (S/
N). With respect to the noise in the
imaging signal, shot noise, or photon
noise, is the most important contribution in that it poses the fundamental
limit to the S/N ratio. Under real
conditions the presence of background shot noise must also be considered. If the signal-to-background
(S/B) ratio is large, nearly all detected
Figure 3. Schematic set-up of a) near-field scanning optical microscopy (NSOM), b) total internal
photons represent image information,
reflection fluorescence microscopy (TIRFM), and c) confocal fluorescence microscopy.
and the S/N ratio approaches the
value n = (for a signal consisting of
n photons). If the background contriinvasiveness and perturbation of the sample owing the
bution is large (small S/B ratio), the desired signal may
nearness of the probe tip. NSOM also covers a huge field of
become lost in the high background shot noise, resulting in an
research and is therefore beyond the scope of this review.
S/N ratio approaching zero. Background stems mainly from
Nevertheless, NSOM finds application in biology and enables
elastic (Rayleigh) and inelastic (Raman) scattering from
simultaneous topographic and single-molecule fluorescence
solvent molecules as well as from autofluorescent impurities.
imaging on live cell membranes.[36–39]
While Rayleigh scattering can be efficiently suppressed by the
use of suitable optical filters, the complete suppression of
In confocal microscopy, a laser beam is brought to near
Raman scattering, which is directly proportional to the
diffraction-limited focus in a sample by using an oil- or waternumber of illuminated solvent molecules, is challenging
immersion, high numerical aperture (NA) objective.[40] The
because it occurs at least partly in the same wavelength
fluorescence is collected through the same objective and
range as the fluorescence signal. Autofluorescence from
emission light is separated from excitation light by a dichroic
impurities strongly depends on the excitation, and detection
mirror (Figure 3 c). A small pinhole (50–100 mm in diameter)
wavelength. Especially in biological samples, luminescent
is placed in the image plane to reject light from out-of-focus
impurities can decrease the sensitivity or even prevent the
regions. This arrangement defines a small spheroidal volume
definite detection of individual fluorescent molecules. Thereof 0.5–1.0 femtoliters in the sample. Rigler and co-workers
fore, it is desirable in single-molecule fluorescence experiwere the first to exploit the use of confocal fluorescence
ments to work with fluorescent dyes that absorb and emit in
microscopy for the detection of single-molecules.[41, 42]
the red region of the electromagnetic spectrum, this is
Using confocal optical arrangements and avalanche
because above 600 nm, there are only a few compounds
photodiodes (APDs) as detectors, individual fluorescent
known which show intrinsic absorption and emission.[33]
molecules can be easily detected with high S/N ratios.
Confocal fluorescence microscopy has even been successfully
Since the contribution of background sources is directly
employed in two-photon excitation (TPE) of individual
proportional to the number of molecules in the excitation
fluorophores.[43–45] The principles of two-photon microscopy
volume, the reduction of the excitation/detection volume is
also key to a high S/N ratio. One SMFS method relies on
were first elucidated by Webb and co-workers.[46] TPE cross
hydrodynamic focusing of a sample stream in sheath-flow
sections are extremely small, typically on the order of
cuvettes down to a layer thickness < 10 mm in combination
1050 cm4 s for most fluorophores, so that high intensity,
with pulsed excitation and time-gated detection to suppress
short laser pulses (pulse widths of approximately 100 fs), for
Rayleigh and Raman scattering.[33] Further reduction of the
example, from a mode-locked Ti:sapphire laser, are required
to achieve efficient TPE.[47, 48] As a second-order, nonlinear
excitation volume can be achieved using laser excitation in a
1) confocal, 2) evanescent, or 3) near-field configuration.
process, the molecular excitation rate depends quadratically
Near-field scanning optical microscopy (NSOM) has been
on the laser intensity. Since efficient excitation occurs only at
developed primarily to break the optical diffraction limit,
the laser-beam focus, photobleaching of out-of-focus molewhich restricts the spatial resolution of conventional optical
cules is reduced. On the other hand, in-focus photobleaching
measurements to approximately half of the wavelength of
is more pronounced so that TPE fluorescence count rates are
light.[34, 35] In NSOM, subdiffraction spatial resolution on the
generally lower. Since high fluorescence count rates are
1
2
2646
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
essential for SMFS, one-photon excitation represents the
method of choice used in most applications. Furthermore, in
some cases, the high excitation intensities required to
generate fluorescence signals by TPE may lead to other
nonlinear processes, for example, continuum generation in
the solvent, which can increase the background and deteriorate fluorescence sensitivity. Nevertheless, TPE can be
successfully utilized in two-color cross-correlation experiments to monitor colocalization of different dyes in the
observation volume at the single-molecule level.[49] The main
advantage of the two-photon technique is that the broadness
of two-photon excitation spectra allows for dyes with different emission wavelengths to be excited by laser light of a
single wavelength (so called spectral multiplexing).
In combination with fluorescence correlation spectroscopy (FCS), confocal fluorescence microscopy has found
broad application, especially in the field of diagnostics and
evolutionary biotechnology.[50] FCS uses Brownian motion to
bring single or a few sample molecules through the confocal
volume. By calculating the autocorrelation of the fluorescence intensity fluctuations, time-dependent dynamics can be
followed in detail. However, since the emission bursts from
many single molecules are used to calculate the autocorrelation function, differences between individuals are impossible
to detect. Since the sensitivity of FCS affords nanomolar
concentrations,[51] target molecules, such as specific DNA or
RNA sequences, have to be amplified prior to the hybridization measurements.
Photon-counting histogram (PCH), and fluorescenceintensity distribution analysis (FIDA) are related techniques
to differentiate between molecules with different fluorescence brightness. Both techniques measure photon counts in
time intervals of fixed duration and fit these measurements to
theoretical distributions.[52–54] Fluorescence-intensity multiple
distributions analysis (FIMDA) uses a similar photon-counting histogram but with time intervals of varying duration, and
can simultaneously extract information about diffusion and
brightness.[55] The most recent autocorrelation method,
photon arrival-time interval distribution (PAID) combines
the autocorrelation approach of FCS with a photon-counting
histogram to reveal information on multiple spectral channels.[56]
Wide-field CCD-array imaging does not have the high
temporal resolution of an APD (ca. 100 ps) and does not
achieve the best S/B ratio because more of the background
from out-of-focus sources is collected and detected. On the
other hand, wide-field imaging offers the advantage of
parallel detection and the ability to acquire statistics rapidly
on the millisecond behavior of hundreds of molecules
individually and simultaneously.[57] This type of detection is
particularly important in experiments where irreversible
reactions are monitored, for example, the unwinding of
immobilized DNA molecules upon addition of a helicase
enzyme.[58] In this experiment, helicase molecules start
unwinding of the DNA almost simultaneously and in a
confocal setup only one molecule could be monitored for
every sample preparation.
One possibility to obtain statistics with a low-temporal
resolution but a high S/B ratio for a large number of
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
molecules is total internal reflection (TIR) excitation of
fluorescence near a dielectric interface (Figure 3 b).[59, 60]
Usually, this evanescent wave excitation is achieved by
prism-[61] or objective-type[62] TIR at a glass–-liquid or a
glass–air interface. At the interface the optical electromagnetic field does not abruptly drop to zero but decays
exponentially into the liquid or air phase. Molecules in a
thin layer of approximately 200 nm thickness immediately
next to this interface can still be excited by the rapidly
decaying optical field. Hence, the illuminated volume is
extremely thin thus providing very good S/N ratios. TIRmicroscopy is particularly applicable for imaging single
molecules at electromagnetic boundaries and has allowed
the observation of fluorescently labeled myosin and kinesin
molecules as well as individual adenosine-5’-triphosphate
(ATP) turnover reactions.[61, 63–65] Evanescent field excitation
has been also successfully employed for two- and threedimensional measurements of the structural dynamics of
single myosin V motor protein complexes.[66]
In principal, single-molecule measurements can be subdivided into different categories depending on the experimental objective. In many single-molecule applications, such
as DNA sequencing, it is necessary to classify the detected
fluorescence signals of individual molecules or distinguish
between single molecules. That is, the identification and
analysis of static properties of freely diffusing or immobilized
molecules. In other applications, individual molecules have to
be observed over a long time to retrieve information about
their mobility and trajectories. Single-molecule experiments
are also utilized to study individual dynamics in a heterogeneous environment or conformational fluctuations of polymers or biomolecules.
Theoretically, several characteristic properties, such as
1) fluorescence intensity 2) fluorescence decay[67, 68] 3) emission spectrum,[10, 69] 4) diffusion coefficient (in solution),[70]
and 5) fluorescence anisotropy,[71] can be used to identify an
individual fluorescent molecule.[70] However, because the
trajectory of each free individual molecule is given by
Brownian motion and environmental factors, the number of
detected fluorescence photons per molecule has a large
statistical distribution. Although the entire fluorescence
intensity distribution can be a well-defined characteristic for
a species,[52, 53, 55, 56] a single event is not. Therefore, fluorescence intensity alone is not an adequate parameter to identify
a single molecule or to follow individual dynamics. There are
many sources of intensity fluctuations other than those that
are being investigated. Among these are orientation changes,
blinking arising from photophysical or photochemical
changes, or uncorrelated changes of the environment of the
fluorophore.[72–79]
Fortunately, fluorescence photons deliver several measurable pieces of information that can help to transmit the
changes of the molecules investigated to the experimenter.
Besides the information that fluorescence lifetime, intensity,
emission spectrum, and fluorescence anisotropy provide on
the dynamics of the investigated system, it is important to
remember that the number of photons emitted from a single
fluorophore is limited. An elegant way to increase the
information obtainable from a single fluorescent molecule is
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2647
Reviews
P. Tinnefeld and M. Sauer
to monitor different fluorescence characteristics, if possible,
in parallel.[80–82] In the static case, the parallel detection of
different parameters can be used to improve the identification
accuracy of single molecules,[71, 82] to discriminate single
fluorophores against background fluorescence, which is of
particular interest in cells and tissues,[83] or to reveal heterogeneity in local environments.[76, 84] The simultaneous recording of various spectroscopic parameters (so-called multiparameter analysis) provides the fundamental basis for
improved single-molecule spectroscopic techniques that
enable a refined understanding of spectroscopic dynamics
and fluorophore interactions.[78, 85–88] Furthermore, in contrast
to ensemble fluorescence spectroscopy, SMFS offers an
additional advantageous parameter, which is the photon
statistic. Implementations for multiparameter SMFS and, in
particular, this fifth, non-classical fluorescence parameter are
explained in more detail in the following.
3. Multiparameter SMFS, Photon Statistics and
Analysis
From the classical point of view, fluctuations in the
number of photons detected cannot be reduced more than in
a light field which obeys Poissonian statistics, that is,
uncorrelated light. However, from the point of view of
quantum mechanics, states are possible for which the photondistribution is narrower than the Poisson distribution. These
states have no classical analogue and are a direct manifestation of the corpuscular aspect of radiation. To achieve a subPoissonian distribution, photons must have the tendency to be
separated in time. A simple system that can provide a light
source for nonclassical or sub-Poissonian light is a single twolevel quantum system, for example, a simplified model of a
fluorophore. A fluorophore requires a certain time after
emission of a photon to repopulate the excited state and emit
the next photon. This can be visualized in a histogram in
which the time lags between photons are represented. Under
pulsed laser excitation with laser pulses that are significantly
shorter than the excited-state lifetime of the fluorophore, no
photon-pairs from the same molecule (coincidences) are
detected within one laser pulse. On the other hand, according
to Poissonian statistics, there is a probability of detecting two
photons in two consecutive laser pulses. Similarly, for two
molecules, the probability of detecting two photons within
one laser pulse is one half that of detecting two photons in two
consecutive laser pulses.[89, 90] Consequently, for three molecules the probability is 0.67, and for four molecules 0.75, and
so on, until the classical limit of one is approached for an
ensemble. Accordingly, the ratio of the probability of detecting two photons within one laser pulse can be used to obtain
information about the number of emitting species. Vice versa,
if it is known that several fluorophores are present, the
absence of any coincidences can indicate annihilation processes and can help to reveal molecular interactions in
multichromophoric systems.[85, 91]
Figure 4 depicts a schematic set-up used for multiparameter SMFS. Usually the sample is excited by a pulsed laser
with a certain repetition rate defining the time-lag between
2648
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
consecutive laser pulses. The fluorescence signal is split by a
dichroic mirror and detected on two single-photon counting
avalanche diodes (APDs). The output TTL-pulses of the
APDs are fed into specialized personal computer (PC) plugin
cards
for
time-correlated-single-photon-counting
(TCSPC), nowadays with multidetector capability, high
time-resolution, and low dead time (e.g. TimeHarp200 from
Picoquant, or SPC-630 from Becker&Hickl). These boards
are specially developed for single-molecule experiments. For
each photon detected, three pieces of information are stored:
1) the macroscopic time, tmacro which is the arrival time after
the beginning of the acquisition (resolution 50 ns), 2) the
microscopic time, tmicro which is the time delay between the
laser pulse and the photon detection (resolution 8 ps), and
3) the channel (detector 1 (det1) or dectector 2 (det2)) from
which each photon was detected. Because TCSPC cards have
dead times in the range of 100–300 ns, it is necessary to apply
an electronic delay to one of the two APD signals so that two
fluorescence photons generated by the same laser pulse can
both be registered. The ratio of fluorescence intensity
detected at both APDs expressed as fractional intensity F2
(F2 = Idet2/(Idet1+Idet2)) is a measure of the wavelength of the
emission maximum of the fluorophore and enables the
monitoring of spectral fluctuations. By exciting with linearpolarized light and monitoring the fluorescence signal in a
perpendicular and parallel detection channel, the fluorescence anisotropy of single molecules can be measured.
Applying linear-polarized laser light rotated in time by the
help of an electrooptical modulator (EOM), delivers information about the number of absorbing species and their
relative orientations.[85, 91] From the time between consecutive
photons registered at both APDs under circular polarization
conditions, coincidence histograms can be constructed providing direct information about the number of emitting
species (photon antibunching measurements).[90]
This arrangement allows the simultaneous construction of
fluorescence intensity and lifetime trajectories for each
detector. Plotting a histogram of the macroscopic arrival
time in pre-selected time intervals, yields a so-called intensity
trajectory of the fluorescence, which is used to visualize the
intensity as a function of time (Figure 5). The histogram of the
microscopic arrival times yields a fluorescence decay. This
photon-by-photon detection mode allows the analysis of
fluorescence intensity and lifetime trajectories in a slidingscale fashion (Figure 5 a),[92, 93] or the construction of statespecific fluorescence decay curves (Figure 5 b).[78] In a slidingscale analysis, the signal trace of microscopic arrival times is
plotted into a series of fluorescence-decay histograms with a
constant number of events (e.g. 500 photon counts) which are
subsequently analyzed by a maximum-likelihood estimator
(MLE).[94–97] State-specific fluorescence decay curves are
constructed from the microscopic arrival times with respect
to the laser pulse by summing all the photon counts associated
with certain intensity levels.
Although two channels give only a poor spectral resolution, this simple geometry can easily reveal spectral
fluctuations on the millisecond time scale, and can substantially improve the identification accuracy of single fluorophores.[78, 98] Two detectors are also required for photon
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
Figure 4. Set-up for multiparameter SMFS including photon antibunching, and polarization-modulated excitation spectroscopy.
antibunching measurement in a Hanbury–Brown and Twisstype experiment.[99] In this case the sample is excited by
circularly polarized light and the signal is split onto two APDs
using either a 50:50 nonpolarizing beam splitter or a dichroic
mirror. The time lag between consecutive photons are
presented in so-called “interphoton times” histograms. The
use of an appropriate delay for the signal of one of the two
detectors enables the detection of quasi simultaneously
arriving photons. This approach circumvents the dead time
of the setup that comes from the detectors and dataacquisition electronics.
The interphoton-times histograms are corrected for the
delay thus directly indicate photons that are associated with
the same laser pulse at around zero interphoton time. As
expected for an ensemble of molecules (a Poissonian light
source), the peak pattern obtained clearly demonstrates the
classical behavior of the radiation field emitted by an
ensemble of molecules, that is, the central peak Nc is identical
in intensity to the lateral (neighboring) peaks N̄l (Figure 6 b).
In contrast, in case of a single molecule the central peak
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
disappears since the probability of emitting more than one
photon per laser pulse is negligible. Only a few background–
background, background–signal, and signal–background
events contribute to the central peak. With a backgroundcount rate of approximately 2 kHz S/B ratios of 30–50 are
obtained for single rhodamine derivatives adsorbed on a dry
cover slide.[89] This situation results in high ratios of the area
of the lateral peaks to the central peak (for example, 20
corresponding to an Nc/N̄l ratio of 0.05 in Figure 6 b). Previous
photon antibunching studies used fluorophores under biologically irrelevant conditions,[100, 101] or had to sample many
molecules to obtain sufficient statistics for photon antibunching measurements.[102, 103] Whereas, pulsed laser excitation
experiments deliver a photon antibunching signal from single
standard fluorophores immobilized on glass coverslides at
room temperature in under one second.[89] The main advantage of pulsed photon antibunching experiments is that the
desired saturation of the molecules singlet transition can be
achieved without high occupation of the triplet state, which is
important as the triplet state is an intermediate for photo-
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2649
Reviews
P. Tinnefeld and M. Sauer
Figure 5. a) Fluorescence intensity (black), and lifetime (red) as a function of time (count rate nc versus time t) for a single oxazine molecule
adsorbed on a cover slide and in the laser focus of a l = 635 nm
pulsed laser diode. Fluorescence lifetime categorization (binning) was
accomplished using a sliding-scale analysis with a stepsize of
500 photon counts.[78] b) State-specific fluorescence decays of the three
different intensity levels shown in (a) with nc > 22 kHz (blue),
10–22 kHz (black), and 2–10 kHz (green).
destruction. Therefore, even dynamics in the radiation field
emitted by the varying number of molecules in the laser focus
can be studied as shown in Figure 6 c (gray squares).[89, 90] The
almost constant Nc/N̄l ratio of under 0.2 demonstrates that the
photon probability distribution exhibits strong sub-Poissonian
character throughout the whole experiment, that is, only one
rhodamine molecule contributed to the detected light field.
Information about the orientations of transition dipoles of
the fluorophores investigated can be obtained using different
arrangements. Rotation of the linear polarization of the
excitation beam using a combination of an EOM and a
quarter-waveplate can reveal the orientation of fixed molecules from the phase shift of the emitted light.[76, 79, 91, 104]
Alternatively, measuring the polarization after excitation
with circularly polarized light, for example, in a two-detector
setup, can provide equivalent information. By using linearly
polarized excitation light, the technique can be used to
measure the polarization anisotropy of single molecules.[105]
As mentioned previously, imaging techniques can be used
advantageously to observe multiple fluorescence parameters
of single molecules in parallel. However, confocal geometries
offer a higher time resolution (usually ms compared to ms in
an imaging technique). Furthermore, not all the parameters
2650
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 6. a) Confocal fluorescence image of a 6 6 mm area with an
average density of approximately 0.5 rhodamine molecules per mm2. By
locating fluorescent molecules in the laser focus, fluorescence photons
are generated at predetermined times, in this case, every 62.5 ns after
excitation with the laser pulse (corresponding to the excitation frequency of 16 MHz. b) Histogram of interphoton times Dt recorded
from an ensemble of rhodamine molecules in solution (108 m), from
a single emitter, and from background. c) Multiparameter analysis of
the data recorded from a single rhodamine molecule. Fluorescence
intensity trajectory (black; count rate nc versus time t), fluorescence
lifetime t (red), and fractional intensity F2 (blue) calculated from the
intensities recorded after a dichroic beam splitter at detector 1 and
detector 2, F2 = Idet2/(Idet1+Idet2). Additionally, the antibunching signature as represented by the ratio of the area of the central peak to the
area of lateral peaks (Nc/N̄l) is shown as dark gray dots. The
stripe (light gray) indicates the statistical error of the Nc/N̄l ratio.
are accessible to imaging, for example fluorescence lifetime
and photon antibunching studies require confocal singlemolecule spectroscopy. Although fluorescence lifetime imaging is possible using a time-gated intensified CCD-camera, it
is in general not very practical at the level of single molecules
because of long measurement times (required to shift the
time-gate over the entire decay) and the limited photostability of conventional fluorophores. The other parameters,
such as color and polarization are accessible by imaging,
based on the dual-view concept originally developed by the
Kinosita group.[106] Splitting the signal with regard to the
property to be measured (e.g. a dichroic mirror for color and a
polarizing beamsplitter for polarization) and subsequent
imaging on different parts of the CCD-chip has proven to
be a versatile approach for multiparameter imaging that has
recently been extended to the simultaneous detection of color
and polarization with a single CCD-Camera.[107]
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
4. SMFS to Study Structure and Dynamics
One trend in SMFS is the investigation of multichromophore systems. These studies can be subdivided into two
categories: exploring the fundamentals of molecular interactions on the level of individual molecules and the exploitation of energy transfer reactions to measure distances
between molecules. To monitor, for example, conformational
dynamics of biomolecules, such as protein folding, distance
changes between certain amino acid or nucleic acid residues
within the polymer chain have to be measured with high time
resolution. From the spectroscopic point of view, any
distance-dependent method, such as energy transfer, proton
transfer, or photoinduced electron-transfer reactions, can be
used to record conformational changes as long as they are
reflected in alterations in the interaction with the probe
molecules, for example, fluorophores. Very prominent in this
context is the application of long-range dipole–dipole interactions, called fluorescence resonance energy transfer
(FRET), to biological systems.
4.1. Fluorescence Resonance Energy Transfer (FRET)
Generally, FRET involves nonradiative transfer of electronic excitation from an excited donor, D* to a ground-state
acceptor molecule A. The efficiency of this dipole–dipole
interaction depends on the distance between the fluorophores, the spectral overlap of the donor emission and the
acceptor absorption (the so-called overlap integral), the
refractive index of the media, the donor quantum yield, and
the relative orientation of the fluorophores. Frster type
resonant energy transfer occurs for allowed singlet–singlet,
and singlet–triplet transitions if the emission of D* and the
absorption of A overlap significantly. For such transitions the
Frster radius R0 ranges from 20 to 80 . The Frster radius
gives the separation at which the transfer efficiency equals
50 %. Owing to the strong separation-dependence (the
efficiency decays proportionally to the sixth power of the
separation) FRET is ideally suited to obtain information
about structure and structural changes of biologically important molecules.[108–114] In combination with SMFS, single-pair
FRET (spFRET) is suited to reveal conformational changes
in protein chains, and finds numerous applications in
“dynamic structural biology”.[115, 116]
spFRET was first demonstrated in 1996 and has become
one of the most popular tools in SMFS because the FRETrange covers the distance in which many important biological
processes take place (Figure 7). The groups of Ha and Weiss
in particular, demonstrated the potential of spFRET to reveal
conformational dynamics in single biomolecules on the
millisecond time scale showing, for example, unzipping of
DNA by a single helicase molecule, magnesium dependent
RNA dynamics, and protein folding.[116–124]
Deniz et al.[117] used spFRET to study protein folding of
chymotrypsin inhibitor 2 (CI2), a small, single-domain protein, at the single-molecule level. CI2 was selected as a model
system to study protein folding at the single-molecule level
because it shows two-state folding. Folded and unfolded
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Figure 7. a) Principle of spFRET to study conformational dynamics in
biomolecules. The molecule investigated is labeled site-specifically
with a donor (green) and an acceptor dye (red). Depending on the
donor–acceptor separation, the excited-state energy of the donor is
transferred to the acceptor. b) Fluorescence trajectory (count rate, nc,
versus time, t) of a single donor–acceptor labeled DNA hairpin (Cy35’-CCCCTA AGTAGTTCCTCAC T(Cy5)AGGGG-3’) immobilized in the
laser focus by biotin–streptavidin binding. Structural changes, in this
case opening and closing of the hairpin structure, are directly reflected
in alterations in the FRET efficiency (e.g. at t 2, and 12 s). c) FRETefficiency, Eeff (Eeff = IA/(IA+ID) histograms (background corrected)
extracted from individual fluorescence bursts of Cy3/Cy5 labeled
double-stranded oligonucleotides (donor–acceptor separation: ca.
3.4 nm, corresponding to 10 base pairs) that are freely diffusing
through the observation volume, and corresponding Gaussian fits. The
large peak in zero energy-transfer efficiency arises from donor-only
labeled DNA (nonhybridized Cy3 oligonucleotides), and premature
photobleaching of the acceptor. The appearance of two maxima in the
FRET-efficiency histogram instead of a broad distribution implies that
conformational fluctuations between sub-states with different donor–
acceptor orientations or separations are on average slower than the
measurement time (typically 0.5–1 ms). If the conformational fluctuation rates were faster than the measurement time (which equals the
diffusion time), the peak corresponding to high FRET efficiency would
at least be broadened or gradually shifted to lower FRET efficiency.
subpopulations were revealed from measurements of freely
diffusing single protein molecules. Different distance distributions between the donor and acceptor in the folded and the
denatured state of the protein cause different FRET efficiencies. These subpopulations can be clearly resolved because
only one molecule is probed at a time. Eaton and co-workers
studied the equilibrium collapse of a polypeptide in the
denatured state by monitoring spFRET efficiencies of
unfolded populations as a function of denaturant concentration.[3] They could show that dynamic information, such as
polypeptide reconfiguration times, can be extracted from the
widths of single-molecule FRET efficiency distributions. The
determination of barriers of reconfiguration times allows, in
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2651
Reviews
P. Tinnefeld and M. Sauer
turn, the heights of free energy barriers of folding to be
estimated. As a result of the availability of hundreds to
thousands of events in only a few minutes, single-molecule
studies in solution can easily differentiate between subpopulations of analyte molecules in heterogeneous ensembles
(Figure 7 c).[92, 125, 126] The kinetics of DNA hairpin–loop fluctuations is one of the fastest conformational fluctuations
observed to date at the single-molecule level. By using a
combination of FRET and FCS experiments, the rates for the
opening and closing of DNA hairpins were determined to a
few tens of microseconds, depending on the size and sequence
of the loop and stem.[127–129]
To gain information about slower conformational dynamics of biopolymers, such as absolute values of transition rates
between differently folded states, time trajectories of individual DNA, RNA, or protein molecules immobilized on a
solid support have to be monitored in real time (Figure 7 b).
However, care must be taken to ensure that minimal
perturbation of the molecules results from the immobilization
on the modified glass surfaces.[130–132] Recently,[124] a new
immobilization method was demonstrated that enables the
observation of individual protein molecules, unperturbed by
surface effects. Rhoades and co-workers monitored adenylate
kinase molecules as they fold and unfold over many seconds.
Individual, site-specific D/A-labeled proteins were trapped
within surface-tethered lipid vesicles, thereby allowing spatial
restriction without inducing any spurious interaction with the
surface. By simultaneous measuring of donor fluorescence
lifetime, intensity, and anisotropy the two main sources of
error in FRET-based distance calculations (local quenching
effects and restricted dye mobility) can be excluded.[105, 133] In
combination with refined labeling strategies, new multiparameter spFRET experiments enabled the direct observation of
the proton-powered subunit rotation in single membranebound F0F1-ATP synthase molecules.[134]
These studies highlight that spFRET is a powerful tool to
unravel the heterogeneity of conformational transitions in
“dynamic structural biology” and to elucidate kinetic and
static properties of biopolymers in both the folded and the
unfolded states. Moreover, the choice of a suitable observable, such as the FRET-efficiency, allows for the motion of the
biopolymer on its energy surface to be followed in real time.
4.2. Photoinduced Electron-Transfer Reactions
Other conformational dynamics with functional significance, for example in enzymes, often appear on much smaller
spatial scales (down to several ngstrms). To monitor subnanometer conformational transitions of biopolymers, photoinduced chemical reactions of fluorescent dyes with molecular quenchers are of great interest. In contrast to FRET,
photoinduced charge-transfer reactions require close proximity or even van der Waals contact between dye and
quencher for fast and efficient conversion of electronic
excited-state energy. To interpret fluorescence-quenching
caused by photoinduced electron transfer, the transfer
mechanism and its distance dependence has to be well
understood. Generally, quenching of fluorophores in the
2652
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
first excited singlet state by electron donors or acceptors
results in the formation of a radical ion pair AC DC+, which
returns to the ground state by charge recombination. The
efficiency of charge separation is controlled by the relationship between the free energy of the reaction, the reorganization energy, and the distance between the donor and
acceptor.[135, 136]
Since efficient electron-transfer reactions usually require
close proximity of donor and acceptor they are ideally suited
to study small conformational changes in biomolecules, and to
monitor, in particular, equilibrium fluctuations in proteins.
For example, Chattopadhyay et al. coupled a fluorescein dye
selectively to a hydrophobic cavity of an intestinal fatty acid
binding protein and used FCS to study fluorescence fluctuations on the microsecond time scale.[137] Since fluorescein is
efficiently quenched by several amino acids, in particular, by
tryptophan residues[138, 139] it is likely that the intensity
fluctuations observed are caused by fluctuations in the
interaction geometry of the fluorescein moiety and a tryptophan residue located in close proximity. Thus, small conformational changes, such as equilibrium fluctuations, can be
directly reflected in the charge separation efficiency. Electron-transfer reactions enable the direct monitoring of subnanometer conformational fluctuations of a protein in its
native state at the single-molecule level with microsecond
time-resolution.
Very recently, Xie and co-workers used photoinduced
electron transfer between a tyrosine residue and the flavin
moiety as a probe for ngstrm-scale structural changes in
single protein molecules.[140] Correlation of the fluorescence
lifetime fluctuations measured from single flavin molecules
enabled the monitoring of conformational dynamics with
submillisecond time resolution. Their data strongly suggest
the existence of multiple interconverting conformers, a result
which helps to explain the fluctuating catalytic activity of
flavin reductase. Furthermore, the monitoring of sub-nanometer conformational changes at the single-molecule level
under equilibrium conditions is advantageous because it
allows the proteins conformational memory to be measured
quantitatively at room temperature.
Alternatively, efficient photoinduced-electron transfer
between fluorescent dyes and certain amino acids or DNA
nucleotides can be used to record conformational fluctuations
in peptides and oligonucleotides at the single-molecule level
with sub-microsecond time-resolution.[5, 141] The basic idea of
the method is that the fluorescence of suitable fluorophores is
selectively quenched by tryptophan or guanosine residues
through photoinduced electron transfer, but only upon
contact formation. Except for tryptophan and guanosine
itself, all other amino acids and DNA bases quench with rates
that are substantially smaller.[138, 139, 142–146] The general principle of this new method for measuring the kinetics of
intramolecular contact formation in biopolymers is depicted
in Figure 8 a. A suitable fluorescent dye and a tryptophan or
guanosine residue are incorporated into a peptide or oligonucleotide. The dynamics of the backbone will lead to
conformations where the dye and the quenching residue
come into contact (right structures in Figure 8 a). After
excitation of the dye molecule, it either emits a fluorescence
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
Figure 8. a) Principle for measuring the kinetics of intramolecular contact formation in peptides or oligonucleotides. The dynamics of the
backbone will lead to conformations where the dye (red) and the tryptophan or guanosine residue (blue) come into contact. This contact
formation results in complex formation and subsequent efficient fluorescence quenching by photoinduced charge separation from the
ground-state tryptophan or guanosine to the first excited singlet state
of the dye. Hence, the structures on the right can be considered to be
non-fluorescent. b) FCS-curves measured from 109 m solutions of
labeled peptides MR121-SQETFSDLWKLLPEN (green), and MR121SQETFSDLFKLLPEN (blue) in PBS pH 7.4. Strong static quenching by
photoinduced electron transfer upon contact formation between the
dye and the tryptophan residue is reflected in the appearance of a
short correlation term in the 10–1000 ns range (gray area). Abbreviations for the amino acid residues are: A alanine, D aspartic acid, E glutamic acid, F phenylalanine, G glycine, I isoleucine, K lysine, L leucine,
N asparagine, P proline, Q glutamine, R arginine, S serine, T threonine,
W tryptophan.
photon (dependent on the fluorescence quantum yield) or is
quenched by an electron transfer reaction with the tryptophan
or guanosine residue. Depending on the dye structure,
quenching is dominated either by dynamic or static quenching.[139] If contact between the dye and the quenching residue
results in complex formation in which there is a nearly
coplanar stacking of quencher and dye, immediate efficient
fluorescence quenching occurs. Thus, the right structure in
Figure 8 b can be considered to be essentially nonfluorescent.
Hence, interphoton-times distribution analysis or FCS can be
used to reveal the rate for contact formation (association
rate), k+, as well as the dissociation rate, k.
To demonstrate the potential of the method, the rate of
contact formation between an oxazine dye (MR121) attached
at the N-terminal end and a tryptophan residue, localized
within the chain of two peptides with lengths of 15 and
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
20 amino acids, were investigated.[5] The peptides are flexible,
linear fragments of the amino terminal transactivation
domain of human p53, which is the target for several protein
partners, such as the TATA box-binding protein (TBP),
MDM2, or antibodies from the humoral immune response of
cancer patients. Fast FCS on two detectors was used to
measure the tryptophan-induced fluorescence quenching
upon contact formation in 109 m peptide solutions. Association (contact formation) and dissociation rates in the range
of a few hundred nanoseconds were measured for both
peptides, consistent with data on the dynamics of unstructured peptides for which a different ensemble methodology
was used.[147] The advantages of this fluorescence-quenchingbased technique over other techniques are the ease of
performing the experiments which can be carried out in
water under air with single-molecule sensitivity. The exchange
of the tryptophan by a phenylalanine residue results in the
disappearance of the short correlation term in the 100 ns time
range (Figure 8 b).
The extraordinary sensitivity of the described fluorescence-based method is especially important for the application of “smart” fluorescent probes, used for the detection of
minute amounts of target structures from biological samples.
Both peptides used in the sub-microsecond kinetic studies
discussed above are the target of p53 autoantibodies. These
so-called epitopes for p53 autoantibodies, are known to be
universal and highly specific tumor markers in cancer
diagnosis.[148] Hence, both fluorescently labeled peptideepitopes can be used as photoinduced electron-transfer
(PET) biosensors to detect single p53 autoantibodies.[149]
Upon specific binding to an p53 autoantibody the peptide
chain adapts to the shape of the antibody cleft. Consequently,
contact formation between tryptophan and MR121 is prevented. The resulting increase in fluorescence intensity can be
used to signal binding events. The single-molecule sensitivity
of the assay format allows for direct monitoring of p53 antibodies, present in blood serum samples of cancer patients.[146]
Selective quenching of oxazine dyes, such as MR121 and
Atto655, by guanosine residues through electron transfer is
also advantageous for the development of single-moleculesensitive DNA or RNA probes, for example, DNA-hairpin
sequences (Figure 9).[143] If quenching interactions between
the dye coupled to one end of the hairpin stem and guanosine
residues, located on the other end, are interrupted, for
example, through binding of a complementary DNA
sequence, or by cleavage by an endo- or exonuclease
enzyme, fluorescence of the DNA hairpin is restored. DNA
hairpin sequences labeled with a single oxazine dye at the 5’end increase their fluorescence upon hybridization by up to
20-fold, which provides the basis for a cost-effective, and
highly sensitive DNA–RNA detection method.[144] Other
results have demonstrated that immobilized DNA hairpins
modified for photoinduced intramolecular electron transfer
from guanosine residues to excited fluorophores can be used
to efficiently detect the presence of single target DNA or
RNA molecules, even in the sub-picomolar concentration
range in a reasonable time.[130] The method is therefore ideally
suited to search for specific sequences in extremely low
concentrations of target sequence.
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2653
Reviews
P. Tinnefeld and M. Sauer
Figure 9. a) Schematic representation of the operation of a DNA hairpin modified for the selective guanosine quenching of fluorophores.
The fluorophore (red) is attached to the cytosine containing 5’-end and
quenched by guanosine residues (blue) by photoinduced electron
transfer. Upon hybridization to the target sequence (complementary to
the loop sequence), the fluorescence is restored as a result of a conformational reorganization that opens the stem. b) 3’-biotinylated hairpins (MR121-C6-5’-CCCCAT20TGGGG-3’) labeled with the oxazine derivative MR121 were immobilized on a cover slide by biotin–streptavidin
binding.[130] Scanning was performed from top left to bottom right
using modulated excitation polarization. The confocal image
(20 20 mm, 50 nm/pixel) shows fluorescence spots from single DNA
hairpin molecules hybridized to their complementary target sequence
(0–16 counts in 3 ms). The arrow indicates the addition of an excess
of target sequence (oligo(dA)30). Most fluorescent spots show a
demodulated fluorescence intensity indicating freely rotating DNA hairpin molecules. The modulation stripes visible in the two spots marked
by circles indicate rotationally stationary absorption dipoles of the fluorophores, that is, the fluorophores are stuck on the surface.
5. Use of SMFS to Study Photophysics in Multichromophoric Systems
With respect to single-molecule spectroscopic techniques,
it should be pointed out that some characteristics, such as
digital photobleaching or blinking, which have been generally
accepted as characteristics of single chromophores have to be
reconsidered. Single-molecule studies of molecular interactions in multichromophoric systems revealed the occurrence
of fluorescence intermittences (off states) of several chromophores at the same time.[104, 150–153] In multichromophoric
systems various types of chromophore–chromophore interactions are possible depending on the interchromophore
distance. Such interactions include Dexter energy transfer,
dimer and excimer formation, electron-transfer reactions,
photochemistry, and intermolecular quenching. Systems
investigated to date include natural chromophores, such as
the light harvesting complex LH2,[154, 155] light harvesting
proteins, such as phycoerythrins and allophycocya-
2654
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
nine,[103, 156, 157] genetically encodable proteins, such as
DsRed,[24, 158–160] and artificial multichromophoric systems
from simple dimers[161–163] to complex systems, such as lightemitting conjugated polymers[150–152, 163, 164] and well defined
dendritic or linear systems.[85, 91, 104, 156, 165, 166]
One prominent finding for many multichromophoric
systems is the observation of collective blinking, that is, the
complete fluorescence intermittency of all fluorophores
simultaneously. Blinking of single molecules was one of the
early phenomena observed and can be attributed to various
processes. Most notable among the possible processes is
triplet blinking, that is, the reversible transition to the triplet
state through inter-system crossing.[167, 168] SMFS has shed light
on some of the mysteries why several molecules seem to
undergo collective transitions to off-states. The common
ground for explanations of collective blinking is that only one
fluorophore undergoes a transition to an off state but that the
nature of this off state is such that it quenches the
fluorescence of the remaining fluorophores.[91, 150, 164] This
principle is, for example, applicable for intersystem crossing
where the molecule in the triplet state “absorbs” all the
excitation energy from the other fluorophores by efficient
Frster-type energy transfer.[91, 104, 156, 163] Thus, the emission
(S1!S0) of the fluorophores is in resonance with the T1!Tn
transition of the fluorophore in the triplet state. This situation
yields a molecule in a higher excited triplet state, Tn. The
energy from Tn is rapidly dissipated thermally thus leaving the
triplet fluorophore in T1 again. The molecule in the triplet
state can be rapidly cycled through T1 and Tn until it returns to
the singlet state by reverse intersystem crossing or phosphorescence.[163, 169] The energy transfer from a fluorophore in the
singlet state to a fluorophore in the triplet state is commonly
called singlet–triplet (ST) annihilation because one S1
excited-state-energy quantum appears to be annihilated.
Besides singlet–triplet annihilation, energy hopping
between identical chromophores has to be considered in
multichromophoric systems as well. Energy hopping is the
transfer of excited-state energy from one molecule to another
through weak dipole–dipole coupling (homo energy transfer),.[104, 156] As a result of the different overlap of the
corresponding spectra, different R0 values and energy-transfer efficiencies are calculated for energy hopping and singlet–
triplet annihilation in multichromophoric systems.
Recently, it was found that a third resonance energy
transfer pathway with a third R0 value has to be taken into
account when studying chromophore interactions, this is
singlet–singlet (SS) annihilation. In this case, two chromophores have to be excited almost simultaneously. Depending
on the spectral overlap of the S1!S0 emission and S1!Sn
absorption of the chromophores, energy transfer can occur.
This transfer yields one chromophore in the ground state and
the other in a higher excited state Sn. Internal conversion
(nonradiative relaxation) rapidly converts the Sn state into the
first excited singlet state. The net reaction is the “annihilation” of one exciton. As has been shown, singlet–singlet
annihilation in multichromophoric systems is easily uncovered by photon antibunching experiments because all simultaneous excitations of more than one molecule in a multichromophoric system are annihilated.[85, 91, 170] Only when the
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
multichromophoric system is significantly larger than the
Frster radius for singlet–singlet annihilation does the
vanished central peak in the interphoton-times histogram
start to reappear.[171]
In summary, SMFS has shown that multichromophoric
systems behave like single emitters in some regards, they
show collective off states and photon antibunching. The basis
of these processes are three Frster-type processes through
which the fluorophores communicate, that is, energy hopping,
singlet–singlet annihilation, and singlet–triplet annihilation
(Figure 10). The efficiency of energy hopping and SS-annihi-
Figure 10. FRET efficiency, Eeff, versus donor–acceptor separation, R,
calculated for three different Frster-type energy-transfer processes
occurring between identical peryleneimide chromophores in rigid multichromophoric dendritic systems: energy hopping, singlet-singlet
(SS)-annihilation, and singlet-triplet (ST)-annihilation.[91] R0 values for
the different transfer pathways were calculated from the absorption
spectra S0–S1, S1–Sn, and T–Tn and the corresponding emission spectra
of the donor assuming an average k2 of 1.92 as computed from molecular mechanics calculations.[91]
lation can be determined by both ensemble and singlemolecule time-resolved fluorescence experiments, for example, by lifetime, anisotropy, and antibunching measurements.
On the other hand, ST-annihilation and its efficiency can only
be measured at the single-molecule level by the analysis of
collective off states.
As a result of the R6 distance dependence, separations in
the range 0.5 R0 to 1.5 R0, that is, with FRETefficiencies Eeff in
the range 0.98 to 0.10, are suitable for FRET measurements.
When using two peryleneimide chromophores this corresponds to distance changes of 2.39–7.05 nm for energy
hopping, 3.52–10.56 nm for SS annihilation, and 3.93–
11.79 nm for ST annihilation (Figure 10).[91] Thus, the use of
two identical dyes enables the measurement of distances
between 2 and 12 nm.[91, 171] In addition, the synthesis of
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
molecules equipped with two identical fluorophores is easier
than the synthesis of compounds carrying two different
fluorophores. In the case of stronger interactions, that is,
beyond long-range FRET-interactions, other effects, such as
alterations in the fluorescence spectrum and lifetime can be
observed.[104, 172]
6. Molecular Photonic Wires
The knowledge obtained from fundamental photophysical
studies represents the basis for the future development of
nanoscale optical devices, such as molecular photonic wires.
In contrast to an electric wire, a photonic wire transports
excited-state energy rather than electrons. On the level of
molecular devices this is of great advantage because it
circumvents the connection problem, that is, the bottleneck
that occurs when trying to connect the molecular with
macroscopic devices. In the case of a photonic wire, excitedstate energy is induced into the wire by means of light,
transported through the wire, and finally emitted at another
wavelength and location or the energy is used for an electrontransfer reaction, that is, the conversion of excited-state
energy into an electric charge. The basic idea of a photonic
wire was first expressed by Lindsay et al.[173] and quite some
effort was put into the synthesis of sophisticated porphyrins
arrays. Besides problems that occur owing to energy sinks
caused by local inhomogeneities throughout the strongly
coupling porphyrins, these chromophores are hard to study by
SMFS because of their low fluorescence and high triplet
quantum yield.
On the other hand, it is clear that nanoscale devices have
to be accessible for investigations on the level of the
individual components. The ideal photonic wire consists of a
very regular dye arrangement which allows for facile incoherent hopping but does not result in the alteration of
photophysical properties of the individual dyes that leads to
the formation of so-called energy sinks. Therefore an alternative strategy has recently been suggested which is based on
1) the use of conventional fluorophores with high quantum
yields, 2) an energy cascade as the driving force for the
excited-state energy to ensure unidirectionality, 3) an
arrangement of the fluorophores such that strong interactions
promoting dimer formation and quenching are prevented.[174, 175] Thus, energy will be transferred through weak
dipole–dipole induced chromophore interactions according to
Frster theory.
From the synthetic point of view, this approach requires
the development of a strategy to label a rigid scaffold in a
well-defined manner with various fluorescent dyes. One
possibility is the use of double-stranded DNA labeled with
different chromophores. These transfer light, that is, energy,
from donor to acceptor through several dipole–dipole
induced energy-transfer steps using spectrally separated
dyes. The chromophores have to form an energy cascade,
that is, the input dye has the highest energy and the output dye
has the lowest energy, ensuring the unidirectionality of the
photonic wire (Figure 11 a). The optimal photonic wire, is the
best compromise between unidirectional driving force and
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2655
Reviews
P. Tinnefeld and M. Sauer
minimized loss of energy, for example, by the exploitation of
homo-FRET processes for some energy-transfer steps.[175, 176]
The unique features of double-stranded DNA make it an
ideal building block in nanoscale molecular devices. With a
persistence length of DNA of 50 nm, DNA constitutes a stiff
scaffold to which chromophores can be attached.[177] In
addition, the use of DNA offers many well developed labeling
and post-labeling strategies for the introduction of a variety of
different chromophores. The strength of this approach is the
synthetic simplicity by which a large number of chromophores
can be put into a well defined order. Starting from a singlestranded DNA fragment counter sequences that carry the
desired chromophores can be hybridized to the single strand
(Figure 11 b). Very recently, we demonstrated the synthesis
and spectroscopic characterization of a unidirectional photonic wire based on four highly efficient energy-transfer steps
between five spectrally different chromophores covalently
attached to double-stranded DNA.[175] SMFS performed on
four spectrally separated detectors uncovered subpopulations
that exhibit overall energy-transfer efficiencies of up to 90 %
across a distance of 13.6 nm and a spectral range of
approximately 200 nm. In the future it should be possible to
define not only the distance between the chromophores but
also their relative orientation. Such control might enable the
use of fluorophores with smaller spacing thus allowing very
fast energy-transfer through the Dexter mechanism.
Approaches include the use of supramolecular chemistry
possibly including template supports such as zeolites.[178]
A photonic wire however, is only one optoelectronic
device complementary to nanoscale electronic devices. The
next level of complexity is the insertion of a gate into a
photonic wire. Recently, a single-molecule fluorescence
photoswitch was developed by Irie et al.[179] They showed in
a two color experiment that a fluorophore (bis(phenylethynyl)anthracene) connected to a switchable quenching unit (a
diarylethene derivative) could be switched on and off by l =
488 nm and 325 nm light, respectively. UV light was used to
activate the quencher while 488 nm light was used for
deactivation of the quenching unit and probing the fluorescence of the fluorophore. Although, in general disadvantageous, the use of the same wavelength (l = 488 nm) for
probing and switching was possible because the deactivation
Figure 11. a) Principle of a unidirectional photonic wire. Excited-state energy is transferred by several energy-transfer steps from one end to the
other. The chromophores used form an energy cascade to ensure the unidirectionality of the photonic wire. b) Synthesis strategy of a unidirectional photonic wire. To introduce different dyes in the correct order, short oligonucleotides labeled with two different dyes are hybridized to the
complementary DNA strand. c) Ensemble excitation (green) and fluorescence emission (red) spectra (relative fluorescence intensity If, versus
wavelength l, of a DNA-based photonic wire carrying three dyes, rhodamine 110 (R110), tetramethylrhodamine (TMR), and ATTO 590 at interchromophore distances of 10 base pairs (ca. 3.4 nm). The transmission curve of the dichroic beam splitter used for SMFS of individual photonic wires
is shown in gray. Calculation of the overall efficiency for energy transfer from the first donor R110 to TMR to the final acceptor ATTO 590 gives
very different values depending on whether it is determined by the decrease in donor intensity (40 %) or the relative intensities of the excitation
spectrum (60 %). On the other hand, it can be anticipated that owing to the size and complexity of the system, different conformational orientations of the chromophores, insufficient hybridization, and/or additional quenching pathways will generate an inhomogeneous broadening of the
transfer efficiency observed form ensemble measurements. d) Fluorescence intensity trajectory recorded from a single photonic wire adsorbed on
a cover slide by using spectrally resolved fluorescence lifetime imaging microscopy (SFLIM) with spectral detection windows from 520–
590 nm (green) and 590–680 nm (red); overall intensity is shown in black.[78] During the first 5 s of the trajectory, the signal is dominated by the
final acceptor ATTO 590 owing to efficient energy transfer. This result is also corroborated by time-resolved measurements. The fluorescence
decay recorded during the first 3 s (gray area) on the short wavelength detector (green) shows distinct quenching while the decay measured on
the long wavelength channel (red) shows a pronounced fluorescence rise time as a result of energy transfer and a long fluorescence decay time of
about 4 ns as expected for the ATTO 590 dye. After 4 s a collective off state appears. After about 5 s the final acceptor ATTO 590 photobleaches
which is reflected in an abrupt change in fractional intensity F2 (blue) After photobleaching of the TMR dye at 7 s the F2 value increases again
slightly as a result of the fluorescence emission of the remaining R110 molecule and the transmission properties of the filter set used.
2656
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
(isomerization) is about 1000 times less efficient than the
activation of the quenching unit. Thus, probing and isomerization can be controlled by the excitation light intensity. In a
perfect photoswitch, however, switching should be completely
independent of probing. Efficient quenching of the fluorophore by electron transfer is a more appropriate quenching
mechanism for photoswitches. It could be possible to selectively address fluorophores within a photonic wire by the
attachment of a switchable electron donor or acceptor
(Figure 12). Ideally, these redox-photoswitches could be
Figure 12. Idealized scheme of a unidirectional photonic wire controlled by a redox photoswitch. Photoisomerization of the linker of the
electron-transfer quencher (Q) results in an altered interaction geometry between the fluorophore and the quencher. Switching of the wire is
accomplished by excitation with a second and third wavelength (ñ1,
and ñ2). In the off state (top) photoinduced electron transfer (PET)
between one of the fluorophores and the quencher (Q) is much faster
than intermolecular energy transfer (ET) to the next fluorophore
(kPET @ kET). This arrangement allows selective switching of the photonic wire between the on and off state.
addressed by two different wavelengths at which the fluorophore does not absorb. A third wavelength is used for
probing the state of the switch. Upon photoisomerization the
unit should change its redox potential or alter the donor–
acceptor interaction geometry such that it becomes a
quencher of the fluorophore in one state while thermodynamics, kinetics, or the conformational state of the redoxphotoswitch forbid the photoinduced electron-transfer reaction in the other (Figure 12).
7. SMFS in Biology
In recent years, considerable progress has been made
towards a better understanding of molecular machines, that is,
highly organized self-assembled nanostructures, and their
functions in living cells.[180–185] Although, rough data about the
composition of biomolecular machines and their molecular
interactions exist, much is still unknown about how nuclear
metabolism, metabolic enzymes, and regulatory factors are
organized. Furthermore, it is generally assumed that the
multitude of cellular building blocks in many of these
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
molecular machines do not act synchronously and might
have different kinetic pathways. Since, for example, transcription factories containing polymerase II have a small size
of approximately 40 nm and contain on average only around
8 active polymerase molecules in mammalian cells,[182] singlemolecule sensitive detection methods are required to study
the dynamic interplay of the cellular building blocks and their
precise location and separations with high spatial and
temporal resolution. Single-molecule fluorescence spectroscopic techniques are advantageous owing to the possibility to
expand the analysis to three dimensions.[40] However, still
single-molecule techniques seem too complicated to many
researchers and in biology the number of genetic and
electron-microscopic studies still outnumbers single-molecule
fluorescence studies.
In the past few decades, in situ hybridization (ISH) and
labeling methods have been optimized to such an extent that
the microscopic detection of specific mRNA or protein
molecules by using fluorescently labeled nucleic acid or
antibody probes is possible.[186–189] Although a lot of effort is
still being put into optimizing these procedures on fixed cells
and tissue materials, there is an increasing demand for
techniques that allow the visualization of proteins and
RNAs in living cells. The reason for this interest in livingcell analysis is the concern that the observed localization
patterns in fixed cells may not reflect the situation in living
cells. Fixation and cell pretreatment may disturb localization
patterns and to some extent deteriorate cell morphology.[190]
In addition, information about the dynamics of DNA, RNA,
or protein synthesis and transport can be obtained best when
these processes are studied in living cells.
7.1. Fluorescence-Labeling Strategies for Cell Imaging
For the investigation of biologically relevant samples,
target molecules have to be labeled with a fluorescent tag. A
central issue for SMFS in cells is fluorescent labels, these have
to fulfill special requirements, such as high biocompatibility,
high photostability, and retention of biological function. The
first problem, however, is site-specific labeling inside a living
cell and some procedures have been reviewed recently.[23, 191]
In the simplest case the molecules to be investigated are
prepared in vitro utilizing standard techniques for labeling
biomolecules. Subsequently cell-loading is carried out by
known chemical, electrical, mechanical, or vesicle-based
procedures. To measure, for example, the localization of
RNA in living cells the target sequence has to be labeled with
a fluorescent complementary sequence. A problem that needs
to be resolved in the fluorescence in vivo hybridization
(FIVH) approach is the efficient delivery of probes to sites
in a cell where they can hybridize with their target sequence.
The most direct method to introduce probe molecules into a
cell is by microinjection using micropipettes. Alternatively,
liposomes can be used to introduce probe molecules into live
cells.[192]
The microinjection technique is based on the use of a
micropipette with a very small diameter tip and the application of a higher pressure at a predetermined time. However,
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2657
Reviews
P. Tinnefeld and M. Sauer
this technique does not exhibit the necessary reproducibility
required for SMFS, that is, the injection of only a few
molecules per cell. Recently, fluorescently labeled oligonucleotides were microinjected into the cytoplasm and nucleus
of living 3T3 mouse fibroblast cells using a micropipette.[83]
Since, the micropipette used has an inner diameter of 500 200 nm at the very end of the tip—comparable to the
diameter of the confocal detection volume applied—almost
all molecules leaving the micropipette were detected.[9, 193, 194]
This situation means that the number of injected molecules is
known. As a consequence, quantitative molecular information on a single cell level can be obtained.
Another very promising method is known as protein
transduction. Several naturally occurring proteins have been
found to enter cells easily, including the TAT protein from
HIV.[195, 196] Specific short sequences within the larger molecule account for the transduction abilities of these proteins.
These arginine-rich peptides allow efficient translocation
through the plasma membrane and subsequent accumulation
in the cell nucleus. Therefore, they could be useful vectors for
the intracellular delivery of various nonpermanent drugs
including antisense oligonucleotides and peptides of pharmacological interest.[197] Cellular uptake of these cationic cellpenetrating peptides has been ascribed to a mechanism that
does not involve endocytosis. Live-cell penetration studies
(on HeLa cells and human foreskin kerantinocytes) without
fixation, using fluorescently labeled, peptidase-resistant boligoarginines, demonstrate that longer-chain b-oligoarginines (8 and 10 residues) enter the cells and end up in the
nuclei, especially in the nucleoli, irrespective of temperature
(37 8C, and 4 8C) or pretreatment with NaN3.[198, 199]
In a related strategy, the high affinity between a target
molecule and a smaller molecule, which can be prepared
in vitro, is exploited. This strategy is more complicated than
in vitro labeling as it requires a particularly high binding
affinity between the partners to avoid unspecific binding
inside the cell. Nevertheless, this method is advantageous
when the investigated molecule is not available as a purified
compound or when the location of the target can not simply
be reproduced. An antibody or another specific ligand can
perform this task of in situ labeling.[200] However, often it still
remains a problem to target structures selectively with analyte
concentrations in the range required for single-molecule
spectroscopy (i.e. 1010 m). Therefore in an improved version
of the indirect labeling strategy the ligand senses the binding
event and the fluorescence of the label is switched on only
upon specific binding.[143, 146] Alternatively, fluorescence can
be activated in specific locations within the cell by using a
label which is bound to a quencher by a photoactive linker.
Cleavage of the quencher through photodissociation of the
linker allows fluorescence of the chromophore. To monitor,
for example, the synthesis of a protease enzyme, nonfluorescent protease substrates which fluoresce upon specific
cleavage by the protease could be used.[201]
However, the most elegant way to specifically label
proteins in vivo is genetic labeling. The fluorophore used is
a genetically encoded protein, such as the green fluorescent
protein (GFP) and its relatives. GFP from the bioluminescent
jellyfish Aequorea victoria has revolutionized many areas of
2658
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
cell biology and biotechnology because it provides direct
genetic encoding of strong visible fluorescence.[202, 203] GFP
can function as a protein tag because it tolerates N- and Cterminal linking to a broad variety of proteins, many of which
retain their native function.[204, 205] According to the method,
the DNA sequence coding for GFP is placed immediately
adjacent to the sequence coding for the protein of interest.
During biosynthesis, the protein will be prepared as a GFPfusion protein. GFP comprises 238 amino acids and has a
barrel-like cylindrical structure where the fluorophore is
highly protected, located on the central helix of the geometric
center of the cylinder. These cylinders have a diameter of
about 3 nm and a length of about 4 nm, that is, significantly
larger than common fluorophores which are less than 1 nm.
Although single-molecule studies of GFP- or yellow fluorescent protein (YFP)-fusion proteins were demonstrated in the
membrane of living cells (GFP, and YFP are the brightest and
most photostable candidates),[206–209] their complicated photophysics and low photostability renders their application in
single-molecule measurements very difficult.[24, 210–212]
The discovery of new fluorescent proteins from nonbioluminescent Anthozoa species,[213] in particular, the redshifted fluorescent protein DsRed, opened new horizons for
multicolor labeling and FRET applications. However, several
major drawbacks, such as slow maturation and residual green
fluorescence, need to be overcome for the efficient use of
DsRed as an in vivo reporter, especially in SMFS applications.[158–160] To improve maturation properties, and to reduce
aggregation, a number of other red fluorescent proteins and
variants of DsRed have been developed, for example,
DsRed2, DsRed-Express, eqFP611, or HcRed1.[214–216]
Recently, the oligomerization problem of DsRed has been
solved by mutagenetic means (mRFP1).[217] An elegant
alternative to common FRET methods to study protein
interactions in living cells is bimolecular fluorescence complementation. The new approach is based on complementation between two nonfluorescent fragments of the yellow
fluorescent protein (YFP) when they are brought together by
interactions between proteins fused to each fragment.[218]
7.2. SMFS in Living Cells
While individual fluorescent molecules can be detected in
solution and on surfaces with good S/N ratios, a key question
is whether single-molecule methods could be developed to
study complex molecular processes in living cells. In contrast
to clean and well-controlled conditions in vitro, the cellular
environment contains a broad collection of biological macromolecules and fluorescent materials, such as porphyrins and
flavins. This complex environment is known to produce
intense background fluorescence, commonly known as autofluorescence. Therefore, often nanomolar or higher concentrations of fluorescently tagged proteins or oligonucleotides
have been applied to enable measurement of intracellular and
intranuclear diffusion by FCS.[219, 220]
However, by using cells with very low expression levels,
fluorescence signals of single diffusing EGFR-EGF fusion
proteins in living cells could be detected.[221] Confocal
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
fluorescence microscopy has been applied as well to detect
fluorescence bursts of rhodamine-labeled oligonucleotide
molecules in living cells.[222] To reduce the autofluorescence
of the cell culture medium arising upon excitation in the green
wavelength range, most measurements were performed in
PBS (phosphate buffered saline). The first intracellular
trajectories of single protein molecules labeled with several
fluorophores was shown by Kubitscheck and co-workers.[223, 224] Employing a laser video-microscope optimized for
high sensitivity and high speed, single ribonucleoproteins
were visualized and tracked at a spatial precision of 35 nm,
and a time resolution of 30 ms.
Recently, Bruchle and co-workers demonstrated the
monitoring of the infection pathway of single viruses using
real-time fluorescence imaging.[225] In a related study, pulsed
laser excitation in the red region (l = 635 nm), in combination
with spectrally resolved fluorescence lifetime imaging microscopy (SFLIM), enabled the discrimination of fluorescence
signals from individual fluorescently labeled oligonucleotides
from autofluorescence signals. This discrimination was possible because of the characteristic fluorescence lifetime and
spectral characteristics of the fluorophores used.[83] SFLIM
images and fluorescence-intensity measurements obtained
from living 3T3 mouse fibroblast cells containing approximately 100–1000 fluorescently labeled oligo(dT) 43 mer
molecules showed that around 10–30 % of the oligonucleotides cannot diffuse freely inside the nucleus but rather are
tethered to elements of the transcriptional, splicing, and/or
polyadenylation machinery. These results demonstrate that
confocal imaging techniques in combination with real-time
imaging microscopy allow transport pathways in intact
biological systems to be studied, enable conclusions to be
drawn from the underlying cellular architecture, and to
analyze individual binding events as well as to observe
mobility restrictions in living cells.
However, when tracking single molecules under native
conditions they rapidly suffer from photobleaching. As a
result, they can only be followed for a limited amount of time,
which is often insufficient to obtain the desired information.[200] For in vitro studies of labeled molecules, methods
have been worked out to increase photostability of the
fluorophores so that they can be followed over minutes. In
particular, the groups of Chu and Ha established the use of
the glucoseoxidase–catalase system to remove oxygen under
in vitro conditions, an approach which increases the photostability of cyanine dyes by orders of magnitude.[58, 226]
Unfortunately, such strategies can not be applied in live
cells. In the future, new labels are required that are more
photostable and still retain the biological functionality of the
biomolecule under investigation. One class of fluorophores
for which there are great expectations are semiconductor
nanocrystals (NCs), such as core–shell CdSe/ZnS NCs.[227]
Their unique optical properties—tunable narrow emission
spectrum, broad excitation spectrum, high photostability, and
long fluorescence lifetime (on the order of tens of nanoseconds)—make these bright probes attractive in experiments
involving long observation times, multicolor, and time-gated
detection. Furthermore, the relatively long fluorescence
lifetime of CdSe nanocrystals could enhance imaging contrast
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
and sensitivity in biological applications using time-gated
detection.[228] Although NC biocompatibility was demonstrated few years ago,[227, 229] the breakthrough in biological
targeting was only achieved very recent.[230–236] These days,
NCs can be easily functionalized, for example with streptavidin, to facilitate mild coupling to biotinylated biomolecules.
Immunofluorescent labeling of cancer markers and other
cellular targets on the surface of fixed and live cancer cells as
well as staining of actin and microtubule fibers in the
cytoplasm, and the detection of nuclear antigens inside the
nucleus have been successfully demonstrated.[235] Due to
these improvements it is anticipated that NCs will soon also
play an important role for in vivo single molecule studies and
first successful results have already been presented.[237, 238]
Nevertheless, the use of NCs in biological applications of
SMFS and fluorescence imaging does not represent the magic
bullet. There are several problems associated with NCs which
should not be underestimated. One disadvantage of NCs is
the difficulty to engineer them with single binding sites that
can be specifically conjugated to only one molecule of
interest. Instead, during the labeling step, NCs tend to bind
to several molecules simultaneously. Another problem is
blinking,[239] which is strongly controlled by the excitation
intensity (Figure 13 a), and often obeys a power law.[240, 241]
Although, blinking of NCs is efficiently reduced in the
presence of 1–10 mm sulfanylethanol (mercaptoethanol), the
addition of reducing agents is not compatible with live cell
imaging.[242] In addition, generally not all semiconductor
particles are luminescent, but they can be photoactivated.[243]
Another issue is whether such particles, which are
composed of toxic material, are well suited for in vivo studies
and whether they impede biological functionality.[244] For
example, besides their core–shell structure, commercially
available NCs have a third layer—an organic surface coating—that gives them chemical and photophysical stability,
inertness in different environments, buffer solubility, and
facilitates the introduction of reactive groups for linking to
biomolecules. Ultimately, this results in particle sizes of 15–
25 nm in diameter, that is, about four to five-times larger than
GFP and its derivatives, and 15 to 25-times larger than
conventional fluorophores used to tag biomolecules. Another
concern to be addressed is whether each of these particles
contains only a single NC. Even though, this situation might
be of minor importance for those users who are basically
interested in the brightness of the fluorescence signal they
obtain from a labeled biomolecule, it seriously renders the
quantitative analysis of the fluorescence signal more complicated, for example, in applications where the number of
molecules bound to a specific target molecule are of interest
(Figure 13 a).
We have used biotinylated polyarginine peptides (Arg10Lys) to afford the intracellular delivery of commercially
available streptavidin-coated NCs (Qdot605 streptavidin
conjugate) into HeLa cells. Fluorescently labeled NC-free
polyarginines (Arg10-Lys-fluorescein) enter the cell and
accumulate in the nuclei, especially in the nucleoli (Figure 13 b).[198, 199] Streptavidin-coated NCs (108 m solution) are
not taken up by HeLa cells neither do they stain the
membranes of the cells even after 24 h exposure. On the
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2659
Reviews
P. Tinnefeld and M. Sauer
molecular machines and for precision measurements
of the distance between them. To close the resolution gap between far-field optical microscopy (>
200 nm) and FRET (ca. 2–8 nm) a single-moleculesensitive colocalization technique with a resolution
of a few tens of nanometers is required.
Recently, a relatively simple procedure, socalled single-molecule high-resolution imaging with
photobleaching, was presented.[249] The method
takes advantage of sequential photobleaching of
two identical fluorophores to determine distances as
small as 10 nm. Unfortunately, the method as
described, can only be used to measure the twodimensional projection of distances between pairs of
Figure 13. a) Fluorescence intensity trajectory (count rate, nc, versus time, t) of a single
semiconductor NC (Qdot605–streptavidin conjugate) adsorbed on a bare glass surface
molecules in the xy plane. Generally, colocalization
under different excitation intensities (excitation at 488 nm using a frequency-doubled
of biomolecular building blocks is determined by
Ti:sapphire laser at a repetition rate of 3.8 MHz). After about 20 s, the shutter was closed
staining the constituents with fluorescent labels that
and the average excitation intensity was changed from 1 to 0.2 kWcm2. Blinking of the NC
can be separated, at least partially, in their emission
is less frequent under lower excitation intensities. The inset shows a normalized histogram
spectra. Quantitative three-dimensional fluoresof the background-corrected ratios of the central to the lateral peaks, Nc/N̄l measured from
cence analysis is performed with multicolor confocal
60 single Qdot605–streptavidin conjugates adsorbed on a dry cover slide. The relatively
high number of Nc/N̄l ratios above 0.3 indicates that more than 50 % of the NCs do not
laser scanning microscopy (CLSM) and advanced
behave as a single emitter, possibly a result of aggregation of NC–streptavidin conjugates.
image analysis techniques.[250, 251] Spherical and chrob) Fluorescence image of a HeLa cell excited at 488 nm in PBS buffer. The inset shows a
matic aberrations in the excitation and detection
phase-contrast image of the cell. Cells were incubated with 108 m solution of Qdot605–
paths diminish the attainable accuracy in established
streptavidin conjugates, 106 m biotinylated polyarginine (Arg10-Lys), and 106 m fluoresmulticolor approaches.[252] To circumvent these
cently labeled polyarginine (Arg10-Lys-fluorescein) for 20 min then washed. Whereas fluoresproblems, Lacoste et al. presented a new method
cein labeled polyarginines (green) enter the cells and accumulate in the nucleoli,[198, 199]
for ultrahigh-resolution multicolor colocalization
streptavidin-coated NCs (red) covered with biotinylated polyarginine only bind on the membrane but do not enter the cells.
which relies on the use of luminescent nanoparticles
that can be excited by a single laser wavelength but
emit at different wavelengths.[253, 254] In combination
with multicolor confocal microscopy with “orthogonal”
other hand, upon addition of an excess of biotinylated
detection channels, the fluorescence of different emitters is
polyarginine peptides (106 m Arg10-Lys-Biotin) NCs bind
separated and recorded independently. Because the same
immediately and efficiently to the cell membrane but do not
laser excites all emitters, chromatic aberrations in the
enter the cell. (Figure 13 b). This result implies that the size
excitation arm are eliminated. The method was demonstrated
and/or charge of the NCs used (15–20 nm) prevents their
using two types of fluorescent nanoparticles: energy-transfer
uptake by peptide specific pathways. Alternatively, the
fluorescent beads (TransFluoSpheres) and semiconductor
peptide recognition sequence (Arg10) might be partly
NCs. The large size (20–40 nm) of energy-transfer beads
hidden in the streptavidin binding pocket.[195, 199]
prevents their use for colocalization experiments in small
molecular machines. As discussed above, NCs are somewhat
smaller, exhibit bright fluorescence, and are far more stable
8. Colocalization and Distance Measurements
than conventional dyes. However, the pronounced blinking of
Between Single Molecules
NCs deteriorates the quality of images and results in a
reduced precision of localization.[254]
SMFS has open new avenues for investigating the threedimensional cellular architecture and function of living cells.
In 1995 Betzig recalled the fact that any optically
The information obtained to date has benefited from the
distinguishable characteristic can be used for identification
small dimensions of a dye molecule. Since fluorophores are
and isolation of a fluorophore.[255] As it has been shown that
much smaller than the wavelength of light they emit, they act
the fluorescence lifetime of single fluorescent dyes can be
as point sources of light. The center of the resulting pointused as an unequivocal identification property,[9, 68, 256] colocspread-function (PSF), that is, the response of the optical
alization of different emitters can also be performed by
system, can be localized with high accuracy using the centerapplying time-resolved fluorescence spectroscopy. The use of
of-mass method or a Gaussian fit. The PSF describes the
fluorescence lifetime as the sole identification criterion, for
three-dimensional energy density in space measured from a
example, two dyes which exhibit similar absorption and
point-like light-emitting object. This localization precision has
emission characteristics but distinct fluorescence lifetimes,
been used to follow the motion of individual motor prosimplifies the experimental set-up drastically because only a
teins,[64, 245] the diffusional trajectories of labeled lipid molesingle detector is required. By the use of appropriately
selected fluorescent dyes, the method allows intermolecular
cules in membranes,[246, 247] and the diffusion of molecules in
distances down to the FRET range to be measured without
gels.[248] The high positioning accuracy is also useful for the
chromatic aberrations.[257]
colocalization of the different building blocks of cellular
2660
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
In such distance measurements, there are two sources of
error: 1) the detected photons have to be assigned to the
corresponding emitter on the basis of their microscopic
arrival time, 2) the two obtained PSFs have to be fitted to
locate the emitters. To decrease the error in assigning the
photons to the correct emitter, a combination of time and
spectral information, such as that obtained by SFLIM, could
be used. This approach would enable distance measurements
to be made with nanometer accuracy between conventional
single fluorescent dyes (Figure 14).
For the assignment of photons as well as for the localization accuracy, the selection of appropriate fluorophores is
essential. Bodipy 630/650 and Cy5.5 are a suitable combination of dyes for high-precision distance measurements
between single molecules as the two dyes exhibit a twofold
difference in fluorescence lifetime (Bodipy 630/650 ca. 4 ns;
Cy5.5 ca. 2 ns) and different fluorescence emission maxima
(Bodipy 630/650 l 660 nm; Cy5.5 l 700 nm). When adsorbed on bare glass surfaces under dry conditions both of these
dyes exhibit short off times in the micro- to millisecond range
(that is, shorter than the integration time per pixel).[78, 257] In
addition, owing to the higher extinction coefficient of Cy5.5,
the two chromophores can be excited at l = 635 nm with
comparable efficiency by using a single pulsed laser diode.
Using a dichroic beam splitter at l = 685 nm, the long
fluorescence lifetime of Bodipy 630/650 is detected predominantly on the short-wavelength detector (det 1) while the
shorter fluorescence lifetime of Cy5.5 is detected exclusively
on the long-wavelength detector (det 2; Figure 14 b). To
demonstrate the feasibility and to evaluate the error of the
method, precision distance measurements between the two
dyes Bodipy 630/650 and Cy5.5 attached covalently to doublestranded DNA at separations varying between 10 and
200 base pairs are currently being carried out (Figure 14 c).
To improve the spatial resolution of optical microscopy
beyond the classical diffraction limit of light, several new
methods have been developed. Among these are 4Pi microscopy,[258, 259] near-field scanning optical microscopy
(NSOM),[260, 261]
standing-wave
fluorescence
microscopy,[262–264] coherent optical dipole coupling of individual
molecules at low temperature,[265] and point-spread-function
(PSF) engineering by stimulated emission depletion
(STED).[266–269] Among the new techniques, PSF engineering
using STED is the most promising because it represents the
first far-field method which, conceptually, has unlimited
spatial resolution. In STED, a diffraction-limited fluorescence
spot is sharpened by inhibiting the fluorescence through
saturation. This is accomplished by quenching of the excited
fluorophore at the rim of the focal spot by stimulated
emission. STED has so far demonstrated a resolution of
28 nm with single molecules.[270] Since the excitation pulse is
followed by a quenching pulse which acts on the excited state
of the same molecule, the single-molecule experiments
reported by Westphal et al. are the first single-molecule
pump-probe experiments.[270] As they rely on saturated
optical transitions, these methods are limited only by the
attainable saturation level. As strong saturation should be
feasible at low light intensities, nanoscale imaging (“nanoscopy”) with focused light may be closer than ever.[268] Very
recently, immunofluorescence imaging with an axial resolution of approximately 50 nm, corresponding to 1/16 of the
irradiation wavelength of l = 793 nm, has been achieved by
stimulated emission depletion through opposing lenses
(STED-4Pi).[269]
9. Summary and Outlook
It is beyond doubt that SMFS has increased our current
understanding of biological macromolecules and their structure–function relations. Furthermore, ongoing single-molecule projects, such as the aim to sequence a single DNA
fragment, have revealed unexpected behavior of biomolecules, such as DNA polymerases. In fact, most natural DNA
polymerases and exonucleases have been found to discriminate against dye-labeled nucleotides; this may be due to
steric hindrance at the active site of the enzyme as a result of
the bulkiness of the label, and/or because the fluorescently
labeled nucleotides typically have a net charge that differs
from that of the natural substrates.[271, 272] However, Brakmann
and co-workers used the well documented power of direct
evolution to identify mutant DNA polymerases that incorporate labeled nucleotides with high efficiency and retain a
sufficient incorporation fidelity.[273] They discovered that a
Figure 14. a) False-color SFLIM image of a double-stranded DNA (148 bp) labeled with Bodipy 630/650 at one terminus and Cy5.5 at the other
and immobilized in agarose gel. Green spots are DNA molecules carrying only Bodipy and red spots are DNA molecules carrying only Cy5.5.
Yellow spots are double-labeled DNA molecules. Photobleaching of Cy5.5 during the recording of the image is visible for the molecule circled
blue. b) Fluorescence decays recorded on the long-wavelength (red), and short-wavelength channel (blue) for the spot circled white in (a).
c) Colocalization of a Bodipy and Cy5.5 molecule with a separation of 34 7 nm. Construction of the two separate fluorescence intensity images
for the Bodipy (bottom left) and Cy5.5 dye (top right) was performed by pattern matching on the basis of the respective characteristic spectroscopic properties (F2 value, and fluorescence lifetime).
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2661
Reviews
P. Tinnefeld and M. Sauer
cloned natural polymerase, the unmodified
exonuclease-deficient Klenow fragment of
Escherichia coli DNA polymerase I polymerized
55 tetramethylrhodamine-4-dUTPs
using an artificial (dA)55 template. Furthermore, the same enzyme was successfully used
in an analogous reaction with a natural
primer template with a length of 2700 base
pairs, in which two of the natural substrates,
dCTP and dTTP, were substituted by their
rhodamine-labeled analogues.[274, 275]
The fact that reaction pathways for different molecules are not predetermined, but still
deliver the same product state represents
another important result obtained by SMFS
which has influenced the way in which we
think about and model molecular systems.
Dynamic structural studies showed that reactions proceed heterogeneously on multidimensional energy landscapes. That is, individual molecules follow different pathways
on the energy landscape. Since these different pathways can occur on a wide range of
time scales, extending from femtoseconds to
seconds, individual reaction rates differ significantly. SMFS can be used to directly track
the pathway of an individual molecule on its
Figure 15. Outlook to possible future experiments correlating conformational dynamics to
potential energy landscape on a broad range
enzyme activity and the topography of energy landscapes. a) Conformational dynamics
are monitored by the fluorescence intensity and lifetime of a fluorophore (red) which is
of time scales, under physiological conditions,
quenched by an amino acid, for example, tryptophan (blue) through electron transfer.
and at equilibrium.[5, 140, 276] This was made
Simultaneously, the enzyme activity is recorded as a function of the conversion rate of a
possible by the use of dipole–dipole induced
nonfluorescent substrate into a highly fluorescent product (green) at another excitation
interactions and short-range electronic interwavelength. b) Alternatively, a combined approach using electron transfer and spFRET
actions between chromophores and naturally
could be applied. Time-resolved monitoring of the fluorescence intensity and lifetime of a
occurring amino acids, that is, photoinduced
red-absorbing fluorophore (red) specifically quenched through electron transfer by an
electron-transfer processes. With a distance
amino acid residue (blue) reveals conformational dynamics. Fast switching of the excitation wavelength between the red fluorophore in the active site and a potential FRET
dependence in the subnanometer range,
donor coupled to the substrate (green) enables the simultaneous monitoring of distance
electron-transfer reactions complement
changes occurring on the nanometer scale between the enzyme’s active site and the subsingle-pair FRET measurements which are
strate.
suited to longer distances. The new techniques enabled, for example, the monitoring of
dynamic fluctuations in electron-transfer efficiency between
the enzyme, each conformation corresponds to a separate
amino acids and fluorophores on a submicrosecond time scale
point on the potential energy surface, thus the conformational
at the single-molecule level.[5] Thus, the monitoring of the
changes give rise to variations in electron-transfer efficiency,
these are directly reflected in fluctuations of fluorescence
conformational dynamics of individual biomolecules at equiintensity and lifetime. Depending on the time resolution
librium, under physiological conditions, with temporal resoachievable in such experiments, enzyme activity could be
lutions comparable to those of new molecular dynamic
monitored simultaneously using, for example, a non-fluoressimulations, is possible.[277]
cent substrate which is converted into a highly fluorescent
These results raise the hope that it might be possible to
product in the active site of the enzyme (Figure 15 a).[278, 279]
relate conformational dynamics directly with enzyme activity,
for example, turnover rates (Figure 15). Such studies would
Are specific reaction pathways related to the binding,
require site-specific labeling of the active site of an enzyme
reaction, and release of the substrate (Figure 15 b)? An
with a fluorophore which is selectively quenched by an amino
even more ambitious study will be to extract topographic
acid, for example, a tryptophan residue, in close proximinformation of the energy landscape by following the
ity.[5, 139, 146, 149] In addition, the fluorophores conformational
temporal order on which different conformations occur.
The application of spFRET experiments to biological
space has to be minimized, otherwise, the observed variation
questions has not exploited the full potential of this method
in electron-transfer efficiency could be masked by conformayet. spFRET is not restricted to interaction measurements
tional fluctuations of the linker between the fluorophore and
between two subunits of molecular machines or assemblies.
the quenching amino acid. Brownian motion results in
Just like FRET was successfully used for the construction of a
subnanometer conformational changes in the active site of
2662
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
photonic wire, SMFS techniques have the potential to
elucidate the interactions of more than two components
simultaneously through multistep-FRET experiments. However, care has to be taken in the interpretation of colocalization and multistep-FRET data obtained from more than two
differently labeled components. First, crosstalk in excitation
and detection has to be reduced to a minimum to obtain
unequivocal information. Second, to assign the lack of a
FRET-signal to the absence of one molecule requires direct
probing (excitation) of each component,[280] otherwise, misinterpretation of the data can easily occur. Nevertheless, by
careful selection of the fluorophores and the excitation and
detection technique used, multicolor experiments could be
used to examine the interaction pathways in molecular
assemblies as well as for the construction of 3D distance
maps of the separations between differently labeled proteins
in biomolecular assemblies.
First successful steps towards application of SMFS in livecell imaging have been made. The pioneering technological
breakthroughs achieved to date provide the basis for a variety
of experiments which were unthinkable a few years ago.
However, besides structural-dynamics studies, the ambition of
single-molecule fluorescence spectroscopy must be to go
beyond pure descriptive experiments and provide quantitative data on the biomolecular machines that are essential for
cell function and survival. It is no longer enough to study just
the colocalization of different biomolecules in specific
compartments, rather we quest for measuring their absolute
separations on the 0.1–200 nm length scale, their orientations,
their subnanometer conformational fluctuations relevant for
functioning, and importantly we are interested in the absolute
number of interacting molecules. Furthermore, single-molecule techniques are at hand that could allow us to probe the
local response of a cell upon specific stimuli (Figure 16).
Using submicrometer pipettes in combination with scanning
ion conductance microscopy (SICM) topographic images of
living cells at high resolution under biological relevant
conditions can be obtained.[281–283] The cell could be stimulated
Figure 16. Possible application of SICM in combination with singlemolecule fluorescence imaging to monitor local cellular dynamics. In
SICM, an electrolyte-filled, glass micropipette is scanned over the surface of a sample bathed in an electrolytic solution.[281] The pipette–
sample separation is maintained at a constant value by controlling the
ion current that flows through the pipette aperture. The measured ion
current is used to generate topographic features and/or images of the
local ion currents flowing through pores on the sample surface. The
spatial resolution achievable using SICM is dependent on the size of
the tip aperture.
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
at well defined positions by injection into the cytoplasm or
nucleus or delivering of individual fluorescently labeled
molecules on the cell membrane while the response of the
cell is measured at different locations by fluorescence
scanning or imaging techniques.[284] Alternatively, SICM
could be used to localize and measure the function of single
ion channels in the membrane[285] while monitoring the
binding of fluorescently labeled molecules at the inner or
outer membrane.[286–288]
To extract quantitative information from SMFS data in
living cells, such as the absolute number of polymerase
molecules located in a transcription or replication foci and the
distances between interacting molecules, several challenges
need to be overcome. First, a technique is required which
gives information about the number of molecules present in
the laser focus. To date, in single-molecule studies in cells,
several indirect methods have been used to confirm that the
fluorescence observed is derived from only a single emitter,
for example, photobleaching in a single, instantaneous step, or
the observation of transient off states. On the other hand,
polarization-modulated excitation spectroscopy or photon
antibunching experiments can provide a more direct proof for
the presence of a single molecule. Furthermore, the ratio of
the number of coincident photon events to the number of
noncoincident photon pairs in photon antibunching experiments gives a measure of the number of independently
emitting fluorophores for a given time interval
(Figure 17).[89, 90, 289] Therefore, photon antibunching measurements are well suited to count the number of fluorescent
molecules present in the laser focus, as long as higher excited
state interactions, such as singlet–singlet annihilation, can be
neglected.[85] Besides the background which influences the
ratio between the central and lateral peaks calculated from
coincidence histograms, thus making counting harder, it will
be more difficult to accurately determine the number of
independent emitters when higher numbers of molecules (>
4) are present.[90] Nevertheless, spectrally resolved fluorescence lifetime imaging microscopy (SFLIM) in combination
with photon antibunching measurements are valuable tools
for imaging and counting single molecules in cells (Figure 17).
In addition, SFLIM is well suited to measure distances
between single molecules although these separations are well
below the optical resolution limit.[257] Consequently, the
techniques are available that would enable a quantitative
comprehension of biomolecular machines.
Since the well-defined labeling of biomolecules, for
example, proteins, with a single fluorophore is the basis for
quantitative measurements, new labeling strategies have to be
developed: In general, proteins, for example, antibodies, are
multiple labeled using fluorophore excess. The excess of free
fluorophores is finally removed by gel filtration. Unfortunately, multiple-labeled antibodies cannot be used for the
counting of molecules by photon antibunching. For example,
the counting of the number of polymerase II molecules
present in a transcription factory requires singly labeled
antibodies. In principle, an average label degree of less than
one can be easily obtained by reducing the fluorophore
concentration in the labeling reaction. However, since the
separation of labeled and unlabeled antibodies is almost
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2663
Reviews
P. Tinnefeld and M. Sauer
Figure 17. Confocal scanning fluorescence images of a fixed 3T3 fibroblast cell treated with 109 m solution of MR121-labeled oligo(dT)
(40 mer) molecules (excitation at l = 635 nm using a pulsed laser
diode). Oligo(dT) can hybridize to the poly(A) tail of mRNAs.[186]
a) Overall fluorescence-intensity image recorded with a SFLIM-microscope showing several bright spots in the nucleus with fluorescence
count rates of up to 30 kHz.[78] b) and c) show the corresponding spectral image (F2 image) and lifetime image (t image) For the calculation
of the F2 and t image, pixels containing less than 15 counts were discarded. Owing to the unequivocal spectroscopic characteristics of the
MR121-labeled oligo(dT) molecules with fluorescence lifetimes of
around 3 ns, and F2 values of approximately 0.5, specific signals can
be easily discriminated against autofluorescence.[83] d) Fluorescenceintensity trajectory (count rate, nc versus time, t in s) of a fluorescent
spot in the nucleus of the fibroblast cell (traces: black overall intensity,
green det 1, red det 2). With a background count rate of approximately
2 kHz, S/B ratios of about 9 are easily obtained in the nucleus. As several oligo(dT) molecules can bind to a single poly(A) tail of mRNA, it
is not known whether the intensity fluctuations observed (compare
gray and green areas) have a photophysical origin or are the result of a
varying number of molecules contributing to the signal. d) The interphoton times (coincidence) histogram of all photons detected during
the first 10 s of the trajectory give a Nc/N̄l ratio of 0.32 0.06. On the
other hand, S/B considerations predict a Nc/N̄l ratio of approximately
0.18 for a single emitter.[90] If however, interphoton-times histograms
are constructed separately from the photons recorded in the higher
intensity (green) and lower intensity parts (gray), two different Nc/N̄l
ratios of 0.55 0.13 (for the green part) and 0.17 0.05 (for the gray
part) are obtained. This result demonstrates that the intensity fluctuations observed in the trajectory are caused by a varying number of fluorescing molecules present in the laser focus.
impossible, unlabeled antibodies would bind to polymerase
molecules as well and prevent any quantitative statement. To
circumvent these problems, Lewinska et al. described a novel
method for the specific labeling of native proteins at a single,
well-defined position using the commercial IgA protease to
attach a non-natural peptidic moiety to the N terminus of the
antibody. This natural peptidic moiety can be selectively
modified.[290] We are currently developing a new labeling
strategy for antibodies which guaranties that practically all
the antibodies present in the sample carry a single fluorophore (with only a small number of double-labeled antibodies). The strategy comprises bridging of the fluorophore
and antibody by a special linker, for example, a peptide, which
binds specifically to its binding partner, for example, another
2664
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
antibody. Using, for example, an oligohistidine (6–12 histidine
residues covalently attached by the N-terminal end to the
fluorophore) as the linker and labeling conditions that ensure
a very low degree of labeling, singly labeled antibodies can be
obtained by the use of a His-tag purification column.
Finally, by combining SMFS measurements with singlemolecule manipulation techniques, such as optical tweezers,[291] or other spectroscopic techniques, such as surface
enhanced Raman scattering (SERS), it could be possible to
monitor simultaneously several different observables of
biological reactions in living cells. In common Raman
spectroscopy the inelastically scattered light is used to identify
molecules through their characteristic vibrational quantum
states. SERS uses the fact that the rather weak Raman effect
can be significantly strengthened (by a factor of up to
14 orders of magnitude) if the molecules are attached to
nanometer-sized metal structures. The apparatus required for
SERS, SMFS, and optical tweezers in cells are similar and
comprise a laser that is focused into the cell by using a high
magnification microscope objective. For example, Kneipp and
co-workers deposited 60-nm gold particles inside cells as
SERS-active nanostructures to measure the enhanced Raman
signals of the chemical constituents of cells.[292] SERS mapping over a cell monolayer with 1 mm lateral resolution
showed different Raman spectra at almost all places which
reflects the very inhomogeneous constitution of the cells. The
new spectroscopic method provides a tool for even more
sensitive and structurally selective detection of chemicals
inside a cell, and for monitoring their intracellular distributions.
Combinations of these new techniques, opens new opportunities for cell biology and biomedical studies. However, to
combine SERS with optical tweezers and SMFS several
challenges need to be overcome. First, optical trapping of
metallic particles required for SERS is complicated.[293]
Furthermore, absorption by the metal particle can cause a
temperature rise and damage to living cells. This process has
become known as opticution.[294] Even for refractory absorbing materials, the effects of heating can be a problem as
bubble formation will disrupt trapping.[295] However, it could
be possible to use standard microbeads for optical trapping if
they are coated with a thin metal layer or some metal clusters
are captured on the surface so that the optical excitation of a
surface plasmon is supported but 3D optical trapping with
minimal heating effects is still enabled. Several nanometers
long water-soluble linkers, such as PEG linkers, have to be
used to attach fluorescence sensors to the metal. This is
because the fluorescence intensities are attenuated depending
on the distance (up to tens of nm) between the fluorophores
and the metal surfaces owing to fluorescence energy transfer
to the metal.[296–298] In the future this combined approach
could provide an elegant method to locally probe forces[299]
and individual binding events with fluorescence sensors.[300]
Simultaneously it could be possible to map the concentration
of specific biomolecules in living cells by SERS.[292] In an
extended version of the method, a single IR-laser line could
be used simultaneously to trap the bead, generate the SERS
signal, and excite the fluorophores by two-photon excitation.[301]
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
Addendum (February 12, 2005)
Since the submission of this Review, scientific development has not been stagnant. The trend to address more
sophisticated biological problems by single-molecule fluorescence spectroscopy continues. Examples include the improvement of the control of chemical approaches to meet the
requirements of biological systems, the development and
application of new techniques to investigate biomolecular
dynamics, and the study of artificial light-harvesting or
electron-transfer systems. In particular, it is clear that in
contrast to previous SMFS studies which often had the
character of feasibility studies more and more attention is
focussed on the reliability of the results obtained, that is,
methods have been further elaborated and characterized by
controls and crosschecks.
Immobilization of biomolecules, for examples, has always
been viewed with skepticism because the impact of the close
surface on the properties of the molecule investigated is hard
to exclude or to evaluate. Recent studies, however, show that
surface perturbations can be minimized in such a way that
they do not dominate the observed biomolecular dynamics.
Okumus and co-workers, for example, compared the dynamics of ribozyme molecules directly tethered to the surface to
those of ribozymes enclosed in immobilized vesicles and
found no differences.[303] In addition, tethering strategies have
been improved by comparing different immobilization
schemes and a star-shaped poly(ethylene oxide) (PEO) has
been found to be the least interfering in nature.[304]
While FRET is now an established SMFS technique which
is applied to more complex problems, such as the investigation of translation,[305, 306] and analyzing the reversible reaction
pathway of the hairpin ribozyme,[307] new territory is entered
when the interaction of more than two chromophores is
investigated. The possibility to resolve FRET between three
or even up to five chromophores, has recently been demonstrated[308–310] which opens up new vistas to visualize the
interaction between more than two molecules or to correlate
movements within a molecule or biomolecular complex. The
understanding of FRET processes in multichromophoric
model systems has led to a comprehensive level understanding of intersystem crossing and photobleaching,[311] so that the
focus is now on the investigations of stronger interactions,
such as the formation of excimer-states[312] and superradiance
in strongly coupled tetraphenoxyperylenediimide trimers.[313]
These new results are of interest in interpreting the photophysics of technological interesting molecules, such as pconjugated polymers. [314]
Recent achievements demonstrate impressively that
SMFS can supply important information on complex processes in living cells. For example, TIRFM has been successfully applied to visualize single-molecule processes in living
cells, such as activation of the G protein Ras,[315] or exocytosis
of IgG as mediated by the FcRn receptor.[316] Furthermore,
transport through nuclear pore complexes could be directly
monitored in permeabilized cells by single-molecule tracking.
[317]
More recently, the direct observation of the real-time
catalysis and substrate kinetics of a single lipase enzyme (a
single enzymatic turnover cycle) could be recorded for up to
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
2 h using nonfluorescent substrate molecules which are
converted into highly fluorescent products upon hydrolysis,
a method similar to that proposed in Figure 15 a.[318, 319]
Insights into the conformational fluctuations of the protein
which are relevant for function could be obtained at the same
time, if the enzyme were labeled site specifically with a
suitable FRET-pair.
Other techniques which supply topographic information
on the nanometer scale or other additional information about
the systems being studied, such as AFM and electrical channel
recording, are currently successfully combined with optical
single-molecule techniques.[320, 321] Biological applications of
nanocrystals (NCs) are maturing with cytotoxic effects[322, 323]
and labeling strategies for live-cell applications at the forefront. [324–326] Recently, Lagerholm and co-workers described a
strategy similar to that outlined in Figure 13 to deliver NCs
into cells using arginine rich peptides on the surface of the
NCs. Interestingly they find that cellular labeling occurs
readily for suspended cells while, in agreement with our
results, with supported cells, mainly the extracellular matrix is
labeled.[325]
Improving the localization and colocalization accuracy of
single biomolecules well below the optical resolution limit,
remain one of the basic concerns of single-molecule spectroscopy. The potential of new methods, such as subsequent
photobleaching of chromophores, STED microscopy, and
spectrally resolved fluorescence lifetime imaging microscopy,
to be of use in DNA mapping with very high resolution and
for studying the motion of molecular motors, such as
myosin V, has just begun to be explored.[327–333] In addition,
scientists worldwide discuss the development of techniques
that will enable the determination of the absolute 3D
localization of individual molecules in living cells. This
requires a kind of global positioning system (GPS) on a
cellular level, a cellular positioning system (CPS) or molecular positioning system (MPS), and might be accomplished,
for example, by labeling specific compartments within the cell
with fluorophores of different spectral signature or by
deposition of cells on 3D nanostructured surfaces at which
at least three fixed points are selectively labeled with different
dyes. Theoretically, this would enable the determination of
the absolute position of each fluorescently labeled biomolecule with respect to the fixed points.
Although, the presented achievements represent a subjective selection they impressively demonstrate that SMFS is
still a dynamic and emerging field that has not lost any of its
fascination.
The authors are indebted to H. Barsch, V. Buschmann, A.
Biebricher, F. C. De Schryver, S. Doose, K. H. Drexhage, J.
Enderlein, M. F. Garcia-Parajo, K. T. Han, M. Heilemann, T.
Heinlein, D.-P. Herten, J. Hofkens, J. P. Knemeyer, N. Marm,
C. Mller, H. Neuweiler, C. Roth, M. Rping, C. W. Park, O.
Piestert, P. Schlter, A. Schulz, D. Seebach, C. A. M. Seidel,
N. F. van Hulst, S. Weiss, K. D. Weston, J. Wolfrum, and C.
Zander for their contributions and active support of the singlemolecule fluorescence spectroscopy research in Heidelberg
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2665
Reviews
P. Tinnefeld and M. Sauer
and Bielefeld. This work was supported by the BMBF, the VWStiftung, and the DFG.
Received: December 29, 2003
[1] G. Binnig, H. Rohrer, Rev. Mod. Phys. 1987, 59, 615 – 625.
[2] G. Binnig, C. F. Quate, C. Gerber, Phys. Rev. Lett. 1986, 56,
930 – 933.
[3] B. Schuler, E. A. Lipman, W. A. Eaton, Nature 2002, 419, 743 –
747.
[4] E. A. Lipman, B. Schuler, O. Bakajin, W. A. Eaton, Science
2003, 301, 1233 – 1235.
[5] H. Neuweiler, A. Schulz, M. Boehmer, J. Enderlein, M. Sauer,
J. Am. Chem. Soc. 2003, 125, 5324 – 5330.
[6] M. J. Levene, J. Korlach, S. W. Turner, M. Foquet, H. G.
Craighead, W. W. Webb, Science 2003, 299, 682 – 686.
[7] I. Braslavsky, B. Hebert, E. Kartalov, S. R. Quake, Proc. Natl.
Acad. Sci. USA 2003, 100, 3960 – 3964.
[8] J. H. Jett, R. A. Keller, J. C. Martin, B. L. Marrone, R. K.
Moyzis, R. L. Ratliff, N. K. Seitzinger, E. B. Shera, C. C.
Stewart, J. Biomol. Struct. Dyn. 1989, 7, 301 – 309.
[9] M. Sauer, B. Angerer, W. Ankenbauer, Z. Foldes-Papp, F.
Gobel, K. T. Han, R. Rigler, A. Schulz, J. Wolfrum, C. Zander,
J. Biotechnol. 2001, 86, 181 – 201.
[10] K. Dorre, S. Brakmann, M. Brinkmeier, K.-T. Han, K.
Riebeseel, P. Schwille, J. Stephan, T. Wetzel, M. Lapczyna, M.
Stuke, R. Bader, M. Hinz, H. Seliger, J. Holm, M. Eigen, R.
Rigler, Bioimaging 1997, 5, 139 – 152.
[11] J. H. Werner, H. Cai, J. H. Jett, L. Reha-Krantz, R. A. Keller,
P. M. Goodwin, J. Biotechnol. 2003, 102, 1 – 14.
[12] W. E. Moerner, T. Basche, Angew. Chem. 1993, 105, 537 – 557;
Angew. Chem. Int. Ed. Engl. 1993, 32, 457 – 476.
[13] X. S. Xie, Acc. Chem. Res. 1996, 29, 598 – 606.
[14] S. M. Nie, R. N. Zare, Annu. Rev. Biophys. Biomol. Struct. 1997,
26, 567 – 596.
[15] X. S. Xie, J. K. Trautman, Annu. Rev. Phys. Chem. 1998, 49,
441 – 480.
[16] W. P. Ambrose, P. M. Goodwin, J. H. Jett, A. Van Orden, J. H.
Werner, R. A. Keller, Chem. Rev. 1999, 99, 2929 – 2956.
[17] W. E. Moerner, S. Weiss, J. K. Gimzewski, A. D. Mehta, Science
1999, 283, 1667 – 1695 (Single Molecule Special Issue).
[18] C. Zander, J. Anal. Chem. 2000, 366, 745 – 751.
[19] A. A. Deniz, T. A. Laurence, M. Dahan, D. S. Chemla, P. G.
Schultz, S. Weiss, Annu. Rev. Phys. Chem. 2001, 52, 233 – 253.
[20] W. E. Moerner, J. Phys. Chem. B 2002, 106, 910 – 927.
[21] Y. Jung, E. Barkai, R. J. Silbey, J. Chem. Phys. 2002, 117,
10 980 – 10 995.
[22] M. Orrit, J. Chem. Phys. 2002, 117, 10 938 – 10 946.
[23] A. N. Kapanidis, S. Weiss, J. Chem. Phys. 2002, 117, 10 953 –
10 964.
[24] W. E. Moerner, J. Chem. Phys. 2002, 117, 10 925 – 10 937.
[25] X. Sunney Xie, J. Chem. Phys. 2002, 117, 11 024 – 11 032.
[26] R. A. Keller, W. P. Ambrose, A. A. Arias, H. Cai, S. R. Emory,
P. M. Goodwin, J. H. Jett, Anal. Chem. 2002, 74, 316A-324A.
[27] U. Kubitscheck, Single Mol. 2002, 3, 267 – 274.
[28] M. Boehmer, J. Enderlein, ChemPhysChem 2003, 4, 792 – 808.
[29] W. E. Moerner, D. P. Fromm, Rev. Sci. Instrum. 2003, 74, 3597 –
3619.
[30] T. Frster, Ann. Phys. 1948, 2, 55 – 75.
[31] W. E. Moerner, L. Kador, Phys. Rev. Lett. 1989, 62, 2535 – 2538.
[32] E. B. Shera, N. K. Seitzinger, L. M. Davis, R. A. Keller, S. A.
Soper, Chem. Phys. Lett. 1990, 174, 553 – 557.
[33] F. Zarrin, N. J. Dovichi, Anal. Chem. 1985, 57, 2690 – 2692.
[34] E. Betzig, R. J. Chichester, Science 1993, 262, 1422 – 1425.
[35] W. P. Ambrose, P. M. Goodwin, J. C. Martin, R. A. Keller,
Science 1994, 265, 364 – 367.
2666
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[36] T. Enderle, T. Ha, D. S. Chemla, S. Weiss, Ultramicroscopy
1998, 71, 303 – 309.
[37] F. De Lange, A. Cambi, R. Huijbens, B. de Bakker, W. Rensen,
M. Garcia-Parajo, N. Van Hulst, C. G. Figdor, J. Cell Sci. 2001,
114, 4153 – 4160.
[38] M. Koopman, B. I. de Bakker, M. F. Garcia-Parajo, N. F.
van Hulst, Appl. Phys. Lett. 2003, 83, 5083 – 5085.
[39] A. Cambi, F. de Lange, N. M. van Maarseveen, M. Nijhuis, B.
Joosten, E. M. H. P. van Dijk, B. I. de Bakker, J. A. M. Fransen,
P. H. M. Bovee-Geurts, F. N. van Leeuwen, N. F. Van Hulst,
C. G. Figdor, J. Cell Biol. 2004, 164, 145 – 155.
[40] J. B. Pawley, Handbook of Biological Confocal Microscopy,
2. Aufl., Plenum, New York, 1995.
[41] R. Rigler, . Mets, Proc. SPIE-Int. Soc. Opt. Eng. Proc. SPIE
1992, 239 – 247.
[42] R. Rigler, . Mets, J. Widengren, P. Kask, Eur. Biophys. J.
Biophys. Lett. 1993, 22, 169 – 175.
[43] C. Xu, W. Zipfel, J. B. Shear, R. M. Williams, W. W. Webb, Proc.
Natl. Acad. Sci. USA 1996, 93, 10 763 – 10 768.
[44] J. Mertz, C. Xu, W. W. Webb, Opt. Lett. 1995, 20, 2532 – 2534.
[45] L. Brand, C. Eggeling, C. Zander, K. H. Drexhage, C. A. M.
Seidel, J. Phys. Chem. A 1997, 101, 4313 – 4321.
[46] W. Denk, J. H. Strickler, W. W. Webb, Science 1990, 248, 73 – 76.
[47] A. Fischer, C. Cremer, E. H. K. Stelzer, Appl. Opt. 1995, 34,
1989 – 2003.
[48] C. Xu, W. W. Webb, J. Opt. Soc. Am. B 1996, 13, 481 – 491.
[49] K. G. Heinze, M. Rarbach, M. Jahnz, P. Schwille, Biophys. J.
2002, 83, 1671 – 1681.
[50] M. Eigen, R. Rigler, Proc. Natl. Acad. Sci. USA 1994, 91, 5740 –
5747.
[51] S. Maiti, U. Haupts, W. W. Webb, Proc. Natl. Acad. Sci. USA
1997, 94, 11 753 – 11 757.
[52] Y. Chen, J. D. Mller, P. T. C. So, E. Gratton, Biophys. J. 1999,
77, 553 – 567.
[53] P. Kask, K. Palo, D. Ullmann, K. Gall, Proc. Natl. Acad. Sci.
USA 1999, 96, 13 756 – 13 761.
[54] P. Kask, K. Palo, N. Fay, L. Brand, . Mets, D. Ullmann, J.
Jungmann, J. Pschorr, K. Gall, Biophys. J. 2000, 78, 1703 – 1713.
[55] K. Palo, . Mets, S. Jger, P. Kask, K. Gall, Biophys. J. 2000, 79,
2858 – 2866.
[56] T. A. Laurence, A. N. Kapanidis, X. Kong, D. S. Chemla, S.
Weiss, J. Phys. Chem. B 2004, 108, 3051 – 3067.
[57] E. S. Yeung, Chem. Rec. 2001, 1, 123 – 139.
[58] T. Ha, I. Rasnik, W. Cheng, H. P. Babcock, G. H. Gauss, T. M.
Lohman, S. Chu, Nature 2002, 419, 638 – 641.
[59] D. Axelrod, Methods Cell Biol. 1989, 30, 245 – 270.
[60] W. P. Ambrose, P. M. Goodwin, J. P. Nolan, Cytometry 1999, 36,
224 – 231.
[61] T. Funatsu, Y. Harada, M. Tokunaga, K. Saito, T. Yanagida,
Nature 1995, 374, 555 – 559.
[62] M. Tokunaga, K. Kitamura, K. Saito, A. H. Iwane, T. Yanagida,
Biochem. Biophys. Res. Commun. 1997, 235, 47 – 53.
[63] A. Yildiz, M. Tomishige, R. D. Vale, P. R. Selvin, Science 2004,
303, 676 – 679.
[64] A. Yildiz, J. N. Forkey, S. A. McKinney, T. Ha, Y. E. Goldman,
P. R. Selvin, Science 2003, 300, 2061 – 2065.
[65] R. D. Vale, T. Funatsu, D. W. Pierce, L. Romberg, Y. Harada, T.
Yanagida, Nature 1996, 380, 451 – 453.
[66] J. N. Forkey, M. E. Quinlan, M. A. Shaw, J. E. T. Corrie, Y. E.
Goldman, Nature 2003, 422, 399 – 404.
[67] C. W. Wilkerson, P. M. Goodwin, W. P. Ambrose, J. C. Martin,
R. A. Keller, Appl. Phys. Lett. 1993, 62, 2030 – 2032.
[68] M. Sauer, J. Arden-Jacob, K. H. Drexhage, F. Gbel, U.
Lieberwirth, K. Mhlegger, R. Mller, J. Wolfrum, C. Zander,
Bioimaging 1998, 6, 14 – 24.
[69] S. A. Soper, L. M. Davis, E. B. Shera, J. Opt. Soc. Am. B 1992, 9,
1761 – 1769.
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
[70] A. Van Orden, R. A. Keller, Anal. Chem. 1998, 70, 4463 – 4471.
[71] J. Schaffer, A. Volkmer, C. Eggeling, V. Subramaniam, G.
Striker, C. A. M. Seidel, J. Phys. Chem. A 1999, 103, 331 – 336.
[72] F. Stracke, C. Blum, S. Becker, K. Mllen, A. J. Meixner, Chem.
Phys. Lett. 2000, 325, 196 – 202.
[73] A. J. Meixner, M. A. Weber, J. Lumin. 2000, 86, 181 – 187.
[74] K. D. Weston, S. K. Buratto, J. Phys. Chem. A 1998, 102, 3635 –
3638.
[75] H. P. Lu, X. S. Xie, Nature 1997, 385, 143 – 146.
[76] K. D. Weston, L. S. Goldner, J. Phys. Chem. B 2001, 105, 3453 –
3462.
[77] P. Tinnefeld, V. Buschmann, D. P. Herten, K.-T. Han, M. Sauer,
Single Mol. 2000, 1, 215 – 223.
[78] P. Tinnefeld, D. P. Herten, M. Sauer, J. Phys. Chem. A 2001, 105,
7989 – 8003.
[79] T. Ha, T. Enderle, D. S. Chemla, P. R. Selvin, S. Weiss, Phys.
Rev. Lett. 1996, 77, 3979 – 3982.
[80] M. Prummer, C. G. Hubner, B. Sick, B. Hecht, A. Renn, U. P.
Wild, Anal. Chem. 2000, 72, 443 – 447.
[81] M. Prummer, B. Sick, A. Renn, U. P. Wild, Anal. Chem. 2004,
76, 1633 – 1640.
[82] D. P. Herten, P. Tinnefeld, M. Sauer, Appl. Phys. B 2000, 71,
765 – 771.
[83] J.-P. Knemeyer, D.-P. Herten, M. Sauer, Anal. Chem. 2003, 75,
2147 – 2153.
[84] D. S. English, E. J. Harbron, P. F. Barbara, J. Chem. Phys. 2001,
114, 10 479 – 10 485.
[85] P. Tinnefeld, K. D. Weston, T. Vosch, M. Cotlet, T. Weil, J.
Hofkens, K. Mllen, F. C. De Schryver, M. Sauer, J. Am. Chem.
Soc. 2002, 124, 14 310 – 14 311.
[86] C. Eggeling, S. Berger, L. Brand, J. R. Fries, J. Schaffer, A.
Volkmer, C. A. M. Seidel, J. Biotechnol. 2001, 86, 163 – 180.
[87] R. Khnemuth, C. A. M. Seidel, Single Mol. 2001, 2, 251 – 254.
[88] A. M. Kelley, X. Michalet, S. Weiss, Science 2001, 292, 1671 –
1672.
[89] P. Tinnefeld, C. Mller, M. Sauer, Chem. Phys. Lett. 2001, 345,
252 – 258.
[90] K. D. Weston, M. Dyck, P. Tinnefeld, C. Mller, D. P. Herten,
M. Sauer, Anal. Chem. 2002, 74, 5342 – 5349.
[91] T. Vosch, M. Cotlet, J. Hofkens, K. Van Biest, M. Lor, K.
Weston, P. Tinnefeld, M. Sauer, L. Latterini, K. Mllen, F. C.
De Schryver, J. Phys. Chem. A 2003, 107, 6920 – 6931.
[92] C. Eggeling, J. R. Fries, L. Brand, R. Gnther, C. A. M. Seidel,
Proc. Natl. Acad. Sci. USA 1998, 95, 1556 – 1561.
[93] J. R. Fries, L. Brand, C. Eggeling, M. Kllner, C. A. M. Seidel,
J. Phys. Chem. A 1998, 102, 6601 – 6613.
[94] J. Tellinghuisen, C. W. Wilkerson, Jr., Anal. Chem. 1993, 65,
1240 – 1246.
[95] J. Enderlein, M. Sauer, J. Phys. Chem. A 2001, 105, 48 – 53.
[96] J. Enderlein, P. M. Goodwin, A. Van Orden, W. P. Ambrose, R.
Erdmann, R. A. Keller, Chem. Phys. Lett. 1997, 270, 464 – 470.
[97] M. Maus, M. Cotlet, J. Hofkens, T. Gensch, F. C. De Schryver, J.
Schaffer, C. A. M. Seidel, Anal. Chem. 2001, 73, 2078 – 2086.
[98] T. Ha, T. Enderle, D. S. Chemla, P. R. Selvin, S. Weiss, Chem.
Phys. Lett. 1997, 271, 1 – 5.
[99] R. Hanbury Brown, R. Q. Twiss, Nature 1956, 178, 1046 – 1048.
[100] T. Basche, W. E. Moerner, M. Orrit, H. Talon, Phys. Rev. Lett.
1992, 69, 1516 – 1519.
[101] L. Fleury, J. M. Segura, G. Zumofen, B. Hecht, U. P. Wild, Phys.
Rev. Lett. 2000, 84, 1148 – 1151.
[102] W. P. Ambrose, P. M. Goodwin, J. Enderlein, D. J. Semin, J. C.
Martin, R. A. Keller, Chem. Phys. Lett. 1997, 269, 365 – 370.
[103] M. Wu, P. M. Goodwin, W. P. Ambrose, R. A. Keller, J. Phys.
Chem. 1996, 100, 17 406 – 17 409.
[104] J. Hofkens, M. Maus, T. Gensch, T. Vosch, M. Cotlet, F. Kohn,
A. Herrmann, K. Mllen, F. De Schryver, J. Am. Chem. Soc.
2000, 122, 9278 – 9288.
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
[105] P. J. Rothwell, S. Berger, O. Kensch, S. Felekyan, M. Antonik,
B. M. Wohrl, T. Restle, R. S. Goody, C. A. M. Seidel, Proc. Natl.
Acad. Sci. USA 2003, 100, 1655 – 1660.
[106] K. Kinosita, Jr., H. Itoh, S. Ishiwata, K. Hirano, T. Nishizaka, T.
Hayakawa, J. Cell Biol. 1991, 115, 67 – 73.
[107] L. Cognet, G. S. Harms, G. A. Blab, P. H. M. Lommerse, T.
Schmidt, Appl. Phys. Lett. 2000, 77, 4052 – 4054.
[108] L. Stryer, R. P. Haugland, Proc. Natl. Acad. Sci. USA 1967, 58,
719 – 726.
[109] L. Stryer, Annu. Rev. Biochem. 1978, 47, 819 – 846.
[110] W. Veatch, L. Stryer, J. Mol. Biol. 1977, 113, 89 – 102.
[111] J. Szollosi, S. Damjanovich, L. Matyus, Cytometry 1998, 34,
159 – 179.
[112] F. Sthmeier, J. B. Welch, A. I. H. Murchie, D. M. J. Lilley,
R. M. Clegg, Biochemistry 1997, 36, 13 530 – 13 538.
[113] K. M. Parkhurst, M. Brenowitz, L. J. Parkhurst, Biochemistry
1996, 35, 7459 – 7465.
[114] T. Tuschl, C. Gohlke, T. M. Jovin, E. Westhof, F. Eckstein,
Science 1994, 266, 785 – 789.
[115] S. Weiss, Nat. Struct. Biol. 2000, 7, 724 – 729.
[116] T. Ha, Methods 2001, 25, 78 – 86.
[117] A. A. Deniz, T. A. Laurence, G. S. Beligere, M. Dahan, A. B.
Martin, D. S. Chemla, P. E. Dawson, P. G. Schultz, S. Weiss,
Proc. Natl. Acad. Sci. USA 2000, 97, 5179 – 5184.
[118] T. Ha, Single Mol. 2001, 2, 283 – 284.
[119] H. D. Kim, G. U. Nienhaus, T. Ha, J. W. Orr, J. R. Williamson, S.
Chu, Proc. Natl. Acad. Sci. USA 2002, 99, 4284 – 4289.
[120] T. Ha, X. W. Zhuang, H. D. Kim, J. W. Orr, J. R. Williamson, S.
Chu, Proc. Natl. Acad. Sci. USA 1999, 96, 9077 – 9082.
[121] X. Zhuang, L. E. Bartley, H. P. Babcock, R. Russell, T. Ha, D.
Herschlag, S. Chu, Science 2000, 288, 2048 – 2051.
[122] E. Tan, T. J. Wilson, M. K. Nahas, R. M. Clegg, D. M. J. Lilley,
T. Ha, Proc. Natl. Acad. Sci. USA 2003, 100, 9308 – 9313.
[123] M. Borsch, M. Diez, B. Zimmermann, R. Reuter, P. Graber,
FEBS Lett. 2002, 527, 147 – 152.
[124] E. Rhoades, E. Gussakovsky, G. Haran, Proc. Natl. Acad. Sci.
USA 2003, 100, 3197 – 3202.
[125] A. A. Deniz, M. Dahan, J. R. Grunwell, T. Ha, A. E. Faulhaber,
D. S. Chemla, S. Weiss, P. G. Schultz, Proc. Natl. Acad. Sci. USA
1999, 96, 3670 – 3675.
[126] M. Sauer, K. H. Drexhage, U. Lieberwirth, R. Mller, S. Nord,
C. Zander, Chem. Phys. Lett. 1998, 284, 153 – 163.
[127] G. Bonnet, O. Krichevsky, A. Libchaber, Proc. Natl. Acad. Sci.
USA 1998, 95, 8602 – 8606.
[128] N. L. Goddard, G. Bonnet, O. Krichevsky, A. Libchaber, Phys.
Rev. Lett. 2000, 85, 2400 – 2403.
[129] M. I. Wallace, L. Ying, S. Balasubramanian, D. Klenerman,
Proc. Natl. Acad. Sci. USA 2001, 98, 5584 – 5589.
[130] O. Piestert, H. Barsch, V. Buschmann, T. Heinlein, J.-P.
Knemeyer, K. D. Weston, M. Sauer, Nano Lett. 2003, 3, 979 –
982.
[131] M. A. Osborne, C. L. Barnes, S. Balasubramanian, D. Klenerman, J. Phys. Chem. B 2001, 105, 3120 – 3126.
[132] D. S. Talaga, W. L. Lau, H. Roder, J. Tang, Y. Jia, W. F.
DeGrado, R. M. Hochstrasser, Proc. Natl. Acad. Sci. USA 2000,
97, 13 021 – 13 026.
[133] M. Margittai, J. Widengren, E. Schweinberger, G. F. Schroeder,
S. Felekyan, E. Haustein, M. Koenig, D. Fasshauer, H.
Grubmller, R. Jahn, C. A. M. Seidel, Proc. Natl. Acad. Sci.
USA 2003, 100, 15 516 – 15 521.
[134] M. Diez, B. Zimmermann, M. Boersch, M. Koenig, E.
Schweinberger, S. Steigmiller, R. Reuter, S. Felekyan, V.
Kudryavtsev, C. A. M. Seidel, P. Graeber, Nat. Struct. Mol.
Biol. 2004, 11, 135 – 141.
[135] D. Rehm, A. Weller, Isr. J. Chem. 1970, 8, 259 – 271.
[136] R. A. Marcus, N. Sutin, Biochim. Biophys. Acta 1985, 811, 265 –
322.
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2667
Reviews
P. Tinnefeld and M. Sauer
[137] K. Chattopadhyay, S. Saffarian, E. L. Elson, C. Frieden, Proc.
Natl. Acad. Sci. USA 2002, 99, 14 171 – 14 176.
[138] R. M. Watt, E. W. Voss, Jr., Immunochem. 1977, 14, 533 – 551.
[139] N. Marme, J.-P. Knemeyer, M. Sauer, J. Wolfrum, Bioconjugate
Chem. 2003, 14, 1133 – 1139.
[140] H. Yang, G. Luo, P. Karnchanaphanurach, T.-M. Louie, I. Rech,
S. Cova, L. Xun, X. S. Xie, Science 2003, 302, 262 – 266.
[141] H. Neuweiler, M. Sauer, Curr. Pharm. Biotechnol. 2004, 5, 285 –
298.
[142] M. Sauer, K. T. Han, R. Mller, S. Nord, A. Schulz, S. Seeger, J.
Wolfrum, J. Arden-Jacob, G. Deltau, N. J. Marx, C. Zander,
K. H. Drexhage, J. Fluoresc. 1995, 5, 247 – 261.
[143] J. P. Knemeyer, N. Marme, M. Sauer, Anal. Chem. 2000, 72,
3717 – 3724.
[144] T. Heinlein, J.-P. Knemeyer, O. Piestert, M. Sauer, J. Phys.
Chem. B 2003, 107, 7957 – 7964.
[145] A. C. Vaiana, H. Neuweiler, A. Schulz, J. Wolfrum, M. Sauer,
J. C. Smith, J. Am. Chem. Soc. 2003, 125, 14 564 – 14 572.
[146] H. Neuweiler, A. Schulz, A. C. Vaiana, J. C. Smith, S. Kaul, J.
Wolfrum, M. Sauer, Angew. Chem. 2002, 114, 4964 – 4968;
Angew. Chem. Int. Ed. 2002, 41, 4769 – 4773.
[147] L. J. Lapidus, W. A. Eaton, J. Hofrichter, Proc. Natl. Acad. Sci.
USA 2000, 97, 7220 – 7225.
[148] T. Soussi, Immunol. Today 1996, 17, 354 – 356.
[149] M. Sauer, Angew. Chem. 2003, 115, 1834 – 1837; Angew. Chem.
Int. Ed. 2003, 42, 1790 – 1793.
[150] D. A. Van den Bout, W. T. Yip, D. H. Hu, D. K. Fu, T. M.
Swager, P. F. Barbara, Science 1997, 277, 1074 – 1077.
[151] W. T. Yip, D. H. Hu, J. Yu, D. A. Van den Bout, P. F. Barbara, J.
Phys. Chem. A 1998, 102, 7564 – 7575.
[152] T. Huser, M. Yan, L. J. Rothberg, Proc. Natl. Acad. Sci. USA
2000, 97, 11 187 – 11 191.
[153] J. D. White, J. H. Hsu, S. C. Yang, W. S. Fann, G. Y. Pern, S. A.
Chen, J. Chem. Phys. 2001, 114, 3848 – 3852.
[154] M. A. Bopp, Y. W. Jia, L. Q. Li, R. J. Cogdell, R. M. Hochstrasser, Proc. Natl. Acad. Sci. USA 1997, 94, 10 630 – 10 635.
[155] M. A. Bopp, A. Sytnik, T. D. Howard, R. J. Cogdell, R. M.
Hochstrasser, Proc. Natl. Acad. Sci. USA 1999, 96, 11 271 –
11 276.
[156] J. Hofkens, W. Schroeyers, D. Loos, M. Cotlet, F. Kohn, T.
Vosch, M. Maus, A. Herrmann, K. Mllen, T. Gensch, F. C. De
Schryver, Spectrochim. Acta Part A 2001, 57, 2093 – 2107.
[157] L. M. Ying, X. S. Xie, J. Phys. Chem. B 1998, 102, 10 399 –
10 409.
[158] B. Lounis, J. Deich, F. I. Rosell, S. G. Boxer, W. E. Moerner, J.
Phys. Chem. B 2001, 105, 5048 – 5054.
[159] M. Cotlet, J. Hofkens, S. Habuchi, G. Dirix, M. Van Guyse, J.
Michiels, J. Vanderleyden, F. C. De Schryver, Proc. Natl. Acad.
Sci. USA 2001, 98, 14 398 – 14 403.
[160] M. F. Garcia-Parajo, M. Koopman, E. M. H. P. Van Dijk, V.
Subramaniam, N. F. Van Hulst, Proc. Natl. Acad. Sci. USA
2001, 98, 14 392 – 14 397.
[161] T. Christ, F. Petzke, P. Bordat, A. Herrmann, E. Reuther, K.
Mullen, T. Basche, J. Lumin. 2002, 98, 23 – 33.
[162] J. Hernando, M. van der Schaaf, E. M. H. P. van Dijk, M. Sauer,
M. F. Garcia-Parajo, N. F. van Hulst, J. Phys. Chem. A 2003,
107, 43 – 52.
[163] P. Tinnefeld, V. Buschmann, K. D. Weston, M. Sauer, J. Phys.
Chem. A 2003, 107, 323 – 327.
[164] J. Yu, D. H. Hu, P. F. Barbara, Science 2000, 289, 1327 – 1330.
[165] T. Gensch, J. Hofkens, A. Heirmann, K. Tsuda, W. Verheijen, T.
Vosch, T. Christ, T. Basche, K. Mllen, F. C. De Schryver,
Angew. Chem. 1999, 111, 3970 – 3974; Angew. Chem. Int. Ed.
1999, 38, 3752 – 3756.
[166] R. Gronheid, J. Hofkens, F. Kohn, T. Weil, E. Reuther, K.
Mllen, F. C. De Schryver, J. Am. Chem. Soc. 2002, 124, 2418 –
2419.
2668
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[167] T. Basche, S. Kummer, C. Bruchle, Nature 1995, 373, 132 – 134.
[168] J. Bernard, L. Fleury, H. Talon, M. Orrit, J. Chem. Phys. 1993,
98, 850 – 859.
[169] D. S. English, E. J. Harbron, P. F. Barbara, J. Phys. Chem. A
2000, 104, 9057 – 9061.
[170] C. G. Hbner, G. Zumofen, A. Renn, A. Herrmann, K. Mllen,
T. Basche, Phys. Rev. Lett. 2003, 91, 093903/1 – 093903/4.
[171] J. Hofkens, M. Cotlet, T. Vosch, P. Tinnefeld, K. D. Weston, C.
Ego, A. Grimsdale, K. Mllen, D. Beljonne, J. L. Bredas, S.
Jordens, G. Schweitzer, M. Sauer, F. C. De Schryver, Proc. Natl.
Acad. Sci. USA 2003, 100, 13 146 – 13 151.
[172] M. Lippitz, C. G. Hbner, T. Christ, H. Eichner, P. Bordat, A.
Herrmann, K. Mllen, T. Basche, Phys. Rev. Lett. 2004, 92,
103001/1 – 103001/4.
[173] R. W. Wagner, J. S. Lindsey, J. Am. Chem. Soc. 1994, 116, 9759 –
9760.
[174] M. Heilemann, V. Buschmann, O. Piestert, P. Tinnefeld, K. D.
Weston, M. Sauer, Proc. SPIE-Int. Soc. Opt. Eng. 2003, 4962,
38 – 46.
[175] M. Heilemann, P. Tinnefeld, G. Sanchez Mosteiro, M. F.
Garcia Parajo, N. F. van Hulst, M. Sauer, J. Am. Chem. Soc.
2004, 126, 6514 – 6515.
[176] S. Vyawahare, S. Eyal, K. D. Mathews, S. R. Quake, Nano Lett.
2004, 4, 1035 – 1039.
[177] C. Bustamante, J. F. Marko, E. D. Siggia, S. Smith, Science 1994,
265, 1599 – 1600.
[178] G. Calzaferri, S. Huber, H. Maas, C. Minkowski, Angew. Chem.
2003, 115, 3860 – 3888; Angew. Chem. Int. Ed. 2003, 42, 3732 –
3758.
[179] M. Irie, T. Fukaminato, T. Sasaki, N. Tamai, T. Kawai, Nature
2002, 420, 759 – 760.
[180] D. L. Spector, Annu. Rev. Cell Biol. 1993, 9, 265 – 315.
[181] A. I. Lamond, W. C. Earnshaw, Science 1998, 280, 547 – 553.
[182] P. R. Cook, Science 1999, 284, 1790 – 1795.
[183] T. Cremer, C. Cremer, Nat. Rev. Genet. 2001, 2, 292 – 301.
[184] D. G. Wansink, O. C. Sibon, F. F. Cremers, R. van Driel, L.
de Jong, J. Cell Biol. 1996, 62, 10 – 18.
[185] A. Pombo, D. A. Jackson, M. Hollinshead, Z. X. Wang, R. G.
Roeder, P. R. Cook, Embo J. 1999, 18, 2241 – 2253.
[186] A. M. Femino, F. S. Fay, K. Fogarty, R. H. Singer, Science 1998,
280, 585 – 590.
[187] J. M. Levsky, S. M. Shenoy, R. C. Pezo, R. H. Singer, Science
2002, 297, 836 – 840.
[188] J. B. Lawrence, R. H. Singer, L. M. Marselle, Cell 1989, 57,
493 – 502.
[189] R. W. Dirks, Histochem. Cell Biol. 1996, 106, 151 – 166.
[190] R. W. Dirks, C. Molenaar, H. J. Tanke, Histochem. Cell Biol.
2001, 115, 3 – 11.
[191] J. Zhang, R. E. Campbell, A. Y. Ting, R. Y. Tsien, Nat. Rev.
Mol. Cell Biol. 2002, 3, 906 – 918.
[192] A. R. Thierry, A. Dritschilo, Nucleic Acids Res. 1992, 20, 5691 –
5698.
[193] C. Zander, K. H. Drexhage, K. T. Han, J. Wolfrum, M. Sauer,
Chem. Phys. Lett. 1998, 286, 457 – 465.
[194] W. Becker, H. Hickl, C. Zander, K. H. Drexhage, M. Sauer, S.
Siebert, J. Wolfrum, Rev. Sci. Instrum. 1999, 70, 1835 – 1841.
[195] V. P. Torchilin, R. Rammohan, V. Weissig, T. S. Levchenko,
Proc. Natl. Acad. Sci. USA 2001, 98, 8786 – 8791.
[196] S. R. Schwarze, A. Ho, A. Vocero-Akbani, S. F. Dowdy, Science
1999, 285, 1569 – 1572.
[197] E. Vives, P. Brodin, B. Lebleu, J. Biol. Chem. 1997, 272, 16 010 –
16 017.
[198] M. Rueping, Y. Mahajan, M. Sauer, D. Seebach, ChemBioChem 2002, 3, 257 – 259.
[199] D. Seebach, K. Namoto, Y. R. Mahajan, P. Bindschdler, R.
Sustmann, M. Kirsch, N. S. Ryder, M. Weiss, M. Sauer, C. Roth,
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
[200]
[201]
[202]
[203]
[204]
[205]
[206]
[207]
[208]
[209]
[210]
[211]
[212]
[213]
[214]
[215]
[216]
[217]
[218]
[219]
[220]
[221]
[222]
[223]
[224]
[225]
[226]
[227]
[228]
[229]
[230]
Chemie
S. Werner, H.-D. Beer, C. Munding, P. Walde, M. Voser, Chem.
Biodiversity 2004, 1, 65 – 97.
G. J. Schtz, V. P. Patuschenko, H. J. Gruber, H. G. Knaus, B.
Pragl, H. Schindler, Single Mol. 2000, 1, 25 – 31.
N. Marme, J. P. Knemeyer, J. Wolfrum, M. Sauer, Angew. Chem.
2004, 116, 3886 – 3890; Angew. Chem. Int. Ed. 2004, 43, 3798 –
3801.
R. Y. Tsien, Annu. Rev. Biochem. 1998, 67, 509 – 544.
G. S. Baird, D. A. Zacharias, R. Y. Tsien, Proc. Natl. Acad. Sci.
USA 2000, 97, 11 984 – 11 989.
S. L. Moores, J. H. Sabry, J. A. Spudich, Proc. Natl. Acad. Sci.
USA 1996, 93, 443 – 446.
K. R. Olson, J. R. McIntosh, J. B. Olmsted, J. Cell Biol. 1995,
130, 639 – 650.
G. S. Harms, L. Cognet, P. H. M. Lommerse, G. A. Blab, T.
Schmidt, Biophys. J. 2001, 80, 2396 – 2408.
G. S. Harms, L. Cognet, P. H. M. Lommerse, G. A. Blab, H.
Kahr, R. Gamsjger, H. P. Spaink, N. M. Soldatov, C. Romanin,
T. Schmidt, Biophys. J. 2001, 81, 2639 – 2646.
R. Iino, A. Kusumi, J. Fluoresc. 2001, 11, 187 – 195.
R. Iino, I. Koyama, A. Kusumi, Biophys. J. 2001, 80, 2667 – 2677.
J. Widengren, B. Terry, R. Rigler, Chem. Phys. 1999, 249, 259 –
271.
A. Zumbusch, G. Jung, Single Mol. 2000, 1, 261 – 270.
M. Cotlet, J. Hofkens, M. Maus, T. Gensch, M. Van der Auweraer, J. Michiels, G. Dirix, M. Van Guyse, J. Vanderleyden,
A. Visser, F. C. De Schryver, J. Phys. Chem. B 2001, 105, 4999 –
5006.
M. V. Matz, A. F. Fradkov, Y. A. Labas, A. P. Savitsky, A. G.
Zaraisky, M. L. Markelov, S. A. Lukyanov, Nat. Biotechnol.
1999, 17, 969 – 973.
B. J. Bevis, B. S. Glick, Nat. Biotechnol. 2002, 20, 1159.
J. Wiedenmann, A. Schenk, C. Rocker, A. Girod, K.-D.
Spindler, G. U. Nienhaus, Proc. Natl. Acad. Sci. USA 2002,
99, 11 646 – 11 651.
N. G. Gurskaya, A. F. Fradkov, A. Terskikh, M. V. Matz, Y. A.
Labas, V. I. Martynov, Y. G. Yanushevich, K. A. Lukyanov,
S. A. Lukyanov, FEBS Lett. 2001, 507, 16 – 20.
R. E. Campbell, O. Tour, A. E. Palmer, P. A. Steinbach, G. S.
Baird, D. A. Zacharias, R. Y. Tsien, Proc. Natl. Acad. Sci. USA
2002, 99, 7877 – 7882.
C.-D. Hu, Y. Chinenov, T. K. Kerppola, Mol. Cell 2002, 9, 789 –
798.
J. C. Politz, E. S. Browne, D. E. Wolf, T. Pederson, Proc. Natl.
Acad. Sci. USA 1998, 95, 6043 – 6048.
P. Schwille, F. J. Meyer-Almes, R. Rigler, Biophys. J. 1997, 72,
1878 – 1886.
R. Brock, G. Vamosi, G. Vereb, T. M. Jovin, Proc. Natl. Acad.
Sci. USA 1999, 96, 10 123 – 10 128.
T. A. Byassee, W. C. Chan, S. Nie, Anal. Chem. 2000, 72, 5606 –
5611.
T. Kues, A. Dickmanns, R. Luhrmann, R. Peters, U. Kubitscheck, Proc. Natl. Acad. Sci. USA 2001, 98, 12 021 – 12 026.
T. Kues, R. Peters, U. Kubitscheck, Biophys. J. 2001, 80, 2954 –
2967.
G. Seisenberger, M. U. Ried, T. Endreß, H. Bning, M. Hallek,
C. Bruchle, Science 2001, 294, 1929 – 1932.
X. Zhuang, H. Kim, M. J. B. Pereira, H. P. Babcock, N. G.
Walter, S. Chu, Science 2002, 296, 1473 – 1476.
M. Bruchez, M. Moronne, P. Gin, S. Weiss, A. P. Alivisatos,
Science 1998, 281, 2013 – 2016.
M. Dahan, T. Laurence, F. Pinaud, D. S. Chemla, A. P.
Alivisatos, M. Sauer, S. Weiss, Opt. Lett. 2001, 26, 825 – 827.
W. C. W. Chan, S. M. Nie, Science 1998, 281, 2016 – 2018.
B. Dubertret, P. Skourides, D. J. Norris, V. Noireaux, A. H.
Brivanlou, A. Libchaber, Science 2002, 298, 1759 – 1762.
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
[231] J. K. Jaiswal, H. Mattoussi, J. M. Mauro, S. M. Simon, Nat.
Biotechnol. 2003, 21, 47 – 51.
[232] B. M. Lingerfelt, H. Mattoussi, E. R. Goldman, J. M. Mauro,
G. P. Anderson, Anal. Chem. 2003, 75, 4043 – 4049.
[233] E. R. Goldman, E. D. Balighian, H. Mattoussi, M. K. Kuno,
J. M. Mauro, P. T. Tran, G. P. Anderson, J. Am. Chem. Soc. 2002,
124, 6378 – 6382.
[234] H. Mattoussi, J. M. Mauro, E. R. Goldman, G. P. Anderson,
V. C. Sundar, F. V. Mikulec, M. G. Bawendi, J. Am. Chem. Soc.
2000, 122, 12 142 – 12 150.
[235] X. Wu, H. Liu, J. Liu, K. N. Haley, J. A. Treadway, J. P. Larson,
N. Ge, F. Peale, M. P. Bruchez, Nat. Biotechnol. 2003, 21, 41 –
46.
[236] D. R. Larson, W. R. Zipfel, R. M. Williams, S. W. Clark, M. P.
Bruchez, F. W. Wise, W. W. Webb, Science 2003, 300, 1434 –
1437.
[237] F. Pinaud, D. King, X. Michalet, S. Doose, S. Weiss, Eur.
Biophys. J. 2003, 32, 261.
[238] M. Dahan, S. Levi, C. Luccardini, P. Rostaing, B. Riveau, A.
Triller, Science 2003, 302, 442 – 445.
[239] M. Nirmal, B. O. Dabbousi, M. G. Bawendi, J. J. Macklin, J. K.
Trautman, T. D. Harris, L. E. Brus, Nature 1996, 383, 802 – 804.
[240] M. Kuno, D. P. Fromm, H. F. Hamann, A. Gallagher, D. J.
Nesbitt, J. Chem. Phys. 2000, 112, 3117 – 3120.
[241] M. Kuno, D. P. Fromm, H. F. Hamann, A. Gallagher, D. J.
Nesbitt, J. Chem. Phys. 2001, 115, 1028 – 1040.
[242] S. Hohng, T. Ha, J. Am. Chem. Soc. 2004, 126, 1324 – 1325.
[243] S. R. Cordero, P. J. Carson, R. A. Estabrook, G. F. Strouse, S. K.
Buratto, J. Phys. Chem. B 2000, 104, 12 137 – 12 142.
[244] W. J. Parak, R. Boudreau, M. Le Gros, D. Gerion, D. Zanchet,
C. M. Micheel, S. C. Williams, A. P. Alivisatos, C. Larabell, Adv.
Mater. 2002, 14, 882 – 885.
[245] I. Sase, H. Miyata, J. E. T. Corrie, J. S. Craik, K. Kinosita,
Biophys. J. 1995, 69, 323 – 328.
[246] T. Schmidt, G. J. Schtz, W. Baumgartner, H. J. Gruber, H.
Schindler, J. Phys. Chem. 1995, 99, 17 662 – 17 668.
[247] T. Schmidt, G. J. Schtz, W. Baumgartner, H. J. Gruber, H.
Schindler, Proc. Natl. Acad. Sci. USA 1996, 93, 2926 – 2929.
[248] R. M. Dickson, D. J. Norris, Y. L. Tzeng, W. E. Moerner,
Science 1996, 274, 966 – 969.
[249] M. P. Gordon, T. Ha, P. R. Selvin, Proc. Natl. Acad. Sci. USA
2004, 101, 6462 – 6465.
[250] A. Patwardhan, E. M. M. Manders, Bioimaging 1996, 4, 17 – 24.
[251] A. Esa, P. Edelmann, G. Kreth, L. Trakhtenbrot, N. Amariglio,
G. Rechavi, M. Hausmann, C. Cremer, J. Microsc. 2000, 199,
96 – 105.
[252] A. M. van Oijen, J. Khler, J. Schmidt, M. Mller, G. J.
Brakenhoff, Chem. Phys. Lett. 1998, 292, 183 – 187.
[253] T. D. Lacoste, X. Michalet, F. Pinaud, D. S. Chemla, A. P.
Alivisatos, S. Weiss, Proc. Natl. Acad. Sci. USA 2000, 97, 9461 –
9466.
[254] X. Michalet, T. D. Lacoste, S. Weiss, Methods 2001, 25, 87 – 102.
[255] E. Betzig, Opt. Lett. 1995, 20, 237 – 239.
[256] C. Zander, M. Sauer, K. H. Drexhage, D. S. Ko, A. Schulz, J.
Wolfrum, L. Brand, C. Eggeling, C. A. M. Seidel, Appl. Phys. B
1996, 63, 517 – 523.
[257] M. Heilemann, D. P. Herten, R. Heintzmann, C. Cremer, C.
Mller, P. Tinnefeld, K. D. Weston, J. Wolfrum, M. Sauer, Anal.
Chem. 2002, 74, 3511 – 3517.
[258] S. Hell, E. H. K. Stelzer, J. Opt. Soc. Am. 1992, 9, 2159 – 2166.
[259] A. Egner, S. Jakobs, S. W. Hell, Proc. Natl. Acad. Sci. USA 2002,
99, 3370 – 3375.
[260] T. Enderle, T. Ha, D. F. Ogletree, D. S. Chemla, C. Magowan, S.
Weiss, Proc. Natl. Acad. Sci. USA 1997, 94, 520 – 525.
[261] H. Gersen, M. F. Garcia-Parajo, L. Novotny, J. A. Veerman, L.
Kuipers, N. F. van Hulst, Phys. Rev. Lett. 2000, 85, 5312 – 5315.
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2669
Reviews
P. Tinnefeld and M. Sauer
[262] P. T. So, H. S. Kwon, C. Y. Dong, J. Opt. Soc. Am. A 2001, 18,
2833 – 2845.
[263] B. Albrecht, A. V. Failla, R. Heintzmann, C. Cremer, J. Biomed.
Opt. 2001, 6, 292 – 299.
[264] V. Failla Antonio, U. Spoeri, B. Albrecht, A. Kroll, C. Cremer,
Appl. Opt. 2002, 41, 7275 – 7283.
[265] C. Hettich, C. Schmitt, J. Zitzmann, S. Kuehn, I. Gerhardt, V.
Sandoghdar, Science 2002, 298, 385 – 389.
[266] T. A. Klar, S. Jakobs, M. Dyba, A. Egner, S. W. Hell, Proc. Natl.
Acad. Sci. USA 2000, 97, 8206 – 8210.
[267] V. Westphal, C. M. Blanca, M. Dyba, L. Kastrup, S. W. Hell,
Appl. Phys. Lett. 2003, 82, 3125 – 3127.
[268] S. W. Hell, Nat. Biotechnol. 2003, 21, 1347 – 1355.
[269] M. Dyba, S. Jakobs, S. W. Hell, Nat. Biotechnol. 2003, 21, 1303 –
1304.
[270] V. Westphal, L. Kastrup, S. W. Hell, Appl. Phys. B 2003, 77,
377 – 380.
[271] Z. Zhu, J. Chao, H. Yu, A. S. Waggoner, Nucleic Acids Res.
1994, 22, 3418 – 3422.
[272] H. Yu, J. Chao, D. Patek, R. Mujumdar, S. Mujumdar, A. S.
Waggoner, Nucleic Acids Res. 1994, 22, 3226 – 3232.
[273] S. Brakmann, P. Nieckchen, ChemBioChem 2001, 2, 773 – 777.
[274] S. Brakmann, S. Lobermann, Angew. Chem. 2002, 114, 3350 –
3352; Angew. Chem. Int. Ed. 2002, 41, 3215 – 3217.
[275] S. Brakmann, S. Lobermann, Angew. Chem. 2001, 113, 1473 –
1476; Angew. Chem. Int. Ed. 2001, 40, 1427 – 1429.
[276] M. Orrit, Science 2003, 302, 239 – 240.
[277] C. D. Snow, H. Nguyen, V. S. Pande, M. Gruebele, Nature 2002,
420, 102 – 106.
[278] L. Edman, Z. Foldes-Papp, S. Wennmalm, R. Rigler, Chem.
Phys. 1999, 247, 11 – 22.
[279] L. Edman, R. Rigler, Proc. Natl. Acad. Sci. USA 2000, 97,
8266 – 8271.
[280] A. N. Kapanidis, N. K. Lee, T. A. Laurence, S. Doose, E.
Margeat, S. Weiss, Proc. Natl. Acad. Sci. USA 2004, 101, 8936 –
8941.
[281] P. K. Hansma, B. Drake, O. Marti, S. A. Gould, C. B. Prater,
Science 1989, 243, 641 – 643.
[282] Y. E. Korchev, C. L. Bashford, M. Milovanovic, I. Vodyanoy,
M. J. Lab, Biophys. J. 1997, 73, 653 – 658.
[283] Y. E. Korchev, M. Raval, M. J. Lab, J. Gorelik, C. R. W.
Edwards, T. Rayment, D. Klenerman, Biophys. J. 2000, 78,
2675 – 2679.
[284] M. Ueda, Y. Sako, T. Tanaka, P. Devreotes, T. Yanagida, Science
2001, 294, 864 – 867.
[285] Y. E. Korchev, Y. A. Negulyaev, C. R. W. Edwards, I. Vodyanoy, M. J. Lab, Nat. Cell Biol. 2000, 2, 616 – 619.
[286] A. Sonnleitner, L. M. Mannuzzu, S. Terakawa, E. Y. Isacoff,
Proc. Natl. Acad. Sci. USA 2002, 99, 12 759 – 12 764.
[287] G. S. Harms, G. Orr, M. Montal, B. D. Thrall, S. D. Colson, H. P.
Lu, Biophys. J. 2003, 85, 1826 – 1838.
[288] T. Ide, Y. Takeuchi, T. Yanagida, Single Mol. 2002, 3, 33 – 42.
[289] C. Brunel, B. Lounis, P. Tamarat, M. Orrit, Phys. Rev. Lett. 1999,
83, 2722 – 2725.
[290] M. Lewinska, C. Seitz, A. Skerra, F. P. Schmidtchen, Bioconjugate Chem. 2004, 15, 231 – 234.
[291] A. Ashkin, J. M. Dziedzic, J. E. Bjorkholm, S. Chu, Opt. Lett.
1986, 11, 288 – 290.
[292] K. Kneipp, A. S. Haka, H. Kneipp, K. Badizadegan, N.
Yoshizawa, C. Boone, K. E. Shafer-Peltier, J. T. Motz, R. R.
Dasari, M. S. Feld, Appl. Spectrosc. 2002, 56, 150 – 154.
[293] M. Gu, D. Morrish, J. Appl. Phys. 2002, 91, 1606 – 1612.
[294] K. Svoboda, S. M. Block, Annu. Rev. Biophys. Biomol. Struct.
1994, 23, 247 – 285.
[295] K. C. Neuman, E. H. Chadd, G. F. Liou, K. Bergman, S. M.
Block, Biophys. J. 1999, 77, 2856 – 2863.
[296] H. Kuhn, J. Chem. Phys. 1970, 53, 101 – 108.
2670
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[297] H. Imahori, H. Norieda, Y. Nishimura, I. Yamazaki, K. Higuchi,
N. Kato, T. Motohiro, H. Yamada, K. Tamaki, M. Arimura, Y.
Sakata, J. Phys. Chem. B 2000, 104, 1253 – 1260.
[298] H. Yokota, K. Saito, T. Yanagida, Phys. Rev. Lett. 1998, 80,
4606 – 4609.
[299] G. V. Soni, F. M. Hameed, T. Roopa, G. V. Shivashankar, Curr.
Sci. 2002, 83, 1464 – 1470.
[300] M. J. Lang, P. M. Fordyce, S. M. Block, J. Biol. 2003, 2:6.
[301] H. Rubinsztein-Dunlop, T. A. Nieminen, M. E. J. Friese, N. R.
Heckenberg, Adv. Quantum Chem. 1998, 30, 469 – 492.
[302] T. A. Laurence, S. Weiss, Science 2003, 299, 667 – 668.
[303] B. Okumus, T. J. Wilson, D. M. J. Lilley, T. Ha, Biophys. J. 2004,
87, 2798 – 2806.
[304] J. Groll, E. V. Amirgoulova, T. Ameringer, C. D. Heyes, C.
Rcker, G. U. Nienhaus, M. Mller, J. Am. Chem. Soc. 2004,
126, 4234 – 4239.
[305] S. C. Blanchard, H. D. Kim, R. L. Gonzalez, Jr., J. D. Puglisi, S.
Chu, Proc. Natl. Acad. Sci. USA 2004, 101, 12 893 – 12 898.
[306] S. C. Blanchard, R. L. Gonzalez, H. D. Kim, S. Chu, J. D.
Puglisi, Nat. Struct. Mol. Biol. 2004, 11, 1008 – 1014.
[307] D. Rueda, G. Bokinsky, M. M. Rhodes, M. J. Rust, X. Zhuang,
N. G. Walter, Proc. Natl. Acad. Sci. USA 2004, 101, 10 066 –
10 071.
[308] S. Hohng, C. Joo, T. Ha, Biophys. J. 2004, 87, 1328 – 1337.
[309] P. Tinnefeld, M. Heilemann, M. Sauer, ChemPhysChem 2005, 6,
217 – 222.
[310] J. P. Clamme, A. A. Deniz, ChemPhysChem 2004, 5, 1 – 4.
[311] P. Tinnefeld, J. Hofkens, D.-P. Herten, S. Masuo, T. Vosch, M.
Cotlet, S. Habuchi, K. Mllen, F. C. De Schryver, M. Sauer,
ChemPhysChem 2004, 5, 1786 – 1790.
[312] S. Masuo, T. Vosch, M. Cotlet, P. Tinnefeld, S. Habuchi,
T. D. M. Bell, I. Oesterling, D. Beljonne, B. Champagne, K.
Mllen, M. Sauer, J. Hofkens, F. C. De Schryver, J. Phys. Chem.
B 2004, 108, 16 686 – 16 696.
[313] J. Hernando, J. P. Hoogenboom, E. M. H. P. van Dijk, J. J.
Garcia-Lopez, M. Crego-Calama, D. N. Reinhoudt, N. F. van Hulst, M. F. Garcia-Parajo, Phys. Rev. Lett. 2004, 93, 236404.
[314] F. Schindler, J. M. Lupton, J. Feldmann, U. Scherf, Proc. Natl.
Acad. Sci. USA 2004, 101, 14 695 – 14 700.
[315] H. Murakoshi, R. Iino, T. Kobayashi, T. Fujiwara, C. Ohshima,
A. Yoshimura, A. Kusumi, Proc. Natl. Acad. Sci. USA 2004,
101, 7317 – 7322.
[316] R. J. Ober, C. Martinez, X. Lai, J. Zhou, E. S. Ward, Proc. Natl.
Acad. Sci. USA 2004, 101, 11 076 – 11 081.
[317] W. Yang, J. Gelles, S. M. Musser, Proc. Natl. Acad. Sci. USA
2004, 101, 12 887 – 12 892.
[318] K. Velonia, O. Flomenbom, D. Loos, S. Masuo, M. Cotlet, Y.
Engelborghs, J. Hofkens, A. E. Rowan, J. Klafter, R. J. M.
Nolte, F. C. de Schryver, Angew. Chem. 2005, 117, 566 – 570;
Angew. Chem. Int. Ed. 2005, 44, 560 – 564.
[319] O. Flomenbom, K. Velonia, D. Loos, S. Masui, M. Cotlet, Y.
Engelborghs, J. Hofkens, A. E. Rowan, R. J. M. Nolte, M.
Van der Auweraer, F. C. de Schryver, J. Klafter, Proc. Natl.
Acad. Sci. USA 2005, 102, ASAP.
[320] G. Harms, G. Orr, H. P. Lu, Appl. Phys. Lett. 2004, 84, 1792 –
1794.
[321] A. Sarkar, R. B. Robertson, J. M. Fernandez, Proc. Natl. Acad.
Sci. USA 2004, 101, 12 882 – 12 886.
[322] A. M. Derfus, W. C. W. Chan, S. N. Bhatia, Nano Lett. 2004, 4,
11 – 18.
[323] C. Kirchner, T. Liedl, S. Kudera, T. Pellegrino, A. M. Javier, H.
E. Gaub, S. Stoelzle, N. Fertig, W. J. Parak, Nano Lett. 2005, 5,
331 – 338.
[324] X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose,
J. J. Li, G. Sundaresan, A. M. Wu, S. S. Gambhir, S. Weiss,
Science 2005, 307, 538 – 544.
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
Angewandte
Single-Molecule Studies
Chemie
[325] B. C. Lagerholm, M. Wang, L. A. Ernst, D. H. Ly, H. Liu, M. P.
Bruchez, A. S. Waggoner, Nano Lett. 2004, 4, 2019 – 2022.
[326] A. M. Derfus, W. C. W. Chan, S. N. Bhatia, Adv. Mater. 2004, 16,
961 – 966.
[327] X. Qu, D. Wu, L. Mets, N. F. Scherer, Proc. Natl. Acad. Sci.
USA 2004, 101, 11 298 – 11 303.
[328] T. Heinlein, A. Biebricher, P. Schlter, C. Roth, D. P. Herten, J.
Wolfrum, M. Heilemann, C. Mller, P. Tinnefeld, M. Sauer;
unpublished results.
[329] E. Y. Chan, N. M. Goncalves, R. A. Haeusler, A. J. Hatch, J. W.
Larson, A. M. Maletta, G. R. Yantz, E. D. Carstea, M. Fuchs,
Angew. Chem. Int. Ed. 2005, 44, 2642 – 2671
[330]
[331]
[332]
[333]
www.angewandte.org
G. G. Wong, S. R. Gullans, R. Gilmanshin, Genome Res. 2004,
14, 1137 – 1146.
L. Kastrup, S. W. Hell, Angew. Chem. 2004, 116, 6814 – 6818;
Angew. Chem. Int. Ed. 2004, 43, 6646 – 6649.
S. W. Hell, Phys. Lett. A 2004, 326, 140 – 145.
S. W. Hell, M. Dyba, S. Jakobs, Curr. Opin. Neurobiol. 2004, 14,
599 – 609.
L. S. Churchman, Z. Okten, R. S. Rock, J. F. Dawson, J. A.
Spudich, Proc. Natl. Acad. Sci. USA 2005, 102, 1419 – 1423.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2671
Документ
Категория
Без категории
Просмотров
1
Размер файла
1 486 Кб
Теги
chemistry, spectroscopy, challenge, fluorescence, branching, molecules, single, biologya, influence
1/--страниц
Пожаловаться на содержимое документа