olympus microscopy resource center _ fluorescence resonance energy transfer (fret) microscopy -...

8/10/2019 Olympus Microscopy Resource Center _ Fluorescence Resonance Energy Transfer (FRET) Microscopy - Introductory

1/12

9/30/2014 Olympus Microscopy Resource Center | Fluorescence Resonance Energy Transfer (FRET) Microscopy - Introductory Concepts

http://www.olympusmicro.com/primer/techniques/fluorescence/fret/fretintro.html 1

Olympus America| Research| Imaging Software| Confocal| Clinical| FAQs

Home Page

Interactive Tutorials

Microscopy Primer

Physics of Light & Color

Microscopy Basic Concepts

Special Techniques

Fluorescence Microscopy

Confocal Microscopy

Confocal Applications

Digital Imaging

Digital Image Galleries

Digital Video Galleries

Virtual Microscopy

Microscopy Web Resources

Olympus Product Brochures

MIC-D Microscope Archive

Photomicrography Archive

Software Downloads

Fluorescence Resonance Energy Transfer (FRET) Microscopy

Introductory Concepts

The precise location and nature of the interactions between specific molecular species in living cells is ofmajor interest in many areas of biological research, but investigations are often hampered by the limitedresolution of the instruments employed to examine these phenomena. Conventional widefield fluorescencemicroscopy enables localization of fluorescently labeled molecules within the optical spatial resolution limitsdefined by the Rayleigh criterion, approximately 200 nanometers (0.2 micrometer). However, in order tounderstand the physical interactions between protein partners involved in a typical biomolecular process,the relative proximity of the molecules must be determined more precisely than diffraction-limited traditionaloptical imaging methods permit. The technique of fluorescence resonance ene rgy transfer(morecommonly referred to by the acronym FRET), when applied to optical microscopy, permits determination ofthe approach between two molecules within several nanometers (see Figure 1), a distance sufficientlyclose for molecular interactions to occur.

Typical fluorescence microscopy techniques rely upon the absorption by a fluorophore of light at onewavelength (excitation), followed by the subsequent emission of secondary fluorescence at a longerwavelength. The excitation and emission wavelengths are often separated from each other by tens to

hundreds of nanometers. Labeling of cellular components, such as the nuclei, mitochondria, cytoskeleton,Golgi apparatus, and membranes, with specific fluorophores enables their localization within fixed andliving preparations. By simultaneously labeling several sub-cellular structures with individual fluorophoreshaving separated excitation and emission spectra, specialized fluorescence filter combinations can beemployed to examine the proximity of labeled molecules within a single cell or tissue section. With thistechnique, molecules that are closer together than the optical resolution limit appear to be coincident, andthis apparent spatial proximity implies that a molecular association is possible. In most cases, however, thenormal diffraction-limited fluorescence microscope resolution is insufficient to determine whether aninteraction between biomolecules actually takes place. Fluorescence resonance energy transfer is aprocess by which radiationless transfer of energy occurs from an excited state fluorophore to a secondchromophore in close proximity. Because the range over which the energy transfer can take place islimited to approximately 10 nanometers (100 angstroms), and the efficiency of transfer is extremelysensitive to the separation distance between fluorophores, resonance energy transfer measurements canbe a valuable tool for probing molecular interactions.

The mechanism of fluorescence resonance energy transfer involves a donorfluorophore in an excitedelectronic state, which may transfer its excitation energy to a nearby acceptorchromophore in a non-radiative fashion through long-range dipole-dipole interactions. The theory supporting energy transfer isbased on the concept of treating an excited fluorophore as an oscillating dipole that can undergo anenergy exchange with a second dipole having a similar resonance frequency. In this regard, resonanceenergy transfer is analogous to the behavior of coupled oscillators, such as a pair of tuning forks vibratingat the same frequency. In contrast, radiative energy transfer requires emission and reabsorption of aphoton and depends on the physical dimensions and optical properties of the specimen, as well as thegeometry of the container and the wavefront pathways. Unlike radiative mechanisms, resonance energytransfer can yield a significant amount of structural information concerning the donor-acceptor pair.

Resonance energy transfer is not sensitive to the surrounding solvent shell of a fluorophore, and thus,produces molecular information unique to that revealed by solvent-dependent events, such asfluorescence quenching, excited-state reactions, solvent relaxation, or anisotropic measurements. Themajor solvent impact on fluorophores involved in resonance energy transfer is the effect on spectralproperties of the donor and acceptor. Non-radiative energy transfer occurs over much longer distancesthan short-range solvent effects, and the dielectric nature of constituents (solvent and hostmacromolecule) positioned between the involved fluorophores has very little influence on the efficacy of
http://www.olympusmicro.com/primer/photomicrography/index.htmlhttp://www.olympusmicro.com/brochures/index.htmlhttp://www.olympusamerica.com/seg_section/product.asp?product=1009http://www.olympusamerica.com/seg_section/product.asp?product=1009http://www.olympusmicro.com/moviegallery/index.htmlhttp://www.olympusmicro.com/galleries/index.htmlhttp://www.olympusmicro.com/primer/techniques/confocal/confocalapplications.htmlhttp://www.olympusmicro.com/primer/techniques/fluorescence/fluorhome.htmlhttp://www.olympusamerica.com/seg_section/product.asp?product=1009http://www.olympusmicro.com/software/index.htmlhttp://www.olympusmicro.com/primer/photomicrography/index.htmlhttp://www.olympusmicro.com/micd/index.htmlhttp://www.olympusmicro.com/brochures/index.htmlhttp://www.olympusmicro.com/primer/webresources.htmlhttp://www.olympusmicro.com/primer/virtual/virtual.htmlhttp://www.olympusmicro.com/moviegallery/index.htmlhttp://www.olympusmicro.com/galleries/index.htmlhttp://www.olympusmicro.com/primer/digitalimaging/index.htmlhttp://www.olympusmicro.com/primer/techniques/confocal/confocalapplications.htmlhttp://www.olympusmicro.com/primer/techniques/confocal/index.htmlhttp://www.olympusmicro.com/primer/techniques/fluorescence/fluorhome.htmlhttp://www.olympusmicro.com/primer/techniques/index.htmlhttp://www.olympusmicro.com/primer/anatomy/anatomy.htmlhttp://www.olympusmicro.com/primer/lightandcolor/index.htmlhttp://www.olympusmicro.com/primer/index.htmlhttp://www.olympusmicro.com/primer/java/index.htmlhttp://www.olympusmicro.com/index.html


2/12



resonance energy transfer, which depends primarily on the distance between the donor and acceptorfluorophore.

The phenomenon of fluorescence resonance energy transfer is not mediated by photon emission, andfurthermore, does not even require the acceptor chromophore to be fluorescent. In most applications,however, both donor and acceptor are fluorescent, and the occurrence of energy transfer manifests itselfthrough quenching of donor fluorescence and a reduction of the fluorescence lifetime, accompanied alsoby an increase in acceptor fluorescence emission. The efficiency of the energy transfer process varies inproportion to the inverse sixth power of the distance separating the donor and acceptor molecules.Consequently, FRET measurements can be utilized as an effective molecular rulerfor determiningdistances between biomolecules labeled with an appropriate donor and acceptor fluorochrome when theyare within 10 nanometers of each other.

A hypothetical example of fluorescence resonance energy transfer between two fluorochromes attached toopposite ends of the same macromolecular protein is presented in Figure 1. In the native conformation(Figure 1(a)), the two fluorophores are separated by a distance of approximately 12 nanometers, too farfor intramolecular resonance energy transfer between the fluorochromes to occur. However, when theprotein is induced to undergo a conformational change (Figure 1(b)), the two fluorochromes are broughtmuch closer together and can now participate in FRET molecular interactions. In the figure, excitation ofthe donor fluorochrome is indicated by a blue glow around the yellow tri-nuclear aromatic molecule, whilethe corresponding acceptor emission (Figure 1(b)) is represented by a green glow surrounding thesecond heterocyclic fluorochrome on the right-hand side of the protein. Energy transfer measurementsare often employed to estimate the distances between sites on a macromolecule and the effects ofconformational changes on these distances. In this type of experiment, the degree of energy transfer isused to calculate the distance between the donor and acceptor and obtain structural information about themacromolecule.

Although fluorescence resonance energy transfer has often been employed to investigate intermolecularand intramolecular structural and functional modifications in proteins and lipids, a major obstacle toimplementation of FRET microscopy techniques in living cells has been the lack of suitable methods forlabeling specific intracellular proteins with appropriate fluorophores. Cloning of the jellyfish greenfluorescent protein(GFP) and its expression in a wide variety of cell types has become a critical key todeveloping markers for both gene expression and structural protein localization in living cells. Severalspectrally distinct mutation variants of the protein have been developed, including a fluorescent proteinthat emits blue light (blue fluorescent protein, BFP). Both the excitation and emission spectra for thenative GFP and BFP mutants are sufficiently separated in wavelength to be compatible with the FRETapproach. Figure 2 illustrates the strategy for detection of protein-protein interactions using fluorescenceresonance energy transfer and mutant fluorescent proteins. If two proteins, one labeled with BFP (thedonor) and the other with GFP (the acceptor), physically interact, then increased intensity at the acceptoremission maximum (510 nanometers) will be observed when the complex is excited at the maximumabsorbance wavelength (380 nanometers) of the donor. Failure of the proteins to form a complex resultsin no acceptor (GFP) fluorescence emission.

Coupled with advances in pulsed lasers, microscope optics, and computer-based imaging technology, thedevelopment of labeling techniques in which the donor and acceptor fluorophores are actually part of thebiomolecules themselves has enabled the visualization of dynamic protein interactions within living cells. Inaddition to the investigation of protein partner interactions, recent applications of fluorescence resonanceenergy transfer include studies of protease activity, alterations in membrane voltage potentials, calciummetabolism, and the conduction of h igh-throughput screening assays, such as for quantification of geneexpression in single living cells.

Principles of Fluorescence Resonance Energy Transfer

The process of resonance energy transfer (RET) can take place when a donor fluorophore in anelectronically excited state transfers its excitation energy to a nearby chromophore, the acceptor. Inprinciple, if the fluorescence emission spectrum of the donor molecule overlaps the absorption spectrum ofthe acceptor molecule, and the two are within a minimal spatial radius, the donor can directly transfer itsexcitation energy to the acceptor through long-range dipole-dipole intermolecular coupling. A theoryproposed by Theodor Frster in the late 1940s initially described the molecular interactions involved inresonance energy transfer, and Frster also developed a formal equation defining the relationship


3/12



between the transfer rate, interchromophore distance, and spectral properties of the involvedchromophores.

Resonance energy transfer is a non-radiative quantum mechanical process that does not require acollision and does not involve production of heat. When energy transfer occurs, the acceptor moleculequenches the donor molecule fluorescence, and if the acceptor is itself a fluorochrome, increased orsensitizedfluorescence emission is observed (see Figure 3). The phenomenon can be observed byexciting a specimen containing both donor and acceptor molecules with light of wavelengths correspondingto the absorption maximum of the donor fluorophore, and detecting light emitted at wavelengths centerednear the emission maximum of the acceptor. An alternative detection method, growing rapidly in popularity,is to measure the fluorescence lifetime of the donor fluorophore in the presence and absence of theacceptor.

Presented in Figure 3 is a Jablonski diagram illustrating the coupled transitions involved between thedonor emission and acceptor absorbance in fluorescence resonance energy transfer. Absorption andemission transitions are represented by straight vertical arrows (green and red, respectively), whilevibrational relaxation is indicated by wavy yellow arrows. The coupled transitions are drawn with dashedlines that suggest their cor rect placement in the Jablonski diagram should they have arisen from photon-mediated electronic transitions. In the presence of a suitable acceptor, the donor fluorophore can transferexcited state energy directly to the acceptor without emitting a photon (illustrated by a blue arrow in Figure3). The resulting sensitized fluorescence emission has characteristics similar to the emission spectrum ofthe acceptor.

Several criteria must be satisfied in order for resonance energy transfer to occur. In addition to theoverlapping emission and absorption spectra of the donor and acceptor molecules, the two involved

fluorophores must be positioned within a range of 1 to 10 nanometers of each other. As described inequations derived by Frster (and discussed below), the energy transfer efficiency between donor andacceptor molecules decreases as the sixth power of the distance separating the two. Consequently, theability of the donor fluorophore to transfer its excitation energy to the acceptor by non-radiative interactiondecreases sharply with increasing distance between the molecules, limiting the FRET phenomenon to amaximum donor-acceptor separation radius of approximately 10 nanometers. At distances less than 1nanometer, several other modes of energy and/or electron transfer are possible. The distancedependence of the resonance energy transfer process is the primary basis for its utility in investigation ofmolecular interactions. In living cell studies involving molecules labeled with donor and acceptorfluorophores, resonance energy transfer will occur only between molecules that are close enough tointeract biologically with one another.

An additional requirement for resonance energy transfer is that the fluorescence lifetime of the donormolecule must be of sufficient duration to permit the event to occur. Both the rate ( K(T)) and the efficiency(E(T)) of energy transfer are directly related to the lifetime of the donor fluorophore in the presence andabsence of the acceptor. According to Frster's theory, and verified experimentally, the rate of energy

transfer is given by the equation:

KT= (1/D) [R0/r]6

where R(0)is the Frster critical distance , (D)is the donor lifetime in the absence of the acceptor, andris the distance separating the donor and acceptor chromophores. The Frster critical distance ( R(0)) isdefined as the acceptor-donor separation radius for which the transfer rate equals the rate of donor decay(de-excitation) in the absence of acceptor. In other words, when the donor and acceptor radius ( r) equalsthe Frster distance, then the transfer efficiency is 50 percent. At this separation radius, half of the donorexcitation energy is transferred to the acceptor via resonance energy transfer, while the other half isdissipated through a combination of all the other available processes, including fluorescence emission.

Conceptually, the Frster critical distance is the maximal separation length between donor and acceptormolecules under which resonance energy transfer will still occur. The critical distance value typically fallswithin a range of 2 to 6 nanometers, which is fortuitously on the order of many protein molecular


4/12



dimensions. In addition, the critical distance range also corresponds to several other biologically significantdimensions, such as cell membrane thickness and the distance separating sites on proteins havingmultiple subunits. The value of R(0)(in nanometers) may be calculated from the following expression:

R0= 2.11 10-2 [2 J() -4 QD]

1/6

in which -squaredis a factor describing the relative orientation in space between the transition dipoles ofthe donor and acceptor, J()is the overlap integral in the region of the donor emission and acceptorabsorbance spectra (with the wavelength expressed in nanometers), represents the refractive index ofthe medium, and Q(D)is the quantum yield of the donor.

The efficiency of energy transfer, E(T), is a measure of the fraction of photons absorbed by the donor that

are transferred to the acceptor, and is related to the donor-acceptor separation distance, r, by theequation:

r = R0 [(1/ET) - 1]1/6

and E(T)is evaluated as:

ET= 1 - (DA/D)

where (DA)is the donor lifetime in the presence of the acceptor and (D)is the donor lifetime in theabsence of the acceptor. Therefore, by measuring the donor fluorescence lifetime in the presence andabsence of an acceptor (which is indicative of the extent of donor quenching due to the acceptor), it ispossible to determine the distance separating donor and acceptor molecules. In many commonly appliedtechniques, the energy transfer efficiency is determined by steady state measurements of the relativeaverage donor fluorescence intensities in the presence and absence of the acceptor (not by measuringthe lifetimes).

In summary, the rate of energy transfer depends upon the extent of spectral overlap between the donoremission and acceptor absorption spectra (see Figure 4), the quantum yield of the donor, the relativeorientation of the donor and acceptor transition dipole moments, and the distance separating the donorand acceptor molecules. Any event or process that affects the distance between the donor and acceptorwill affect the resonance energy transfer rate, consequently allowing the phenomenon to be quantified,provided that artifacts can be controlled or eliminated.

Presented in Figure 4 are the absorption and emission spectra of cyan fluorescent protein (CFP, thedonor) and red fluorescent protein (RFPor DsRed, the acceptor) when compared for their potentialapplication as a fluorescence resonance energy transfer pair. Absorption spectra for both biologicalpeptides are illustrated as red curves, while the emission spectra are presented as blue curves. Theregion of overlap between the donor emission and acceptor absorption spectra is represented by a gray

area near the base of the curves. Whenever the spectral overlap of the molecules is increased too far, aphenomenon known as spectral bleed-throughor crossoveroccurs in which signal from the excitedacceptor (arising from excitation illumination of the donor) and the donor emission are detected in theacceptor emission channel. The result is a high background signal that must be extracted from the weakacceptor fluorescence emission.

The basic theory of non-radiative energy transfer is directly applicable to a donor-acceptor pair separatedby a fixed distance, in which case the rate of energy transfer is a function of the Frster distance, R(0),which in turn depends upon -squared, J(), , and Q(D). If these factors are known, the distancebetween the donor and acceptor can be calculated. More complex formulations are required to describesituations such as multiple acceptor chromophores and distance distributions. Presented in Table 1 are aseries of experimentally measured Frster cr itical distances, which were ascertained from the spectraloverlap of several popular donor-acceptor fluorophore pairs. Because the variable includes the donorquantium yield and the degree of spectral overlap, both of which depend on the localized environmentalconditions, Frster distance values should be determined under identical experimental conditions as thoseemployed to investigate resonance energy transfer.


5/12



The refractive index of the energy transfer medium is generally known from the solvent composition, orcan be estimated for a particular macromolecule, and is commonly taken to be 1.4 in aqueous solution.The quantum yield of the donor is determined by comparison to standard fluorophores with knownquantum yield. Because Q(D)appears as the sixth-root in the calculation of R(0), small errors oruncertainties in the value of Q(D)do not have a large effect on the Frster distance calculation. Also dueto the sixth-root dependence, R(0)is not largely affected by the variations in J(), but the overlap integralmust still be evaluated for each donor-acceptor pair. In general, higher degrees of overlap between thedonor emission spectrum and the acceptor absorption spectrum yield larger Frster critical distancevalues.

Frster Critical Distance forCommon RET Donor-Acceptor Pairs

Donor Acceptor Frster Distance(Nanometers)

Tryptophan Dansyl 2.1

IAEDANS (1) DDPM (2) 2.5 - 2.9

BFP DsRFP 3.1 - 3.3

Dansyl FITC 3.3 - 4.1

Dansyl Octadecylrhodamine 4.3

CFP GFP 4.7 - 4.9

CF (3) Texas Red 5.1

Fluorescein Tetramethylrhodamine 4.9 - 5.5

Cy3 Cy5 >5.0

GFP YFP 5.5 - 5.7

BODIPY FL (4) BODIPY FL (4) 5.7

Rhodamine 6G Malachite Green 6.1

FITC Eosin Thiosemicarbazide 6.1 - 6.4

B-Phycoerythrin Cy5 7.2

Cy5 Cy5.5 >8.0

(1) 5-(2-iodoacetylaminoethyl)aminonaphthalene-1-sulfonic acid

(2) N-(4-dimethylamino-3,5-dinitrophenyl)maleim ide

(3) carboxyfluorescein succinimidyl ester

(4) 4,4-difluoro-4-bora-3a,4a-diaza-s-indacene

Table 1

The uncertainty in evaluating the orientation factor (

-squared) has been discussed extensively in theliterature, and in spite of experimental evidence that the Frster theory is valid and applicable to distancemeasurement, this var iable has continued to be somewhat controversial. It is important to recognize thatFrster distances are usually given for an assumed value of -squared, typically the dynamically averagedvalue of 2/3 (0.67). This assumed value results from randomization of donor and acceptor orientation byrotational diffusion prior to energy transfer. The orientation factor depends upon the relative orientationsin space of the donor emission dipole and the acceptor absorption dipole, and can range from zero to 4. Avalue of 1 corresponds to parallel transition dipoles, while a value of 4 results from dipoles that are bothparallel and collinear.

Because of the sixth-root relationship to the Frster distance, a variation in the orientation factor from 1 to4 produces only a 26 percent change in the calculated distance, and a maximum error of 35 percent ispossible when the customarily assumed value of 0.67 is applied. The most serious potential error results ifthe dipoles are oriented exactly perpendicular to each other and the corresponding -squared valuebecomes equal to zero. Several techniques for dealing with the uncertainty have been employed, includingthe assumption that a range of static or ientations exist and do not change during the excited-state lifetimeof the fluorophore. Measurements of fluorescence anisotropy for the donor and acceptor can allow limitsto be determined for -squared variation. In addition, utilization of fluorophores with low fluorescencepolarization (due to emission from several overlapping transitions) reduces the uncertainty in orientationfactor. Limitation of the possible values of -squared in this fashion reduces the potential distancecalculation error to perhaps 10 percent.

In many cases, the orientation factor is difficult, if not impossible, to determine and the exact value of thevariable is often regarded as an insurmountable problem. However, some evidence indicates a limitationfor the importance of the factor in resonance energy transfer calculations. Comparison of donor andacceptor distances using resonance energy transfer spectroscopy and x-ray diffraction techniques largelysupports the validity of assuming a value of 0.67 for the factor (as proposed by Frster theory), at least forsmall peptides and proteins. More uncertainty exists for larger proteins. The use of this value for theorientation factor is valid under the assumption that both donor and acceptor probes are free to undergounrestricted isotropic motion. Further justification is gained from experimental evidence that forfluorophores attached by a single or double bond to macromolecules, segmental motions of the donor andacceptor tend to result in dynamically randomized orientations.


6/12



For loosely bound fluorochromes, free rotational motion around single bonds should enable use of anaverage orientation value, but unrestricted motion of molecules bound through multiple linkage sitesprobably does not occur. On the other hand, the extreme values of zero and 4 for -squared requirecomplete fluorescence polarization of the donor and acceptor, a condition that is unlikely to be achieved.Statistical calculations have been presented by some investigators that argue the donor-acceptordistribution distances and their orientations determine the observed average distance. Provided that thereis some distribution in observed distance (and this is not limited by the donor and acceptor being too closerelative to R(0)), the average distance between fluorophores can be reliably obtained and the uncertaintydue to orientation factor assessed.

The dependence of the orientation factor (-squared) on the relative orientations of the donor emissiondipole and the acceptor absorption dipole (illustrated in Figure 5) is given by the equation:

2= (cos T- 3cos Dcos A)2

= (sin Dsin Acos - 2cos Dcos A)2

where (T)is the angle between the emission transition dipole of the donor and the absorption transitiondipole of the acceptor, (D)and (A)are the angles between these dipoles and the vector joining thedonor and acceptor, and is the angle between the planes containing the two transition dipoles.

Energy transfer efficiency is most sensitive to distance changes when the donor-acceptor separationlength approaches the Frster distance (R(0)) for the two molecules. Figure 6 illustrates the exponentialrelationship between transfer efficiency and the distance separating the donor and acceptor. Theefficiency rapidly increases to 100 percent as the separation distance decreases below R(0), andconversely, decreases to zero when ris greater than R(0). Because of the strong (sixth-power)dependence of transfer efficiency on distance, measurements of the donor-acceptor separation distanceare only reliable when the donor and acceptor radius lies within the Frster distance by a factor of two.When ris approximately 50 percent of R(0), the resonance energy transfer efficiency is near the maximumand shorter distances cannot be reliably determined. When the donor-acceptor distance exceeds the R(0)value by 50 percent, the slope of the curve is so shallow that longer separation distances are notresolved.

The practical significance of knowing the critical Frster distance is that the value provides a guide to therange of separation distances that can be determined by FRET for a given pair of probes (see Table 1).Because energy transfer measurement is very sensitive to distance variation when donor-acceptordistances are close to the Frster distance, the approximate dimensions of the target molecular interactionis the most important factor in selecting a fluorescent dye pair. Other factors that should be considered,depending upon whether steady state or time-resolved measurements are being conducted, includechemical stability, quantum yield, and fluorophore decay lifetimes. Since no internal distance referenceexists for common fluorescence resonance energy transfer techniques, distances calculated by measuringtransfer efficiencies are relative to a Frster distance, which is derived from spectroscopic data measuredon the donor-acceptor pairs.

The phenomenon of resonance energy transfer by the Frster mechanism is complex in some aspects,but simple and dependable in its resulting effect. Frster distances are accurately predictable fromspectral properties of the donor and acceptor, and since no exceptions to the theory have yet been


7/12



identified, resonance energy transfer can be assumed to occur under any conditions that place the donor-acceptor molecule pair in close proximity. The complexity in the theory describing dipole transfer arises,not because of the transfer mechanism itself, but because of the occurrence of distance distributions(including nonrandom distributions), and diffusion of the donor and acceptor molecules. When steps aretaken to average the distance dependence of the energy transfer over a range of geometries andtimeframes, FRET is a reliable technique for study of the spatial distributions between interactingmolecules.

Application of FRET Techniques in Optical M icroscopy

Microscope configurational parameters for fluorescence resonance energy transfer investigations varywith the requirements of the fluorophores, specimen, and imaging mode(s), but vir tually any upright or

inverted microscope can be retrofitted for FRET microscopy (see Figure 7). In general, the microscopeshould be equipped with a high-resolution (12-bit) cooled and intensified CCD camera system coupled toquality interference filters having low levels of crosstalk (minimum blocking level) and bandpass regionscorresponding closely to the fluorophore spectra. The detector sensitivity determines how narrow the filterbandpass can be and still enable data acquisition to proceed at acceptable speeds with a minimum ofspectral bleed-through noise. In most cases, a single dichromatic mirror coupled to excitation and emissionfilter wheels or sliders should be used to acquire images in order to minimize or eliminate image shifts.

Widefield fluorescence microscopy suffers from fluorophore emission originating above and below thefocal plane to yield images with significant out-of-focus signal that reduces contrast and leads to imagedegradation. This problem is compounded in FRET microscopy because of the inherently low signal levelsproduced as a result of resonance energy transfer. Digital deconvolution techniques can be coupled tooptical sectioning in order to reduce or eliminate signals away from the focal plane, but the process iscomputationally intensive and may not be fast enough for many dynamic FRET imaging experiments.Laser scanning confocal techniques can be applied to FRET microscopy to produce a significantimprovement in lateral resolution, while enabling the collection of serial optical sections at intervalsapproaching real time. The major drawback of confocal microscopy is the limitation of excitationwavelengths to the standard laser lines available for a particular system, which restricts the choice offluorophore donor and acceptor pairs in resonance energy transfer experiments. Multiphoton excitationcan also be employed in combination with FRET techniques and is less damaging to cells due to thelonger excitation wavelengths involved. In addition, autofluorescence artifacts and photobleaching of thespecimen are less likely to occur within the restricted excitation volume characteristic of multiphotonexcitation.

A typical microscope configuration capable of observing living cells in culture with several fluorescenceresonance energy transfer imaging motifs is presented in Figure 7. The inverted tissue culture microscopeis equipped with a standard tungsten-halogen lamphouse on the pillar in order to examine and record thecells using standard brightfield, phase contrast, or differential interference contrast (DIC) illumination. Notethat the latter two contrast enhancing techniques can be employed in combination with fluorescence toreveal the spatial location of fluorophores within the cellular architecture. A standard Peltier-cooled CCDcamera system is attached to the microscope trinocular head for widefield fluorescence and brightfieldimage capture.

Resonance energy transfer experiments are conducted with the multispectral microscope illustrated inFigure 7 using either widefield illumination (arc discharge lamp) or a real-time scanning confocalattachment equipped with a high-speed Nipkow disk system. The argon-krypton laser beam is first filteredthrough an acousto-optic tunable wavelength device to select specific excitation wavelengths beforepassing to the confocal scan head. Images are collected using two high-resolution Gen III-intensifiedcooled CCD cameras reading separate channels and spooled to a host computer. Scanning the specimenin the lateral (xand y) and axial (z) planes enables collection of optical sections for three-dimensionalimage reconstruction. A variety of image processing software programs are compatible with the illustratedmicroscope configuration.

Based upon the fundamental principles of the phenomenon, a number of important practical points shouldbe considered when fluorescence resonance energy transfer measurements are conducted with an optical


8/12



microscope:

The concentrations of donor and acceptor fluorophores must be closely controlled. Thestatistically highest probability of achieving fluorescence resonance energy transfer occurs whena number of acceptor molecules surround a single donor molecule.

Photobleaching must be eliminated because the artifact can alter the donor-to-acceptormolecular ratio, and therefore, the measured value of the resonance energy transfer process.

The donor fluorescence emission spectrum and the acceptor absorption spectrum should have asubstantial overlap region.

There should be minimal direct excitation of the acceptor in the wavelength region utilized toexcite the donor. A common source of error in steady state FRET microscopy measurements isthe detection of donor emission with acceptor filter sets.

The emission wavelengths of both the donor and acceptor must coincide with the maximumsensitivity range of the detector.

The donor absorption and emission spectra should have a minimal overlap in order to reducethe possibility of donor-to-donor self-transfer.

The donor molecule must be fluorescent and exhibit sufficiently long lifetime in order forresonance energy transfer to occur.

The donor should exhibit low polarization anisotropy to minimize uncertainties in the value of theorientation (-squared) factor. This requirement is satisfied by donors whose emission results

from several overlapping excitation transitions.

When using antibody labeling techniques, reagents conjugated with donor and acceptorfluorochromes should not be altered in their biological activity. Any reduction in activity willseriously affect the validity of resulting resonance energy transfer measurements.

Because fluorescence resonance energy transfer requires the donor and acceptor molecules tohave the appropriate dipole alignment and be positioned within 10 nanometers of each other, thetertiary structure of the reagents to which the molecules are attached must be considered. Forexample, when donor-acceptor molecules can be attached to different structural locations (suchas the carboxy or amino terminus) on a protein, it is possible that FRET will not be observedeven though the proteins do interact, because the donor and acceptor molecules are located onopposite ends of the interacting molecules.

Living cells labeled with green fluorescent protein mutants for FRET investigations should beanalyzed using traditional immunohistochemical techniques to verify that the tagged protein

adopts the same intracellular habitat and properties as the native counterpart.

In order for the fluorescence resonance energy transfer phenomenon to provide meaningful data as a toolin optical microscopy, both specimen preparation and imaging parameters must be optimized. Theselection of appropriate donor and acceptor probes and the manner in which they are employed asmolecular labels is a major challenge. In addition, once a labeling strategy that permits energy transfer hasbeen elucidated, a wide spectrum of techniques may be used to perform the measurement itself. Amajority of quantitative fluorescence microscopy investigations are conducted by measuring the intensityof fluorescence emission. Fluorescence intensity-based detection of FRET is typically achieved bymonitoring changes in the relative amounts of emission intensity at the two wavelengths corresponding tothe donor and acceptor chromophores. When conditions are appropriate for fluorescence resonanceenergy transfer to occur, an increase in acceptor emission (I(A)) is accompanied by a concomitantdecrease in donor emission (I(D)) intensity.

Although a change in the relative emission intensity of either donor or acceptor can be taken as indicativeof resonance energy transfer, the customary approach is to utilize the ratio of the two values, I(A)/I(D), as

a measure of FRET. The value of the ratio depends upon the average distance between donor-acceptorpairs and is insensitive to differences in the path length and volume accessed by the exciting light beam.

Any specimen condition that induces a change in the relative distance between the molecular pairsproduces a change in the ratio of donor and acceptor emission. Consequently, FRET can be observed inthe microscope by preferential excitation of a donor fluorophore and detection of the increased emissionof an interacting acceptor fluorophore, accompanied by a reduction in donor fluorescence produced byquenching due to energy transfer. Measurement of FRET employing the intensity-monitoring approach istermed steady statefluorescence resonance energy transfer imaging.


9/12



Appropriate donor and acceptor probes are selected on the basis of their absorption and emissionspectral characteristics. For maximum resonance energy transfer, the donor emission spectrum shouldsubstantially overlap the absorption spectrum of the acceptor. In addition, there should be minimal directexcitation of the acceptor fluorophore at the excitation maximum of the donor, and there should not besignificant emission overlap between the donor and acceptor in the wavelength region at which acceptoremission occurs. In practice, it can be difficult to identify donor-acceptor pairs that satisfy theserequirements. The situation is often complicated by the fact that the commercially available fluorescencefilter sets are not completely effective in passing only the desired wavelengths, and a small percentage oflight outside the design passband may be transmitted. Unless very well characterized and controlledexpression systems are used, the precise concentration of the donor and acceptor fluorophores may bedifficult to determine. Additional corrections may be also be required for autofluorescence,photobleaching, and background fluorescence.

A typical investigation of intracellular protein-protein association in a living cell culture is illustrated inFigure 8 for events associated with apoptosis, a physicological process of cellular death resulting from anintricate cascade of sequential interactions. Gene products directly involved in the chain of events can belabeled by fusion to appropriate members of the fluorescent protein family (in this case, BFP and GFP) forco-expression in the same cell in order to probe specific associations by FRET. The proteins involved withapoptosis interact within the mitochondria and display a gradual decrease in binding as programmed celldeath proceeds. Thus, an image of donor emission (Figure 8(a)) contains only fluorescence from the BFP-labeled proteins, while the corresponding acceptor emission profile (Figure 9(b)) illustrates signals due toproteins labeled with GFP (and some contribution from donor emission). A FRET filter (Figure 8(c)), asdescribed below, reveals fluorescence derived from resonance energy transfer between the two proteins

Among the factors that may potentially affect the accuracy of fluorescence resonance energy transfermeasurements in general, several are highly specific to the optical microscope. A primary target inmicroscopy investigations is to obtain high resolution images, and this requires particular attention to thequality and performance of optical filters employed to spectrally discriminate among the absorption andemission wavelengths of the donor and acceptor. In order to maximize the signal-to-noise ratio (withoutdeleteriously affecting the specimen or the process being investigated), it is necessary to carefully balancethe intensity and time of exposure to excitation light with the concentration of donor and acceptorfluorophores and the detector efficiency. If the concentration of donor-acceptor fluorophores is excessive,

self-quenching can occur, affecting the accuracy of FRET measurements. Photobleaching is a problemwith all fluorophores, and can affect the donor-acceptor rat io, altering fluorescence measurements. Excessillumination intensity can also damage specimens, particularly those containing living cells or tissues.

A technique known as donor photobleaching fluorescence resonance ene rgy transfer(pbFRET),which exploits the photobleaching process to measure FRET, is often applied in the study of fixedspecimens. Based on pixel-by-pixel analysis, the method has been applied to measure proximityrelationships between cell surface proteins labeled with fluorophore-conjugated monoclonal antibodies.Photobleaching FRET is founded on the theory that a fluorophore is sensitive to photodamage only whenit is elevated to an excited state. Statistically, only a small proportion of molecules are in an excited state atany one time, and therefore, fluorophores with longer fluorescence lifetimes have a higher probability ofsuffering photodamage and exhibit a higher rate of photobleaching.

Experimental evidence supporting this concept has demonstrated that the photobleaching time of afluorophore varies inversely with its excited-state lifetime. The occurrence of resonance energy transferreduces the fluorescence lifetime of the donor molecule, effectively protecting it against photobleaching.

Calculations of pbFRET are based on the decreased rate of donor photobleaching relative to thatmeasured for the donor in the absence of resonance energy transfer. The measurement ofphotobleaching in FRET studies requires a relatively long timeframe, and therefore is most applicable tofixed cell specimens in which temporal data is not important and the effect on cell function fromphotobleaching is not an issue. In some respects, the donor photobleaching technique is less complicatedthan sensitized emission measurement, although fitting of time constants to photobleaching curvesinvolving multiple components presents some additional difficulties.

The energy transfer efficiency can also be determined by acceptor photobleachingtechniques, inwhich the change in donor emission quenching is measured by comparing the value before and afterselectively photobleaching the acceptor molecule. Analysis of the change in donor fluorescence intensityin the same specimen regions, before and after removal of the acceptor, has the advantage of requiringonly a single specimen preparation, and directly relates the energy transfer efficiency to both donor andacceptor fluorescence.


10/12



Accurate measurement of fluorescence resonance energy transfer in the microscope requirescompensation for all of the potential error sources. A straightforward technique to correct for detection ofdonor fluorescence with the acceptor emission filter and acceptor fluorescence with the donor emissionfilter (due to crossover or spectral bleed-through) has been developed. The method also corrects for thedependence of FRET on the concentrations of the donor and acceptor fluorophores. The measurementstrategy, which requires a minimum of spectral information, utilizes a combination of three filter sets andcan be readily implemented. The donor, FRET, and acceptor filter sets are designed to isolate andmaximize three specific signals: donor fluorescence, acceptor fluorescence attributable to FRET, and thedirectly excited acceptor fluorescence, respectively. In practice, three different specimens, containing justdonor, just acceptor, and both donor and acceptor are examined with each of the three filter sets, and theresulting data manipulated arithmetically to correct for crossover and for uncontrolled variations in donor-acceptor concentrations.

Presented in Figure 9 are schematic illustrations of crossover (spectral bleed-through) and filter crosstalk,two significant problems that must be overcome in order to achieve quantitative results in fluorescenceresonance energy transfer experiments. Crossover or bleed-through is manifested by an overlap of thedonor fluorescence emission spectrum with the bandpass region of the acceptor emission interferencefilter in Figure 9, resulting in donor emission signal (unwanted wavelengths) being transmitted through theemission filter. In contrast, filter crosstalk describes the minimum attenuation (blocking) level over aspecific range of two filters placed together in series, and is of concern when matching excitation andemission filters for fluorescence sets. Dichromatic mirrors are often included in crosstalk evaluation offluorescence filter combinations. Although two emission filters are rarely placed in the light path at thesame time, the spectra are drawn together in Figure 9 to simultaneously illustrate both concepts. Note thatthe two filter spectra (blue and red curves) represent light transmittance by the interference filters,whereas the donor emission curve (green) is a plot of intensity versus wavelength.

Additional factors, which can potentially introduce significant errors, also require correction when steadystate FRET measurement techniques are employed. Furthermore, careful control of the donor and

acceptor fluorophore concentrations is desirable. Fluorophore concentration determinations can bepartially avoided through the application of time-resolvedfluorescence measurements, which provide amethod of obtaining average lifetimes without a precise knowledge of donor concentrations. The techniqueenables quantitative determination of donor-acceptor separation distances, and is based onmeasurements of the donor lifetime in the presence and absence of the acceptor. Measuring fluorescenceintensity decay as a function of time elucidates the emission dynamics of the excited-state molecule, andconsequently, more detailed information about the nature of the donor-acceptor interaction may beobtained. Graphical plots of intensity decay illustrate time-averaged details of the fluorescence decayprocess (see Figure 10(a)), which are unresolved when employing steady state techniques.Measurements indicating the same value for average lifetime, when recorded as steady state intensitynormalized to absorption, may correspond to significantly different decay curve shapes in time-resolveddata plots, indicating differences in the intermolecular processes involved.

The fluorescence lifetime () of a fluorophore is the characteristic time that a molecule exists in the excitedstate prior to returning to the ground state. Representing fluorescence decay in a simplified singleexponential form following a br ief pulse of excitation light, the fluorescence intensity as a function of time

(t) is given by the equation:

I(t) = I0exp (-t/)

where I(0)is the initial fluorescence emission intensity immediately after the excitation light pulse, and I(t)is the fluorescence intensity measured at time t. The fluorescence lifetime () is defined as the timerequired for the intensity to decay to 1/eof its initial value (approximately 37 percent of I(0); Figure 10(a)),and is the reciprocal of the rate constant for fluorescence decay from the excited state to the ground state.

The primary overall advantage of time-resolved versus steady state FRET measurements is that donor-acceptor separation distances can be mapped with greater quantitative accuracy. This results in partbecause fluorescent lifetimes do not depend upon local intensity or concentration, and are largelyunaffected by photobleaching of the fluorophores. Fluorescence lifetimes are, however, highly sensitive tofluorophore environment, and even molecules with similar spectra may display distinct lifetimes underdifferent environmental conditions. Because scattering does not affect fluorophore lifetimes,measurements of lifetime variation can provide information that is specifically related to local molecular


11/12



processes.

The lifetime of a fluorophore is subject to modification by numerous variables in the localmicroenvironment, including factors such as hydrophobicity, oxygen concentration, ionic strength of othermedia components, binding to macromolecules, and the proximity to acceptor molecules that can depletethe excited state by resonance energy transfer. It is a significant practical advantage that lifetimemeasurements can serve as absolute indicators of molecular interactions, and are independent offluorophore concentration.

Two general techniques commonly employed to measure fluorescent lifetimes are categorized as thetime-domain (pulsed, see Figure 10(a)) and frequency-domain(also termed phase-resolved; Figure10(b)) methods. Time-domain lifetime measurements employ pulsed excitation light sources, and thefluorescent lifetime is obtained by directly measuring the emission signal or by photon-counting detection.The frequency-domain approach utilizes sinusoidal modulation of the excitation light source (obtained frompulsed or modulated laser systems), and lifetimes are determined from the phase shift and demodulationdepth of the fluorescence emission signal. Each of these approaches to fluorescence lifetime imaging hasspecific advantages and disadvantages, and both have been widely applied to conventional widefield,confocal, and multiphoton microscopy.

Illustrated in Figure 10 are schematic diagrams representing the time domain and frequency domaintechniques for determination of fluorescence lifetimes. In the time domain approach (Figure 10(a)), thespecimen is excited by a brief pulse of laser light having a duration much shorter than the lifetime of theexcited species, and the exponential decay profile is measured as a function of time. Fluorescence decayis usually a monoexponential function for a single fluorophore, but can display far more complex characterif the excited state has numerous relaxation pathways available in the environment. Sinusoidally modulatedlight from a continuous wave laser coupled to an acousto-optical modulator is employed to excite thefluorophore in frequency domain experiments (Figure 10(b)). The resulting fluorescence emission issinusoidally modulated at the same frequency as the excitation, but is accompanied by a phase shift and

reduction in the modulation depth. In the case of a single exponential decay, the fluorescence lifetime canbe calculated by determining either the degree of phase shift (), or the modulation ratio (M), using theequations presented in Figure 10(b). If the two values are identical, fluorescence decay is truly composedof a single exponential function. When more than one fluorescent species is present (or a singlefluorophore experiences a complex environment), the phase shift and modulation lifetimes should beevaluated over a wide range of frequencies.

The time-domain technique for measuring fluorescence lifetime relies essentially on single-photoncounting and requires a detection system with sufficient temporal resolution to collect nearly 100 percentof the photons generated by each excitation pulse. Although phase-resolved techniques are relatively lessdemanding to perform, they are not generally as sensitive as the photon-counting approach. When phasemodulation is employed to resolve complicated multifluorophore lifetimes, long exposure times to damagingexcitation illumination can prove to be excessive for some specimens, and also may not provide sufficienttemporal resolution for live-cell processes. The preferred technique depends upon both the informationrequired from the investigation and the type of specimen being studied.

Fluorescence lifetime measurements have proven to be a sensitive indicator of FRET, and have particularadvantages in live-cell studies because of the independence of lifetime measurements upon factors suchas concentration and light path length, which are difficult to control in living specimens. A primaryadvantage of performing FRET studies by fluorescence lifetime measurement lies in the fact that it ispossible to distinguish energy transfer even between donor-acceptor pairs with similar emission spectra.When fluorescence lifetimes are measured directly (in contrast to the use of steady state values), adetermination of FRET is possible without the photodestruction of the donor or acceptor fluorophores.Because FRET reduces the fluorescence lifetime of the donor molecule through energy transfer to theacceptor, a direct comparison of the donor lifetime in the presence of the acceptor ((DA)) to that in theabsence of the acceptor ((D)), enables the calculation of a FRET efficiency value (E(T)) for each imagepixel.

Depending upon the technique, fluorescence lifetime measurements require the specimen to be exposedto either high-frequency repetitive pulses of excitation light, or to continuous sinusoidally modulated light.In studies with living cells, the effect of high-intensity illumination must always be evaluated. Regardless ofthe method, the reference lifetime of the donor without acceptor must be determined under experimental


12/12


Contributing Authors

Brian Hermanand Victoria E. Centonze Frohlich- Department of Cellular and Structural Biology, University of

Texas Health Science Center, 7703 Floyd Curl Drive, San Antonio, Texas 78229.Joseph R. Lakowicz- Center for Fluorescence Spectroscopy, Department of Biochemistry and Molecular

Biology, University of Maryland and University of Maryland Biotechnology Institute (UMBI), 725 West Lombard

Street, Baltimore, Maryland 21201.

Thomas J. Fellersand Michael W. Davidson- National High Magnetic Field Laboratory, 1800 East Paul Dirac

Dr., The Florida State University, Tallahassee, Florida, 32310.

conditions identical to those of the donor-acceptor measurement. One means of accomplishing this with asingle specimen is to measure the donor-only lifetime after photobleaching destruction of the acceptorfollowing the energy transfer experiment.

Conclusions

In biological investigations, the most common applications of fluorescence resonance energy transfer arethe measurement of distances between two sites on a macromolecule (usually a protein or nucleic acid) orthe examination of in vivointeraction between biomolecular entities. Proteins can be labeled with syntheticfluorochromes or immunofluorescent fluorophores to serve as the donor and acceptor, but advances influorescent protein genetics now enable researchers to label specific target proteins with a variety ofbiological fluorophores having differing spectral characteristics. In many cases, the amino acid tryptophan

is used as an intrinsic donor fluorophore, which can be coupled to any number of extrinsic probes servingas an acceptor.

If macromolecules are labeled with a single donor and acceptor, and the distance between the twofluorochromes is not altered during the donor excited state lifetime, then the distance between the probescan be determined from the efficiency of energy transfer through steady state measurements, asdiscussed above. In cases where the distance between the donor and acceptor fluctuates around adistribution curve, such as protein assemblies, membranes, single-stranded nucleic acids, or unfoldedproteins (see the scenarios presented in Figure 11), FRET can still be employed to study the phenomena,but time-resolved lifetime measurements are preferred. Several biological applications that fall into bothcases are illustrated in Figure 11, including conformational changes, dissociation or hydrolysis, fusion ofmembrane-like lipid vesicles, and ligand-receptor interactions.

Although various methods are available for the measurement of fluorescence resonance energy transferin the optical microscope, none are completely without disadvantages. Some techniques require moreelaborate and expensive instrumentation, while others are based on assumptions that must be carefullyvalidated. Certain approaches are appropriate for fixed specimens, but cannot be applied to living cellsystems, while other methods must incorporate significant corrective calculations or data analysis

algorithms. It is certain, however, that FRET analysis shows great promise for further development in theutility and scope of biological applications. Dramatic improvements in instrumentation have occurred inrecent years, particularly with respect to time-resolved techniques.

Fluorescence lifetime measurements that were only accomplished with extreme difficultly in the past arenow aided by mature picosecond and nanosecond technologies. Advances in fluorescent probedevelopment have produced smaller and more stable molecules with new mechanisms of attachment tobiological targets. Fluorophores have also been developed with a wide range of intrinsic excited statelifetimes, and a significant effort is being placed on development of a greater diversity in genetic variationsof fluorescent proteins. Entirely new classes of fluorescent materials, many of which are smaller thanprevious fluorophores and allow evaluation of molecular interactions at lower separation distances,promise to improve the versatility of labeling and lead to new applications of the FRET technique.

BACK TO SPECIALIZED MICROSCOPY TECHNIQUES

2012 Olympus America Inc.

All rights reserved
http://www.olympusmicro.com/primer/techniques/index.html

olympus microscopy resource center _ fluorescence resonance energy transfer (fret) microscopy -...

Documents