Académique Documents
Professionnel Documents
Culture Documents
The light microscope has long been used During acquisition of the digital image, accuracy is obvious. Precision is equally
to document the localization of fluores the photons that are detected at each important in quantitative fluorescence
cent molecules in cell biology research. pixel are converted to an intensity value microscopy because we are often forced
With advances in digital cameras and that is correlated to, but not equal to, the to make only one measurement (for ex-
the discovery and development of geneti number of detected photons (Pawley, ample, one time-point in a live-cell
cally encoded fluorophores, there has 2006c). In fluorescence microscopy, the time-lapse experiment). In addition, we
THE JOURNAL OF CELL BIOLOGY
been a huge increase in the use of fluor intensity value of a pixel is related to the are usually measuring biological speci-
escence microscopy to quantify spatial number of fluorophores present at the cor- mens that have some level of natural
and temporal measurements of fluores responding area in the specimen. We can variability, so variance seen in measure-
cent molecules in biological specimens.
Figure 2. The importance of SNR in intensity and spatial measurements. (A) A digital image taken with a cooled CCD camera (ORCA-AG; Hamamatsu
Photonics), with no light sent to the camera. Using MetaMorph software, a line (shown in red) was drawn across the bead and a line-scan graph was gener-
ated to show the intensity value of the pixels along the line. The graph shows line-scans of two similar images, taken in quick succession. The intensity values
in the images fluctuate (range, 195–205) around the camera digital offset value of 200. Notice that the fluctuation in intensity values changes at each pixel
from one image to the next. This variance is due primarily to thermal and readout noise from the CCD camera, and the extent of the variance will differ
The specimen. Fluorophores Giepmans et al., 2006; Johnson, 2006) excite the fluorophore and collecting as
vary greatly in their intrinsic brightness to make the best choice of fluorophore much of the emission light as possible
and the rate at which they photobleach; and anti-photobleaching reagent for your (Ploem, 1999; Rietdorf and Stelzer, 2006).
an easy way to maximize signal is to specimen and experiment. Goodwin (2007) There are several useful online tools avail-
choose a brighter and more photo-stable provides a complete discussion of the able for matching fluorophores to filters
fluorophore (Diaspro et al., 2006; Tsien importance of mounting medium choice (for example, as of the date of this publi-
et al., 2006). The brightness of a fluoro- to both signal intensity and resolution. cation Invitrogen has a very useful tool
phore is determined primarily by its ex- The microscope. To get the on their website: http://www.invitrogen
tinction coefficient and quantum yield, brightest signal while minimizing speci- .com/site/us/en/home/support/Research-
properties that are dependent on the fluor men damage, it is important to use illu- Tools/Fluorescence-SpectraViewer.html).
ophore’s environment (Diaspro et al., mination wavelengths that will optimally In an epifluorescence microscope,
2006). It should be noted that new fluor excite the fluorophore and to collect as the objective lens both illuminates the
escent proteins are routinely introduced many of the emission photons as possi- specimen and collects photons emitted
that outperform their predecessors; it is ble (Ploem, 1999; Rietdorf and Stelzer, from fluorophores to form the optical
therefore advisable to search the current 2006). Fluorescence spectra that show image. The numerical aperture (NA) of
scientific literature for the latest variants. the absorption and emission efficiency of the objective lens (marked on the barrel
Fixed specimens should be mounted fluorophores are available from the man- of the lens after the magnification; Keller,
in a glycerol-based mounting medium ufacturer or in the scientific literature 2006) is an important determinant of the
(Egner and Hell, 2006; Goodwin, 2007) (for example, see Shaner et al., 2005), brightness of the optical image. The num-
that contains an anti-photobleaching in- and filter manufacturers provide spectra ber of photons an objective can collect
hibitor (Diaspro et al., 2006). No one anti- online that show the percent transmis- from a specimen (and therefore the bright-
photobleaching reagent is the best, as sion of their filters across wavelength. It ness of the image) increases with NA2.
each reagent is more or less effective for is important to compare the spectra for Brightness of an objective is also deter-
a given fluorophore (Diaspro et al., 2006). the fluorophore you are imaging to spec- mined by properties such as transmission
Review the fluorophore manufacturer’s tra for the fluorescence filter sets (and/or and correction for aberration (Keller,
product information or the relevant scien- laser illumination line) to ensure you are 2006). Spherical aberration caused by the
tific literature (Shaner et al., 2005; using the correct wavelengths of light to objective lens (Hell and Stelzer, 1995;
Choose a bright (high quantum yield, high extinction coefficient) and photo-stable fluorophorea
Image through a clean No. 1.5 coverslipb
Mount specimen as close to the coverslip as possiblec
Use high NA clean objective lens with lowest acceptable magnificationd
Choose fluorescence filter sets that match fluorophore spectrad
Align arc lamp for Koehler illuminationd
For fixed specimens, use a glycerol-based mounting medium containing anti-photobleaching inhibitors3
Remove DIC Wollaston prism and analyzer from light pathe
Use a cooled CCD camera with at least 60% quantum efficiencyd
Use camera binningd
Decrease noise:
Use a cooled CCD camera with less than 8 electrons readout noise and negligible dark noisef
Use amplification (e.g., EM-CCDs) only when signal is limitingf
Increase signal (see above) to reduce relative contribution of Poisson noisef
Decrease background:
Clean coverslips and opticse
Perfect fluorophore labeling protocol to minimize nonspecific labelingg
Mount specimens in minimally fluorescent medium (e.g., without phenol red)d
Use band-pass filter sets that block autofluorescenced
Goodwin, 2007) or introduced by the the immersion medium (e.g., mounting Increasing the exposure time al-
specimen (Egner and Hell, 2006) decreases medium with a high concentration of lows the flux of photons coming from the
image intensity (North, 2006; Waters, glycerol will have a refractive index specimen to accumulate (as electrons)
2007; Waters and Swedlow, 2007). Spher- close to that of standard immersion oil). in the detector, increasing the intensity
ical aberration occurs when there is a rel- The detector. The number of values in the image—up to a point
atively large difference in refractive index photons reaching the detector that are (Moomaw, 2007; Spring, 2007; Waters,
between the specimen and the lens im- collected and contribute to the intensity 2007). Detectors have a limited capacity
mersion medium; for example, when an values in a digital image depends on the to hold electrons; if this capacity is
oil immersion lens is used to image a quantum efficiency (QE) of the detec- reached, the corresponding pixel will be
specimen in an aqueous solution such as tor, and how long the signal is allowed “saturated” and any photons reaching the
cell culture medium (Egner and Hell, to integrate on the detector (usually re- detector after saturation will not be
2006). Spherical aberration caused by ferred to as the exposure time). QE is a counted. The linearity of the detector is
refractive index mismatch generally in- measure of the percentage of photons therefore lost, and saturated images can-
creases with distance from the coverslip reaching the detector that are counted not be used for quantitation of fluores-
(Joglekar et al., 2008). Spherical aber- (Moomaw, 2007). The QE of research- cence intensity values. Choosing to “crop
ration can be addressed using water im- grade CCD cameras most often used for out” saturated areas is not acceptable
mersion objective lenses (Keller, 2006), by quantitation of fluorescence images (unless they can be shown to be irrele-
using an objective lens with a correction ranges from 60% to over 90%, whereas vant to the experimental hypothesis) be-
collar (Keller, 2006; Waters, 2007), or by the QE of PMTs used in point-scanning cause it will select for the weaker
immersion oil refractive index matching confocals is much lower, usually 10– intensity parts of the specimen. Satura-
(Goodwin, 2007). For fixed specimens, 20% (although the effective QE is sig- tion should be avoided by using image
spherical aberration is reduced by mount- nificantly less; see Pawley, 2006b). QE acquisition software to monitor intensity
ing fixed specimens in a mounting medium values are available online from the de- values when setting up the acquisition
with a refractive index similar to that of tector manufacturer. parameters (Table II).
• Set up specimen and imaging system for optimal signal detection, low background, and low noise (Table I)
3. Store images
• Always save the raw imagesc
• Use either no compression or lossless compressionc
4. Process images
• Use flat-field correction to correct for uneven illuminationd
• Be sure any other image processing used prior to quantitation preserves relative intensity valuesc,d
5. Analyze images
• Subtract local background value from intensity measurementse
In most live biological specimens, ments of intensity, it is also very important Poisson noise. Subtracting a constant
saturation is much less of a problem than to first reduce background as much as background value from intensity measure-
collecting enough signal to get adequate possible (Fig. 1, Table I). Background in ments does not change the variance due to
SNR images for quantitation. Many an image effectively reduces both the dy- Poisson noise; the presence of background
research-grade cooled cameras allow bin- namic range and the SNR. Dynamic range therefore reduces image SNR.
ning of adjacent pixels on the CCD chip. of a CCD camera is defined as the full A common source of background
With all other acquisition parameters be- well capacity of the photodiodes (i.e., the in biological specimens is out-of-focus
ing equal, binning on the CCD chip in- number of photons that can be detected fluorescence. In fluorescence micros-
creases the intensity of the pixels without per pixel before saturation) divided by the copy, the illuminating light is focused at
increasing readout noise, resulting in a detector noise (Moomaw, 2007; Spring, the image focal plane by the objective
higher SNR digital image (Moomaw, 2007). High dynamic range is particularly lens, such that maximum excitation of
2007; Spring, 2007; Waters, 2007). How- important for collecting an adequate num- fluorophores occurs at the focal plane
ever, because the resulting pixels repre- ber of signal photons from both dim and (Hiraoka et al., 1990). However, illumi-
sent a larger area of the specimen (i.e., 4x bright areas of the specimen. Photons nating light above and below the image
larger with a 2 × 2 bin), binning decreases from background sources fill the detector, focal plane excites fluorophores above
the resolution of the digital image (Fig. 3). limiting the number of signal photons that and below the image focal plane. Light
In many low-light imaging experiments, can be collected before the detector satu- emitted from these out-of-focus fluoro-
however, the decrease in resolution is well rates (Fig. 1) and effectively decreasing phores is collected by the objective lens,
worth the increase in SNR (Table II). dynamic range. In addition, recall that the and appears as out-of-focus background
Background fluorescence re- number of photons counted defines the in the in-focus image of the specimen. In
duces dynamic range and de- Poisson noise level in an image. Poisson wide-field epifluorescence microscopy,
creases SNR. Although it’s true that noise is equal to the square root of the sig- adjusting the diameter of the field dia-
background fluorescence can and must be nal photons plus background photons; phragm to match the visible field of view
subtracted from quantitative measure- higher background therefore means higher minimizes the illumination of out-of-focus
fluorophores (Hiraoka et al., 1990) and re- methods used to remove out-of-focus fluor or j (pixels in the background), obj is the
duces background (Waters, 2007). escence has limitations, and may contrib- object of interest, bkg is the background,
There are several microscopy tech- ute additional noise to the image (Murray and N is the number of pixels in the object
niques that serve to reduce the amount of et al., 2007). In specimens with low levels of interest or the background. This equa-
Read noise is usually expressed in the choice of detector and acquisition set- equations define the theoretical resolu-
manufacturer’s technical specifications tings (Table I). tion limits in most cases; it should be
as a number of electrons, meaning that noted that a handful of very talented
the measured voltage will have a vari- Resolution microscopists have found it possible to
ance equal to that number of electrons In digital microscopy, spatial resolution surpass these limits using specialized
(i.e., the lower the value, the lower the is defined by both the microscope and “super-resolution” imaging techniques (for
noise). Detectors that use signal amplifi- the detector, and limits our ability to ac- review see Evanko, 2009).
cation (e.g., PMTs and electron multi- curately and precisely locate an object It is a common misconception
plying [EM] CCDs) introduce additional and distinguish close objects as separate among cell biologists that confocal mi-
noise during the amplification process. from one another (Inoue, 2006; Rasnik croscopy should be used to obtain the
For example, EM-CCD cameras amplify et al., 2007; Waters and Swedlow, 2007). highest resolution images. Although
signal differences sufficiently to reveal Objects that cannot be detected in an confocal microscopy is very effective at
clock-induced charging—stochastic vari- image cannot be resolved, so spatial increasing contrast in specimens with sig-
ations in the transfer of charge from one resolution is dependent on image SNR nificant out-of-focus fluorescence (Pawley,
pixel to another during read operations (Pawley, 2006c). When imaging a dy- 2006b; Murray et al., 2007), the obtain-
(Robbins and Hadwen, 2003; Moomaw, namic specimen over time, accuracy able resolution of confocal microscopy
2007). When possible, collecting more of quantitation may be further limited is essentially the same as conventional
photons from the specimen to increase by temporal resolution (the rate of wide-field fluorescence microscopy
the signal (see Table I) will result in a image acquisition; see Dorn et al., 2005; (Inoue, 2006). In addition, Murray et al.
higher SNR image than amplifying a Jares-Erijman and Jovin, 2006 for (2007) recently demonstrated that the
smaller number of collected photons. an example). photon collection efficiency and SNR of
Figure 4. Non-uniform illumination results in nonuniform fluorescence. All images were collected using a microscope (model TE2000E; Nikon), a Plan-
Apochromat 20x 0.75 NA objective lens, a camera (ORCA-AG; Hamamatsu Photonics), and MetaMorph software. (A) An image of a field of fluorescent
beads, using wide-field illumination. Individual beads contain a similar concentration of fluorophore (clumps of beads appear brighter, as is seen near the
center of the image). A pseudo-color displaying the range of intensity values (see inset) was applied. Note that beads in the top left have different intensity
values than the beads in the bottom right. (B) An image of a uniform field of fluorophore taken with the same microscope optics and conditions as A, show-
ing uneven illumination across the field of view. This nonuniform illumination explains the nonuniform fluorescence from the beads of similar fluorophore
concentration shown in A. (C) After flat-field correction (Zwier et al., 2004; Wolf et al., 2007), the image intensity values more accurately reflect the real
fluorescence in the specimen. This image was obtained using the image arithmetic function in image processing software (in this case, MetaMorph) to
lost in the digital image (Fig. 3 C). In live- of one object can be accurately deter- indicator dyes, are also reliable for
cell imaging, it is often favorable to give up mined; then intensity values can be used quantitation (Johnson, 2006). Quantita-
some resolution (by binning pixels, for to count multiple objects that are too close tion of live versus fixed cells is gener-
example) to increase image SNR and/or to one another to spatially resolve. These ally preferable because the fixation and
decrease photobleaching and photo dam- types of measurements are very challeng- extraction process can remove tagged
age (Waters, 2007). ing to perform with accuracy, and require proteins, quench the fluorescence of
Optical resolution need not a thorough understanding of, and atten- fluorescent proteins, and change the
limit accuracy in localization or tion to, every possible pitfall (Pawley, size and shape of cells (Allan, 2000;
counting. How does the resolution 2000)—but they are possible. For exam- Straight, 2007).
limit affect our ability to quantitate ple, the measured intensity of proteins One should use caution when using
using fluorescence microscopy? Clearly, conjugated to fluorescent proteins has immunofluorescence to measure the
the size of an object that is below the been used to accurately and precisely local concentration of a protein of inter-
resolution limit cannot be accurately count the number of labeled proteins lo- est, particularly with soluble proteins.
measured with the light microscope. calized to the kinetochore (Joglekar et al., The fixation and extraction necessary to
However, objects that are below the res- 2006) and proteins involved in cytokine- get antibodies into cells can change the
olution can be detected and an image of sis (Wu and Pollard, 2005). quantity and localization of detectable
the object formed by the microscope, if epitopes (Melan and Sluder, 1992). Multi
the imaging system is sensitive enough Additional threats to valent antibodies also bind with higher
and the object is bright enough. Al- accuracy and precision in affinity to multiple epitopes, which can
though the size of the object in the im- quantitative microscopy make areas with high concentration of
age will be inaccurate, the centroid of a Specimen preparation. Fluorescence the epitope label more efficiently than
high SNR image of the object can be from a fluorophore tagged to a molecule areas of low concentration (Mason and
used to locate the object with nanome- of interest is often used to measure the Williams, 1980). In addition, penetration
ter precision, far beyond the resolution quantity of the tagged molecule. Fluor of the antibody may not be consistent in
limit (Churchman et al., 2005; Yildiz escent proteins expressed in live cells different areas of the tissue, cells, and
and Selvin, 2005; Huang et al., 2008; are excellent for quantitation; because subcellular compartments (Allan, 2000).
Manley et al., 2008). there is a constant number of fluoro- Therefore, although accurate quantita-
In fluorescence microscopy, the res- phores per labeled protein, the number tion of the emission photons from fluor
olution limit does not limit our ability to of photons emitted can be an accurate escently labeled antibodies is possible,
accurately count fluorescently labeled ob- measure of the quantity of fluorescently that number of photons may not accu-
jects, even if the objects are below the labeled protein (Shaner et al., 2005; rately reflect the number of epitopes in
resolution limit. If the objects are all of Straight, 2007; Joglekar et al., 2008). the specimen. It is possible to use rigor-
similar size, are all labeled with the same Small molecules that bind with high ous controls to demonstrate that an im-
number of fluorophores, and the intensity affinity to their target, such as calcium munofluorescence protocol is an accurate
measure of a particular epitope of inter- microscope (using a liquid light guide, (Aubin, 1979; Tsien et al., 2006).
est (Mortensen and Larsson, 2001). for example) can increase the uniformity For quantitative measurements, bleed-
Measuring fluorescence deep into of illumination across the field of view through and autofluorescence should
specimens can also be problematic for (Nolte et al., 2006). In many cases, com- be avoided when possible, and measured
measurements of signal intensities. In pletely uniform illumination is impossi- and subtracted from measurements when
wavelengths (see Murray et al., 2007 for in 3D is almost always necessary for ac- on the rate of photobleaching, exciting
an example). curately determining intensity of objects fluorescent specimens before collect-
Focus. Focus is critical to accu- larger than the diffraction limit, and ing images that will be used for
rate and precise quantitation of fluores- when tracking the movements of objects quantitation may introduce error in
cence intensity values (Murray, 2005; that occur in 3D (De Mey et al., 2008). measurements of signal intensity. Ini-
Table II). The distribution of intensity Photobleaching. Almost all tial focusing and scanning of the speci-
values along the z-axis of the optical fluorophores photobleach when exposed men is best done using techniques such
image depends on the size of the fluores- to excitation light in the microscope, as phase or differential interference
cently labeled object and the point spread including all of the fluorescent pro- contrast (DIC) microscopy (Inoué and
function of the microscope. For small teins. The rate of photobleaching is Spring, 1997) because the halogen light
objects imaged with high resolution op- specific to the fluorophore, its environ- sources typically used for transmitted
tics, small changes in focus can have ment, and the intensity of the illuminat- light illumination usually will not bleach
large effects on measured intensity val- ing light (Diaspro et al., 2006). For fluorophores. To accurately measure
ues. Joglekar et al. (2008) describe a some specimens and fluorophores, anti- fluorescence intensity in the same field
method of determining the error intro- photobleaching reagents can be added of view over time, one should measure
duced by imprecise focus when measur- to the mounting medium to reduce the and correct for photobleaching that
ing objects that are thinner than the rate of photobleaching (Diaspro et al., occurs while imaging (Rabut and
diffraction limit of the optics. Measuring 2006; Tsien et al., 2006). Depending Ellenberg, 2005).
Image processing and pitfalls to accuracy and precision in quan- can occur only when at least three condi-
storage. Some types of image process- titative microscopy measurements de- tions are met: (1) the emission spectrum
ing and storage can change the relative scribed throughout this review (and in of the donor fluorophore overlaps with
intensity values in a digital image, ren- Pawley, 2000; North, 2006; Wolf et al., the absorption spectrum of the acceptor
dering them unusable for quantitative 2007). In this section, I will address some fluorophore, (2) the donor and acceptor
measurements (Russ, 2007). For example, of the additional specific issues surround- fluorophores are within 10 nm or less of
pseudo-coloring, bit-depth conversion, ing these methods, and will refer the one another, and (3) the emission dipole
and some types of image compression interested reader to more thorough treat- of the donor and the absorption dipole of
(e.g., JPEG) can all compromise the in- ments of each subject. the acceptor are orientated in the correct
tensity values in digital images (Table Colocalization. In its simplest position (i.e., not perpendicular) relative
II). Before using any image-processing and least informative form, colocalization to one another (Stryer, 1978; Schaufele et
algorithm for a quantitative study, be analysis is performed by pseudo-coloring al., 2005). In cell biology research, FRET
sure to understand how it affects image and merging two or more fluorescence experiments usually involve tagging two
intensity values. For example, image- images together, and looking for visual molecules of interest with two different
processing software packages refer to cues that the different wavelengths are fluorophores that are capable of FRET.
many different types of algorithms as present in the same pixels; for example, The presence or absence of FRET is then
“deconvolution”, but not all of these al- yellow pixels in a merged image of red used to make conclusions about the prox-
gorithms are appropriate for quantitation and green fluorophores are often used to imity of the two molecules to which the
(see Wallace et al., 2001). In general, conclude that the two fluorophores “co- different fluorophores are attached. There
analysis of pixel intensity values should localize”. These types of qualitative ob- are many excellent reviews and books
be done on raw images stored without servations show, at best, that both available on quantitative FRET micros-
Table III. Information to include in the Materials and methods or figure legend
Follow the journal’s instructions for authors, and/or include the following:
intense focused illumination is used to of the effect of reversibility of photo- for the experiment (Table III; Waters and
photobleach fluorophores in a select re- bleaching on FRAP measurements de- Swedlow, 2007). Even with this informa-
gion of a specimen. If the fluorescently pends on the rate of image acquisition, tion, it may be difficult for the reader to
labeled molecules are mobile, un- and can be controlled by collecting a assess whether your system was opti-
bleached molecules will move into the number of pre-bleach images to reach a mized and operated to obtain the best
Allan, V.J. 2000. Protein Localization by Fluorescence In Handbook of Biological Confocal Micros Murray, J.M. 1998. Evaluating the performance
Microscopy: A Practical Approach. Oxford copy. J.B. Pawley, editor. Springer-Verlag of fluorescence microscopes. J. Microsc.
University, New York. 256 pp. New York, Inc., New York. 347–353. 191:128–134.
Art, J. 2006. Photon detectors for confocal micros- Hiraoka, Y., J.W. Sedat, and D.A. Agard. 1990. Murray, J.M. 2005. Confocal microscopy, deconvo-
copy. In Handbook of Biological Confocal Determination of three-dimensional imag- lution and structured illumination methods.
Microscopy. J.B. Pawley, editor. Springer- ing properties of a light microscope system. In Live Cell Imaging: A Laboratory Manual.
Verlag New York, Inc., New York. 251–262. Partial confocal behavior in epifluorescence R.D. Goldman and D.L. Spector, editors.
Aubin, J.E. 1979. Autofluorescence of viable microscopy. Biophys. J. 57:325–333. Cold Spring Harbor Laboratory, Cold Spring
cultured mammalian cells. J. Histochem. Hoffman, D.B., C.G. Pearson, T.J. Yen, B.J. Howell, Harbor, NY. 239–279.
Cytochem. 27:36–43. and E.D. Salmon. 2001. Microtubule- Murray, J.M., P.L. Appleton, J.R. Swedlow, and
Axelrod, D., N.L. Thompson, and T.P. Burghardt. dependent changes in assembly of micro- J.C. Waters. 2007. Evaluating performance
1983. Total internal inflection fluorescent tubule motor proteins and mitotic spindle in three-dimensional fluorescence micros-
microscopy. J. Microsc. 129:19–28. checkpoint proteins at PtK1 kinetochores. copy. J. Microsc. 228:390–405.
Mol. Biol. Cell. 12:1995–2009. Nolte, A., J. Pawley, and L. Horing. 2006. Non-laser
Bolte, S., and F.P. Cordelieres. 2006. A guided tour
into subcellular colocalization analysis in Huang, B., W. Wang, M. Bates, and X. Zhuang. light sources for three-dimentional micros-
light microscopy. J. Microsc. 224:213–232. 2008. Three-dimensional super-resolution copy. In Handbook of Biological Confocal
imaging by stochastic optical reconstruction Microscopy. J.B. Pawley, editor. Springer-
Cardullo, R.A., and E.H. Hinchcliffe. 2007. Digital microscopy. Science. 319:810–813. Verlag New York, Inc., New York. 126–144.
manipulation of brightfield and fluorescence
images: noise reduction, contrast enhance- Inoue, S. 2006. Foundations of confocal scanned North, A.J. 2006. Seeing is believing? A beginners’
ment, and feature extraction. Methods Cell imaging in light microscopy. In Handbook guide to practical pitfalls in image acquisition.
Biol. 81:285–314. of Biological Confocal Microscopy. J.B. J. Cell Biol. 172:9–18.
Pawley, editor. Springer-Verlag New York,
Chen, H., H.L. Puhl III, S.V. Koushik, S.S. Vogel, Inc., New York. 1–19. Patterson, G.H., and D.W. Piston. 2000. Photo
and S.R. Ikeda. 2006. Measurement of bleaching in two-photon excitation micros-
FRET efficiency and ratio of donor to ac- Inoué, S., and K.R. Spring. 1997. Video Microscopy: copy. Biophys. J. 78:2159–2162.
ceptor concentration in living cells. Biophys. The Fundamentals. Plenum Press, New
York. 770 pp. Pawley, J. 2000. The 39 steps: a cautionary tale of
J. 91:L39–L41. quantitative 3-D fluorescence microscopy.
Churchman, L.S., Z. Okten, R.S. Rock, J.F. Dawson, Jares-Erijman, E.A., and T.M. Jovin. 2003. FRET Biotechniques. 28:884–886, 888.
and J.A. Spudich. 2005. Single molecule imaging. Nat. Biotechnol. 21:1387–1395.
Pawley, J.B. 1994. Sources of noise in 3D micro-
Jares-Erijman, E.A., and T.M. Jovin. 2006. Imaging
Salmon, E.D., and J.C. Canman. 2001. Proper application to molecular motors. Acc. Chem.
alignment and adjustment of the light mi- Res. 38:574–582.
croscope. Curr. Protoc. Cell Biol. Chapter Zwier, J.M., G.J. Van Rooij, J.W. Hofstraat, and
4:Unit 4.1. G.J. Brakenhoff. 2004. Image calibration
Schaufele, F., I. Demarco, and R.N. Day. 2005. in fluorescence microscopy. J. Microsc.
FRET imaging in the widefield microscope. 216:15–24.
In Molecular Imaging: FRET Microscopy
and Spectroscopy. A. Periasamy and R.N.
Day, editors. Oxford University Press.
72–94.
Sekar, R.B., and A. Periasamy. 2003. Fluorescence
resonance energy transfer (FRET) micros-
copy imaging of live cell protein localiza-
tions. J. Cell Biol. 160:629–633.
Shaner, N.C., P.A. Steinbach, and R.Y. Tsien. 2005.
A guide to choosing fluorescent proteins.
Nat. Methods. 2:905–909.
Siegel, R.M., F.K. Chan, D.A. Zacharias, R.
Swofford, K.L. Holmes, R.Y. Tsien, and
M.J. Lenardo. 2000. Measurement of mo-
lecular interactions in living cells by fluor
escence resonance energy transfer between
variants of the green fluorescent protein. Sci.
STKE. 2000:PL1.
Snapp, E.L., N. Altan, and J. Lippincott-Schwartz.
2003. Measuring protein mobility by photo-
bleaching GFP chimeras in living cells. Curr.
Protoc. Cell Biol. Chapter 21:Unit 21.1.
Sprague, B.L., and J.G. McNally. 2005. FRAP analy