Improved Blue, Green, and Red Fluorescent Protein
Tagging Vectors for S. cerevisiae
Sidae Lee1,2, Wendell A. Lim1,2,3,4, Kurt S. Thorn5*
1UCSF Center for Systems and Synthetic Biology, University of California San Francisco, San Francisco, California, United States of America, 2Department of Cellular and
Molecular Pharmacology, University of California San Francisco, San Francisco, California, United States of America, 3Howard Hughes Medical Institute, University of
California San Francisco, San Francisco, California, United States of America, 4California Institute for Quantitative Biomedical Research, San Francisco, California, United
States of America, 5Department of Biochemistry and Biophysics, University of California San Francisco, San Francisco, California, United States of America
Fluorescent protein fusions are a powerful tool to monitor the localization and trafficking of proteins. Such studies are
particularly easy to carry out in the budding yeast Saccharomyces cerevisiae due to the ease with which tags can be
introduced into the genome by homologous recombination. However, the available yeast tagging plasmids have not kept
pace with the development of new and improved fluorescent proteins. Here, we have constructed yeast optimized versions
of 19 different fluorescent proteins and tested them for use as fusion tags in yeast. These include two blue, seven green, and
seven red fluorescent proteins, which we have assessed for brightness, photostability and perturbation of tagged proteins.
We find that EGFP remains the best performing green fluorescent protein, that TagRFP-T and mRuby2 outperform mCherry
as red fluorescent proteins, and that mTagBFP2 can be used as a blue fluorescent protein tag. Together, the new tagging
vectors we have constructed provide improved blue and red fluorescent proteins for yeast tagging and three color imaging.
Citation: Lee S, Lim WA, Thorn KS (2013) Improved Blue, Green, and Red Fluorescent Protein Tagging Vectors for S. cerevisiae. PLoS ONE 8(7): e67902.
Editor: Amy S. Gladfelter, Dartmouth College, United States of America
Received April 26, 2013; Accepted May 22, 2013; Published July 2, 2013
Copyright: ? 2013 Lee et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funding for this project was provided by an NIGMS Systems Biology Center grant, number P50 GM081879. The funders had no role in study design,
data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
The ability to directly modify the genome of the budding yeast
Saccharomyces cerevisiae by homologous recombination is a major
advantage of this model system. In particular, PCR-based
recombination methods allow the targeting of any region in the
genome by amplifying a cassette with primers containing short
(40 bp) regions homologous to the desired integration site. PCR-
based recombination has been used for both deletion of yeast
genes and fusion of tags to those genes (reviewed in ). Such
gene-tagging approaches are particularly powerful as they leave
the tagged gene in its native chromosomal context, expressed
under its native promoter. Furthermore, in a haploid yeast cell, the
tagged gene will be the only copy of that gene present, allowing for
easy assessment of whether the tagged gene has a phenotype.
A wide variety of tagging vectors are available for fusing
different fluorescent proteins to yeast proteins. These allow
imaging of the tagged protein by fluorescence microscopy so that
its spatial distribution and transport can be determined. As the
tagged gene will be the sole copy present in a haploid cell, this also
allows the measurement of protein abundance by fluorescence
intensity measurements or by fluorescence correlation spectrosco-
py . Spectrally separated fluorescent proteins allow multiple
tagged proteins to be imaged and protein interactions to be
monitored by resonance energy transfer . Tagging vectors are
available for constructing gene fusions to a wide variety of
fluorescent proteins: green fluorescent protein and its blue, cyan,
and yellow variants, red fluorescent proteins, and the photo-
activatible proteins PA-GFP and mEos2 [3–9].
Recent advances in fluorescent protein engineering have
produced many fluorescent proteins with desirable properties.
Fluorescent proteins now span a wide range of colors, with bright
blue fluorescent proteins complementing green and red fluorescent
proteins. Significant improvements in green and red fluorescent
protein performance have been described with the generation of
brighter, more photostable, and faster maturing fluorescent
proteins. However, these newer proteins have not been system-
atically tested in S. cerevisiae, leaving it unclear which of these
proteins will perform best in yeast.
Here, we have optimized and systematically tested a number of
blue, green, and red fluorescent proteins for use in yeast protein
tagging. We have also optimized and tested two long Stokes shift
fluorescent proteins, and one far-red protein reported to be
fluorescent when excited at 640 nm. We have assessed these
proteins for brightness, photostability, and perturbation of fusion
proteins, and have recommendations for optimal blue, green, and
red fluorescent proteins for imaging tagged proteins in yeast. In
particular, we identify red fluorescent proteins that are several-fold
brighter than the commonly used mCherry, and a bright blue
fluorescent protein for imaging in yeast.
Materials and Methods
The plasmid backbones were derived from pFA6a-link-tdimer2-
SpHIS5 (pKT146), pFA6a-link-tdimer2-SpURA3 (pKT176), and
pFA6a-link-tdimer2-Kan (pKT178) . Protein sequences for
PLOS ONE | www.plosone.org1 July 2013 | Volume 8 | Issue 7 | e67902
fluorescent proteins were taken from the literature and the
corresponding DNA sequences were optimized for S. cerevisiae
expression by DNA2.0 . All protein sequences are shown in
Sequences S1 and are identical to the literature sequences except
for GFPc, which contains the additional mutations S72A (known
to improve folding) and L231H. The resulting sequences were
tailed with PacI and AscI sites and synthesized by DNA2.0. The
resulting fluorescent proteins were subcloned into the pFA6a-link
backbones using the PacI and AscI sites, replacing tdimer2 with
the desired fluorescent protein.
All plasmids are available from Addgene (www.addgene.org)
except for TagBFP, TagBFP2, TagRFP-T, TagRFP657, LSS-
mKate2, mKate2, and PA-TagRFP. These incorporate sequences
sold by Evrogen and cannot be distributed by Addgene.
Yeast Gene Tagging
Genes were tagged in S. cerevisiae strain BY4741 by PCR-
mediated transformation . Tagging cassettes were amplified
with KOD Hotstart PCR (EMD Millipore) using the forward
primer (gene-specific sequence)-GGTGACGGTGCTGGTTTA
(10 mL) were diluted into 100 mL of fresh media, grown to OD
0.7–1.0, washed twice with 0.1 M lithium acetate/16 TE and
resuspended in 2 mL 0.1 M lithium acetate/16TE. 20 ml of the
PCR product was then incubated with 200 ml of washed cells,
10 ml of ssDNA, and 1.2 ml of 0.1 M lithium acetate/16 TE/
50% PEG3350 and incubated at 30uC for 30 min. Cells were then
heat shocked at 42uC for 15 min after adding 154 ml of DMSO.
The cells were then pelleted, resuspended in 100 ul of water and
spread on selective media. Tagging of the targeted gene of interest
was confirmed by colony PCR to verify the presence of both
integration junctions and the absence of the unmodified gene .
For imaging, cells were grown overnight in low fluorescence SC
media , diluted 1:20–1:100 in fresh media and then grown three
hours before imaging. Cells were immobilized on concanavalin A-
coated glass bottom 35 mm dishes. Widefield microscopy was
performed on a Nikon Ti microscope with a Photometrics
Coolsnap HQ2 camera, using 606/1.4 NA or 1006/1.4 NA oil
immersion lenses. Illumination was provided by a Lambda XL
lamp (Sutter Instrument Company, Novato CA). Chroma
fluorescence filter sets 89021 and 89000 were used, with the
specific channels as follows: GFP: ET470/206, ET525/50m;
mCherry: ET572/356, ET632/60m;
ET455/50m; Cy3: ET555/256, ET605/52m; Cy5: ET645/
306, ET705/72m; mKeima: ET402/156, ET605/52m.
Spinning disk confocal imaging was performed on a Nikon Ti-E
equipped with a Yokogawa CSU-22 spinning disk confocal and a
Photometrics Evolve EMCCD camera, using a 1006/1.4 NA oil
immersion lens. Laser illumination was at 405 nm (BFP), 491 nm
(GFP), or 561 nm (RFP), and detection filters were ET460/50m
(BFP), ET525/50m (GFP), or ET610/60m (RFP).
Image analysis was performed in NIS-Elements (Nikon Instru-
ments Inc.). The background was estimated for each image from a
region free of cells and subtracted. For time lapse images, this
subtraction was performed at each time point. For brightness
measurements, the image was thresholded to identify cells, and the
intensity was calculated for each cell. The mean intensity of all
cells from three or more images was recorded (typically 100s of
cells). For time lapse photobleaching images, the same procedure
was followed at each time point. The time point at which point the
mean intensity dropped below 50% was then determined, and the
sum of the mean intensity at all prior time points (the integrated
intensity) was calculated.
Construction of Novel Yeast Fluorescent Protein Tagging
We set out to systematically test recently developed fluorescent
proteins for use in protein tagging in yeast. To do so, we first
collected a list of bright fluorescent proteins recently published in
the literature, as well as those recommended by our colleagues.
This list includes both commercially and academically developed
proteins. We focused on proteins compatible with the common
four color filter set used for imaging DAPI/FITC/Cy3/Cy5 and
with 405 nm, 488 nm, 561 nm, and 640 nm lasers on a confocal
microscope. The complete list of proteins tested along with their
photophysical properties is provided in Table 1, and the protein
sequence of each protein is provided in Sequences S1. Specifically,
we chose two blue fluorescent proteins, seven green fluorescent
proteins, and seven red fluorescent proteins. We also included one
far-red protein, TagRFP657, that is reported to be detectable
under 640 nm excitation, and two long-Stokes shift proteins that
we hoped might allow five-color imaging by exciting at 405 nm
and detecting in the Cy3 emission filter. We additionally
Table 1. Fluorescent Proteins Tested.
lem QY ECBrightness Reference
Blue Fluorescent Proteins:
mTagBFP4024570.63 5200032.8 
mTagBFP2 399454 0.64 5060032.4 
Green Fluorescent Proteins:
EGFP488 507 0.6 56000 33.6 [29,30]
Clover 505 515 0.76111000 84.4
Emerald487509 0.68 57500 39.1
Superfolder GFP485 510 0.6583300 54.1 
mWasabi493 509 0.80 70000 56.0 
Red Fluorescent Proteins:
mCherry 5876100.2272000 15.8 
mApple 568 592 0.49 75000 36.8 
mKate2 588 6330.4 62500 25.0 
mKO2551 565 0.62 6380039.6 
mRuby558 605 0.35 11200039.2
mRuby2559 600 0.38 113000 42.9 
TagRFP-T555584 0.41 81000 33.2 
TagRFP657 6116570.103400 0.34 
LSS-mKate2 460605 0.17260004.4
lex and lem are the peak excitation and emission wavelengths of the
fluorescent protein, respectively. QY is the quantum yield and EC the extinction
coefficient in M21cm21. Brightness is the product of QY and EC, divided by
1000. Data was taken from the literature and is not available for GFPc or
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org2 July 2013 | Volume 8 | Issue 7 | e67902
constructed, but did not test, five photoactivatible and photo-
convertible fluorescent proteins (Table S1).
Because codon usage has been shown to significantly affect
fluorescence intensity of fusion proteins , and additional factors
such as RNA secondary structure can affect expression level ,
we had each protein optimized by DNA2.0 for expression in S.
cerevisiae. This ensures that differences in codon usage between
proteins do not affect our comparison and that the sequences we
are testing are optimized for yeast expression. The resulting yeast
optimized fluorescent proteins (denoted by yo followed by the
fluorescent protein name; e.g. yoEGFP) were then cloned into the
pFA6a-link tagging vectors we have previously published . The
proteins were cloned into vectors with the selectable markers
CaUra3, SpHis5, and KanR (G418 resistance), giving a set of 72
vectors (Figure 1 and Table S2). These vectors share the same
quence)-TCGATGAATTCGAGCTCG) as our previous vectors
and can be used interchangeably with them.
Fluorescent Protein Brightness
To test the performance of these proteins, we first fused each
fluorescent protein to the highly abundant metabolic gene Tdh3.
As the resulting fusions are identical except for the sequence of the
fluorescent protein, we expect these fusions to accurately reflect
the performance of these tags. The brightness of different tags can
differ for a number of reasons, including differences in the
photophysical properties of the tag (e.g. quantum yield or
extinction coefficient) or poor expression or folding of the tag.
However, we expect that these properties should be independent
of the protein being tagged and so that this should be a reliable
reporter of the tag performance in other applications. While all the
versions tested are yeast optimized, we have omitted the ‘yo’ prefix
below for clarity.
We first measured the relative detectability of each protein by
comparing the brightness of tagged cells to that of untagged cells as
imaged on a widefield microscope. This is a measure of the signal-
to-background ratio (SBR) for each protein. By this metric, the
commonly used proteins EGFP and mCherry have SBRs of ,180.
Several of the tested performed very poorly in this assay. The two
long-Stokes shift proteins, mKeima and LSS-mKate2 and the far-
red fluorescent protein TagRFP657, had SBRs ,5 and were not
studied further. In the case of the long-Stokes shift proteins this
poor performance likely reflects both their low intrinsic brightness
and that our filters were poorly matched to their spectra; we
excited with light centered at 402 nm and these proteins are
optimally excited at 440–460 nm. The poor matching of our filters
results from the fact that we were using a four-band filter set
optimized for DAPI/FITC/Cy3/Cy5. A filter set designed for
imaging CFP/YFP/RFP might perform better with these proteins.
The poor performance of TagRFP657 likely results from both low
intrinsic brightness and poor matching to filters designed for Cy5;
nevertheless, as the longest-wavelength intrinsically fluorescent
protein identified to date, we wanted to determine if this protein
was bright enough to be useful for yeast imaging. The blue
fluorescent protein mTagBFP also had an SBR less than 5. The
improved mTagBFP2 is about ten times brighter and is only ,5-
fold less detectable than EGFP, making it a viable tag for imaging
with DAPI filters and 405 nm excitation.
We next systematically compared the multiple green and red
fluorescent proteins we had produced. We first assessed their
brightness by comparing the relative brightness of each green
fluorescent protein to EGFP and each red fluorescent protein to
mCherry. Because the red fluorescent proteins have varying
excitation spectra we evaluated their brightness using two different
commonly used filter sets, one designed for imaging mCherry and
one designed for imaging Cy3. The results of this comparison are
shown in Figure 2 and Table S3. Strikingly, most of the green
fluorescent proteins perform no better than EGFP, with the
exception of GFPc, which is approximately 50% brighter. Despite
its dimness, Wasabi may be useful for certain experiments as,
unlike other GFPs, it is not excited in the near UV and can be
multiplexed with the UV-excited GFP T-Sapphire [13,14]. The
improved EGFP variants Clover and Emerald [15,16], which are
reported to be substantially brighter than EGFP, perform
comparably to it under these conditions. These proteins derive
their high brightness in part because of optimization for folding at
37uC; it is possible that the mutations conferring improved folding
at 37uC in bacteria and mammalian cells as free protein are
unimportant for folding at 30uC in yeast as a C-terminal fusion
protein. Furthermore, the observed brightness in S. cerevisiae is
poorly correlated with the photophysical brightness (product of
quantum yield and extinction coefficient), suggesting that factors
other than the intrinsic chromophore brightness are important for
the measured brightness. In addition to protein folding, this could
include rapid photobleaching or interactions with the ionic or
redox environment in the cell [17–19].
We find many red fluorescent proteins that are brighter than
mCherry. Furthermore, because of the wide spectral range
spanned by red fluorescent proteins, the optimal choice of protein
depends on the choice of filter set used to view it. Figures 2B and C
and Table S4 show the relative brightness of these seven red
fluorescent proteins as measured through a Cy3 filter set and an
mCherry filter set. Not surprisingly, the results differ substantially,
with mCherry performing much more poorly when imaged with
the Cy3 filter set. In both cases, however, we find a number of
proteins that outperform mCherry. We also compared the
brightness of each protein as measured through the Cy3 filter
set with that of mCherry as measured through the mCherry filter
set to assess what the best protein imaged through either filter set is
(Table S4). For the mCherry filter set, the brightest protein is
mKate2, 2.36 brighter than mCherry. For the Cy3 filter set,
mRuby2 and mKO2 are 2.36 and 3.16 brighter, respectively,
than mCherry in the mCherry filter set. Overall, these are the
three brightest red fluorescent proteins.
We also compared the brightness of these green and red
proteins when imaged under laser illumination with a spinning
disk confocal. Under these conditions, the brightness of the green
fluorescent proteins was very similar to that observed in the
widefield measurements above (Table S3). For red fluorescent
proteins, we see that the proteins which perform well when imaged
with the Cy3 cube also perform well when imaged with the
spinning disk confocal, although the relative performance
improvement compared with mCherry is larger when imaged
with the spinning disk confocal (Table S4).
Figure 1. Schematic design of tagging plasmids. The overall
design of these plasmids is identical to the pFA6a-link tagging plasmids
previously published . yoFP is one of the 24 yeast optimized proteins
cloned here and S.M. is the yeast selectable marker, either SpHis5,
CaUra3, or KanR. These tagging sequences can be amplified with the
forward primer (gene-specific sequence)-GGTGACGGTGCTGGTTTA and
reverse primer (gene-specific sequence)-TCGATGAATTCGAGCTCG. A
complete list of plasmids constructed in this study is in Table S2.
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org3 July 2013 | Volume 8 | Issue 7 | e67902
Fluorescent Protein Stability
Brightness is not the only important
choosing a fluorescent protein. For time lapse imaging, a
critical parameter is the photostability of the protein. Photo-
bleaching occurs when a fluorophore in the excited state
undergoes a chemical reaction leading to its irreversible
destruction. Accordingly, photobleaching limits the amount of
data that can be recorded in a timelapse acquisition. To
measure the photobleaching rate we captured sequential images
of each fluorescent protein tagged to Tdh3 under continuous
illumination. We then quantified the time required to bleach to
50% of the initial intensity and the integrated intensity of the
cell during this time. This latter measurement is probably the
most relevant for assessing the performance of fluorescent
proteins as it measures the total amount of photons that can be
detected from a fluorescent protein until it drops to half of its
initial intensity. It also takes into account the intrinsic brightness
of the protein and partially corrects for illumination intensity
changes: if the illumination brightness decreases, the brightness
decreases but so does the photobleaching rate.
The integrated intensities measured during bleaching to 50%
of initial intensity for both the green and red fluorescent
proteins are shown in Figure 3 and Tables S3 and S5.
Surprisingly, none of the green fluorescent proteins tested
perform better than EGFP; even those proteins brighter than
EGFP are substantially less photostable. For red fluorescent
proteins, imaged through the mCherry filter, we find three
RFPs, mRuby2, mKate2, and TagRFP-T, that have significant-
ly higher integrated intensities than mCherry. mRuby2 actually
bleaches slightly more rapidly than mCherry, but its higher
brightness more than compensates for its rapid bleaching.
mKate2 is brighter than mCherry, but bleaches at about the
same rate, while TagRFP-T is about the same brightness as
mCherry but bleaches much more slowly. TagRFP-T was
selected for photostability so this is not surprising .
Perturbation of Fusion Protein Function
It is difficult to systematically assess whether a fluorescent
protein will perturb the function of the protein it is fused to,
because this perturbation depends on the molecular details of the
interactions made by the protein. However, to partially assess the
potential for perturbation of fusion protein function, we fused
mTagBFP2 and the green and red fluorescent proteins to the
septin Cdc12. We have previously observed that this protein is
sensitive to C-terminal fusions. Fusions which disrupt the function
of Cdc12 cause mislocalization of the protein, elongation of the
yeast cell, or both. In Figure 4, we show images of yeast cells
carrying each of these fusions. In general, the green fluorescent
proteins perform well, with minimal effect on the localization of
Cdc12. mTagBFP2 shows moderate perturbation to Cdc12
function, with some mislocalized Cdc12 and misshapen cells.
The red fluorescent proteins have highly variable effects on
Cdc12. In particular, a large fraction of Cdc12-mKO2 and
Cdc12-mKate2 cells show mislocalized Cdc12, while fusions to the
other red fluorescent proteins appear to function normally. This
suggests that these two proteins may perturb other proteins as well.
However, mKate2 has been successfully expressed in fusions to
many mammalian proteins , so it may be worth trying in other
We have constructed a set of yeast optimized fluorescent protein
tagging vectors expressing multiple blue, green, and red fluores-
cent proteins, as well as far-red and long-Stokes shift proteins. We
have systematically expressed these as yeast fusions and assessed
their brightness, photostability, and function as fusion proteins. We
find that the blue fluorescent protein mTagBFP2, while ,5-fold
less detectable than EGFP, is bright enough to use as a third color
in fluorescence microscopy. This is particularly useful for imaging
with laser-based imaging systems with a 405 nm laser. The long-
Stokes shift proteins and the far-red protein TagRFP657 are not
bright enough to be useful tags. Somewhat surprisingly, we find
Figure 2. Brightness of green and red fluorescent proteins. Yeast expressing fusions of each of the optimized fluorescent proteins to the
TDH3 protein were imaged, and the mean fluorescence of each strain was calculated. Data from each day was normalized to EGFP (for green
proteins) or mCherry (red proteins) to compensate for day-to-day fluctuations in lamp brightness and detection efficiency. The measurement was
repeated on at least three days and the mean and standard error for each strain is plotted. * indicates a protein significantly brighter than EGFP or
mCherry as determined by a one-sided t-test with 5% significance threshold. A. Green fluorescent proteins. B. Red fluorescent proteins imaged with
an mCherry filter set. C. Red fluorescent proteins imaged with a Cy3 filter set.
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org4 July 2013 | Volume 8 | Issue 7 | e67902
that none of the green fluorescent proteins we have tested
outperform EGFP. However we find a number of red fluorescent
proteins that are brighter and more photostable than mCherry.
The set of proteins we have constructed also includes the new
Clover/mRuby2 FRET pair, which is an improved green/red
replacement for CFP/YFP variants .
Our recommendations for fluorescent proteins are summarized
in Figure 5, broken down by the filter set used and the
requirements of the experiment. When using imaging systems
designed for DAPI/FITC/Cy3/Cy5, mTagBFP2, EGFP, and
mRuby2 are an excellent set of proteins for three-color imaging. If
an additional color is needed, it should be possible to multiplex T-
Sapphire , mTagBFP2, mWasabi, and mRuby2; mWasabi is
not excited at 405 nm, and T-Sapphire is a UV-excited, green
emitting GFP variant that should not crosstalk with mTagBFP2 or
mWasabi. When using a filter set with a Cy5 channel, it should be
possible to image iFP1.4 or iRFP  in the Cy5 channel,
although these proteins have not been tested in yeast to our
knowledge. The new red fluorescent proteins described here also
offer improved options for imaging with CFP/YFP/RFP filter sets
or GFP/mCherry filter sets. There are a number of improved CFP
and YFP variants [24,25] that may offer improved performance
although these have not yet been tested in yeast. Additionally, the
recently reported novel green fluorescent protein mNeonGreen
 may be a brighter GFP replacement.
For experiments where photobleaching is not a concern, such as
single time point imaging and flow cytometry, the green
fluorescent protein of choice is either EGFP or GFPc, which is
,1.56brighter. The choice of red fluorescent protein depends on
the filter set used: for longer wavelength filter sets designed for
mCherry, mKate2 (2.36 mCherry) is the brightest fluorescent
protein. However, it perturbs Cdc12 when fused to it, so mRuby2
(1.76 mCherry and non-perturbative) may be preferred. For
shorter wavelength filter sets designed for Cy3 or rhodamine,
mKO2 (7.46 mCherry) is the brightest fluorescent protein.
However, it also perturbs Cdc12 when fused to it, and mRuby2
is again the second-brightest protein (4.26mCherry).
For experiments where photobleaching is a concern, such as
time-lapse imaging, no green fluorescent protein outperforms
EGFP. The most photostable red fluorescent protein is TagRFP-
T, which we were able to collect 4.16 more light from than
mCherry, before bleaching to 50% of its initial intensity. It is
equally bright to mCherry in the mCherry channel and is brighter
than mCherry in the Cy3 channel. It also appears to be non-
perturbative in fusions. mKate2 and mRuby2 also outperform
mCherry in photostability (2.66and 1.56, respectively).
Overall, fornew tagging
mTagBFP2 as the best blue fluorescent protein, EGFP as the best
green fluorescent protein and TagRFP-T or mRuby2 as the best
red fluorescent protein, depending on the requirements for
photostability and brightness. mKate2 is also promising but
fusions to it should be carefully assessed for perturbation. TagRFP-
T and mRuby2 are also blue-shifted compared to mCherry and so
perform better when used with shorter wavelength filters or
Figure 3. Photostability of red and green fluorescent proteins. Yeast expressing fusions of each of the optimized fluorescent proteins to the
TDH3 protein were imaged continuously until their intensity dropped below 50% of the initial intensity. The intensity of each cell integrated over the
time until 50% bleaching occurred was then calculated, and the mean integrated intensity for each strain on each day was normalized to EGFP (for
green proteins) or mCherry (red proteins) to compensate for day-to-day fluctuations in lamp brightness and detection efficiency. The measurement
was repeated on at least two days and the mean and standard error for each strain is plotted. * indicates a protein with significantly larger integrated
intensity than mCherry as determined by a one-sided t-test with 5% significance threshold.
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org5 July 2013 | Volume 8 | Issue 7 | e67902
561 nm excitation. Together with EGFP and mTagBFP2, these
provide a set of fluorescent proteins for three color imaging with
405 nm/488 nm/561 nm laser systems or common DAPI/
FITC/Cy3 filter sets.
Figure 4. Perturbation of protein function. Yeast expressing fusions of each of the indicated proteins to the C-terminus of Cdc12 were imaged
to assess whether they perturb its function. Perturbation of Cdc12 function manifests as misshapen yeast cells and/or mislocalized Cdc12. The green
fluorescent proteins show minimal perturbation; mKate2 and mKO2 show major perturbation; mTagBFP2 is intermediate. Brightness has been
normalized separately for each image so it is not comparable from image to image.
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org6 July 2013 | Volume 8 | Issue 7 | e67902
generated in this study.
Plasmids generated in this study.
Brightness and photostability of green fluo-
Brightness of red fluorescent proteins.
Photostability of red fluorescent proteins.
proteins generated in this study.
Protein sequences of the fluorescent
All imaging data was acquired in the Nikon Imaging Center at UCSF/
Conceived and designed the experiments: KST WAL. Performed the
experiments: SL KST. Analyzed the data: SL KST. Contributed reagents/
materials/analysis tools: SL. Wrote the paper: KST WAL.
1. Petracek ME, Longtine MS (2002) PCR-based engineering of yeast genome.
Meth Enzymol 350: 445–469.
2. Slaughter BD, Schwartz JW, Li R (2007) Mapping dynamic protein interactions
in MAP kinase signaling using live-cell fluorescence fluctuation spectroscopy and
imaging. Proc Natl Acad Sci USA 104: 20320–20325. doi:10.1073/
3. Hailey DW, Davis TN, Muller EGD (2002) Fluorescence resonance energy
transfer using color variants of green fluorescent protein. Meth Enzymol 351:
4. Wach A, Brachat A, Alberti-Segui C, Rebischung C, Philippsen P (1997) AID-
5. Reid RJD, Lisby M, Rothstein R (2002) Cloning-free genome alterations in
Saccharomyces cerevisiae using adaptamer-mediated PCR. Meth Enzymol 350:
6. Sheff MA, Thorn KS (2004) Optimized cassettes for fluorescent protein tagging
in Saccharomyces cerevisiae. Yeast 21: 661–670. doi:10.1002/yea.1130.
7. Janke C, Magiera MM, Rathfelder N, Taxis C, Reber S, et al. (2004) A versatile
toolbox for PCR-based tagging of yeast genes: new fluorescent proteins, more
markers and promoter substitution cassettes. Yeast 21: 947–962. doi:10.1002/
8. Vorvis C, Markus SM, Lee WL (2008) Photoactivatable GFP tagging cassettes
for protein-tracking studies in the budding yeast Saccharomyces cerevisiae. Yeast
25: 651–659. doi:10.1002/yea.1611.
9. Young CL, Raden DL, Caplan JL, Czymmek KJ, Robinson AS (2012) Cassette
series designed for live-cell imaging of proteins and high-resolution techniques in
yeast. Yeast 29: 119–136. doi:10.1002/yea.2895.
10. Welch M, Govindarajan S, Ness JE, Villalobos A, Gurney A, et al. (2009) Design
parameters to control synthetic gene expression in Escherichia coli. PLoS ONE
4: e7002. doi:10.1371/journal.pone.0007002.
11. Wach A, Brachat A, Po ¨hlmann R, Philippsen P (1994) New heterologous
modules for classical or PCR-based gene disruptions in Saccharomyces
cerevisiae. Yeast 10: 1793–1808.
12. Kudla G, Murray AW, Tollervey D, Plotkin JB (2009) Coding-sequence
determinants of gene expression in Escherichia coli. Science 324: 255–258.
13. Ai H, Olenych SG, Wong P, Davidson MW, Campbell RE (2008) Hue-shifted
monomeric variants of Clavularia cyan fluorescent protein: identification of the
molecular determinants of color and applications in fluorescence imaging. BMC
Biol 6: 13. doi:10.1186/1741-7007-6-13.
14. Zapata-Hommer O, Griesbeck O (2003) Efficiently folding and circularly
permuted variants of the Sapphire mutant of GFP. BMC Biotechnol 3: 5.
15. Cubitt AB, Woollenweber LA, Heim R (1999) Understanding structure-function
relationships in the Aequorea victoria green fluorescent protein. Methods Cell
Biol 58: 19–30.
16. Lam AJ, St-Pierre F, Gong Y, Marshall JD, Cranfill PJ, et al. (2012) Improving
FRET dynamic range with bright green and red fluorescent proteins. Nat
Methods 9: 1005–1012. doi:10.1038/nmeth.2171.
17. Bogdanov AM, Mishin AS, Yampolsky IV, Belousov VV, Chudakov DM, et al.
(2009) Green fluorescent proteins are light-induced electron donors. Nat Chem
Biol 5: 459–461. doi:10.1038/nchembio.174.
18. Shaner NC, Steinbach PA, Tsien RY (2005) A guide to choosing fluorescent
proteins. Nat Methods 2: 905–909. doi:10.1038/nmeth819.
19. Griesbeck O, Baird GS, Campbell RE, Zacharias DA, Tsien RY (2001)
Reducing the environmental sensitivity of yellow fluorescent protein. Mechanism
and applications. J Biol Chem 276: 29188–29194. doi:10.1074/
Figure 5. Recommended fluorescent protein combinations for yeast imaging. Recommended fluorescent protein combinations for
imaging in yeast are broken down by filter set (horizontal axis) and experimental requirement (vertical axis). All proteins mentioned here are available
in yeast tagging vectors either from this paper or from [6,9] and most are available from Addgene. Recommended proteins are listed first, with
alternatives given in parentheses. It is likely that iFP1.4  or iRFP  can be used to image in the far-red (Cy5) channel, but this has not been tested
in yeast. mWasabi is dimmer that EGFP or GFPc, but is not excited at 405 nm, allowing it to be multiplexed with T-Sapphire .
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org7 July 2013 | Volume 8 | Issue 7 | e67902
20. Shaner NC, Lin MZ, McKeown MR, Steinbach PA, Hazelwood KL, et al. Download full-text
(2008) Improving the photostability of bright monomeric orange and red
fluorescent proteins. Nat Methods 5: 545–551. doi:10.1038/nmeth.1209.
21. Shcherbo D, Murphy CS, Ermakova GV, Solovieva EA, Chepurnykh TV, et al.
(2009) Far-red fluorescent tags for protein imaging in living tissues. Biochem J
418: 567–574. doi:10.1042/BJ20081949.
22. Shu X, Royant A, Lin MZ, Aguilera TA, Lev-Ram V, et al. (2009) Mammalian
expression of infrared fluorescent proteins engineered from a bacterial
phytochrome. Science 324: 804–807. doi:10.1126/science.1168683.
23. Filonov GS, Piatkevich KD, Ting LM, Zhang J, Kim K, et al. (2011) Bright and
stable near-infrared fluorescent protein for in vivo imaging. Nat Biotechnol 29:
24. Kremers GJ, Goedhart J, van Munster EB, Gadella TWJ Jr (2006) Cyan and
yellow super fluorescent proteins with improved brightness, protein folding, and
FRET Fo ¨rster radius. Biochemistry 45: 6570–6580. doi:10.1021/bi0516273.
25. Goedhart J, von Stetten D, Noirclerc-Savoye M, Lelimousin M, Joosen L, et al.
(2012) Structure-guided evolution of cyan fluorescent proteins towards a
quantum yield of 93%. Nat Commun 3: 751. doi:10.1038/ncomms1738.
26. Shaner NC, Lambert GG, Chammas A, Ni Y, Cranfill PJ, et al. (2013) A bright
monomeric green fluorescent protein derived from Branchiostoma lanceolatum.
Nat Methods. doi:10.1038/nmeth.2413.
27. Subach OM, Gundorov IS, Yoshimura M, Subach FV, Zhang J, et al. (2008)
Conversion of red fluorescent protein into a bright blue probe. Chem Biol 15:
28. Subach OM, Cranfill PJ, Davidson MW, Verkhusha VV (2011) An enhanced
monomeric blue fluorescent protein with the high chemical stability of the
chromophore. PLoS ONE 6: e28674. doi:10.1371/journal.pone.0028674.
29. Cormack BP, Valdivia RH, Falkow S (1996) FACS-optimized mutants of the
green fluorescent protein (GFP). Gene 173: 33–38.
30. Yang TT, Cheng L, Kain SR (1996) Optimized codon usage and chromophore
mutations provide enhanced sensitivity with the green fluorescent protein.
Nucleic Acids Res 24: 4592–4593.
31. Zhang C, Konopka JB (2010) A photostable green fluorescent protein variant for
analysis of protein localization in Candida albicans. Eukaryotic Cell 9: 224–226.
32. Pe ´delacq JD, Cabantous S, Tran T, Terwilliger TC, Waldo GS (2006)
Engineering and characterization of a superfolder green fluorescent protein. Nat
Biotechnol 24: 79–88. doi:10.1038/nbt1172.
33. Shaner NC, Campbell RE, Steinbach PA, Giepmans BNG, Palmer AE, et al.
(2004) Improved monomeric red, orange and yellow fluorescent proteins derived
from Discosoma sp. red fluorescent protein. Nat Biotechnol 22: 1567–1572.
34. Sakaue-Sawano A, Kurokawa H, Morimura T, Hanyu A, Hama H, et al. (2008)
Visualizing spatiotemporal dynamics of multicellular cell-cycle progression. Cell
132: 487–498. doi:10.1016/j.cell.2007.12.033.
35. Kredel S, Oswald F, Nienhaus K, Deuschle K, Ro ¨cker C, et al. (2009) mRuby, a
bright monomeric red fluorescent protein for labeling of subcellular structures.
PLoS ONE 4: e4391. doi:10.1371/journal.pone.0004391.
36. Morozova KS, Piatkevich KD, Gould TJ, Zhang J, Bewersdorf J, et al. (2010)
Far-red fluorescent protein excitable with red lasers for flow cytometry and
superresolution STED nanoscopy. Biophys J 99: L13–15. doi:10.1016/
37. Kogure T, Karasawa S, Araki T, Saito K, Kinjo M, et al. (2006) A fluorescent
variant of a protein from the stony coral Montipora facilitates dual-color single-
laser fluorescence cross-correlation spectroscopy. Nat Biotechnol 24: 577–581.
38. Piatkevich KD, Hulit J, Subach OM, Wu B, Abdulla A, et al. (2010) Monomeric
red fluorescent proteins with a large Stokes shift. Proc Natl Acad Sci USA 107:
Improved Yeast Fluorescent Protein Tagging Vectors
PLOS ONE | www.plosone.org8 July 2013 | Volume 8 | Issue 7 | e67902