Download the PDF version

1. Introduction

The application of fluorescence polarization offers unique advantages over conventional
fluorescence imaging and quantitation. In fluorescence polarization studies, fluorophores tied to
biological samples or chemical compounds help elucidate the underlying mechanism of interest.
As opposed to the standard fluorescence techniques, fluorescence polarization allows for fast
and accurate quantitative measurements with relatively simple instrumentation (compared to
sophisticated research instruments). Therefore, applications such as High Throughput
Screening (HTS), a key technique in drug discovery, have made widespread use of
fluorescence polarization. Progressively more complicated studies at the molecular level are
being conducted by the application of microscopy-based fluorescence polarization studies.
Excellent technical information on the theory and applications of fluorescence polarization is
available in the literature [1-6]. However many of these resources provide a limited perspective
on applications. This article is a concise review of various applications in life sciences that are
based on fluorescence polarization.
2. The Fundamentals

Polarization is a fundamental property of light [7]. Light is considered to be linearly polarized
when the orientation of its electric field vector does not change during propagation.
Fluorescence polarization studies utilize linearly polarized light for the illumination of the sample.

The interaction of polarized light with a fluorophore can be best described by considering the
concept of an electric dipole. A dipole refers to the formation of separated positive and negative
charges and forms the basis of a simple classical model of the interaction of light with matter.
The dipole moment (a vector) denotes the strength of the dipole and its direction. Fluorophores
tend to have a dipole moment, and the excitation dipole moment of a fluorophore can be
different from its emission dipole moment. The dipole moment of a fluorophore is also
influenced by the environmental factors. When the electric field of the excitation light is parallel
to a fluorophore’s absorption dipole moment the fluorophore has the greachance of
absorbing a photon, which leads to its preferential excitation among the population of
illuminated fluorophores. On the other hand, when the absorption dipole moment is
perpendicular to the electric field of excitation then the fluorophore cannot be excited.
Furthermore, when the excited fluorophore reemits light, the emission is polarized parallel to the
plane of its dipole. These phenomena are the basis for fluorescence polarization studies. They
occur in all fluorescence applications at the molecular level, but are generally not apparent in
conventional fluorescence studies. Due to the random orientations of individual fluorophores in
a sample that is typically excited by unpolarized light, the polarization effect is lost due to
averaging.
3. Principle of Application

Figure 1 illustrates the basic principle employed in conventional fluorescence polarization
studies. A fluorescently labeled sample is illuminated with linearly polarized light at the
wavelength of the fluorophore absorption and the longer wavelength emission from the sample
passes through a rotatable linearly polarizing filter (analyzer) before detection. Since the degree
of polarization cannot be detected directly by conventional detectors such as photomultiplier
tubes (PMTs), an observation of the change in the fluorescence intensity is used to indirectly
study polarization.

Figure 1: Fluorescence polarization measurement. A monochromatic beam illuminates the
fluorophores within the sample. A linear polarizer located in the illumination beam attenuates the
randomly oriented polarization states except for those in one plane, thereby generating linearly
polarized light. This plane of polarized light also becomes the reference plane. Depending upon the
orientation of their absorption dipoles, individual fluorophore molecules are preferentially excited
(see Section 2). The emission from the fluorescent sample can be considered as another light
source, composed of the combination of signals from the emission dipoles of individual fluorophore
molecules. If the absorption and emission dipole moments of all the fluorophores were to be aligned
with the electric field vector of the illumination beam (plane polarized light) then the emission signal
would also be plane polarized. However, this is typically not the case. Therefore, the emission signal
is partially depolarized due to the random orientation of the fluorophores, even if the fluorophores
are completely static. As the molecules move during the time window of detection, the emission
becomes even more depolarized. A polarizer placed in front of the detector is used to detect the
intensity of emission light in a given plane of polarization. The intensity measurement made with the
emission polarizer oriented orthogonal to the electric field of illumination beam (the reference plane)
is denoted as 
I and when the emission polarizer is rotated by 90 degrees with respect to the I|| measurement is made.

A ratio derived using two different fluorescence intensity measurements (see Fig. 1)
provides a measure of the degree of polarization of the emission. Two expressions that are
used interchangeably to study polarization are “polarization ratio” ( p ) and “emission
anisotropy” ( r ). These dimensionless quantities are calculated as follows [1]:

where I||  and I⊥ are the intensity measurements of the emission signal made parallel to or
orthogonal to, respectively, the direction of the electric field of illumination light (determined by
the polarizer in the excitation light path). Note that I||  and I measurements can also be made by
rotating the polarizer in the excitation light path instead of rotating the emission polarizer. If all of
the emission dipoles were to reradiate light parallel to the electric field of the illumination light
each absorbed, p would equal 1 (and r would also equal 1), whereas if all the emission
dipoles were to reradiate light perpendicular to the illumination light electric field then p would
be -1 (and r would be -0.5). Therefore in a properly calibrated experiment, the values of p and
r can vary within these limits. In particular, and r do not reach their limiting values
because the dipole moments of the fluorophore molecules are not completely aligned in any one
direction.

The relationship between and r is given by r = 2 p /(3 − p) or p = 3r /(2 + r) .
Regarding the use of  versus r as the measured value of polarization, many biological
applications use emission anisotropy ( r ) because it offers advantages over polarization ratio
p ) in the physical interpretation of measured data. However, historically, polarization ratio ( )
has been used in the clinical chemistry and drug discovery. One of the advantages of
polarization studies is that since the measurements are ratiometric in nature they are not
dependent on the actual light intensity values or the fluorophore concentration.

What is the significance of relative strength of I||  and I? The majority of FP studies are based
on the premise that rotational diffusion of fluorophores is the dominant cause of fluorescence
depolarization (a change in the degree of polarization). Depolarization can be caused by several
factors such as the optical system itself used for polarization measurement, or fluorescence
resonance energy transfer (FRET), or a change in the orientation of the dipole between
excitation and emission events. Note that (partial) depolarization is intrinsically present in all
fluorescence polarization applications due to the random orientation of the population of
fluorophores that behave as spatially incoherent light sources (see Fig.1). Therefore, in real life
experiments I||  and I are only relative measurements that refer to the maximum or minimum
signals.

Fluorescence polarization was first described by Perrin [8] and is elucidated by the following
mathematical equation called the Perrin equation:

In this equation τ is the fluorescence lifetime (the average time a fluorophore stays in the
excited state before emitting a photon), θ is a rotational correlation time, r is the measured
anisotropy and r0 is the fundamental anisotropy (in the absence of any rotational diffusion).
Also note that the rotational correlation time of the fluorophore (θ) can be described as [2]:

where η is the viscosity of the sample, R is the gas constant, T is absolute temperature, and V
is the volume of the rotating unit (i.e. the fluorescently labeled molecule). This relation implies
that molecular size affects the polarization of fluorescence emission. Therefore slowly tumbling
molecules in solution generate a higher r (or p ) value compared to molecules that are
tumbling faster. However when θ <<τ, the orientations of individual fluorophores are rerandomized
by the time measurements are made and the net emission is unpolarized ( r →0 ).
And when θ >>τ then immediately following excitation the measured anisotropy approaches
fundamental anisotropy ( r → r0).
4. Applications of Fluorescence Polarization

The measurement of fluorescence polarization (or anisotropy) adds another dimension of
observation to a fluorescence-based experiment. This added dimension can provide information
on the local environment, fluorescence lifetime and molecular mass. A variety of instruments are
utilized in fluorescence polarization studies. These instruments are based on the design of
existing fluorescence spectroscopy or microscopy techniques. Depending upon the
requirements of an application a suitable instrument and a compatible assay may be selected.

Spectroscopy based fluorescence polarization studies

Traditionally spectroscopic instruments such as a fluorometer (see Fig. 2) have been used
for fluorescence polarization studies. In these instruments the sample is placed in a cuvette or a
tube. As shown in this figure, the excitation and emission wavelengths of light are selected
using monochromators. The signal intensity from the fluorophores is recorded on a nonpixelated
detector such as a PMT. This allows for the measurement of bulk properties of a
population of fluorophores.

Figure 2: Spectrofluorometer for polarization measurement. One of the polarizers (typically the polarizer
in the emission light path) is rotated along parallel or orthogonal directions to obtain I||  and I
measurements. This is an L-format (or single channel) spectrofluorometer. In a T-format (or two channel)
spectrofluorometer (see the principle in the emission light path of Figs. 3 & 4B) the emission signal from
the sample is split into two channels, each of which is used to measure different polarization states, I||
and I simultaneously.

This measurement approach has been in use for several decades for applications targeted
towards biochemical research or clinical diagnostics. For example the measurement of
fundamental anisotropy ( r0 ) and an understanding of the electronic properties of fluorophores
can be obtained by anisotropy measurements at different wavelengths. Perrin plots (based upon
the Perrin equations, above) of labeled macromolecules have been extensively used to
determine the apparent hydrodynamic volumes [1]. Historically biological membranes have also
been studied using fluorescence anisotropy measurements. Other biological applications
include the study of association of proteins with larger molecules.

Fluorescence polarization studies for HTS

One of the bottlenecks in the application of the above measurement configuration in life
sciences has been the use of a cuvette or a tube for sample preparation. For example, the
sample preparation conventions of HTS require analysis in microplates. HTS plays a key role in
the small-molecule drug discovery process. In this technique, a large number of synthetic
compounds are tested against biological targets (such as a gene or a protein affecting a
disease, for example) in in vitro assays. The objective of this screening is to identify a lead, or a
promising molecule that can interact with the target and hence has potential therapeutic
application. Since it might be necessary to screen millions of compounds against the target, the
time required for each assay is crucial. In this regard an advantage of fluorescence polarization
assays is that they are homogeneous; i.e., these assays do not require separation of bound and
free species, thereby making it easier to automate the experiments so that results can be
obtained almost instantaneously and hence enhance the overall throughput. This reason has
led to significant advances in the application of fluorescence polarization assays in HTS.

Figure 3: A modern HTS instrument essentially employs the principle of operation of a
conventional fluorescence microscope. Typically excitation and emission filters are used
to select monochromatic wavelengths of light in the excitation and emission channel.
The emission signal is split into two channels each of which is utilized to simultaneously
measure I||  and I components using a non-pixelated detector such as a PMT. The
sample is placed in microwells that have 96, 384 or 1536 format. This allows for screening
of a large number of compounds in a given experiment.

Figure 3 shows the principle of operation of a modern HTS instrument that utilizes optical
filters for distinguishing excitation light from emission signal. Plane polarized excitation light is
used to illuminate a fluorescently labeled sample and the emission from the sample is split into
two channels, each of which is used to measure different polarization states, (i.e. ||  and I
intensities). Simultaneous acquisition of two orthogonal polarization components has the
advantage of minimizing artifacts resulting from photobleaching and movement of the sample,
and is also faster than two sequential measurements (see Figs.1 & 2). Depending upon the
flexibility of requirements, monochromators may be used in the illumination and emission paths.
Monochromators provide greater flexibility in working with different wavelengths and therefore
with different colored fluorophores. However the use of optical filters has the advantage of
providing higher blocking of excitation light into the emission channel and therefore higher
signal-to-noise ratio leading to enhanced throughput.

The first generation HTS instruments launched in mid 90’s enabled significant throughput
advancement in the drug development process [2-3]. In order to further optimize the cost of
screening, the assays needed to be miniaturized, leading to the development of second
generation HTS instruments that provide much higher throughput and can handle a larger
number of wells (384 or even 1536, compared to the original standard of 96 wells) in an
automated fashion.

HTS fluorescence polarization assays are generally performed under equilibrium conditions.
For example, the affinity of a protein-DNA interaction can be determined by monitoring the
steady-state (equilibrium condition) fluorescence anisotropy changes associated with the
formation of the complex. An example of a direct binding assay that is commonly employed in
HTS might look like [2, 6]:

   F-Ligand + Receptor ⇔ F-Ligand:Receptor

where F refers to the fluorophore that is covalently bound to the Ligand forming a F-Ligand
complex. This complex and the Receptor have a non-covalent interaction. Owing to a significant
difference in their sizes, under equilibrium condition, the emission signal from the F-Ligand
complex has much lower polarization value compared to that of the F-Ligand:Receptor.

However, the most popular HTS assay utilizing fluorescence polarization is the competition
binding assay:

F-Ligand + Receptor ⇔ F-Ligand:Receptor
I + Receptor ⇔ I:Receptor

In this assay, I is an inhibitor (for example, another ligand which is not fluorescently labeled) that
competes with the F-Ligand complex in binding with the Receptor. In these experiments the
affinity of binding of F-Ligand with that of the Receptor is known and therefore this technique
enables determination of the binding affinity of a ligand without fluorescently labeling it, thus
simplifying the sample preparation step. Some of the classes of targets that utilize this
technique are protein-protein interaction, GPCR (G-protein coupled receptors that are involved
in signaling pathways across cellular membranes), nuclear receptors, protein kinase, and
phosphatase.

Other types of assay modes for HTS include enzyme assays (direct or competitive
immunodetection, transferase) and protease assays (based on size reduction or detection by
protein binding). The applications of these assays include receptor and ligand binding studies,
protein-peptide interactions and DNA-protein interactions.

Fluorescence polarization experiments are not limited to only equilibrium binding studies
but are also applicable to “real-time” experiments. Such experiments can measure fluorescence
polarization as a function of time and therefore help understand the kinetics of the assay.

Microscopy based fluorescence polarization studies

Depending upon the requirements, fluorescence polarization assays can be performed in
steady state or by utilizing time-resolved measurements (such as fluorescence lifetime imaging
microscopy or FLIM). Such measurements can be done in the time or the frequency domain
utilizing widefield or scanning methods (see Fig. 4). Since microscopy techniques utilize fairly
sophisticated optics and instrumentation they allow for high sensitivity as well as high resolution
imaging of a fluorescently labeled sample. For example, several microscopy techniques utilize
high numerical aperture (NA) objectives which work with a high refractive index medium
(typically oil) between the coverglass and the objective lens. While high NA objectives allow for
better resolution, they can significantly alter the polarization state of the light and therefore
require careful calibration. Additionally, the high cost of such instruments prohibits their
widespread application, in contrast to HTS, and therefore they are primarily used for research
applications.

Figure 4: Microscopy based fluorescence polarization. (A) Schematic of a widefield imaging
system. (B) Scanning instrument based fluorescence imaging. Depending upon the application,
a pixelated or a nonpixelated detector such as an avalanche photodiode (APD) can be used to collect
the emission signal. Additionally advanced instrumentation may be utilized for the measurement
of fluorescence decay and lifetime imaging using the time-domain or the frequency-domain measurements.

We have already noted that fluorescence polarization studies uniquely allow the
determination of the chemical and physical environments of macromolecules in intracellular
environments. Even within a given fluorescence experiment, polarization measurement provides
complimentary information to that of the original fluorescence experiment.

For example, Fluorescence Resonance Energy Transfer (FRET) is used to study
molecular interactions at very small distances (generally less than 10 nm). Typically FRET
refers to hetero-FRET, i.e., FRET occurring between two (or more) different types of
fluorophores consisting of a donor-acceptor pair. Several approaches are available to study
hetero-FRET that measure a change in the property (such as, lifetime or intensity) of the donor
and acceptor. However, another type of FRET that occurs between the same species of a
fluorophore, termed homo-FRET, can only be studied using fluorescence polarization
measurements. Neither observed fluorescent intensity nor the lifetime of the fluorophore
changes in the presence of homo-FRET. Fluorescence polarization allows identification of
homo-FRET because fluorophores that are directly excited by polarized illumination light exhibit
highly polarized emission signals compared to the sub-population of fluorophores that were
excited by homo-FRET. Therefore in the absence of other depolarization factors, homo-FRET
leads to a decrease in fluorescence anisotropy. An advantage in the study of homo-FRET is that
a large spectral window is available (since only one type of fluorophore is used) therefore
benefitting multiplexing assays. The study of homo-FRET using fluorescence polarization has
been utilized for several investigations in biological environments. For example, homo-FRET
studies have enabled determination of the oligomerization states of membrane proteins and the
study of heterogeneity in lipid-order in the plasma membrane of cells. Such studies elucidate the
complex associations of states of molecules and how they relate to biological functions [4].

Even hetero-FRET measurements using fluorescence polarization offer potentially higher
dynamic range and temporal resolution [9-10] compared to conventional FRET measurement
techniques. Another example is time-domain FLIM [4] combined with polarization measurement
which is an important tool for the investigation of molecular rotation, binding reactions, and
protein-protein interactions in living cells. FLIM based polarization studies remain one of the few
methods that can differentiate among monomers, dimers, trimers and higher order assemblies
in live cells [5]. However due to the complexity of implementation and analysis, time-resolved
fluorescence anisotropy applications based on FLIM are still not widely prevalent.

Polarized Total Internal Reflection Fluorescence (polTIRF) microscopy (Fig. 4A) is
another variant of an existing fluorescence microscopy technique (i.e. TIRF microscopy) that
has been used to study the motility properties of motor proteins such as myosin along actin
filaments. This technique even allows for the study of the orientation of individual molecules.
With the illumination of the sample at various polarization angles it has become possible to
understand the tilting and rotational motions of individual molecules [11].
5. Conclusion

Given the unique advantages offered by fluorescence polarization, it has found widespread
application in disciplines such as in HTS. However, its full potential still remains untapped for
research applications.
References

[1] Lakowicz, J.R., Principles of Fluorescence Spectroscopy, Springer, New York, USA, 1999.
[2] Owicki, J.C., Fluorescence Polarization and Anisotropy in High Throughput Screening: Perspectives
and Primer, Journal of Biomolecular Screening, 5 (5), 297-306, 2000.
[3] Nasir, M.S., Jolley, M.E., Fluorescence Polarization: An Analytical Tool for Immunoassay and Drug
Discovery, Combinatorial Chemistry & High Throughput Screening, 1999, 2, 177-190, 1999.
[4] Levitt, J.A., Matthews, D.R, Ameer-Beg, S.M. and Suhling, K., Fluorescence Lifetime and Polarization-
Resolved Imaging in Cell Biology, Current Opinion in Biotechnology, 20:28-36, 2009.
[5] Vogel, S.S., Thaler, C, Blank, P.S., Koushik, S., Time Resolved Fluorescence Anisotropy, Chapter 10.
FLIM Microscopy in Biology and Medicine, CRC Press Taylor & Francis Group, USA, 2009.
[6] Fluorescence Polarization Technical Resource Guide (4th Edition), www. Invitrogen.com
[7] Erdogan, T., Understanding Polarization, Semrock White Paper Series, www.semrock.com
[8] Perrin, M.F. Polarization de la luniere de fluorescence. Vie moyenne de molecules dans l’etat excite”,
J. Phys. Radium, 7, 390-401, 1926.
[9] Mattheyses A.L., Hoppe A.D., Axelrod D., Polarized fluorescence resonance energy transfer
microscopy, Biophysical Journal, 87(4):2787-97, 2004.
[10] Rizzo M.A., Piston D.W., High-contrast Imaging of Fluorescent Protein FRET by Fluorescence
Polarization Microscopy, Biophysical Journal, 88(2):L14-6, 2005.
[11] Rosenberg, S.A., Quinlan, M.E., Forkey, J.N., and Goldman, Y.E., Rotational Motions of Macromolecules
by Single-Molecule Fluorescence Microscopy, Accounts of Biochemical Research, 38(7), 583-
593, 2005.

Authors

Prashant Prabhat, Ph.D. and Turan Erdogan, Ph.D., Semrock, Inc., A Unit of IDEX Corporation.
E-mail: pprabhat@idexcorp.com; Tel: (585) 594-7064; Fax: (585) 594-7095