DNA polyfluorophores for real-time multicolor tracking of dynamic biological systems

Article abstract of DOI:10.1002/anie.201201928

Dye-ing to live: Spectral limitations of common organic dyes make it difficult or impossible to visualize and follow multiple biological components in rapidly moving systems. The development of a multispectral set of improved DNA-scaffolded fluorophores is described. Their use in multicolor cellular imaging (see scheme) and in tracking of biological motions on the subsecond timescale is demonstrated. Copyright

Full text of DOI:10.1002/anie.201201928

Angewandte
Chemie
DOI: 10.1002/anie.201201928
Fluorescence Imaging
DNA Polyfluorophores for Real-Time Multicolor Tracking of Dynamic
Biological Systems**
Shenliang Wang, Jia Guo, Toshikazu Ono, and Eric T. Kool*
It is becoming widely appreciated that complex biological
networks and interaction events need to be understood not
only in the three spatial dimensions but also in the fourth
dimension, time. To address this need, fluorescence imaging
with light microscopy has entered a dynamic era. A series of
time-lapsed images or an aggregated video of a dynamic event
imparts much more information than a single still image,
better revealing the natural organization and flow of biolog-
ical processes. The widespread application of fluorescent
proteins,[1] quantum dots,[2] and small-molecule fluoro-
phores[3] in time-lapsed imaging, along with state-of-the-art
optical technologies, such as super-resolution microscopy,[4]
and integrated systems, such as intravital microscopy,[5] have
aided enormous advances in understanding cellular and
organismal biology. Tracking the movements of molecular
and cellular targets has helped reveal processes such as
metastasis,[6] secretory pathway cargo transfer,[1] and many
more. The timeframe of the studied events can vary from
hours to minutes, or even seconds.
Although it is possible to track sub-second motions by
light microscopy, simultaneously tracking vectors of multiple
species is hindered by the spectral limitations of common
organic dyes and proteins, which require separate excitation
wavelengths and emission filters. This drawback prevents the
experimenter from recording visualization of multiple colors
in real time, and thus most current multicolor labeling studies
are based on post-image merging of multiple false-colored
images. In moving systems, more complex imaging setups are
required, including rapid filter wheel changes or multichannel
imaging systems, increasing the complexity and cost of the
instrumentation. Even with this added sophistication, time-
scales and numbers of colors are limited. Inorganic quantum
dots can address this problem in part by taking advantage of
a single UV excitation;[7] however, they present some of their
own limitations as biological labels, including large size,
multivalency, toxicity, and limited cellular permeability.[8]
Ideally, small-molecule, water-soluble organic dyes could
be useful in multicolor dynamic imaging if they could be
excited at one wavelength. In previous studies, we developed
a new class of fluorescent dyes (oligodeoxyfluorosides
(ODFs)) in which fluorescent aromatic species replace
nucleobases in short DNA-like oligomers.[9] This molecular
design has the advantage of rapid automated synthesis of
thousands of possible composite dyes from a few components.
In addition, the short DNA-like oligomers retain small size,
are water-soluble, and are easily conjugated to small mole-
cules and biomacromolecules.[10] An early study of a set of
ODF fluorophores demonstrated multicolor emissions (in
excess of ten colors) with single long wavelength UV
excitation.[9] However, this first-generation set of dyes
exhibited some limitations, such as low quantum yields of
some monomers and oligomers, the chemical instability of
one monomer, the rapid photooxidation of a benzopyrene
component, and high toxicity of the benzopyrene starting
material.[11] The dyes were not tested in dynamic systems. For
more general multispectral applications, a set of single-
excitation dyes might ideally exhibit similar levels of bright-
ness, chemical and biological stability, and cell permeability.
To begin to address these issues, we designed and
synthesized two new fluorescent deoxyriboside monomers.
The extended pyrene monomer V (Figure 1a) was designed
to substitute for a previous benzopyrene monomer by offering
similar spectral characteristics, as well as the ability to form
excimers in an aqueous environment.[12] The synthesis is
shown in Scheme S1 of the Supporting Information, in
addition details and spectral characterization are given. The
superior fluorescent properties of a previous dicyanomethyl-
ene aminostyryl pyran monomer allowed us to retain the
fluorophore while installing a more chemically stable link to
the deoxyribose, giving the new monomer K (Figure 1a and
Scheme S2 of the Supporting Information).
Incorporating these components, we prepared a candidate
set of 18 ODF dyes using a DNA synthesizer (Figure 1c), and
their absorption and fluorescence emission spectra were
recorded (Figure 2). These ODFs can all be excited at 355 nm
and cover the visible light range from blue to red. Quantum
yields were above 25% for 13 of 18 dyes; fluorescence
lifetimes ranged from 2.2 ns to more than 40 ns for the entire
set (Table S1). The new V fluorophore (as SV; sequence given
5’- > 3’ with analogy to DNA) has a high quantum yield and
emits strongly in the blue region. The monomers Y, E, and V
are found to form excimers/exciplexes from cyan to green
color (Figure 2 and Table S1) involving various combinations
of the three aromatic monomers.
[*] Dr. S. Wang, Dr. J. Guo, Dr. T. Ono, Prof. Dr. E. T. Kool
Department of Chemistry, Stanford University
Stanford, CA 94305 (USA)
E-mail: kool@stanford.edu
[**] This work was supported by the U.S. National Institutes of Health
(GM067201). We thank Dr. Lindsey E. McQuade and Prof. James K.
Chen for assistance and advice with zebrafish experiments, and
Prof. Ingmar H. Riedel-Kruse for helpful advice on Paramecium
culture.
Supporting information for this article such as details of synthetic
procedures, optical measurements, additional data, details of
fluorescent staining procedures, and additional imaging and video
data is available on the WWW under http://dx.doi.org/10.1002/
anie.201201928.
Incorporation of K monomer into the composite dyes
greatly increased the ODF emission color range to yellow,
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Figure 1. Molecular formulas and colors of ODF dyes in water. a) Monomer formulas; b) scheme of a representative ODF dye (SSEYK shown);
c) images of aqueous solutions of ODFs under UV excitation (sequences 5’->3’): 1) SY, 2) SV, 3) SE, 4) SSSYYYY, 5) SSEY, 6) SSYV, 7) SSEYE,
8) SSVEV, 9) SSVE, 10) SSYVE, 11) SSVV, 12) SSVVE, 13) SSEVE, 14) SSEYK, 15) SSVVK, 16) SSYKY, 17) SSVKV, 18) SSVKVK. One or more
monomers lacking a fluorophore (S) are added to increase water solubility and inhibit aggregation. Illumination by UV lamp (>365 nm).
orange, and red wavelengths. Sequences containing both V
and K monomers emit magenta hues, an effect of color
addition of the blue and red wavelengths. For visual
distinction in subsequent imaging experiments we chose the
sequences SV (blue), SSEY (cyan), SSVV (green), SSEYK
(yellow), and SSVKV (magenta/red) as a test set for further
imaging experiments (see below).
Tracking labeled species over time requires that dyes
possess sufficient photostability and biostability across the
experimental timeframe. We measured spectra over a 2 h
experiment (more than 20 minutes illumination time), eval-
uating the above five dyes along with FITC and Alexa 350 in
phosphate buffered saline (PBS), as well as 10% human
serum. One dye (SSVV) showed evidence of photobleaching
(Figure S1), similar in rate to that of FITC. The remaining
four ODF dyes exhibited little or no photobleaching.
Interestingly, despite their phosphodiester backbones, none
of the ODF dyes showed any evidence of biological degra-
dation (e.g. by nucleases) in the serum solution over two
hours (Figure S2), which suggests that the unnatural nucleo-
base structures are sufficient to confer nuclease resistance.
Also, the dye SSVV displayed altered emission in serum
relative to PBS, suggesting an effect of protein binding on the
dye conformation.
While the V monomer is photophysically quite stable (see
data for the SV dye, Figure S3), we detected an interesting
bleaching process of the VV dimer in the sequence SSVV.
The 510 nm green emission, characteristic of an excimer state,
is bleached at a rate similar to that of FITC. The monomer
emission near 400 nm is unchanged. Neither pyrene alone (as
Y) nor pyrene oligomers (as YY) exhibit this phenomenon,
which suggests a major contribution of the ethynyl group in
a photobleaching reaction. More work will be needed to shed
light on this unusual bleaching mechanism. However, the
SSEY sequence offers a photophysically stable alternative
dye with a similar emission wavelength.
The most stable dyes (SV, SSEY, SSEYK, and SSVKV)
were incubated with HeLa cells to evaluate their membrane
permeability. The cultured cells were exposed to 2 mm dye in
DMEM medium for 2 h at 378C (Figure 3). The results
confirm our earlier observations[9] that ODF dyes can
successfully penetrate into the cells (showing predominantly
cytoplasmic localization). The fact that they retain their
emission colors in the intracellular environment (Figure 3)
establishes that they are not degraded appreciably by
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
To provide a further test of the
use of ODF dyes for monitoring
separate moving targets, ODFs were
applied to the invertebrate shrimp
Artemia salina. The shrimp 1–2 days
post-hatching were incubated with
10 mm ODFs in 1% NaCl solution
for 4 h. After rinsing with 1% NaCl
solution, the shrimp stained with
differently colored dyes were mixed,
and their motion was observed under
an
epifluorescence
microscope
(Figure 5 and Supporting Video 2).
Although the shrimp were intention-
ally confined in their motions for
imaging, their positions in each frame
can be tracked and accurate moving
traces could be recorded. Figure 5c
shows four selected individuals with
four different dye colors processed by
the software Volocity, demonstrating
the possibility of tracking four differ-
ent targets by visual distinction as
well as trace recording.
Next we moved to single-cell
living targets for labeling, with the
aim of attempting to track a greater
number of moving components syn-
chronously. We chose the ciliate pro-
tozoan Paramecium caudatum, which
swims rapidly with spiral motion by
way of coordinated cilia movements.
The outer pellicle of the organism did
not absorb ODF dyes, so we used an
indirect staining approach, by dyeing
their food. Yeast (S. cerevisiae) were
heat-killed and separately stained
Figure 2. Normalized absorption (a) and emission (b) spectra of ODFs in this study. Emission
spectra were recorded in water with 355 nm excitation.
nucleases over the time of the experiment. It is worth noting
that the color of SSVKV in the HeLa cells shifted slightly
closer to a true red color (as compared with its magenta hue in
water, Figure 2).
with four different ODFs; the yeast were then fed to the
protozoans for one hour. Under high magnification, the
individual yeast cells are then visible in the cytoplasm of the
Paramecium within food vacuoles (Figure 6). Using the lower-
magnification 4 ꢀ obective, approximately seventy moving
Paramecium individuals, with four label colors, were present
in the field of view, many moving over one body length on the
100 millisecond timescale (Supporting Information Video 3).
Of these, eight Paramecium cells were selected and success-
fully tracked in an automated fashion. Their moving traces
were processed, and are shown in Figure 5 f.
The experiments confirmed that the dye set SV, SSEY,
SSEYK, and SSVKV allows the experimenter to visualize
four label colors simultaneously with simple equipment. This
not only captures motion in several colors very easily, but also
simplifies multicolor still imaging. Still images such as in
Figure 3 and 4 (top) could be seen directly by eye because
they require no filter changes. This feature allows for the
judgement of true colors and relative intensities for each
species labeled with a given color. The relative brightness of
this four-dye set is reasonably balanced (within about
threefold across the series); this contrasts with inorganic
As an initial test of applying ODF dyes in a dynamic
system, we chose zebrafish (Danio rerio) embryos at one day
post-fertilization. Embryos were separately stained in differ-
ent ODF solutions (10 mm), rinsed with buffer and then mixed
together for imaging. Although the oligomeric dyes did not
penetrate the embryo chorion (except the blue SV dye), the
outside of the chorion was vividly stained by the ODFs. The
embryos were observed under an epifluorescence microscope
and motions of developing fish inside the chorion were
recorded in full color at 7 frames per second (fps; see Figure 4
and Supporting Information Video 1). In addition to the ODF
staining, the embryos showed some yellowish autofluores-
cence background signal under the 340–380 nm wavelength
excitation. Using a single excitation filter and capturing
emission with a single, greater than 420 nm long-pass emission
filter, the synchronous motions of embryos within the labeled
chorion were monitored, showing the natural motions clearly
on a subsecond timescale.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
Figure 4. Still and video imaging of zebrafish embryos (Danio rerio),
showing four-color (true color) labeling. Top: still image from the
video showing four ODF dye colors in a single frame, without the need
for recombination of separate false-colored images. Below: the three
rows show six consecutive frames of the motion of each embryo
within its chorion. Timespan of six frames is about 800 ms. Video was
taken using 340–380 nm excitation wavelength and a single long-pass
filter (greater than 420 nm) on an epifluorescence microscope. Dye
sequences were: SV (blue), SSEY (cyan), SSEYK (yellow-orange), and
SSVKV (red). See Supporting Information for the full video. Scale
bar=200 mm.
Figure 3. Epifluorescence microscope images of Hela cells stained
with single ODF dyes (2 mm, 2 h incubation). Dye sequences are as
shown (5’->3’). Images were taken with 340–380 nm excitation wave-
length and a single long-pass filter (greater than 420 nm).
quantum dots (QD), which vary strongly in brightness across
the color range.[13] Although inorganic QDs can exhibit high
photostability, our data show that a set of ODFs can be stable
against bleaching, even without antifade reagents, for over
twenty minutes of cumulative UV illumination in air-satu-
rated solutions.
Recent experiments in other groups have explored the
unusual photophysical characteristics of multiple fluoro-
phores organized by DNA.[14] Those approaches involve
larger double-helical DNA structures, and few have been
tested yet in biological systems. The current ODF molecules
are distinct by being considerably smaller and single-
Figure 5. Tracking the motion of Artemia salina (a–c) and Paramecium caudatum (d–f) stained in four colors with ODF dyes. Artemia salina were
stained with four ODFs (SV, SSEY, SSEYK, SSVKV) at 10 mm, 4 h. a) The first and b) last frame of a 14 s video and c) shows the tracking of the
motions of four individuals, which are constrained by crowding. d–f) Paramecium caudatum were fed with yeast stained with ODFs (the same four
dyes; see Figure 6 and Supporting Information for details). d) The first and e) last frame of a 10 s video and f) shows the automated tracking of
the motion of eight individuals. Scale bar=200 mm. Video was taken with 340–380 nm excitation wavelength and a single long-pass filter (greater
than 420 nm) on an epifluorescence microscope. For full videos see Supporting Information.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
P. R. Brink, G. R. Gaudette, I. S. Cohen, Stem Cells 2007, 25,
2128 – 2138; b) X. Gao, Y. Cui, R. M. Levenson, L. W. Chung, S.
Nie, Nat. Biotechnol. 2004, 22, 969 – 976.
[3] a) S. T. Laughlin, J. M. Baskin, S. L. Amacher, C. R. Bertozzi,
Science 2008, 320, 664 – 667; b) J. M. Baskin, K. W. Dehnert, S. T.
Laughlin, S. L. Amacher, C. R. Bertozzi, Proc. Natl. Acad. Sci.
USA 2010, 107, 10360 – 10365; c) N. Y. Kang, H. H. Ha, S. W.
Yun, Y. H. Yu, Y. T. Chang, Chem. Soc. Rev. 2011, 40, 3613 –
3626.
[4] a) U. V. Nagerl, K. I. Willig, B. Hein, S. W. Hell, T. Bonhoeffer,
Proc. Natl. Acad. Sci. USA 2008, 105, 18982 – 18987; b) J.
Tønnesen, F. Nadrigny, K. I. Willig, R. Wedlich-Soldner, U. V.
Nagerl, Biophys. J. 2011, 101, 2545 – 2552.
[5] M. J. Pittet, R. Weissleder, Cell 2011, 147, 983 – 991.
[6] E. Sahai, Nat. Rev. Cancer 2007, 7, 737 – 749.
[7] A. Sergꢁ, N. Bertaux, H. Rigneault, D. Marguet, Nat. Methods
2008, 5, 687 – 694.
[8] a) U. Resch-Genger, M. Grabolle, S. Cavaliere-Jaricot, R.
Nitschke, T. Nann, Nat. Methods 2008, 5, 763 – 775; b) A. M.
Smith, S. Nie, Nat. Biotechnol. 2009, 27, 732 – 733; c) R. J. Byers,
E. R. Hitchman, Prog. Histochem. Cytochem. 2011, 45, 201 – 237.
[9] a) Y. N. Teo, J. N. Wilson, E. T. Kool, J. Am. Chem. Soc. 2009,
131, 3923 – 3933; b) Y. N. Teo, Ph.D. Thesis, Stanford University
(USA), 2010.
[10] a) O. Khakshoor, E. T. Kool, Chem. Commun. 2011, 47, 7018 –
7024; b) N. Dai, E. T. Kool, Chem. Soc. Rev. 2011, 40, 5756 –
5770; c) J. Guo, S. Wang, N. Dai, Y. N. Teo, E. T. Kool, Proc. Natl.
Acad. Sci. USA 2011, 108, 3493 – 3498.
[11] a) J. Gao, S. Watanabe, E. T. Kool, J. Am. Chem. Soc. 2004, 126,
12748 – 12749; b) J. Gao, Ph.D. Thesis, Stanford University
(USA), 2004.
[12] I. V. Astakhova, A. D. Malakhov, I. A. Stepanova, A. V. Ustinov,
S. L. Bondarev, A. S. Paramonov, V. A. Korshun, Bioconjugate
Chem. 2007, 18, 1972 – 1980.
[13] Y. Xing, Q. Chaudry, C. Shen, K. Y. Kong, H. E. Zhau, L. W.
Chung, J. A. Petros, R. M. OꢂRegan, M. V. Yezhelyev, J. W.
Simons, M. D. Wang, S. Nie, Nat. Protoc. 2007, 2, 1152 – 1165.
[14] a) E. Mayer-Enthart, H. A. Wagenknecht, Angew. Chem. 2006,
118, 3451 – 3453; Angew. Chem. Int. Ed. 2006, 45, 3372 – 3375;
b) V. L. Malinovskii, F. Samain, R. Hꢃner, Angew. Chem. 2007,
119, 4548 – 4551; Angew. Chem. Int. Ed. 2007, 46, 4464 – 4467;
c) J. Chiba, S. Takeshima, K. Mishima, H. Maeda, Y. Nanai, K.
Mizuno, M. Inouye, Chem. Eur. J. 2007, 13, 8124 – 8130; d) A. L.
Benvin, Y. Creeger, G. W. Fisher, B. Ballou, A. S. Waggoner,
B. A. Armitage, J. Am. Chem. Soc. 2007, 129, 2025 – 2034; e) I. V.
Astakhova, V. A. Korshun, K. Jahn, J. Kjems, J. Wengel,
Bioconjugate Chem. 2008, 19, 1995 – 2007; f) D. Baumstark,
H. A. Wagenknecht, Chem. Eur. J. 2008, 14, 6640 – 6645; g) N. A.
Grigorenko, C. J. Leumann, Chem. Eur. J. 2009, 15, 639 – 645;
h) R. Varghese, H. A. Wagenknecht, Chem. Eur. J. 2009, 15,
9307 – 9310; i) A. L. Stadler, J. O. Delos Santos, E. S. Stensrud,
A. Dembska, G. L. Silva, S. Liu, N. I. Shank, E. Kunttas-Tatli,
C. J. Sobers, P. M. Gramlich, T. Carell, L. A. Peteanu, B. M.
McCartney, B. A. Armitage, Bioconjugate Chem. 2011, 22,
1491 – 1502; j) F. Garo, R. Hꢃner, Angew. Chem. 2012, 124,
940 – 943; Angew. Chem. Int. Ed. 2012, 51, 916 – 919.
Figure 6. Epifluorescence microscope images of single P. caudatum
cells fed with yeast stained with different ODF dyes (sequences
shown). Scale bar=20 mm. Images were taken with 340–380 nm
excitation and a single long-pass filter (greater than 420 nm). Small
groups of yeast cells are visible within food vacuoles of the Para-
mecium. See Figure S5 for bright-field image of P. caudatum.
stranded, which may confer benefits in cost, stability, and cell
permeability.
The selected ODF dye set, SV, SSEY, SSEYK, and
SSVKV, can be used as a fluorescent toolbox for blue, cyan,
yellow, and red color labeling, respectively. Although the dyes
were applied by direct staining without covalent conjugation,
ODFs can be easily conjugated to small molecules and to
antibodies if desired.[10] To our knowledge, the current results
are an unprecedented demonstration of the use of small-
molecule fluorescent dyes for multicolor simultaneous obser-
vation and tracking of rapid motions. Although we tested the
application in a limited set of biological systems, we expect
that such dyes could be used more broadly in biology and in
other fields as well.
Received: March 11, 2012
Revised: April 16, 2012
Published online: && &&, &&&&
Keywords: fluorescence · imaging agents · multicolor ·
.
oligonucleotides · real-time tracking
[1] J. F. Presley, N. B. Cole, T. A. Schroer, K. Hirschberg, K. J. Zaal,
J. Lippincott-Schwartz, Nature 1997, 389, 81 – 85.
[2] a) A. B. Rosen, D. J. Kelly, A. J. Schuldt, J. Lu, I. A. Potapova,
S. V. Doronin, K. J. Robichaud, R. B. Robinson, M. R. Rosen,
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Fluorescence Imaging
Dye-ing to live: Spectral limitations of
common organic dyes make it difficult or
impossible to visualize and follow multi-
ple biological components in rapidly
moving systems. The development of
a multispectral set of improved DNA-
scaffolded fluorophores is described.
Their use in multicolor cellular imaging
(see scheme) and in tracking of biological
motions on the subsecond timescale is
demonstrated.
S. Wang, J. Guo, T. Ono,
E. T. Kool*
&&&& — &&&&
DNA Polyfluorophores for Real-Time
Multicolor Tracking of Dynamic
Biological Systems
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!