Analysis

Quantitative assessment of fluorescent proteins

Paula J Cranfill1,2,5, Brittney R Sell1,5, Michelle A Baird1,5, John R Allen1,5, Zeno Lavagnino2,3,
H Martijn de Gruiter4, Gert-Jan Kremers4, Michael W Davidson1,6, Alessandro Ustione2,3 & David W Piston2,3

The advent of fluorescent proteins (FPs) for genetic labeling of focused on improving maturation of fluorescence at 37 °C
molecules and cells has revolutionized fluorescence microscopy. (refs. 7,12), enhancing protein maturation and stability13,14 and
Genetic manipulations have created a vast array of bright and improving stability in response to chloride, acidity and photoexci-
stable FPs spanning blue to red spectral regions. Common to tation15. Beyond improvement of the photophysical properties of
autofluorescent FPs is their tight b-barrel structure, which FPs, silent mutations have been introduced to optimize expression
provides the rigidity and chemical environment needed for in host organisms16. Finally, almost all FPs are oligomeric (either
© 2016 Nature America, Inc. All rights reserved.

effectual fluorescence. Despite the common structure, each dimeric or tetrameric) in their natural environment, so muta-
FP has unique properties. Thus, there is no single ‘best’ FP tions have been made toward creating monomeric FPs derived
for every circumstance, and each FP has advantages and from jellyfish17 and other species6,18. After more than 20 years
disadvantages. To guide decisions about which FP is right of research, there is still no perfect FP. Each has advantages and
for a given application, we have quantitatively characterized disadvantages, so there is no universally ‘best’ choice. To pro-
the brightness, photostability, pH stability and monomeric vide consistent data for comparison, we characterized more than
properties of more than 40 FPs to enable straightforward and 40 FPs for their brightness, photostability, pH stability and mono­
direct comparison between them. We focus on popular and/or meric properties. We report values for all FPs characterized
top-performing FPs in each spectral region. but focus our discussion on some of the more popular and/or
best-performing FPs in various spectral regions.
Live-cell fluorescence microscopy offers a powerful way to exam-
ine tissues, cells and subcellular components at functionally RESULTS
important scales of time and length1. The cloning of GFP from The FP color palette
the jellyfish Aequorea victoria2 and its expression in non-jellyfish A major advantage of fluorescence is the ability to resolve multiple
systems3 empowered fluorescence microscopy in cell biology and labels by color. FP variants span the visible spectrum, broadly
physiology. A major leap in FP technology came with the dis- categorized into blue, cyan, green, yellow, orange, red and far red.
covery of homologous FPs from corals and other anthozoans4. Within a given color range, the simplest delineation is bright-
Genetic manipulations performed on A. victoria GFP5 as well as
npg

ness, defined as the combination of light absorbance (extinction

non-jellyfish proteins6 have led to FPs covering spectral regions coefficient) and fluorescence quantum yield (ratio of fluorescent
from blue to red with improved brightness and stability5,7,8. photons per absorbed photon). To compare brightness, we puri-
Detailed discussions of FP mutations are found elsewhere1,5,8. fied FPs and measured their extinction coefficients, quantum
Some general properties of FPs are critical for understanding yields and spectral properties7,19. Enhanced GFP (EGFP) was the
their function and utility. FPs are ~25 kD, which is large compared first robust and reliable FP12, and it remains the ‘gold standard’
to organic fluorophores, and the entire protein structure appears for FPs. Excitation (Fig. 1a) and fluorescence (Fig. 1b) spectra
to be essential to its fluorescence9. FP structures consist of 11 of FPs can be normalized to EGFP to quantify relative absorb-
β-strands surrounding a central α-helix10, where the fluorophore ance and brightness. Beyond its excitation and emission maxima
arises from a few amino acids. During protein biogenesis, a reac- and brightness, an FP’s specific spectral shape is also impor-
tion occurs among FP residues to form an extended conjugation5, tant; for example, the excitation peaks for mKate2 and mCar-
and fluorescence depends on rigidity and chemical environment dinal (Fig. 1a) are separated by 15 nm, but the broader spectral
within the protein structure11. Within the β-barrel fold, mutations width of mCardinal gives substantial absorption at ~635 nm
can change the local fluorophore environment and lead to vari- (a typical laser line), whereas the absorption of mKate2 at 635 nm
ations in spectral characteristics, photostability, acid resistance is negligible. Thus, knowledge of spectral shapes is important
and other physical properties. Mutational work on FPs initially for choosing an FP.

1National High Field Magnet Lab, Florida State University, Tallahassee, Florida, USA. 2Department of Molecular Physiology and Biophysics, Vanderbilt University,

Nashville, Tennessee, USA. 3Department of Cell Biology and Physiology, Washington University, St. Louis, Missouri, USA. 4Erasmus Optical Imaging Centre, Josephine
Nefkens Institute, Erasmus MC, Rotterdam, the Netherlands. 5Present addresses: H. Lee Moffitt Cancer Center, Tampa, Florida, USA (P.J.C.); Department of Biochemistry
and Molecular Medicine, The George Washington University, Washington, DC, USA (B.R.S.); US National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland,
USA (M.A.B.); Nikon Instruments, Melville, New York, USA (J.R.A.). 6Deceased. Correspondence should be addressed to D.W.P. (piston@wustl.edu).
Received 18 March; accepted 7 May; published online 30 May 2016; doi:10.1038/nmeth.3891

nature methods | VOL.13 NO.7 | JULY 2016 | 557

 Analysis
Figure 1 | Excitation and emission spectra of several commonly used FPs a
across the visible region. (a,b) Excitation (a) and emission (b) spectra 2.0
for various FPs. Peak values are normalized to EGFP = 1 by the extinction
coefficient (a) or to EGFP = 1 by their relative brightness (extinction 1.5

Excitation
coefficient × quantum yield) (b). All spectra were acquired with
at least three separate preparations of the purified FPs, and spectral 1.0
standard deviations were all <1%.
0.5

0
FP brightness peaks in the middle of the visible spectrum 300 350 400 450 500 550 600 650
with the yellow and orange FPs (mVenus and mKO in Fig. 1), as Wavelength (nm)
follows from general fluorophore properties. Optical transi-
tion energy depends on charge separation in the fluorophore, b
2.5 TagBFP2
with larger separations leading to lower energy transitions. Blue mTurquoise2
2.0 EGFP
fluorophores (higher energy) are necessarily smaller than red mVenus

Emission
mKO
fluorophores and thus have smaller extinction coefficients. 1.5 mApple
mCherry
However, quantum yield is usually inversely correlated with fluoro­ 1.0 mKate2
phore size. High quantum yield depends on fluorophore rigidity; mCardinal

larger molecules have more degrees of freedom and are thus less 0.5

rigid. Therefore, despite the higher absorption of red FPs, their 0

brightness is reduced by lower quantum yields. The extinction 400 500 600 700 800
© 2016 Nature America, Inc. All rights reserved.

coefficients, quantum yields, relative brightness and pKa values of Wavelength (nm)

the FPs studied here are listed in Supplementary Table 1.

comparison between microscopies (Supplementary Table 2).
Photobleaching Under identical conditions (80-µW laser illumination), we meas-
A major consideration in fluorescence microscopy is fluorophore ured bleaching half-times ranging from 530 s (for mCardinal) to
photostability. We measured photobleaching rates of FP variants 2.7 s (for DsRed2). We also used different illumination intensities
embedded in polyacrylamide20 or microdroplets19,21 to enable and modalities (wide-field metal halide lamp or LED and laser
scanning). We verified that fluorescence intensity is a linear func-
tion of excitation power, and we expected, on the basis of 40 years
a 1
mCitrine = 0.93 of literature22,23, that photobleaching rates would also be linear
0.9 mNeonGreen = 1.34
mEGFP = 1.29
(i.e., at twice the intensity, the photobleaching rate doubles).
0.8 EGFP = 1.23 Instead, we found that almost all FPs show supralinear (‘acceler-
mTurquoise = 1.89
0.7
ated’) photobleaching (Fig. 2). This acceleration of photobleaching
is similar to what is measured under two-photon excitation21.
0.6
We could fit these data to log(F) = −αlog(P) + c, which yields a
log(t1/2) (s)

1.0
0.5 photobleaching rate of kbleach = bIα, where F is the fluorescence
0.4 intensity, P is the illumination power and b and c are constants.
The slope of the log–log plot gives the value of α, and these values
npg

0.9
0.3
are listed in Supplementary Table 2 for each modality.
0.2
0.8
Accelerated FP photobleaching has significant implications,
0.1 1.60 1.70 1.80 as even a small α can make a big difference in the total pho-
0 tobleaching, depending on the imaging method. Wide-field
1.60 1.80 2.00 2.20 2.40 2.60
microscopy uses low power spread over every pixel for the full
log(power) (uW)
duration of the image, but laser-scanning confocal microscopy
b 1 focuses all of the light on a single pixel for a short period of time
0.9
mKate2 = 0.98
mCherry = 1.38
and then raster scans the pixels. Using identical total power for
mApple = 1.14 wide-field and confocal imaging, the instantaneous power (i.e.,
mKO2 = 1.53
0.8
intensity) is much higher in laser scanning. For a 1-s 512 × 512
0.7 image, the laser-scanning intensity is 250,000-fold higher than
0.6
log(t1/2) (s)

1.0
0.5 Figure 2 | Log–log plot of photobleaching rates versus illumination
power as measured by laser-scanning microscopy. (a) Excitation intensity
0.4
dependence for a few common cyan, green and yellow FPs. (b) Behavior
0.9
0.3 of orange and red FPs. Similar power dependence is seen with wide-field
0.2
lamp or LED illumination (Supplementary Table 2). Data are plotted
0.8 as mean ± s.d. (Supplementary Note). For each FP, the fit value of
0.1 1.60 1.70 1.80 the exponent α is given on the legend. All photobleaching experiments
0 were repeated a minimum of 20 times with at least 10 separate FP
1.60 1.80 2.00 2.20 2.40 2.60 purifications. Photobleaching rates were fit as described in the text;
log(power) (uW) means ± s.d. are reported in Supplementary Table 2.

558 | VOL.13 NO.7 | JULY 2016 | nature methods

 Analysis
Figure 3 | Wide-field fluorescence images of
FP–CytERM fusion proteins expressed in live
a b c d
cells. (a–l) CytERM-fused mEGFP (A206K) (a),
mCherry (b), mCitrine (A206K) (c), mOrange2
(d), mNeonGreen (e), mKate2 (f), EGFP (A206)
(g), mKO2 (h), mTagBFP2 (i), mTagRFP-T (j),
Citrine (A206) (k) and DsRed2 (l) expressed in
HeLa cells. Scale bars, 10 µm. All images are
representative of at least 7,000 cells analyzed
for each FP.
e f g h

in wide-field illumination. For α = 1.07

(monomeric EGFP (mEGFP)), the laser-
scanning bleaching rate (kbleach) is propor-
tional to 250,0001.07 ≈ 596,750, but each
pixel is illuminated for only 1/250,000 of
i j k l
the imaging time, which yields bleaching
(kbleach × time) ~2.4-fold greater than in
wide-field microscopy. Photobleaching
differences between laser-scanning and
wide-field illumination grow rapidly as a
© 2016 Nature America, Inc. All rights reserved.

function of the exponent α: α = 1.2 leads

to >10-fold faster photobleaching in laser
scanning, and α = 1.35 leads to >100-fold
faster bleaching. Despite the disproportionate effects of accel- vary as a function of the incubation medium26. Similar differ-
erated photobleaching on laser-scanning approaches, confocal ences can be expected for FPs in various intracellular compart-
imaging is superior for many studies of intact tissue because of its ments, but slowly photobleaching FPs are generally more slowly
3D signal discrimination24, and many long-term confocal imaging photobleached in all compartments. Photobleaching can also be
studies have been successful with FPs. Accelerated photobleach- exacerbated during multicolor imaging, as some FPs photobleach
ing is not as bad for spinning-disk approaches (where multiple faster in the presence of additional illumination wavelengths,
spots are imaged simultaneously, thus lowering the intensity at even when those additional wavelengths would not excite the FP
each spot), but photobleaching in these instruments can still be on their own27.
several-fold faster than in wide-field imaging25. The mechanism Several consequences arise from this accelerated photobleach-
underlying accelerated photobleaching remains unknown, but we ing phenomenon. First, it is difficult to choose between FPs on
hypothesize that additional photons interacting with the excited- the basis of any single photobleaching rate. An FP that bleaches
state fluorophore increase the photobleaching probability. too fast at low excitation intensity may be preferable for higher-
The data described above (Fig. 2 and Supplementary Table 2) intensity experiments if it demonstrates minimally accelerated
were acquired in vitro, but accelerated photobleaching of FPs also photobleaching. Second, the most common approach for time-
npg

occurred in cells. We measured laser-scanning confocal bleach- lapse, 3D-resolved measurements is confocal microscopy, but the
ing rates of the widely used FPs mCerulean (α = 1.12), mEGFP instruments used are susceptible to problems arising from accel-
(α = 1.29), mVenus (α = 1.13), and mCherry (α = 1.38) at 60 µW erated photobleaching. Several alternatives are less susceptible to
and 240 µW. Free FPs were expressed and bleaching was measured photobleaching given supralinear bleaching. These alternatives
over entire cells (>100 cells per condition). Linear photobleach­ do not require scanned illumination or a pinhole for detection,
ing would yield fourfold faster bleaching at 240 µW than at and they include total internal reflection (TIRF), selective plane
60 µW, so deviations from the factor of four yield a value for α. illumination (SPIM), deconvolution and structured illumination
For mCerulean, the bleaching half-time was 108 s with 60 µW and microscopies. Some of these approaches do not create out-of-focus
24 s with 240 µW, an apparent α = 1.09. For mEGFP, bleaching background (TIRF and SPIM), whereas others (deconvolution and
half-times were 239 s at 60 µW and 46 s at 240 µW (α = 1.19); structured illumination) utilize the out-of-focus light to generate
for mVenus, 58 s at 60 µW and 12 s at 240 µW (α = 1.14); for a more complete view of a 3D data volume. TIRF is ideal for cell
mCherry, 348 s at 60 µW and 49 s at 240 µW (α = 1.41), all of imaging28, particularly for events involved in apoptosis, but can
which were consistent with our in vitro measurements. Absolute measure things only within ~100 nm of the surface. Both decon-
photobleaching rates and accelerated photobleaching appear to be volution29 and structured illumination30 microscopies can achieve
independent. In cells, mCherry was bleached ~50% more slowly high lateral and axial resolution (100 nm in x and y and <500 nm
than mEGFP with 60 µW, but their photobleaching rates were in z, respectively), but for scattering samples, such as cell clus-
nearly identical at 240 µW. ters, these approaches lose effectiveness at <10 µm into the sam-
Photobleaching is a complex phenomenon that depends on ple31,32. SPIM permits imaging deep into samples with minimal
many factors, especially oxidizing or reducing environments, photobleaching33. These alternatives are widely used in cell biology
but accelerated photobleaching appears to be independent of the research, and with the potential severity of accelerated photo­
local environment21. However, photobleaching rates in cells can bleaching, even more researchers should consider using them.

nature methods | VOL.13 NO.7 | JULY 2016 | 559

 Analysis
Monomeric quality Table 1 | Percentage of cells scored without visible OSER whorls as a
Virtually all FPs are oligomeric (either dimeric or tetrameric) function of FP–CytERM fusion protein, ranked from most monomeric
in their natural environment. Wild-type A. victoria GFP is part (100%) to most strongly oligomeric (0%)
of a heterotetrameric complex with aequorin34, whereas most FP Normal cells (%) s.d. (%)
coral and anemone FPs occur naturally as tetramers6. To prevent mEGFP (L221K) 98.8 1.2
oligomerization, mutations have been developed to eliminate mEGFP (A206K) 98.1 1.6
dimerization of jellyfish-derived FPs17, and though elimination mEmerald (A206K) 96.6 1.1
of tetramers in coral FPs has proven more difficult, substantial mRFP1 95.8 1.1
progress has been made. For applications such as cell-type iden- mT-Sapphire (A206K) 95.5 0.6
mApple 95.3 1.7
tification, monomeric proteins are not required, so a bright FP mPapaya 95.1 1.1
of any color can be used. Monomerized FPs work well in most mCherry 95.0 0.8
applications, even though some can form low-affinity oligo- CyPet 94.0 2.4
mers when expressed at high levels or constrained within subcel- mKate2.5 93.9 1.7
lular compartments. To evaluate FP oligomeric tendencies, we mCitrine (A206K) 93.8 2.6
used FPs fused onto an endoplasmic reticulum (ER) membrane mTurquoise2 (A206K) 93.8 1.0
mTurquoise (A206K) 93.3 1.2
protein (CytERM)35. These ER-associated FP-fusion proteins
mRuby 93.1 2.1
can be expressed at high effective concentrations, where homo- mCerulean (A206K) 92.6 2.0
oligomerization can lead to cross-linking between FPs and recon- mTFP1 92.0 1.5
figuration of ER from a tubular network into an organized smooth mTopaz (A206K) 91.9 2.0
ER (OSER) whorl structure. We evaluated live cells to eliminate mOrange2 91.8 1.4
© 2016 Nature America, Inc. All rights reserved.

gross overexpression artifacts and then conducted an OSER assay FusionRed 91.5 3.0
of different FPs (Fig. 3). Approximately 10,000 cells were ana- mPlum 91.5 2.4
mOrange 91.4 1.7
lyzed for each FP (using 7,000–20,000 cells), and the percentage mCerulean3 (A206K) 91.0 1.8
of observed cells exhibiting OSER whorls was determined over at mStrawberry 90.6 2.5
least 15 preparations by three individual scorers. mClover (A206K) 90.5 4.5
We performed the OSER assay on 63 FPs, including known oli- mNeonGreen 90.4 2.1
gomeric FPs. Many FPs appeared strongly monomeric, although mEos3d.2G16 90.4 2.4
others that have been previously reported as monomeric were mAzamiGreen 90.3 4.7
mAmetrine 90.0 4.5
found to oligomerize (Table 1). The two monomeric forms of
mRaspberry 89.6 4.0
EGFP (mEGFP containing either the A206K or the L221K muta- mPapaya2 87.6 3.2
tion)17 were the best-performing FPs in the OSER assay. There is a mRuby2 87.4 5.8
continuum of monomeric quality, but we suggest that FPs scoring mWasabi 86.4 2.5
~90% or higher can be reasonably assumed to be monomeric. We mVenus (A206K) 83.9 4.4
expect these FPs to perform well in most fusion constructs, but mKate2 81.1 6.1
Cerulean (A206) 78.3 4.8
their behavior should be evaluated in each specific construct and
EGFP (A206) 76.5 6.9
cell type. Many widely used and promising FPs had scores below Clover (A206) 72.9 5.5
90%. In the same way that a high score in the OSER assay does not mKO2 68.4 4.2
npg

mean an FP will always perform well, a lower score does not mean Topaz (A206) 63.8 8.0
that an FP can never be used. We have had success using mKO2 tdTurboRFP 62.5 9.4
and mCardinal in protein fusions, although each has also failed in YPet (A206) 62.5 5.2
some circumstances. Such results would be consistent with their Superfolder GFP (A206V) 62.4 6.3
Emerald (A206) 58.9 5.5
OSER assay scores of 68% and 41%, respectively.
mTagRFP 57.7 7.5
Beyond brightness, pH stability, photostability and mono- tdTomato 57.6 4.9
meric quality, other factors can affect FP performance in specific EBFP2 (A206) 57.0 2.7
situations, but these are difficult to assess across all FPs. Some mKate 54.5 4.8
experiments require a fast maturation rate or high maturation mEos3dG16 53.3 7.8
efficiency of the FP fluorophore, for which some FPs are optimal mTagBFP2 49.8 1.9
mKO 46.9 6.1
(for example, sfGFP14 and mVenus36). However, this property can
MiCy 41.7 10.7
be cell-type dependent and is difficult to recapitulate accurately mCardinal 41.3 3.6
in vitro. Another property that is difficult to assess broadly is mTagRFP-T 41.2 4.0
protein stability. Some FPs have been designed to degrade37 or mNeptune 40.5 3.8
change color38, but protein stability and color switching can be Venus (A206) 36.5 3.5
confounding factors. The photoactivation properties intrinsic to Citrine (A206) 36.2 3.9
the original GFP7 have been leveraged for lineage-tracing and tdKatushka 30.6 5.9
acGFP 29.5 12.4
super-resolution microscopy, but many FPs show photoswitching TurboRFP 0 0
properties that complicate experimental results39,40. Not every FP dKatushka 0 0
exhibits photoswitching, and those that do have properties that DsRed2 0 0
can depend on both excitation intensity and pH40. Interference zsYellow 0 0
is also another consideration; fidelity of any FP in fusions relies AsRed 0 0

560 | VOL.13 NO.7 | JULY 2016 | nature methods

 Analysis
on more than oligomerization, as FPs can interfere with proper improvements are likely for orange and red FPs and for far-red
trafficking of the target protein. Such interference is often specific FPs especially, which are less bright because of their low quantum
to an FP and cell type. Finally, FPs can have different expres- yields. Because tissue absorption of light and autofluorescence
sion levels under identical conditions41. These differences might are minimal in the red spectral region, these FPs are particu-
reflect mRNA stability, translation efficiency or FP stability and larly important for animal and tissue imaging. Current FP devel-
can be critical for FP use in transgenic animals. opment is focused on the discovery and creation of improved
red and far-red FPs.
DISCUSSION
As demonstrated by these data, there is no single best FP for every Methods
circumstance. It is notable that even though EGFP was the first Methods and any associated references are available in the online
generally useful FP, mEGFP remains one of the best and most version of the paper.
versatile FPs available. Considering how the growth of FP use
Note: Any Supplementary Information and Source Data files are available in the
depended on this early variant, it seems fortuitous that it is one online version of the paper.
of the best performers. When choosing an FP, it is important to
note that every FP measured here (as well as previous variants Acknowledgments
This work was supported in part by NIH grants R01DK085064, R01DK098659,
that were dimmer and less photostable) can be used effectively S10OD010681 and P20GM072048 to D.W.P.
in many experiments. For this reason, it may not be worth the
effort to switch to a better-performing FP if the one currently AUTHOR CONTRIBUTIONS
in use is yielding reliable data. For new projects, however, our D.W.P., A.U. and M.W.D. designed the research; P.J.C., B.R.S., M.A.B., J.R.A., Z.L.,
H.M.d.G. and G.-J.K. performed experiments and analyzed data; D.W.P. wrote and
comprehensive data set provides a guide to which FPs are likely edited the paper with comments from all authors.
© 2016 Nature America, Inc. All rights reserved.

to be the best choice. On the basis of overall performance, mTag-

BFP2 and mTurquoise2 were the best of the blue and cyan FPs, COMPETING FINANCIAL INTERESTS
The authors declare no competing financial interests.
respectively; however, unless these colors are specifically needed,
it is usually better to use more red-shifted FPs to avoid the greater Reprints and permissions information is available online at http://www.nature.
autofluorescence in these spectral regions. In particular, blue FPs com/reprints/index.html.
have proven challenging and should be used with caution. In the 1. Goldman, R.D., Swedlow, J. & Spector, D.L. Live Cell Imaging: A Laboratory
green region, EGFP was the first generally reliable FP, and because Manual 2nd edn. (Cold Spring Harbor Laboratory Press, 2010).
of its combination of positive attributes, mEGFP remains the gold 2. Prasher, D.C., Eckenrode, V.K., Ward, W.W., Prendergast, F.G. & Cormier,
M.J. Primary structure of the Aequorea victoria green-fluorescent protein.
standard to which the performance of other FPs is compared. For Gene 111, 229–233 (1992).
the yellow spectrum, mVenus and mCitrine have proven reliable, 3. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W. & Prasher, D.C. Green fluorescent
but because mCitrine is more strongly monomeric, it is likely to protein as a marker for gene expression. Science 263, 802–805 (1994).
4. Matz, M.V. et al. Fluorescent proteins from nonbioluminescent Anthozoa
be better for the most demanding situations. species. Nat. Biotechnol. 17, 969–973 (1999).
The orange FPs are the brightest owing to their combination of 5. Tsien, R.Y. The green fluorescent protein. Annu. Rev. Biochem. 67,
large extinction coefficients and relatively high quantum yields, 509–544 (1998).
but choosing an orange FP is not straightforward. mKO and 6. Shaner, N.C. et al. Improved monomeric red, orange and yellow fluorescent
proteins derived from Discosoma sp. red fluorescent protein. Nat.
mKO2, the brightest orange FPs, show a propensity to oligomer- Biotechnol. 22, 1567–1572 (2004).
ize and are prone to accelerated photobleaching. mOrange2 is not 7. Patterson, G.H., Knobel, S.M., Sharif, W.D., Kain, S.R. & Piston, D.W. Use
npg

as bright as the other orange FPs, but it shows excellent photosta- of the green fluorescent protein and its mutants in quantitative
fluorescence microscopy. Biophys. J. 73, 2782–2790 (1997).
bility and is strongly monomeric and may therefore be the best
8. Kremers, G.J., Gilbert, S.G., Cranfill, P.J., Davidson, M.W. & Piston, D.W.
choice, even though it also exhibits accelerated photobleaching. Fluorescent proteins at a glance. J. Cell Sci. 124, 157–160 (2011).
In the red spectral region, mCherry is widely used for many 9. Shu, X., Shaner, N.C., Yarbrough, C.A., Tsien, R.Y. & Remington, S.J. Novel
applications but has been reported to aggregate when expressed in chromophores and buried charges control color in mFruits. Biochemistry
45, 9639–9647 (2006).
some fusions42, despite its excellent performance as a monomer in 10. Ormö, M. et al. Crystal structure of the Aequorea victoria green fluorescent
the OSER assay. mApple can be used effectively as a brighter sub- protein. Science 273, 1392–1395 (1996).
stitute for mCherry in most cases. mApple is strongly monomeric 11. Follenius-Wund, A. et al. Fluorescent derivatives of the GFP chromophore
and shows low accelerated photobleaching. However, the mApple give a new insight into the GFP fluorescence process. Biophys. J. 85,
1839–1850 (2003).
emission is blue-shifted from that of mCherry by approximately 12. Cormack, B.P., Valdivia, R.H. & Falkow, S. FACS-optimized mutants of the
18 nm, which increases spectral overlap with the orange FP vari- green fluorescent protein (GFP). Gene 173, 33–38 (1996).
ants. In the far-red region, mKate2 is currently the brightest and 13. Cubitt, A.B. et al. Understanding, improving and using green fluorescent
proteins. Trends Biochem. Sci. 20, 448–455 (1995).
most monomeric FP. However, mCardinal can be a better choice 14. Pédelacq, J.D., Cabantous, S., Tran, T., Terwilliger, T.C. & Waldo, G.S.
because its red-shifted excitation profile allows it to be used with Engineering and characterization of a superfolder green fluorescent
635-nm (or Cy5 cube) excitation, which is unique among cur- protein. Nat. Biotechnol. 24, 79–88 (2006).
rent autofluorescent FPs. mCardinal is photostable, but it shows 15. Griesbeck, O., Baird, G.S., Campbell, R.E., Zacharias, D.A. & Tsien, R.Y.
Reducing the environmental sensitivity of yellow fluorescent protein.
accelerated photobleaching and is not strongly monomeric. Mechanism and applications. J. Biol. Chem. 276, 29188–29194 (2001).
Green and yellow FPs have reached near-maximal perform- 16. Yang, T.T. et al. Improved fluorescence and dual color detection with
ance, as many have very high quantum yields and are strongly enhanced blue and green variants of the green fluorescent protein.
J. Biol. Chem. 273, 8212–8216 (1998).
monomeric. Some improvements in photostability may still be
17. Zacharias, D.A., Violin, J.D., Newton, A.C. & Tsien, R.Y. Partitioning of
forthcoming, but many of these FPs are sufficiently photostable lipid-modified monomeric GFPs into membrane microdomains of live cells.
for long-term live-cell imaging. Looking forward, significant Science 296, 913–916 (2002).

nature methods | VOL.13 NO.7 | JULY 2016 | 561

 Analysis
18. Shaner, N.C. et al. Improving the photostability of bright monomeric 31. Hiraoka, Y., Sedat, J.W. & Agard, D.A. Determination of three-dimensional
orange and red fluorescent proteins. Nat. Methods 5, 545–551 (2008). imaging properties of a light microscope system. Partial confocal behavior
19. Kremers, G.-J. & Piston, D. Photoconversion of purified fluorescent in epifluorescence microscopy. Biophys. J. 57, 325–333 (1990).
proteins and dual-probe optical highlighting in live cells. J. Vis. Exp. 40, 32. Carlton, P.M. et al. Fast live simultaneous multiwavelength four-
1995 (2010). dimensional optical microscopy. Proc. Natl. Acad. Sci. USA 107,
20. Piston, D.W., Patterson, G.H. & Knobel, S.M. Quantitative imaging of the 16016–16022 (2010).
green fluorescent protein (GFP). Methods Cell Biol. 58, 31–48 (1999). 33. Tomer, R., Khairy, K. & Keller, P.J. Light sheet microscopy in cell biology.
21. Patterson, G.H. & Piston, D.W. Photobleaching in two-photon excitation Methods Mol. Biol. 931, 123–137 (2013).
microscopy. Biophys. J. 78, 2159–2162 (2000). 34. Ward, W.W. Biochemical and physical properties of green fluorescent
22. Hirschfeld, T. Quantum efficiency independence of the time integrated protein. Methods Biochem. Anal. 47, 39–65 (2006).
emission from a fluorescent molecule. Appl. Opt. 15, 3135–3139 (1976). 35. Costantini, L.M., Fossati, M., Francolini, M. & Snapp, E.L. Assessing the
23. Song, L., Hennink, E.J., Young, I.T. & Tanke, H.J. Photobleaching kinetics tendency of fluorescent proteins to oligomerize under physiologic
of fluorescein in quantitative fluorescence microscopy. Biophys. J. 68, conditions. Traffic 13, 643–649 (2012).
2588–2600 (1995). 36. Nagai, T. et al. A variant of yellow fluorescent protein with fast and
24. Sandison, D.R., Piston, D.W., Williams, R.M. & Webb, W.W. Quantitative efficient maturation for cell-biological applications. Nat. Biotechnol. 20,
comparison of background rejection, signal-to-noise ratio, and resolution 87–90 (2002).
in confocal and full-field laser scanning microscopes. Appl. Opt. 34, 37. Li, X. et al. Generation of destabilized green fluorescent protein as a
3576–3588 (1995). transcription reporter. J. Biol. Chem. 273, 34970–34975 (1998).
25. Wu, Y. et al. Spatially isotropic four-dimensional imaging with dual-view 38. Terskikh, A. et al. “Fluorescent timer”: protein that changes color with
plane illumination microscopy. Nat. Biotechnol. 31, 1032–1038 (2013). time. Science 290, 1585–1588 (2000).
26. Bogdanov, A.M. et al. Cell culture medium affects GFP photostability: a 39. Valentin, G. et al. Photoconversion of YFP into a CFP-like species
solution. Nat. Methods 6, 859–860 (2009). during acceptor photobleaching FRET experiments. Nat. Methods 2, 801
27. Dempsey, W.P. et al. In vivo single-cell labeling by confined primed (2005).
conversion. Nat. Methods 12, 645–648 (2015). 40. Kremers, G.J., Hazelwood, K.L., Murphy, C.S., Davidson, M.W. & Piston,
28. Axelrod, D. Chapter 7: Total internal reflection fluorescence microscopy. D.W. Photoconversion in orange and red fluorescent proteins. Nat. Methods
© 2016 Nature America, Inc. All rights reserved.

Methods Cell Biol. 89, 169–221 (2008). 6, 355–358 (2009).

29. Swedlow, J.R., Sedat, J.W. & Agard, D.A. Multiple chromosomal 41. Chen, S.X. et al. Quantification of factors influencing fluorescent protein
populations of topoisomerase II detected in vivo by time-lapse, three- expression using RMCE to generate an allelic series in the ROSA26 locus in
dimensional wide-field microscopy. Cell 73, 97–108 (1993). mice. Dis. Model. Mech. 4, 537–547 (2011).
30. Gustafsson, M.G. et al. Three-dimensional resolution doubling in wide- 42. Katayama, H., Yamamoto, A., Mizushima, N., Yoshimori, T. & Miyawaki, A.
field fluorescence microscopy by structured illumination. Biophys. J. 94, GFP-like proteins stably accumulate in lysosomes. Cell Struct. Funct. 33,
4957–4970 (2008). 1–12 (2008).
npg

562 | VOL.13 NO.7 | JULY 2016 | nature methods

 ONLINE METHODS 49. Ai, H.W., Henderson, J.N., Remington, S.J. & Campbell, R.E. Directed
Expression, purification and characterization of FPs. The evolution of a monomeric, bright and photostable version of Clavularia
cyan fluorescent protein: structural characterization and applications in
methods used in this work were as described4,6,12,14,15,17,18,36,43–65. fluorescence imaging. Biochem. J. 400, 531–540 (2006).
Detailed protocols are presented in the Supplementary Note. 50. Zapata-Hommer, O. & Griesbeck, O. Efficiently folding and circularly
These protocols cover the expression and purification of FPs, permuted variants of the Sapphire mutant of GFP. BMC Biotechnol. 3, 5
(2003).
including the use of SDS–PAGE gels, protein dialysis, and the 51. Cubitt, A.B., Woollenweber, L.A. & Heim, R. Understanding structure-
BCA protein assays needed to verify and quantify DNA and pro- function relationships in the Aequorea victoria green fluorescent protein.
tein levels. The protocols developed to assess purified FPs for their Methods Cell Biol. 58, 19–30 (1999).
photophysical properties (extinction coefficient, quantum yield, 52. Nguyen, A.W. & Daugherty, P.S. Evolutionary optimization of
fluorescent proteins for intracellular FRET. Nat. Biotechnol. 23, 355–360
and fluorescence pKa) are also described. (2005).
53. Lam, A.J. et al. Improving FRET dynamic range with bright green and red
FP photobleaching and OSER assays. Protocols for embedding fluorescent proteins. Nat. Methods 9, 1005–1012 (2012).
54. Shaner, N.C. et al. A bright monomeric green fluorescent protein
the purified FPs in polyacrylamide and measuring their photo­
derived from Branchiostoma lanceolatum. Nat. Methods 10, 407–409
bleaching in wide-field and confocal microscopy are described (2013).
in the Supplementary Note. Protocols for use of the established 55. Karasawa, S., Araki, T., Nagai, T., Mizuno, H. & Miyawaki, A. Cyan-
cytoplasmic end of ER signal anchor membrane protein (CytERM) emitting and orange-emitting fluorescent proteins as a donor/acceptor
pair for fluorescence resonance energy transfer. Biochem. J. 381, 307–312
oligomerization assay are also described in the Supplementary (2004).
Note. 56. Sakaue-Sawano, A. et al. Visualizing spatiotemporal dynamics of
multicellular cell-cycle progression. Cell 132, 487–498 (2008).
Statistical analysis. Data were collected in replicates as described 57. Merzlyak, E.M. et al. Bright monomeric red fluorescent protein
© 2016 Nature America, Inc. All rights reserved.

with an extended fluorescence lifetime. Nat. Methods 4, 555–557

in the Supplementary Note and fit to the equations given in the (2007).
text. Results are presented as mean ± s.d. 58. Haas, J., Park, E.C. & Seed, B. Codon usage limitation in the expression
of HIV-1 envelope glycoprotein. Curr. Biol. 6, 315–324 (1996).
59. Kredel, S. et al. mRuby, a bright monomeric red fluorescent protein for
43. Ai, H.W., Shaner, N.C., Cheng, Z., Tsien, R.Y. & Campbell, R.E. Exploration labeling of subcellular structures. PLoS One 4, e4391 (2009).
of new chromophore structures leads to the identification of improved 60. Campbell, R.E. et al. A monomeric red fluorescent protein. Proc. Natl.
blue fluorescent proteins. Biochemistry 46, 5904–5910 (2007). Acad. Sci. USA 99, 7877–7882 (2002).
44. Subach, O.M., Cranfill, P.J., Davidson, M.W. & Verkhusha, V.V. An enhanced 61. Shemiakina, I.I. et al. A monomeric red fluorescent protein with low
monomeric blue fluorescent protein with the high chemical stability of cytotoxicity. Nat. Commun. 3, 1204 (2012).
the chromophore. PLoS One 6, e28674 (2011). 62. Shcherbo, D. et al. Far-red fluorescent tags for protein imaging in living
45. Goedhart, J. et al. Bright cyan fluorescent protein variants identified by tissues. Biochem. J. 418, 567–574 (2009).
fluorescence lifetime screening. Nat. Methods 7, 137–139 (2010). 63. Lin, M.Z. et al. Autofluorescent proteins with excitation in the optical
46. Goedhart, J. et al. Structure-guided evolution of cyan fluorescent proteins window for intravital imaging in mammals. Chem. Biol. 16, 1169–1179
towards a quantum yield of 93%. Nat. Commun. 3, 751 (2012). (2009).
47. Rizzo, M.A., Springer, G.H., Granada, B. & Piston, D.W. An improved cyan 64. Chu, J. et al. Non-invasive intravital imaging of cellular differentiation
fluorescent protein variant useful for FRET. Nat. Biotechnol. 22, 445–449 with a bright red-excitable fluorescent protein. Nat. Methods 11, 572–578
(2004). (2014).
48. Markwardt, M.L. et al. An improved cerulean fluorescent protein with 65. Wang, L., Jackson, W.C., Steinbach, P.A. & Tsien, R.Y. Evolution of new
enhanced brightness and reduced reversible photoswitching. PLoS One 6, nonantibody proteins via iterative somatic hypermutation. Proc. Natl.
e17896 (2011). Acad. Sci. USA 101, 16745–16749 (2004).
npg

doi:10.1038/nmeth.3891 nature methods