New fluorinated rhodamines for optical microscopy and nanoscopy

Article abstract of DOI:10.1002/chem.200903272

New photostable rhodamine dyes represented by the compounds 1a-r and 3-5 are proposed as efficient fluorescent markers with unique combination of structural features. Unlike rhodamines with monoalkylated nitrogen atoms, N′,N-bis(2,2,2-trifluoroethyl) derivatives 1e, 1i, 1j, 3-H and 5 were found to undergo sulfonation of the xanthene fragment at the positions 4′ and 5′. Two fluorine atoms were introduced into the positions 2′ and 7′ of the 3′,6′-diaminoxanthene fragment in compounds 1a-d, 1i-l and 1m-r. The new rhodamine dyes may be excited with λ= 488 or 514 nm light; most of them emit light at λ = 512-554 nm (compounds 1q and 1r at λ=576 and 589 nm in methanol, respectively) and have high fluorescence quantum yields in solution (up to 98%), relatively long excited-state lifetimes (>3ns) and are resistant against photobleaching, especially at high laser intensities, as is usually applied in confocal microscopy. Sulfonation of the xanthene fragment with 30% SO₃ in H₂SO₄ is compatible with the secondary amide bond (rhodamine-CON(Me)CH₂CH ₂COOH) formed with MeNHCH₂CH₂COOCH₃ to providing the sterically unhindered carboxylic group required for further (bio)conjugation reactions. After creating the amino reactive sites, the modified derivatives may be used as fluorescent markers and labels for (bio)molecules in optical microscopy and nanoscopy with very-high light intensities. Further, the new rhodamine dyes are able to pass the plasma membrane of living cells, introducing them as potential labels for recent live-cell-tag approaches. We exemplify the excellent performance of the fluorinated rhodamines in optical microscopy by fluorescence correlation spectroscopy (FCS) and stimulated emission depletion (STED) nanoscopy experiments.

Full text of DOI:10.1002/chem.200903272

FULL PAPER
DOI: 10.1002/chem.200903272
New Fluorinated Rhodamines for Optical Microscopy and Nanoscopy
Gyuzel Yu. Mitronova, Vladimir N. Belov,* Mariano L. Bossi, Christian A. Wurm,
Lars Meyer, Rebecca Medda, Gael Moneron, Stefan Bretschneider, Christian Eggeling,*
Stefan Jakobs, and Stefan W. Hell*[a]
Abstract: New photostable rhodamine
dyes represented by the compounds
1a–r and 3–5 are proposed as efficient
fluorescent markers with unique com-
bination of structural features. Unlike
rhodamines with monoalkylated nitro-
gen atoms, N’,N-bis(2,2,2-trifluoroeth-
yl) derivatives 1e, 1i, 1j, 3-H and 5
were found to undergo sulfonation of
the xanthene fragment at the positions
4’ and 5’. Two fluorine atoms were in-
troduced into the positions 2’ and 7’ of
the 3’,6’-diaminoxanthene fragment in
compounds 1a–d, 1i–l and 1m–r. The
new rhodamine dyes may be excited
with l=488 or 514 nm light; most of
them emit light at l=512–554 nm
(compounds 1q and 1r at l=576 and
589 nm in methanol, respectively) and
have high fluorescence quantum yields
in solution (up to 98%), relatively long
excited-state lifetimes (>3 ns) and are
resistant against photobleaching, espe-
cially at high laser intensities, as is usu-
ally applied in confocal microscopy.
Sulfonation of the xanthene fragment
with 30% SO3 in H2SO4 is compatible
with the secondary amide bond (rhoda-
mine-CON(Me)CH2CH2COOH)
carboxylic group required for further
(bio)conjugation reactions. After creat-
ing the amino reactive sites, the modi-
fied derivatives may be used as fluores-
cent markers and labels for (bio)mole-
cules in optical microscopy and nano-
scopy with very-high light intensities.
Further, the new rhodamine dyes are
able to pass the plasma membrane of
living cells, introducing them as poten-
tial labels for recent live-cell-tag ap-
proaches. We exemplify the excellent
performance of the fluorinated rhoda-
mines in optical microscopy by fluores-
cence correlation spectroscopy (FCS)
and stimulated emission depletion
(STED) nanoscopy experiments.
formed with MeNHCH2CH2COOCH3
to providing the sterically unhindered
Keywords: dyes/pigments · fluores-
cence
· fluorine · microscopy ·
rhodamines
Introduction
let or dark-state formation, high photostability, good solubil-
ity in water and high labeling capability of, for example, bio-
molecules determine the usability of a fluorophore.[1] Photo-
stable and bright organic dyes served well in recent ad-
vancements of optical microscopy, such as the detection of
single molecules[2] or fluorescence nanoscopy.[3] Single-mole-
cule based detection (SMD) technologies, such as fluores-
cence correlation spectroscopy (FCS)[4] or fluorescence in-
tensity distribution analysis (FIDA) or photon-counting his-
togram analysis (PCH)[5] may efficiently determine molecu-
lar heterogeneities and have even successfully been applied
in high-throughput drug discovery,[6] but they require the ac-
quisition of as much fluorescence signal as possible in the
shortest time. That is why large laser intensities are often
applied in SMD, demanding a new degree of photostability
of the dyes in use, as such large intensities promote the pop-
ulation of highly excited electronic states and, consequently,
the very efficient photobleaching of the fluorescent dyes.[7]
High laser intensities are also required in imaging tech-
niques, such as stimulated depletion emission (STED) nano-
Fluorescence microscopy is one of the most versatile tools
in modern life sciences. Its sensitivity mainly depends on the
brightness of the applied fluorescent markers. Photophysical
and -chemical parameters, such as a high-absorption cross-
section, high fluorescence quantum yield, low degree of trip-
[a] Dr. G. Y. Mitronova, Dr. V. N. Belov, Dr. M. L. Bossi,
Dr. C. A. Wurm, Dr. L. Meyer, Dr. R. Medda, Dr. G. Moneron,
Dr. S. Bretschneider, Dr. C. Eggeling, Dr. S. Jakobs,
Prof. Dr. S. W. Hell
Department of Nano-Biophotonics
Max Planck Institute for Biophysical Chemistry
Am Fassberg 11, 37077 Gçttingen (Germany)
Fax : (+49)551-2012505
E-mail: vbelov@gwdg.de
ceggeli@gwdg.de
shell@gwdg.de
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200903272.
Chem. Eur. J. 2010, 16, 4477 – 4488
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4477

scopy,[3a,d] a fluorescence microscopy technique that has
truly surpassed the spatial resolution limit of conventional
far-field microscopy given by diffraction.[8] Here, strong ad-
ditional laser light allows for the spatially-selective inhibi-
tion of the markerꢁs fluorescent state by stimulated emis-
sion, thereby restricting the remaining fluorescence signal to
sub-diffraction sized areas. Although, the additional laser
light is not really absorbed by the dye, these large intensities
also call for photostable markers with a low degree of dark
(triplet) state population.[9]
Further, cellular immunolabeling and microscopy applica-
tions, such as multicolor detection and colocalization experi-
ments require a flexible choice of reagents, that is, the avail-
ability of the whole gamut of photostable labels. Rhodamine
dyes bring along all of the desired photophysical and -chem-
ical characteristics and may serve as a basis for further, ver-
satile improvements. Rhodamines have thus proven advan-
tageous in modern fluorescence microscopy or SMD appli-
cations.[10] Here, we present new bright and photostable flu-
orinated rhodamine dyes based on the structure of Rhoda-
Results and Discussion
The route to new fluorinated rhodamines: So far, several
classes of the photostable organic substances have been
used as scaffolds for the design of improved fluorescent
dyes, for example, pyrenes, terrylenes and other condensed
polycyclic aromatic compounds, coumarines, rhodols, rhoda-
mines and BODIPY derivatives. Further, attempts to device
and improve photostable dyes of other classes, especially
emitting in the red-spectral region, are being actively under-
taken first of all for the water-soluble terrylenediimides[11]
and dicyanomethylene dihydrofuranes.[12] However, not all
of the above mentioned dyes can easily be chemically modi-
fied and improved. For example, it was rather difficult to
modify the BODIPY residue in such a way, that it becomes
water soluble.[13] On the contrary, many rhodols or rhoda-
mines may easily be decorated with two sulfonic acid groups
simply by sulfonation with SO3 in H2SO4. Sulfonation is
known not only to improve solubility in aqueous media, but
it also considerably reduces aggregation and quenching of
fluorescence in water or aqueous buffers. Introduction of
the sulfonic acid residues is known to often increase the
photostability and fluorescence quantum yield of a fluores-
cent dye in solution and after attaching it to a biological
macromolecule.[14] In the case of fluorescein, substitution of
the aromatic hydrogen atoms at the positions 2’ and 7’ of
the xanthene fragment is known to increase photostability,
without changing the absorption and emission spectra.[15]
The same may be true for rhodamines: fluorination of the
xanthene fragment at these positions is very likely to in-
crease the photoresistance of a dye, keeping other spectral
properties intact. Surprisingly, no data on 2’,7’-difluorosub-
stituted rhodamines have been reported in the literature.[16]
However, fluorinated rhodol derivatives have been dis-
closed.[17]
Rhodols—3-amino-6’-hydroxyxanthenes—have
lower absorption coefficients than the corresponding fluo-
rescein or rhodamine derivative and their fluorescent quan-
tum yields (Ffl) are lower than these of fluorescein deriva-
tives or rhodamines.[18] Substituted rhodols are used as
sensor for metal ions, but not as bright and photostable fluo-
rescent labels in microscopy and optical imaging. Another
new and important feature of the rhodamines introduced in
the present study is a combination of the two 2,2,2-trifluor-
oethyl amino groups at the positions 3’ and 6’ with the two
sulfonic acid residues at the positions 4’ and 5’ of the xan-
thene ring. In 1989 and 1991 a very limited number of N,N’-
bis(2,2,2-trifluoroethyl)-3’,6’-diaminoxanthenes have been
already disclosed as photostable analogues of Rhodami-
ne 6G but these patents[19] have very narrow claims and do
not deal with any of the compounds presented here. We
chose the 2,2,2-trifluoroethyl residue as a substituent for the
amino group for the following reasons.[3r] In some respects,
it behaves like a hydrogen atom. For example, 2,2,2-trifluoro-
propionic acid (CF3CH2COOH) is as strong as formic acid
(HCOOH). This means that the electronic inductive effect
of the CF3CH2-group is equal to the inductive effect of the
hydrogen atom. Fluorination is known to increase the lipo-
Scheme 1. New fluorinated rhodamines prepared in this study.
mine 110 (see Scheme 1). The new rhodamines are
efficiently excited between l=488 and 515 nm and emit
light at around l=520 nm with high fluorescence quantum
yields close to one. They have long fluorescence lifetimes of
about 4 ns, low triplet yield and feature an extreme photo-
stability, especially at large laser intensities and above all,
their water solubility allows an efficient coupling to biomol-
ecules. The STED nanoscopy images with superior spatial
resolution over conventional microscopy, single-molecule
based FCS measurements and the capability to penetrate
into live-cells prove the new rhodamines to be versatile
markers for modern microscopy and nanoscopy.
4478
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4477 – 4488

Rhodamines for Optical Microscopy and Nanoscopy
FULL PAPER
philicity and reduce the solubil-
ity in water. Sulfonation com-
pensates this effect and restores
the hydrophilic properties of
dyes, providing a fairly good
solubility in water. Sulfonation
is known to shift very slightly
both the absorption and the
emission bands to the blue (up
to ꢀ5 nm),[3p] but it does not
alter the spectral properties
considerably. We found that the
presence of one CH2CF3 group
at the nitrogen atom in rhoda-
mine does not inhibit the sulfo-
nation with SO3 in H2SO4. On
the
other
hand,
the
NHCH2CF3-groups in the deriv-
atives of amino acid are resist-
ant to acylation (in the pres-
ence of “normal” primary
amines).[20] Thus, this residue
provides a sufficient chemical
protection for the amino group
and it also reduces the number
of NH atoms prone to photo-
bleaching reactions.
Scheme 2. Synthesis of N,N’-bis(2,2,2-trifluoroethyl)rhodamines 1e and 1 f, amides 3-R and 5, sulfonated
amide 4-OH, as well as NHS-esters 1 f-6’-CONHS and 4-NHS: a) 1608C, 3 h, then second half of 2, 85% aq.
H3PO4, 1608C, 3 h; b) N-hydroxysuccinimide, O-(7-azabenzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium·PF6
ꢁ
(HATU), DMF, RT, overnight; c) POCl3, 1,2-dichloroethane, reflux, 4 h; then a secondary amine, Et3N,
MeCN, RT, 20 h; d) 1m aq. NaOH, MeOH, THF, 08C!RT, 2 h; e) 30% SO3 in H2SO4, 08C, 2–3 days;
ꢁ
f) N,N,N’,N’-tetramethyl-O-(N-succinimidyl)uranium·BF4 (TSTU), Et3N, DMF, RT, 24 h.
Rhodamine 110 (Rh110) is a
convenient reference and start-
ing compound for the introduc-
tion of the structural modifications mentioned above. Rh110
is a well known laser dye,[21] with an absorption maximum at
l=496 nm and emission at l=520 nm (in water). It exhibits
a bright-green fluorescence (Ffl =0.92 in basic ethanol).[22]
For the reasons mentioned above, all derivatives of Rh110
depicted in Scheme 1 (compounds 1a–l) are expected to
have the same absorption and emission spectra as the
parent dye.
ester 3-Me smoothly gave the acid 3-H, which was trans-
formed into the NHS-ester 3-NHS. N,N’-Disuccinimidylcar-
bonate in the presence of a base transforms alcohol 5 into
the corresponding NHS-carbonate. Rhodamines with two
carboxylic groups (e.g., 1 f-6’-COOH) react with one equiva-
lent of N-hydroxysuccinimide and O-(7-azabenzotriazol-1-
ꢁ
yl)-N,N,N’,N’-tetramethyluronium·PF6 (HATU), or with 1
equivalent of N,N,N’,N’-tetramethyl-O-(N-succinimidyl)uro-
nium tetrafluoroborate (TSTU) in the presence of a base
and yield mono NHS-esters with the sterically less hindered
carboxylic group activated by the formation of an ester with
a good leaving group (e.g., 1 f-6’-CONHS).
We observed that the secondary amide group in rhoda-
mines is quite stable under sulfonation conditions (30% SO3
in H2SO4, 08C) for several days. Thus, sulfonation of the
acid 3-H affords disulfonic acid 4-OH with an intact amide
bond (Scheme 2). It was found to be much easier to obtain
the corresponding rhodamine amide first and then to sulfo-
nate it, than to perform the reversed procedure, which re-
quires two tedious reversed-phase chromatographic separa-
tions instead of one and in which the yield at the amidation
step is lower.
Synthesis of the fluorinated and sulfonated rhodaminic dyes
and their NHS esters: Synthetic routes to the new fluorinat-
ed rhodamines are given in Schemes 2–6. The synthesis and
amidation of rhodamine 1e have been already described.[3r]
Condensation of aminophenol 2 with 4-carboxyphthalic an-
hydride led to the mixture of 5’- and 6’-carboxyrhodamines
1 f-5’/6’-COOH (Scheme 2).
By the repeated recrystallization of this mixture from
aqueous iPrOH, the nearly pure regioisomer 1 f-5’-COOH
may be isolated (with less than 4% of 1 f-6’-COOH). Evap-
oration of the combined mother liquids followed by re-
versed-phase chromatography with MeOH/H2O mixture
(1:4) gave the second pure diastereomer (1 f-6’-COOH). De-
rivatisation of the monocarboxy rhodamine 1 f is necessary
for bioconjugation.[23] For that, rhodamine 1 f was converted
into the corresponding acid chloride and the latter reacted
with methyl 3-[(N-methyl)amino]propionate or 2-[N-(meth-
ylamino)]ethanol (Scheme 2). Saponification of the methyl
Commercially available 2-fluoro-5-nitroaniline (7) was
used as a starting material for the synthesis of other fluori-
nated rhodamines (Scheme 3). It was transformed to N-
methyl- (8-NO2-Me) and N,N-dimethylaniline 6-NO2 and
then nitro groups in these compounds and in the intermedi-
Chem. Eur. J. 2010, 16, 4477 – 4488
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4479

V. N. Belov, C. Eggeling, S. W. Hell et al.
be even less than that of 2,2,2-trifluoroethyl groups. Proba-
bly, N-alkylamino groups increase the basicity of nitrogen
atoms (compared with unsubstituted or mono 2,2,2-trifluoro-
ethyl substituted ones) and an excess of SO3 in H2SO4
forms a very stable complex only with mono N-alkylamino
groups, inhibiting the aromatic SE-reaction. Sulfonation was
found to be always regioselective: only C-4’ an C-5’ atoms
were sulfonated. This fact may be explained by the lower lo-
calization energy for the position 1 than for the position 2 in
Scheme 3. Preparation of 2-fluoro-5-nitroanilines 6-NO2 and 8-NO2:
a) 35% aq. H2CO, NaBH4, THF, 3m aq. H2SO4, RT; b) (CF3CO)2O,
CH2Cl2, Et3N, 0 8C!RT, overnight; c) MeI, K2CO3, DMF, 908C, 24 h.
ate 8-NO2-H were reduced (Schemes 4 and 5). The next im-
portant step included the diazotation of the anilines 8-NH2-
R (Scheme 4) and 6-NH2 (Scheme 5) followed by the trans-
formation of diazonium salts into phenols 8-OH-R and 6-
OH under very mild conditions.[24] Compound 6-OH was di-
rectly used in the synthesis of rhodamine 13-Me-Me-H (1q)
with one carboxylic acid group and the mixture of re-
gioisomers 13/14-Me-Me-COOH (1r) with two carboxylic
acid groups (Scheme 6).
N-Trifluoroacetylated
m-aminophenols
8-OH-R
(Schemes 4 and 5) were the key intermediates for the fol-
lowing transformations. They were either reduced to 3-hy-
droxy-N-(2,2,2-trifluoroethyl)anilines 10-R (Scheme 4), or
deprotected to anilines 12-R1 (Scheme 5). m-Aminophenols
10-R and 12-R1 were condensed with phthalic or trimellitic
anhydrides and rhodamines 9-H-H (1i), 9/11-H-COOH (1j),
9-Me-H (1o), 9/11-Me-COOH (1p), 13-H-H-H (1a), 13/14-
H-H-COOH (1b), 13-Me-H-H (1m) and 13/14-Me-H-
COOH (1n) were obtained. All rhodamines in Schemes 4
and 5 were obtained in the course of the condensation reac-
tion in the presence of fused anhydrous ZnCl2 and isolated
from the reaction mixtures by column chromatography on
silica gel. Unoptimized yields were found to be moderate.
The stepwize condensation reaction exemplified in
Scheme 2 can be applied to two various m-hydroxyanilines
chosen from the set of compounds 2, 10-R (Scheme 4) and
12-R (Scheme 5). This approach
Scheme 4. Synthesis of 2,7-difluoro-N,N’-bis(2,2,2-trifluoroethyl)rhoda-
mines 9 and 11 (1i–j, 1o–p): a) 10% Pd/C, H2, EtOAc, RT, 7 h; b) aq.
NaNO2, 35% aq. H2SO4, 08C, then aq. CuACHTUNGTRENGN(U NO3)2·3H2O, Cu2O, RT; c) 1 m
BH3·THF, THF, 0!908C, 24 h; d) fused ZnCl2, 180–2008C, 3–4 h.
may enable the synthesis of the
unsymmetrical
rhodamines
(3’,6’-diaminoxanthenes) with
various combinations of the
substituents.
The feasible sulfonation reac-
tions are given in Scheme 6. As
in the case of compound 3-H
(Scheme 2),
N-(2,2,2-trifluor-
oethyl) substituents did not pre-
vent the sulfonation and the
corresponding disulfonic acids
1k–l and 1h-5’-COOH were ob-
tained (Scheme 6). The pres-
ence of methyl groups at the ni-
trogen atoms in rhodamines in-
hibited the sulfonation reac-
tions completely, though the
steric hindrance caused by
Scheme 5. Synthesis of 2,7-difluororhodamines 13 and 14 (1a–b, 1m–n, 1q–r): a) 30% aq. NaOH, MeOH,
908C, 24 h; b) fused ZnCl2, 180–2008C, 3–4 h; c) 10% Pd/C, H2, EtOAc, RT, 7 h; d) aq. NaNO2, 35% aq.
methyl groups is supposed to H2SO4, 08C, then aq. CuACHTUNGTRNE(NUG NO3)2·3H2O, Cu2O, RT.
4480
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4477 – 4488

Rhodamines for Optical Microscopy and Nanoscopy
FULL PAPER
minor changes in the spectra
are observed in the biologically
relevant pH region of 6–9. Fur-
ther, most of the fluorinated
rhodamines presented here
have fluorescence lifetimes of
3–4 nanoseconds that are typi-
cal for rhodamines. The spectral
properties, extinction coeffi-
cients and fluorescence lifetime
values of the fluorinated rhoda-
mines are summarized in
Table 1.
A
strong absorption maxi-
mum at l=290–300 nm, charac-
teristic for the compounds 1o
and 1p, indicates the presence
of closed non-fluorescent forms
with a spirolactone fragment.
For example, compound 1o
gives colorless solutions in
methanol, water and neutral
buffers. In acidic solutions,
Scheme 6. Sulfonation of N,N’-bis(2,2,2-trifluoroethyl)rhodamines 1k–l, 1 f-5’-COOH, as well as N,N’-unsubsti-
tuted 2,7-difluororhodamines 1j, 1a–b and alcohol 5: a) 30% SO3 in H2SO4, 08C, 2–3 days; b) N-
hydroxysuccinACHTUNGTRENNUNGimide, HATU, DMF, RT, overnight.
9,10-disubstituted anthracenes (if the reaction is thermody-
namically controlled).
Activation of the sterically available carboxylic group in
rhodamines may be achieved by the formation of the active
esters (e.g., NHS-esters 1 f-6’-CONHS, 3-NHS and 4-NHS in
Scheme 2). Together with compound 15-H-CH2CF3-H-5’-
CONHS in Scheme 6, these esters were used in bioconju-
gation reactions described in one of the following sections.
Figure 1. Spectral properties. Absorption and fluorescence emission spec-
tra of compound 13-H-H-H (1a) in the pH range 1–7.
Spectral properties: Most of dyes of the present study are
highly fluorescent and all of them are photostable com-
pounds in water, aqueous buffers, alcohols and typical
mounting media used in standard fluorescence microscopy
techniques, such as polyvinyl alcohol (PVA), or its mixtures
with water and glycerol. The majority of the compounds
form colored and fluorescent solutions in a wide pH range
from pH 1 to 12. They have absorption maxima between l=
480 and 515 nm with high extinction coefficients of up to e
ꢀ9ꢂ104 mꢁ1 cmꢁ1 and emit in the l=510–535 nm region with
high emission efficiencies, that is, fluorescence quantum
yields of up to 0.95. Their solutions are yellow or light
orange in daylight and have a yellow-greenish or yellow
emission when excited with blue light. Compounds 9-Me-H
(1o), 13-Me-Me-H (1q) and 13/14-Me-Me-COOH (1r)
absorb and emit at l=525–560 nm and 555–580 nm, respec-
tively. Their solutions have a pink or red color. An example
of the absorption and emission spectra for compound 13-H-
H-H (1a) is shown in Figure 1.
compound 1o exists in an open acidic form with relatively
low fluorescence efficiency. The acid-base equilibrium (see
Scheme S1 in the Supporting Information), which is respon-
sible for the absence of color and fluorescence of compound
1o in neutral solutions, was not observed for the other com-
pounds.
Photostability at low laser intensities: The photostability of
a fluorophore determines the observation time of fluores-
cence microscopy experiments and is mainly characterized
by the chemical “fragility” of the excited states.[1,25] As men-
tioned, the photostability of rhodamines depends on the
substitution pattern in the 3’,6’-diaminoxanthene fragment.
Comparison of Rhodamine 6G (with disubstituted nitrogen
groups), Tetramethyl rhodamine (TMR; with trisubstituted
nitrogen atoms) and methyl ester of Rhodamine 110 (with
monosubstituted nitrogen groups) showed that at low light
intensities (up to 103 Wcmꢁ2) TMR was the most stable dye.
Photobleaching quantum yields in water were found to be
12, 3.3 and 6.6ꢂ10ꢁ7, respectively.[7] It is interesting that
In acidic solutions (pHꢂ5), a small bathochromic shift in
absorption and emission maxima (typical 5 nm) was com-
monly observed (Figure 1). It may be caused by the proto-
nation of the benzoic acid moiety. However, none or only
Chem. Eur. J. 2010, 16, 4477 – 4488
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4481

V. N. Belov, C. Eggeling, S. W. Hell et al.
Table 1. Spectral properties of the fluorinated rhodamines.
step of the bleaching reaction
also involves radical species as
Compound
Absorption
Emission
lmax [nm] Ffl
[a]
lmax [nm]
e [ꢂ10ꢁ4 mꢁ1 cmꢁ1
]
tfl [ns][a]
intermediates. According to the
Hammondꢁs postulate, the ge-
ometry of a transition state of
an endothermic reaction lead-
ing to a radical intermediate is
better approximated by the ge-
ometry of a radical than by the
geometry of the initial sub-
stance (reactant). This means
that any effects (de)stabilizing
the intermediate radical species
will (de)stabilize the transition
state, as well. The fluorine
atoms with their strong nega-
tive inductive effect (ꢁI-effect)
not only substitute hydrogen
atoms and reduce the probabili-
ty of their abstraction in the
course of bleaching reactions,
but also destabilize radical cen-
ters. In this case, the ꢁI-effect
of fluorine atoms in the inter-
mediate radicals should in-
crease the activation energy of
their formation (compared with
the similar reaction of a non-
fluorinated substrate) and slow-
down photobleaching.
1a (13-H-H-H)
494,[a] 498[b]
4.1[b]
520,[a]
522[b]
522,[a]
523[b]
512,[a]
514[b]
514,[a]
517[b]
522[a]
525,[a]
524[b]
523,[a]
522[b]
520,[a]
524[b]
524,[a]
524[b]
521,[a]
519[b]
522,[a]
520[b]
531[a]
533,[a]
541[b]
539,[a]
537[b]
543,[a]
538[b]
0.87
0.92
0.89
0.97
3.9
4.0
4.0
4.1
1b (13/14-H-H-COOH)
1c (15-H-H-F-H)
496,[a] 498[b]
488,[a] 498[b]
490,[a] 502[b]
5.0[b]
4.2[b]
3.2[b]
1d (15-H-H-F-COOH)
1e
499[a]
8.1[a]
8.5[a]
0.88
0.95
3.9
4.1
1 f-5’-COOH
501,[a] 500[b]
1 f-6’-COOH
501,[a] 497[b]
499,[a] 507[b]
8.8[a]
8.7[b]
8.6[b]
4.1[b]
5.3[a]
0.93
0.88
0.96
1.00
0.89
–
1g (15-H-CH2CF3-H-H)
3.8
3.9
3.8
3.8
1h-5’-COOH (15-H-CH2CF3-5’-COOH) 501,[a] 511[b]
1i (9-H-H)
494,[a] 500,[b]
AHCTUNGTRENNUNG
(292)[2b]
1j (9/11-H-COOH)
496,[a] 498[b]
1k (15-H-CH2CF3-F-H)
1 L (15-H-CH2CF3-F-COOH)
506[a]
5.4[a]
4.2[b]
0.66
0.86
3.2
3.3
509,[a] 517[b]
1m (13-Me-H-H)
513,[a] 511[b]
515,[a] 515[b]
5.6[b]
5.6[b]
0.87
0.86
4.0
4.1
1n (13/14-Me-H-COOH)
1o (9-Me-H)
1p (9/11-Me-COOH)
1q (13-Me-Me-H)
A
A
A
(0.20)[1] (1.2)[1]
ACHUTGTNRENNUG CAHTUNGTRENNUGN
–
0.02
E
–
–
–
3.3
554,[a] 546[b]
5.0[a]
586,[a]
576[b]
554,[a]
589[b]
534[b]
531,[a]
534[b]
530,[a]
536[b]
531,[a]
534[b]
530[a]
1r (13/14-Me-Me-COOH)
525,[a] 558[a+b]
3.1[a]
0.28
1.6
Our new fluorinated rhoda-
mines display an outstanding
photostability when irradiated
for a prolonged time with low
laser intensities (<kWcmꢁ2).
Figure 2 shows the emission in-
tensity of the stirred solutions
with micromolar concentrations
of some exemplary compounds,
when irradiated with l=488 nm
light in a 1 cm-path quartz cuv-
3-Me
3-H
512[b]
8.7[b]
5.7[a]
–
0.86
–
4.2
510,[a] 512[b]
4-OH
5
512,[a] 517[b]
510,[a] 512[b]
512[a]
7.4[b]
7.1[b]
5.2[a]
0.86
0.86
0.86
4.1
4.2
4.1
5-SO3H
[a] In an aqueous buffer at pH 7. [b] In methanol. [a+b] Mixture. [1] Values measured at pH 1. [2] Value for
the closed form.
under these conditions, methyl ester of Rh110 bleached
slower than Rhodamine 6G, though the degree of substitu-
tion at the nitrogen atoms of latter compound is higher. This
fact may be explained by the presence of two methyl groups
at the positions 2’ and 7’ of Rhodamine 6G: they block the
carbon atoms of the xanthene heterocycle, but, at the same
time, these methyl groups may give rise to the relative
stable benzylic radicals—good carriers in the chain radical
processes during photobleaching. Another possible explana-
tion relies on the relative polarities of the dyes: Rhodami-
ne 6G is much less polar and poorly solvated in water than
the methyl ester of Rh110.
ette, at an intensity of 3 Wcmꢁ2. The initial absorption at
the irradiation wavelength was the same for all solutions.
The compounds are far more stable than the common dyes
excitable at l=488 nm, such as Fluoresceine and Oregon
Greenꢃ 488 and have similar or even better performance
than Rhodamine Green or Alexa Fluor 488. For example,
aqueous solutions of compounds 1g, 4-OH and 1h can be ir-
radiated for several hours with the l=488 nm line of an
Argon laser without any significant changes in their emis-
sion. After two hours of irradiation, compounds 1e and 1 f
retain more than 50% of the initial emission intensity and
compound 3-H has about 90% of the initial signal left.
Further, photobleaching of rhodamines involves energy-
rich radical intermediates (which are electron-deficient, as
they lack one electron of the stable electron octet).[26] There-
fore, it is quite reasonable to assume that the rate-limiting
Performance at high laser intensities: Maximizing the sensi-
tivity of fluorescence microscopy often requires maximizing
the fluorescence rate, that is, the fluorescence emission
4482
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4477 – 4488

Rhodamines for Optical Microscopy and Nanoscopy
FULL PAPER
Figure 3. Photostability at high laser intensities. FCS curves obtained
from Rh110-CO2H (left) and compound 1h (B) in aqueous solution at
ꢁ2
relatively low ( ; 0.1 MWcm ) and high ( ; 2 MWcmꢁ2) excitation in-
tensity (dots: experimental data; lines: best fits). The new rhodamine 1h
shows about the same triplet buildup (decay in ms time range) but much
less photobleaching (smaller shift of ms decay to shorter correlation
times at larger laser intensities) than Rh110.
*
*
Figure 2. Photostability at low excitation intensities. Decrease of fluores-
cence signal over time due to photobleaching with 3 Wcmꢁ2 of l=
488 nm laser light of selected new rhodamines (1e–h, 3-H and 4-OH)
and, for comparison, of some commercially available fluorescent dyes
with the similar spectral properties (Alexa Fluor 488, Rhodamine Green,
Oregon Green 488 and Fluorescein) in aqueous solutions. At low laser
intensities, some of the new rhodamines allow for observation times of
several hours, superior over some other commercially available dyes.
between compound 1h and Rh110. The second decay of the
correlation curve in the ms time range characterizes the ob-
servation time of the fluorophore, which is usually given by
the diffusion time through the laser focus. However, at large
laser intensities, photobleaching may dominate the observa-
tion time, thus resulting in a shortening of this decay.[7,30b]
The increased photostability of compound 1h becomes obvi-
ous.
Fitting of the FCS data resulted in values of the intersys-
tem crossing rate kisc, of the triplet de-excitation rate con-
stant kT (inverse triplet lifetime tT) and of the effective pho-
tobleaching cross-section sbl from higher excited electronic
states (see the Supporting Information).[7,26a,30b,31] The lower
kisc, kT and sbl, the better is the performance of the dye at
high laser intensities, that is, the population of the dark trip-
let state is lower and the photobleaching reactions from
higher excited electronic states are less probable. Table 2 de-
within a certain time span, which is usually approached by
applying large excitation intensities of laser light. For exam-
ple, in common microscopy, such as confocal or single-mole-
cule based techniques, fluorescent markers are subject to ex-
citation intensities much higher than those mentioned
above, and are in the range of a few kWcmꢁ2. The compara-
tive photostabilities presented in Figure 2 may be incorrect
under these irradiation conditions, as new photobleaching
pathways open up. At high laser intensities, the population
of long-lived dark (triplet) states, as well as higher excited
electronic states, is much more pronounced.[7,27] These
states, on one hand, temporally dislodge the dye from the
excitation-emission cycle, thus leading to a loss in brightness.
On the other hand, they are usually much more reactive,
thus introducing much more profound photobleaching.[7,28]
For example, in aqueous solution, higher excited electronic
states couple quite efficiently with highly-reactive radical
states.[7,29]
Table 2. Photostabilities and triplet parameters of the fluorinated rhoda-
mines in an aqueous buffer at pH 7 at large laser intensities.
Compound
Triplet parameters
Photobleaching
[a]
[b]
kisc [106 sꢁ1
]
kT [105 sꢁ1
]
sbl [10ꢁ21 cm2][c]
Rh110
0.8
0.7
0.9
1
1.5
2
3
1.5
5
2
5
5
2
1.5
0.4
0.9
0.45
We applied fluorescence correlation spectroscopy (FCS)[4]
to get further insights on the triplet-state population and
photobleaching of the new rhodamines at large excitation
intensities.[7,26a,30] Figure 3 shows correlation data of com-
pound 1h and for comparison of Rh110 in aqueous solution
at relatively low (115 kWcmꢁ2) and at very large intensities
(2 MWcmꢁ2) of l=488 nm laser light. On one hand, the
quality of the correlation data depicts that our compounds
perform equally well in single-molecule based experiments,
such as FCS, like the well-established dye Rh110. On the
other hand, the curves are characterized by two decays. The
first one in the ms time range is given by population and de-
population of the triplet state. The decay time at low inten-
sities characterizes the triplet-state lifetime, whereas the am-
plitude reveals the relative triplet population.[30a] No signifi-
cant difference in triplet lifetime and population is observed
Alexa 488
1 f-5’-COOH
1h-5’-COOH
5
5-SO3H
2.5
[a] Intersystem crossing rate. [b] Triplet de-excitation rate (inverse triplet
lifetime). [c] Effective photobleaching cross-section; all parameters deter-
mined by FCS with standard deviations of about 10–20%.
picts the values of kisc, kT and sbl of the new rhodamine com-
pounds and, for comparison, of the commercial rhodamines,
Rh110 and Alexa Fluor 488. The values confirm that triplet
population of the new compounds is comparable to other
rhodamines, but photobleaching is efficiently reduced. Note
that sulfonation of compound 1 f leads to an improved pho-
tostability of compound 1h.
Chem. Eur. J. 2010, 16, 4477 – 4488
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4483

V. N. Belov, C. Eggeling, S. W. Hell et al.
Clearly, the new rhodamines are advantageous in fluores-
cence microscopy, both for long observation times at low
laser intensities and for maximum brightness and photosta-
bility at high laser intensities, that is, under conditions, such
as single-molecule based experiments for which the maxi-
mum number of photons per time is crucial.
Labeling characteristics: The amino reactive derivatives of
the dyes presented here can easily be attached to antibodies
(AB) or other bio-molecules in the course of the standard
labeling and immunostaining procedures. The labeling of
the AB with a reactive (fluorescent) dye involves the disso-
lution of both components (e.g., in phosphate buffer-saline
[PBS] at pHꢀ8.5) followed by incubation for 1–2 h at room
temperature. For conjugation, a 15–20-fold molar excess of
a dye dissolved in anhydrous DMF is used. Poor solubility
of a dye in aqueous buffers may result in precipitation and
low labeling efficiency (low average number of dye mole-
cules per AB), or also some unspecific binding may occur.
Another problem, which may arise, is a decrease in the fluo-
rescence quantum yield of the dye-AB adducts. Lipophilic
fluorescent dyes are often less bright in polar environments
or after binding with proteins. The studied dyes are suitable
for conjugation with commercial ABs, yielding highly fluo-
rescent adducts with only minor alterations of the spectral
properties. For instance, the absorption and emission spectra
of the dye 1h and related adduct with an AB are shown in
Figure 4.
The spectra are very similar, except for the strong absorp-
tion band observed for the 1h-AB conjugate at l=280 nm,
which is typical for proteins. Moreover, the dye retained
about 75% of its emission efficiency after coupling with the
AB.
From the Figure 4, an average labeling efficiency of 4.6
molecules of the fluorescent dye 1h per antibody is calculat-
ed. For Immunoglobulin G, the optimal labeling with rhoda-
mines is achieved with 3–7 residues of a dye attached to one
molecule of an AB. The immunostained samples of the AB
labeled with fluorinated rhodamines were very bright and
stable and gave excellent images in standard fluorescence
microscopy and nanoscopy. Examples of these results are
given in the next sections.
Figure 4. Spectral properties after bioconjugation: Normalized absorption
and emission spectra of compound 1h (a) and its conjugate with anti-
bodies (c) in PBS solution. No significant shift of the spectra is ob-
served after bioconjugation.
STED nanoscopy: switch-off characteristics: The new fluori-
nated rhodamines are excellent markers for STED nanosco-
py. The additional laser light produced by STED nanoscopy
is used to spatially and selectively depopulate the markerꢁs
fluorescent state by stimulated emission, thereby restricting
the fluorescence emission to sub-diffraction sized areas.[3a,8]
An important prerequisite is that strong intensities of the
STED laser are applied to saturate the switch-off transition.
Although not really absorbing the additional light, the dye
has to withstand large intensities without being photobleach-
ed too fast. Low saturation intensities, that is, low STED in-
tensities at which 50% of the fluorescence is switched off
are favorable, since, in this case, lower laser intensities are
required for the same degree of saturation. Figure 5 shows
the relative fluorescence switch-off with increasing focal
peak intensity of the STED light. Figure 5A and B show
data of the dyes in spin-coated PVA layers with 250 kHz
pulsed irradiation, whereas Figure 5C shows data of the
dyes embedded in mowiol with continuous-wave (CW) light.
Excitation was performed at l=488 nm, whereas STED
wavelengths were tuned to lꢀ600 nm in the pulsed mode
and l=592 nm in the CW case. In all cases, the new com-
Figure 5. STED nanoscopy and fluorescence switch-off characteristics. Relative fluorescence switch-off with increasing intensity of the STED light for
rhodamines in spin-coated PVA layers with 250 kHz pulsed irradiation (A,B) and embedded in mowiol with CW light (C). The new compounds show ex-
cellent switch-off characteristics by the STED light with a large saturation degree and low saturation intensities.
4484
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4477 – 4488

Rhodamines for Optical Microscopy and Nanoscopy
FULL PAPER
pounds showed excellent switch-off characteristics by the
STED light with a large saturation degree and time-aver-
aged saturation intensities of ꢀ1–5 kWcmꢁ2 in the 250 kHz
case, which are very common values (see for example
Alexa 488 in Figure 5A).[9b] In the case of CW excitation,
much larger average STED intensities (ꢀ3 MWcmꢁ2) are
necessary to achieve the same switch-off efficiency as in the
250 kHz case, which is in line with the theoretically expected
increase in saturation intensity 1/(ln2·f·tfl)ꢀ1450 (with fluo-
rescence lifetime tfl ꢀ4 ns and pulse repetition rate f=
250 kHz).[32] The photobleaching was found to be low
enough to allow accurate recording of these switch-off data,
which are as expected from the low photobleaching yields at
high-excitation intensities (compare Table 2).
For comparison of the STED images with the correspond-
ing conventional confocal images of cellular samples immu-
nostained with our compounds, see Figure 6: Microtubulin
of PtK2 cells labeled with compound 1 f (Figure 6A), vimen-
tin of PtK2 cells labeled with compound 1h (Figure 6B) and
doublecortin of redifferentiated human neuroblastoma cells
SH-SY5Y labeled with compound 15 (Figure 6C). Images
Figure 6A and B were recorded with CW irradiation (STED
intensity 80 MWcmꢁ2 at l=592 nm), whereas those of Fig-
ure 6C were recorded with 250 kHz pulsed irradiation (with
observable aggregation. Further, the spatial resolution of
the STED images was significantly improved over that of
the confocal counterparts, allowing an improved disclosure
of the structural details.
Labeling of lipids: The use of fluorescent lipid analogues for
the study of membrane organization helps to get detailed in-
sight into cellular processes and most importantly into the
role of lipid dynamics and the formation of lipid nanodo-
mains or “rafts”.[33] Small organic fluorophores as markers
are preferable to minimize the influence of the label on the
dynamics of the small lipid molecules. Using a combination
of STED and FCS, we have previously shown detailed in-
sight of nanoscale dynamics of small organic dye labeled
lipids in the plasma membrane of living cells, depicting tran-
sient and cholesterol assisted molecular complexes.[34] The
new fluorinated rhodamines can also be successfully at-
tached to lipids. For example, the hydrophilic and water-
soluble dye 1h-5’-COOH (or as the NHS ester 15-H-
CH2CF3-H-5’-CONHS) was successfully used for labeling of
the polar carbohydrate domain in the derivative of the natu-
ral glycosphingolipid GM1 (Scheme 7).[3q]
The starting compound 16 was obtained from the natural
GM1 ganglioside (R1 =CH3CO, R2 =CH
3ACHTNUGTRNE(UGN CH2)16CO) by al-
time-averaged STED intensity of ꢀ100 kWcmꢁ2 at
l
kaline hydrolysis, in the course of which the acetyl group in
the neuraminic acid residue and the stearic acid fragment
had been cleaved off. Owing to the higher activity of the pri-
mary amino group in the sphingosine moiety, diamine 16
can be selectively labeled by two various markers. The first
ꢀ600 nm). In all cases, the images were of high quality with
respect to brightness and signal-to-background ratio. Note
that the dye 1 f is not a sulfonated rhodamine derivative.
Nevertheless, it neither produced any background, nor any
Figure 6. STED nanoscopy and imaging. STED and according conventional confocal images of immunostained cellular samples: A) Microtubules of
PtK2 cells labeled with compound 1 f, B) vimentin of PtK2 cells labeled with compound 1h, and C) doublecortin of redifferentiated human neuroblasto-
ma cells SH-SY5Y labeled with the compound 15. Images A) and B) were recorded with CW irradiation (STED intensity 80 MWcmꢁ2 at l=592 nm),
whereas those of C) were recorded with 250 kHz pulsed irradiation (STED intensity 2.1 GWcmꢁ2 at lꢀ600 nm). In all cases, the images were of high
quality with respect to brightness and signal-to-background ratio, that is, the immunolabeling is excellent and the spatial resolution of the STED images
is significantly improved over the confocal counterparts, allowing for an improved disclosure of structural details.
Chem. Eur. J. 2010, 16, 4477 – 4488
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4485

V. N. Belov, C. Eggeling, S. W. Hell et al.
Scheme 7. Derivatives of the natural GM1 ganglioside with fluorescent labels: a) ATTO 647N NHS ester, DMSO, NEt3, RT, 24 h; b) 1h-5’-COOH,
DMSO, HATU, NEt3, RT, 24 h (or 15-H-CH2CF3-H-5’-CONHS in DMSO at RT).
one has to be preferably lipophilic, to mimic the natural
stearic acid fragment. Therefore, the unpolar fluorescent
dye (ATTO 647N) was used at the first step. The second
marker has to be hydrophilic, as it labels the polar carbohy-
drate domain. Compound 18 was successfully used in two-
color FCS experiments).[3q]
Studies on the intracellular availability of the new fluores-
cent dyes: Microscopic imaging of protein dynamics and dis-
tribution in living cells is one of the most important tasks in
cell biology. For imaging of specifically labeled proteins in
living cells, the genetically created chimeric proteins consist-
ing of a host protein and a reporter protein are used in most
cases. In addition to the available fluorescent fusion pro-
teins, several strategies, such as SNAP-, CLIP- or Halo-tags
were established over the last years to label proteins in
living cells specifically with organic fluorophores.[35] Modi-
fied cellular enzymes are applicable for covalent addition of
a dye. However, only a limited number of dyes are currently
available for these labeling strategies.
The most important prerequisite for the dyes used in
these labeling strategies is that a sufficient amount of it
needs to pass through cellular membrane(s) in order to
reach the target protein. To asses the permeability of the
cell membrane to the new fluorinated rhodamines, the fol-
Figure 7. Membrane-permeation experiments of the new rhodamine
lowing protocol was used. The living mammalian cells
dyes: Wide-field fluorescence images of living mammalian PtK2 cells
(PtK2) had been grown overnight on cover slips and then
they were subjected to either 500 or 50 mg of a dye in 1 mL
of the growth medium for 15 min. (The amount of a dye
used for the labeling of SNAP-fusion proteins is below 5 mg
in 1 mL.) After that, the cells were washed with a pure
growth medium (without a dye) for 15 min and imaged with
an epifluorescence microscope (Figure 7).
after 15-minute incubation with aqueous buffer solution of the dyes.
A) 15-H-CH2CF3-H-H (1g); B) 1 f-5’-COOH; C) 1 f-5’/6’-COOMe;
D) 13-Me-Me-H (1q); E) 1e-methylester; and F) 9-H-H (1i). The dyes
cross the cellular membranes and may thus be used for new labeling
strategies in living cells.
Upon treatment with 500 mg of a dye in 1 mL, all the dyes
were able to permeate the cellular membrane(s) of living
mammalian cells (see Table S2 in the Supporting Informa-
tion), except two compounds (15-H-H-F-5’/6’-COOH (1d)
and 15-H-H-F-H (1c)). Only 15 min of incubation were
enough to non-specifically label intracellular structures like
protein aggregates (visible as dots), vesicles or mitochon-
dria. Therefore, this (high) concentration is expected to be
sufficient to label proteins specifically, using the labeling
strategies based on SNAP-, CLIP- or Halo-tags. The dyes
still reached the interior of living cells (Table S3 and
Figure 11 in the Supporting Information) upon treatment
with 50 mg of a dye in 1 mL of the solution. However, the
nonspecifically labeled structures were dimmer, in accord-
ance with the lower concentration of a dye inside the cells.
Interestingly, two methyl esters (1 f-5’/6’-COOMe and
1e(14)-methyl ester) were found among the dyes that pene-
trate through the cell membrane and stained the organelles
even at a lower concentration. This fact indicates that prob-
ably the esterification of one or two carboxy groups in rhod-
4486
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4477 – 4488

Rhodamines for Optical Microscopy and Nanoscopy
FULL PAPER
Joo, H. Balci, Y. Ishitsuka, C. Buranachai, T. Ha, Annu. Rev. Bio-
chem. 2008, 77, 51–76.
amines facilitates the ability to cross membranes of live
cells.
[3] For reviews, see: a) S. W. Hell, Science 2007, 316, 1153–1158; b) M.
Fernꢆndez-Suꢆrez, A. Y. Ting, Nat. Rev. Mol. Cell Biol. 2008, 9,
929–943; c) D. Evanko, Nat. Methods 2009, 6, 19–20; d) S. W. Hell,
Nat. Methods 2009, 6, 24–32; e) X. Zhuang, Nat. Photonics 2009, 3,
365–367; f) J. Lippincott-Schwartz, S. Manley, Nat. Methods 2009, 6,
21–23; g) B. Huang, M. Bates, X. Zhuang, Annu. Rev. Biochem.
2009, 78, 993–1016; for recent reports, see: h) J. A. Fitzpatrick, Q.
Yan, J. J. Sieber, M. Dyba, U. Schwarz, C. Szent-Gyorgyi, C. A.
Woolford, P. B. Berget, A. S. Waggoner, M. P. Bruchez, Bioconjugate
Chem. 2009, 20, 1843–1847; i) E. Betzig, G. H. Patterson, R. Sou-
grat, O. W. Lindwasser, S. Olenych, J. S. Bonifacino, M. W. David-
son, J. Lippincott-Schwartz, H. F. Hess, Science 2006, 313, 1642–
1645; j) M. J. Rust, M. Bates, X. Zhuang, Nat. Methods 2006, 3, 793–
795; k) M. Bates, B. Huang, G. T. Dempsey, X. Zhuang, Science
2007, 317, 1749–1753; l) S. T. Hess, T. P. K. Girirajan, M. D. Mason,
Biophys. J. 2006, 91, 4258–4272; m) J. Fçlling, V. Belov, R. Kunet-
sky, R. Medda, A. Schçnle, A. Egner, C. Eggeling, M. Bossi, S. W.
Hell, Angew. Chem. 2007, 119, 6382–6386; Angew. Chem. Int. Ed.
2007, 46, 6266–6270; n) J. Fçlling, V. Belov, D. Riedel, A. Schçnle,
A. Egner, C. Eggeling, M. Bossi, S. W. Hell, ChemPhysChem 2008,
9, 321–326; o) M. Bossi, J. Fçlling, V. N. Belov, V. P. Boyarskiy, R.
Medda, A. Egner, C. Eggeling, A. Schçnle, S. W. Hell, Nano Lett.
2008, 8, 2463–2468; p) V. P. Boyarskiy, V. N. Belov, R. Medda, B.
Hein, M. Bossi, S. W. Hell, Chem. Eur. J. 2008, 14, 1784–1792;
q) S. M. Polyakova, V. N. Belov, C. Eggeling, S. F. Yan, C. Ringe-
mann, G. Schwarzmann, A. de Meijere, S. W. Hell, Eur. J. Org.
Chem. 2009, 5162–5177; r) V. N. Belov, M. L. Bossi, J. Fçlling, V. P.
Boyarskiy, S. W. Hell, Chem. Eur. J. 2009, 15, 10762–10776; s) H.
Shroff, C. G. Galbraith, J. A. Galbraith, H. White, J. Gilette, S. Ole-
nych, M. W. Davidson, E. Betzig, Proc. Natl. Acad. Sci. USA 2007,
104, 20308–20313; t) B. Huang, W. Wang, M. Bates, X. Zhuang, Sci-
ence 2008, 319, 810–813; u) B. Huang, S. A. Jones, B. Brandenburg,
X. Zhuang, Nat. Methods 2008, 5, 1047–1052.
In summary, the new dyes are able to cross the cellular
membrane(s) and reach the interior of living cells. They may
be used as an addition to the currently available dyes for
the new labeling strategies in living cells (e.g., N,N,N’,N’-tet-
ramethylrhodamine).
Conclusion and Outlook
New fluorinated rhodamines serving as excellent fluorescent
markers in (live) cell microscopy and nanoscopy were intro-
duced. They can be excited between l=470 to 515 nm with
common solid sate or argon ion lasers, have large absorption
cross-sections, emit light at about l=520 nm with high fluo-
rescence quantum yields in solution (in a free state and
after bioconjugation) and possess relatively long fluores-
cence lifetimes of 3–4 ns. Their intersystem crossing rates
are relatively low and they are extremely photostable, allow-
ing a very long observation time at low excitation intensities
and excellent performance at large excitation intensities.
Consequently, these rhodamines perform outstanding in
single-molecule based microscopies, such as FCS. The com-
pounds are available in a hydrophilic (sulfonated) and a lip-
ophilic form (with the same chromophore) and amino reac-
tive sites may easily be introduced, utilizing specific labeling
of bio-molecules, such as proteins and lipids. Further, the
compounds can be used in STED nanoscopy delivering a
significantly improved spatial resolution of immunostained
cellular structures. Live cell applications may be feasible
since the new fluorinated rhodamines have been shown to
penetrate the cellular membrane.
[4] a) D. Magde, E. Elson, W. W. Webb, Phys. Rev. Lett. 1972, 29, 705–
708; b) M. Ehrenberg, R. Rigler, Chem. Phys. 1974, 4, 390–401;
c) E. Hausteinand, P. Schwille, Methods 2003, 29, 153–166.
[5] Y. Chen, J. D. Mꢄller, P. T. C. So, E. Gratton, Biophys. J. 1999, 77,
553–567; P. Kask, K. Palo, D. Ullmann, K. Gall, Proc. Natl. Acad.
Sci. USA 1999, 96, 13756–13761.
[6] C. Eggeling, L. Brand, D. Ullmann, S. Jaeger, Drug Discovery Today
2003, 8, 632–641.
[7] C. Eggeling, J. Widengren, R. Rigler, C. A. M. Seidel, Anal. Chem.
1998, 70, 2651–2659.
The analogue of Rhodamine B with four 2,2,2-trifluor-
oethyl groups would be an interesting extension of the pres-
ent “dye gallery”. It is one of the most lipophilic “green”
fluorescent dyes. However, it is hard to predict, whether this
rhodamine could be sulfonated.
[8] a) S. W. Hell, J. Wichmann, Opt. Lett. 1994, 19, 780–782; b) T. A.
Klar, S. Jakobs, M. Dyba, A. Egner, S. W. Hell, Proc. Natl. Acad.
Sci. USA 2000, 97, 8206–8210.
[9] a) M. Dyba, S. W. Hell, Appl. Opt. 2003, 42, 5123–5129; b) G. Don-
nert, J. Keller, R. Medda, M. A. Andrei, S. O. Rizzoli, R. Lꢄhrmann,
R. Jahn, C. Eggeling, S. W. Hell, Proc. Natl. Acad. Sci. USA 2006,
103, 11440–11445.
Acknowledgements
This work was supported by the Leibnitz Prize Fund of the DFG (S. W.
H.) and the Max-Planck-Gesellschaft. The authors are grateful to Donald
Ouw for the optical measurements. We highly appreciate the assistance
of Reinhard Machinek, Dr. Holm Frauendorf, and their co-workers who
recorded numerous NMR and mass spectra at the Institut fꢄr Organische
und Biomolekulare Chemie (Georg-August-Universitꢅt Gçttingen).
[10] For a general description and spectral range of rhodamines as fluo-
rescent dyes, see: a) M. Beija, C. A. M. Afonso, J. M. G. Martinho,
Chem. Soc. Rev. 2009, 38, 2410–2433; M. S. T. GonÅalves, Chem.
Rev. 2009, 109, 190–212; b) R. P. Haugland, A Guide to Fluorescent
Probes and Labeling Technologies, Invitrogen, Carlsbad, 2005,
pp. 11–37; c) for the use of rhodamines in bioconjugation tech-
niques, see: G. T. Hermanson, Bioconjugate Techniques, Academic
Press, Amsterdam, 2008, pp. 415–427; for laser dyesꢁ properties of
rhodamines, see: d) F. P. Schꢅfer in Dye Lasers (Ed.: F. P. Schꢅfer),
Springer, Berlin, 1990, pp. 3–4; e) K.-H. Drexhage in Dye Lasers
(Ed.: F. P. Schꢅfer), Springer, Berlin, 1990, p. 158; f) K.-H. Drexhage
in Dye Lasers (Ed.: F. P. Schꢅfer), Springer, Berlin, 1990, p. 175–
180; g) K.-H. Drexhage, Top. Appl. Phys. 1973, 1, 155–200; for the
most recent report on the new red-emitting rhodamines, see: h) K.
Kolmakov, V. N. Belov, J. Bierwagen, C. Ringemann, V. Mꢄller, C.
Eggeling, S. W. Hell, Chem. Eur. J. 2010, 16, 158–166.
[1] “Fluorophores for Confocal Microscopy: Photophysics and Photo-
chemistry”: R. Y. Tsien, L. Ernst, A. Waggoner in Handbook of Bio-
logical Confocal Microscopy, 3 ed. (Ed.: J. B. Pawley), Springer,
New York, 2006, pp. 338–352.
[2] a) W. E. Moerner, M. Orrit M, Science 1999, 283, 1670–1676;
b) W. P. Ambrose, P. M. Goodwin, J. H. Jett, A. Van Orden, J. H.
Werner, R. A. Keller, Chem. Rev. 1999, 99, 2929–2956; c) X. S. Xie,
J. K. Trautman, Annu Rev. Phys. Chem. 1998, 49, 441–480; d) W. E.
Moerner, Proc. Natl. Acad. Sci. USA 2007, 104, 12596–12602; e) C.
Chem. Eur. J. 2010, 16, 4477 – 4488
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
4487

V. N. Belov, C. Eggeling, S. W. Hell et al.
[11] a) K. Peneva, G. Mihov, F. Nolde, S. Rocha, J. Hotta, K. Braeck-
mans, J. Hofkens, H. Uji-i, A. Hermann, K. Mꢄllen, Angew. Chem.
2008, 120, 3420–3423; Angew. Chem. Int. Ed. 2008, 47, 3372–3375;
b) C. Jung, B. K. Mꢄller, D. C. Lamb, F. Nolde, K. Mꢄllen, C.
Brꢅuchle, J. Am. Chem. Soc. 2006, 128, 5283–5291.
[24] T. Cohen, A. G. Dietz, Jr., J. Miser, J. Org. Chem. 1977, 42, 2053–
2058.
[25] C. Eggeling, J. Widengren, R. Rigler, C. A. M. Seidel in Applied
Fluorescence in Chemistry, Biology and Medicine (Eds.: W. Rettig,
B. Strehmel, M. Schrader, H. Seifert), Springer, Berlin, 1999, p. 193.
[26] For a short review on the photobleaching kinetics of Rhodamine 6G,
see: a) C. Eggeling, A. Volkmer, C. A. M. Seidel, ChemPhysChem
2005, 6, 791–804 and the references therein; b) V. E. Korobov, A. K.
Chibisov, J. Photochem. 1978, 9, 411–424; c) radicals, which were
formed by irradiation of Rhodamine 6G in solution, were detected
by ESR spectroscopy: R. Zondervan, F. Kulzer, S. B. Orlinskii, M.
Orrit, J. Phys. Chem. A 2003, 107, 6770–6776.
[12] Z. Lu, N. Liu, S. J. Lord, S. D. Bunge, W. E. Moerner, R. J. Twieg,
Chem. Mater. 2009, 21, 797–810.
[13] a) G. Ulrich, R. Ziessel, A. Harriman, Angew. Chem. 2008, 120,
1202–1219; Angew. Chem. Int. Ed. 2008, 47, 1184–1201. For a very
recent report on the hydrophilic BODIPY fluorophores decorated
with sulfonic acid residues, see: b) S. L. Niu, G. Ulrich, R. Ziessel,
A. Kiss, P.-Y. Renard, A. Romieu, Org. Lett. 2009, 11, 2049–2052.
[14] F. Mao, W.-Y. Leung, R. P. Haugland (Molecular Probes), US Patent
6130101, 2000.
[15] This conclusion follows from the comparison of the stabilities of the
Oregon Green dyes—fluorescein derivatives bearing two fluorine
atoms at the positions 2’ and 7’ of the xanthene ring—with that of
the parent dye; see the Supporting Information for the details.
[16] S. W. Hell, V. N. Belov, G. Mitronova, M. Bossi, G. Moneron, C. A.
Wurm, S. Jakobs, C. Eggeling, J. Bierwagen, L. Meyer (Max-Planck-
Gesellschaft zur Fçrderung der Wissenschaften e.V.), PCT/EP 2009/
004650, 2009.
[27] M. Kasha, Rad. Res. 1960, 2, 243–275.
[28] a) A. K. Chibisov, H. Gçrner, J. Photochem. Photobiol. A 1997, 105,
261–267; b) G. Donnert, C. Eggeling, S. W. Hell, Nat. Methods 2007,
4, 81–86.
[29] a) M. Anbar, E. J. Hart, J. Am. Chem. Soc. 1964, 86, 5633–6537;
b) E. V. Khoroshilova, D. N. Nikogosyan, J. Photochem. Photobiol.
B 1990, 5, 413–427.
[30] a) J. Widengren, U. Mets, R. Rigler, J. of Phys. Chem. 1995, 99,
13368–13379; b) J. Widengren, R. Rigler, Bioimaging 1996, 4, 149–
157; c) P. S. Dittrich, P. Schwille, Appl. Phys. B 2001, 73, 829–837.
[31] C. Ringemann, A. Schçnle, A. Giske, C. von Middendorff, S. W.
Hell, C. Eggeling, ChemPhysChem 2008, 9, 612–624.
[17] K. R. Gee, M. Poot, D. H. Klaubert, W.-C. Sun, R. P. Haugland, F.
Mao (Molecular Probes), WO97/39064, 1997.
[18] The “hybrid” of Rhodamine B (labs. =545, lem. =569 nm, e=114000,
Ffl =52%) and fluoresceine (labs. =497 lem. =518 nm, e=84000,
Ffl =100%) was reported to absorb at l=517 nm and emit at l=
542 nm (e=39000, Ffl =58%): P. R. Hammond, J. Photochem. 1979,
10, 467–471.
[19] P. R. Hammond, J. F. Feeman (US Dep. Energy), US Patent
4945176, 1989; P. R. Hammond, J. F. Feeman (US Dep. Energy), US
Patent 5047559, 1991; P. R. Hammond, J. F. Feeman (US Dep.
Energy), US Patent 5111472, 1991.
[20] S. Bigotti, A. Volonterio, M. Zanda, Synlett 2008, 958–962.
[21] For the structure of Rhodamine 110 and other similar and commer-
cially available “green” fluorescent dyes, see Figure 1S in the Sup-
porting Information.
[32] K. I. Willig, B. Harke, R. Medda, S. W. Hell, Nat. Methods 2007, 4,
915–918.
[33] a) K. Simons, E. Ikonen, Nature 1997, 387, 569–572; b) Lipid Rafts
and Caveolae (Ed.: C. J. Fielding), Wiley-VCH, Weinheim, 2006;
c) K. Jacobson, O. G. Mouritsen, R. G. W. Anderson, Nat. Cell Biol.
2007, 9, 7–14; d) M. F. Hanzal-Bayer, J. F. Hancock, FEBS Lett.
2007, 581, 2098–2104.
[34] C. Eggeling, C. Ringemann, R. Medda, G. Schwarzmann, K. Sandh-
off, S. Polyakova, V. N. Belov, B. Hein, C. von Middendorff, A.
Schçnle, S. W. Hell, Nature 2009, 457, 1159–1163.
[35] a) H. M. OꢁHare, K. Johnsson, A. Gautier, Curr. Opin. Struct. Biol.
2007, 17, 488–494; b) T. Gronemeyer, G. Godin, K. Johnsson, Curr.
Opin. Biotechnol. 2005, 16, 453–458.
[22] R. F. Kubin, A. N. Fletcher, J. Lumin. 1982, 27, 455–462.
[23] To prevent the formation of the cyclic non-fluorescent amides
formed from the sterically hindered carboxy group of the mono-car-
boxy rhodamines and primary amines, an additional linker with a
secondary amide group should be introduced to this carboxy group;
for an example, see reference [3p].
Received: November 30, 2009
Please note: Minor changes have been made to this publication in
Chemistry–A European Journal Early View. The Editor.
Published online: March 22, 2010
4488
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 4477 – 4488