An effective minor groove binder as a red fluorescent marker for live-cell DNA imaging and quantification

Article abstract of DOI:10.1002/anie.201007386

Seeing red: DEAB-TO-3 shows a large fluorescence enhancement upon binding to native DNA and a distinct selectivity for double-stranded DNA over total RNA. As a red fluorescent live-cell-permeant DNA minor groove binder, DEAB-TO-3 is promising for highly sensitive DNA detection in vitro and nucleus-specific imaging and DNA quantification in vivo (see picture). Copyright

Full text of DOI:10.1002/anie.201007386

Communications
DOI: 10.1002/anie.201007386
Fluorescent Probes
An Effective Minor Groove Binder as a Red Fluorescent Marker for
Live-Cell DNA Imaging and Quantification**
Xiaojun Peng,* Tong Wu, Jiangli Fan, Jingyun Wang, Si Zhang, Fengling Song, and Shiguo Sun
Since the understanding of DNA organization and structure
in vivo is so important, fluorescent staining techniques using
organic DNA-binding molecules are required for biological
research and medical diagnosis, including cellular imaging
and DNA quantification.[1–5] However, membrane-permeable
DNA-specific stains are uncommon. The minor groove
binders 4’,6-diamidino-2-phenylindole (DAPI) and Hoechst
33258 are presently used for DNA-specific staining,[6] but they
require ultraviolet excitation, which can lead to cellular
damage due to lengthy irradiation.[7] SYTO stains do provide
cell-permeable dyes excitable by visible and near-infrared
radiation. Unfortunately they are not specific nuclear stains,[6]
and moreover are of undisclosed chemical structures.
Although the cell-permeant anthraquinone dye DRAQ5
shows red fluorescence emission and DNA-specific labeling,
it is a DNA intercalator, which seriously interferes with the
structure and function of nuclear DNA, in contrast to the
minor groove binders such as SYTO17.[8]
finding has been reported by Thomas et al.,[5] namely that a
dinuclear ruthenium(II) polypyridyl complex (lem(DNA)
=
680 nm) can be used as a nuclear DNA stain for both
luminescence and transmission electron microscopy. Despite
the high hydrophilicity and charge of the molecule, it was still
taken up by live cells, but only if a relatively high concen-
tration of more than 200 mm was used. In fluorescence
microscopic imaging studies, it is desirable for DNA staining
to involve large fluorescence enhancements, and for nuclei to
stain rapidly even when low concentrations of dye are used, as
this minimizes dye toxicity to live cells.
Herein, we report a novel red fluorescent dye DEAB-TO-
3 (Scheme 1, DEAB = (diethylamino)butyl)), a TO-3 ana-
logue with long-wavelength excitation and emission
Therefore, a pressing need exists to develop fluorescent
dyes satisfying the multiple criteria of long-wavelength
excitation/emission, high DNA selectivity, and live-cell per-
meability. Recently, a terbium complex [Tb·L2]+ [2] was
described which stained the nuclear DNA in mitotic cells
using very low dye concentrations (< 1 mm). Unfortunately,
intense irradiation of 350 nm is needed as excitation light.
BENA435,[3] with an N,N-dimethylpropane-1,3-diamine
group, was reported to stain the nucleus in live cells, but
still with excitation/emission located at the cyan region
(lex(DNA) = 435 nm; lem(DNA) = 485 nm). Chang et al.[4] have
discovered an excellent DNA-selective probe, C61, with
relatively long emission wavelength (lem(DNA) = 540 nm) for
live-cell nuclear imaging and DNA quantification. C61 shows
a 19.9-fold fluorescence increase when bound to double-
stranded (ds) DNA (FFDNA = 0.0675). Another promising
Scheme 1. Chemical structures of DEAB-TO-3, TEAB-TO-3, TO-PRO-3,
E-TO-3, and TO-3.
(labs(DNA) = 626 nm and lem(DNA) = 649 nm), to meet the
above demands. Notably, besides avoiding cellular autofluor-
escence interference, another advantage of a red fluorescent
dye is its applicability with a red semiconductor laser (energy
line 633 nm) as light source. This laser is much smaller in size
and more stable than an argon-ion laser (energy lines 488,
514 nm). DEAB-TO-3 shows a very low intrinsic fluorescence
in aqueous solution (FFfree = 0.0037), which is of prime
importance for a fluorescent probe for DNA detection.
Upon binding to calf thymus (CT) DNA, the quantum yield of
DEAB-TO-3 increased 97.3-fold (FFDNA = 0.36), much larger
than the 13.5-fold increase of the commercially available dye
ethidium bromide (EB) under the same conditions (Figure 1).
Furthermore, DEAB-TO-3 fluoresces four times more
strongly in the presence of AT sequences (80.3-fold increase
of fluorescence) than GC sequences (18.9-fold), although it
has very similar affinity for poly(dA-dT)2 and poly(dG-dC)2
(see Figure S1 in the Supporting Information). Previous work
has shown that intramolecular twisting is an efficient quench-
ing pathway of unsymmetrical cyanine dyes in unconstrained
environments, and the huge increase of fluorescence quantum
[*] Prof. X. Peng, T. Wu, J. Fan, S. Zhang, F. Song, S. Sun
State Key Laboratory of Fine Chemicals
Dalian University of Technology
2 Linggong Road, 116024 Dalian (China)
E-mail: pengxj@dlut.edu.cn
Prof. J. Wang
School of Life Science & Biotechnology
Dalian University of Technology (China)
[**] We thank the NSF of China (Nos. 20725621, 20876024, 21076032,
21072024, and 21006009), the National Basic Research Program of
China (2009CB724706), and the Cultivation Fund of the Key
Scientific and Technical Innovation Project (707016) for support of
this work. We also thank Dr. Richard W. Horobin for advice on the
estimation of logP values.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201007386.
4180
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4180 –4183

Figure 3. CD spectra during the titration of a 14 mm solution of DEAB-
TO-3 with CT DNA at 208C in buffer. The [base pair]/[DEAB-TO-3]
molar ratios are 0.33, 0.63, 1.25, 1.67, and 2.5 in (a) and 15, 20, 30,
and 45 in (b). The black dashed line refers to DEAB-TO-3 without CT
DNA. The arrows indicate how the CD bands respond to the increases
in the CT DNA concentration.
Figure 1. a) Fluorescence emission spectra of DEAB-TO-3 (1 mm) and
EB (1 mm) in the absence and presence of CT DNA in buffer at 208C.
b) Fluorescence emission spectra and enhancement of fluorescence
intensities during the titration of a solution of DEAB-TO-3 (1 mm) with
poly(dA-dT)2 and poly(dG-dC)2 at 208C. Inset: The [base pair]/[DEAB-
TO-3] molar ratios are from 0.5 to 7. Quantum yield was measured in
Tris–HCl buffer (pH 7.4). Rhodamine B in methanol was used as a
standard.
CD band at 618 nm and positive CD band at 654 nm, then
finally the single positive CD band at 652 nm (Figure 3b). It is
noteworthy that, besides the hypochromicity and red-shift
phenomena of the absorption band, the obvious increases of
viscosity and Tm of CT DNA (see Figures S3 and S4 in the
Supporting Information) caused by the addition of DEAB-
TO-3 well support our suggestion that DEAB-TO-3 displays
multiple and cooperative associations with the genomic
DNA.
yield upon binding to DNA originates from the loss of
mobility in the constrictive DNA environment.[9] So the light-
up difference between binding to AT and GC segments can
mainly be ascribed to the different deactivation pathways of
the excited state (probably resulting from different binding
modes with AT and GC segments) of DEAB-TO-3. Specif-
ically, fluorescence can be quenched in GC-rich regions by
photoinduced electron transfer (PET) with the guanine
bases,[3,10] which also accounts for the much lower fluores-
cence increase of DEAB-TO-3 in the presence of GC
sequences.
The shapes of both the electronic bands and induced
circular dichroism (ICD) bands seen during the titration of
DEAB-TO-3 with CT DNA indicate that DEAB-TO-3 binds
with a great contribution of minor groove binding to native
DNA. Based on the work reported by Hannah and Armit-
age,[11] the blue-shifted band as complex I (lmax = 586 nm)
shown in Figure 2 can be explained by the face-to-face
stacking of two molecules of DEAB-TO-3, forming an “H”
dimer. When two dimers of DEAB-TO-3 are assembled
adjacent to one another in the minor groove, a secondary
coupling arises because of the end–end interaction between
dimers.[11] This is normally manifested as a split blue-shifted
band displayed as the negative circular dichroism (CD) band
at 580 nm and positive CD band at 616 nm in the CD
spectrum (Figure 3a). For complex II (lmax = 626 nm), a
corresponding transition is seen as the rather small negative
Unlike DEAB-TO-3, the methyl-substituted TO-3
(Scheme 1) shows a half-intercalation binding (no minor
groove binding was seen) with dsDNA at a high [base
pair]/[TO-3] molar ratio (see Figures S5 and S6 in the
Supporting Information). Similar results are also obtained
from the ethyl-substituted E-TO-3 (Scheme 1). So for groove
binding to occur with this type of conjugated structure,
inhibition of intercalation by steric hindrance resulting from
bulky substituents (such as the bulky (diethylamino)butyl
group in DEAB-TO-3) on the DNA probe is necessary.
Early studies had reported that although staining of cell
nuclei was observed for TO-PRO-3, TO-PRO-1, YO-PRO-3,
and propidium iodide (PI), only TO-PRO-3 showed specific
staining of this structure.[12] Moreover, all of these highly
hydrophilic dyes are impermeable to live cells. TEAB-TO-3
(Scheme 1, logPcation = À5.2) with a cationic side chain showed
a very similar staining result to that of TO-PRO-3 (logPcation
À5.0). On the other hand, E-TO-3 (Scheme 1, logPcation
=
=
À0.3) was monocationic and only slightly hydrophilic.
Whilst the protonated dicationic species of DEAB-TO-3 is
more strongly hydrophilic (logPdication = À4.1), the nonproto-
nated monocationic species, which will also be present under
physiological conditions, is less hydrophilic (logPmonocation
=
À2.0) and so DEAB-TO-3 will be membrane permeable. Of
these two dyes only DEAB-TO-3 gave strong staining of
nuclear chromatin, with faint cytoplasmic staining in two cell
lines (Figure 4). E-TO-3, on the other hand, was less selective,
giving stronger staining of cytoplasm. This is perhaps a
consequence of the much more bulky nature of the substitu-
ent of DEAB-TO-3, as discussed above.
Both dual staining and ribonuclease (RNase) digestion
experiments (see Figures S8 and S9 in the Supporting
Information) indicated that the cytoplasmic staining seen
with E-TO-3 resulted from dye binding to ribosomal RNA. It
may be noted in passing that DEAB-TO-3 (both the
Figure 2. Absorption spectra during the titration of a 5 mm solution of
DEAB-TO-3 with CT DNA at 208C in buffer. The [base pair]/[DEAB-TO-
3] molar ratios are 0.33, 0.63, 1.25, 1.67, 2.5, 3.3, and 5 in (a) and 6.5,
8, 10, 12.5, 15, 20, 30, 45, and 60 in (b). The black dashed line refers
to DEAB-TO-3 without CT DNA. The arrows indicate how the absorp-
tion bands respond to the increases in the CT DNA concentration.
Angew. Chem. Int. Ed. 2011, 50, 4180 –4183
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4181

Communications
Moreover, an RNase digestion experiment was performed to
further confirm the DNA selectivity of DEAB-TO-3 in live
cells. As shown in Figure 5b, the nuclear fluorescence
intensity was hardly altered by RNAse digestion. Although
not a 100% DNA-selective probe, DEAB-TO-3 has proved
superior to commercially available DNA probes such as cell-
impermeant EB, PI, thiazole orange homodimer (TOTO),
and the cell-permeant SYTO dyes, which always need treat-
ment with nucleases to distinguish between RNA and DNA in
cells.[6,12]
Since cell cycle analysis by flow cytometry is a significant
application of DNA-specific probes, DEAB-TO-3 was tested
for quantitation of DNA without RNase digestion. We
prepared two distinct cell populations: 1) normal PC12 cells
and 2) PC12 cells induced by glutamate. Since glutamate-
induced apoptosis[15] can lead to the breakdown of the nucleus
and result in decreased cellular DNA content, apoptotic peak
sub-G1 can be observed before the G0 or G1 peak. The results
in Figure 6a and b clearly show that for DEAB-TO-3, little
Figure 4. Live-cell staining with DEAB-TO-3 and E-TO-3. Both dyes
were tested at a 5 mm concentration for HeLa cells and 2 mm for MCF-
7 cells. A 1000ꢀ magnification was utilized in the imaging. Scale bars:
20 mm. DEAB-TO-3 and E-TO-3 (red: Cy5 channel) are shown. The
white light and the merge images were obtained by enlarging one cell
shown in the fluorescence image of MCF-7 cells.
monocationic and dicationic species) fits all the criteria for a
selective nuclear probe in terms of size of the planar aromatic
system, base strength, and logP as specified in the quantitative
structure–activity relationship (QSAR) model of Horobinꢀs
group.[13] Generally, we were able to get an excellent
fluorescence signal of the stained cells at 2 mm for DEAB-
TO-3. Moreover, it is noteworthy that DEAB-TO-3 works in
both live and fixed cells, which allows various experimental
options for cellular imaging.
For many DNA dyes, avoiding the nonspecific binding to
cellular RNA, protein, and other biomacromolecules is a
critical problem in DNA labeling and measurement.[6,12]
Therefore, a set of solution assays was carried out to
investigate the selectivity of DEAB-TO-3 to DNA in vitro.
As shown in Figure 5a, the fluorescence of DEAB-TO-3
Figure 6. Cell cycle and apoptosis analysis. Normal PC12 cells (a,b)
and glutamate-induced PC12 cells (c) were used in this experiment.
The final concentration of DEAB-TO-3 was 1.12 mgmLÀ1 (2 mm). Values
at vertical axes: number of cells.
difference (G2/G1, and peak shape) is observed between
digestion with and without RNase for cell cycle analysis of
normal PC12 cells. Even for the glutamate-induced apoptotic
cells, the results shown in Figure 6c are acceptable (narrow
peak), despite the difficulties arising from nuclear degrada-
tion. By contrast, PI was not capable of giving precise analysis
unless the cells were digested with RNase (see Figure S10 in
the Supporting Information). Note also that the final con-
centration of DEAB-TO-3 is 2 mm (1.12 mgmLÀ1), which
indicates a much higher sensitivity of DEAB-TO-3 than PI
(120 mm; 50 mgmLÀ1) and DRAQ5[16] (10 mm) in flow cytom-
etry.
In summary, we have developed a TO-3 analogue DEAB-
TO-3 (logPmonocation = À2.0) with a (diethylamino)butyl sub-
stituent group playing a significant role in the whole structure.
DEAB-TO-3 shows great fluorescence enhancement (97.3-
fold) when bound to native DNA and a distinct selectivity for
dsDNA over total RNA (six times). DEAB-TO-3, as a red
fluorescent live-cell-permeant DNA minor groove binder, is a
promising candidate for highly sensitive DNA detection in
vitro and nucleus-specific imaging and DNA quantification in
vivo.
Figure 5. a) Fluorescence emission spectra of DEAB-TO-3 in the
absence and presence of CT DNA, S. cerevisiae RNA, and BSA in buffer
at 208C. b) RNase digestion experiment of DEAB-TO-3 (2 mm). Live
MCF-7 cell images of DEAB-TO-3 (2 mm) are shown as a control; equal
exposure was used for dye imaging. A 1000ꢀ magnification was
utilized in the imaging. Scale bars: 20 mm. DEAB-TO-3 (red: Cy5
channel) is shown.
increases 45.5-fold upon binding to the CT DNA, which is
about six times more than that when bound to the same
amount of Saccharomyces cerevisiae RNA (eightfold increase
of fluorescence). Bovine serum albumin (BSA) was tested as
a representative protein to investigate the possibility of
protein binding in the cytoplasm and membrane of subcel-
lular structures.[14] As shown in Figure 5a, DEAB-TO-3
showed almost no change of fluorescence response to BSA.
Received: November 24, 2010
Published online: April 6, 2011
4182
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 4180 –4183

Patterson, Cytometry 2000, 40, 280 – 291; b) K. Wojcik, J. W.
Dobrucki, Cytometry Part A 2008, 73, 555 – 562.
Keywords: cell cycle · DNA · fluorescent probes ·
imaging agents · nucleus staining
.
[9] a) A. Fꢁrstenberg, M. D. Julliard, T. G. Deligeorgiev, N. I.
Gadjev, A. A. Vasilev, E. Vauthey, J. Am. Chem. Soc. 2006,
128, 7661 – 7669; b) G. L. Silva, V. Ediz, D. Yaron, B. A.
Armitage, J. Am. Chem. Soc. 2007, 129, 5710 – 5718.
[10] S. O. Kelley, J. K. Barton, Science 1999, 283, 375 – 381.
[11] K. C. Hannah, B. A. Armitage, Acc. Chem. Res. 2004, 37, 845 –
853.
[12] a) K. Bink, A. Walch, A. Feuchtinger, H. Eisenmann, P. Hutzler,
H. Hꢂfler, M. Werner, Histochem. Cell Biol. 2001, 115, 293 – 299;
b) R. M. Martin, H. Leonhardt, M. C. Cardoso, Cytometry Part
A 2005, 67, 45 – 52.
[13] R. W. Horobin, J. C. Stockert, F. Rashid-Doubell, Histochem.
Cell Biol. 2006, 126, 165 – 175.
[14] a) P. J. Smith, L. H. Patterson, US 6468753B1, 2002; b) P. Panda,
Z. G. Li, M. Szczepanik, W. F. Patton, US 20090226954A1, 2009.
[15] X. Chena, J. Liua, X. S. Gua, F. Ding, Brain Res. 2008, 1238, 189 –
198.
[1] L. J. Kricka, P. Fortina, Clin. Chem. 2009, 55, 670 – 683.
[2] G.-L. Law, C. Man, D. Parker, J. W. Walton, Chem. Commun.
2010, 46, 2391 – 2393.
[3] A. Erve, Y. Saoudi, S. Thirot, C. Guetta-Landras, J.-C. Florent,
C.-H. Nguyen, D. S. Grierson, A. V. Popov, Nucleic Acids Res.
2006, 34, e43.
[4] S. Feng, Y. K. Kim, S. Yang, Y. T. Chang, Chem. Commun. 2010,
46, 436 – 438.
[5] M. R. Gill, J. Garcia-Lara, S. J. Foster, C. Smythe, G. Battaglia,
J. A. Thomas, Nat. Chem. 2009, 1, 662 – 667.
[6] R. P. Haugland, The Handbook: A Guide to Fluorescent Probes
and Labeling Technologies, 2005, chap. 8.1. For more detailed
information see the website: www.invitrogen.com/site/us/en/
home/References/Molecular-Probes-The-Handbook.html.
[7] a) S. K. Davis, C. J. Bardeen, Photochem. Photobiol. 2003, 77,
675 – 679; b) G. P. Pfeifer, Y.-H. You, A. Besaratinia, Mutat. Res.
2005, 571, 19 – 31.
[16] P. J. Smith, N. Blunt, M. Wiltshire, T. Hoy, P. Teesdale-Spittle,
M. R. Craven, J. V. Watson, W. B. Amos, R. J. Errington, L. H.
Patterson, Cytometry 2000, 40, 280 – 291.
[8] a) P. J. Smith, N. Blunt, M. Wiltshire, T. Hoy, P. Teesdale-Spittle,
M. R. Craven, J. V. Watson, W. B. Amos, R. J. Errington, L. H.
Angew. Chem. Int. Ed. 2011, 50, 4180 –4183
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4183