Red-emitting rhodamines with hydroxylated, sulfonated, and phosphorylated dye residues and their use in fluorescence nanoscopy

Article abstract of DOI:10.1002/chem.201201168

Fluorescent dyes emitting red light are frequently used in conventional and super-resolution microscopy of biological samples, although the variety of the useful dyes is limited. We describe the synthesis of rhodamine-based fluorescent dyes with absorption and emission maxima in the range of 621-637 and 644-660 nm, respectively and demonstrate their high performance in confocal and stimulated emission depletion (STED) microscopy. New dyes were prepared by means of reliable chemical transformations applied to a rhodamine scaffold with three variable positions. They feature polarity, water solubility, variable net charges, improved stabilities of N-hydroxysuccinimidyl (NHS) esters, as well as large fluorescence quantum yields in dye solutions and antibody conjugates. The photophysical and imaging properties of dyes containing three different polar groups, namely primary phosphate, sulfonic acid (in two different positions), and hydroxyl were compared. A dye with two primary phosphate groups was explored as a valuable alternative to dyes with "classical" sulfonic acid groups. Due to the increased net charge of the phosphorylated dye (q=-4 at pH 8), it demonstrated a far better electrophoretic mobility compared with analogues with two sulfonic acid groups (q=-2). As an example, one fluorescent dye was designed to be especially convenient for practical use. It is characterized by sufficiently high chemical stability of the NHS ester, its simple isolation, handling, and solubility in aqueous buffers, as well as in organic solvents. All these features, accompanied by a zero net charge in conjugates, were accomplished by the introduction of hydrophilic groups of two types: two hydroxyl groups and one sulfonic acid residue. A study in scarlet: Rhodamines that absorb at 621-637 nm and emit at 644-660 nm were prepared and used in fluorescence microscopy and nanoscopy. They are photostable and feature variable polarities and net charges, high stabilities of N-hydroxysuccinimidyl (NHS) esters, large fluorescence quantum yields in aqueous solutions, and antibody conjugates. Copyright

Full text of DOI:10.1002/chem.201201168

FULL PAPER
DOI: 10.1002/chem.201201168
Red-Emitting Rhodamines with Hydroxylated, Sulfonated, and
Phosphorylated Dye Residues and Their Use in Fluorescence Nanoscopy
Kirill Kolmakov,[a] Christian A. Wurm,*[a, b] Renꢀ Hennig,[c] Erdmann Rapp,[c]
Stefan Jakobs,[a, b] Vladimir N. Belov,*[a] and Stefan W. Hell*[a]
Abstract: Fluorescent dyes emitting
red light are frequently used in conven-
tional and super-resolution microscopy
of biological samples, although the va-
riety of the useful dyes is limited. We
describe the synthesis of rhodamine-
based fluorescent dyes with absorption
and emission maxima in the range of
621–637 and 644–660 nm, respectively
and demonstrate their high perform-
ance in confocal and stimulated emis-
sion depletion (STED) microscopy.
New dyes were prepared by means of
reliable chemical transformations ap-
plied to a rhodamine scaffold with
three variable positions. They feature
polarity, water solubility, variable net
charges, improved stabilities of N-hy-
droxysuccinimidyl (NHS) esters, as
well as large fluorescence quantum
yields in dye solutions and antibody
conjugates. The photophysical and
imaging properties of dyes containing
three different polar groups, namely
primary phosphate, sulfonic acid (in
two different positions), and hydroxyl
were compared. A dye with two pri-
mary phosphate groups was explored
as a valuable alternative to dyes with
“classical” sulfonic acid groups. Due to
the increased net charge of the phos-
phorylated dye (q=À4 at pH 8), it
demonstrated a far better electropho-
retic mobility compared with analogues
with two sulfonic acid groups (q=À2).
As an example, one fluorescent dye
was designed to be especially conve-
AHCTUNGTRENNnGUN ient for practical use. It is character-
ized by sufficiently high chemical sta-
bility of the NHS ester, its simple isola-
tion, handling, and solubility in aque-
ous buffers, as well as in organic sol-
vents. All these features, accompanied
by a zero net charge in conjugates,
were accomplished by the introduction
of hydrophilic groups of two types: two
hydroxyl groups and one sulfonic acid
residue.
Keywords: chromophores · conju-
gation · electrophoresis · fluores-
cence · microscopy · rhodamines
Introduction
cific labeling. However, for practical reasons (e.g., for pro-
viding convenient and reliable protocols for labeling), the
properties of the reactive marker such as chemical stability,
solubility in aqueous media, the possibility to afford high de-
grees of labeling, and simple handling are also important
but not easy to achieve. Due to their good photochemical
and photophysical properties, rhodamine dyes are widely
used in modern far-field optical microscopy and nanosco-
py.[3] Despite their long history and wide use, rhodamine
dyes, their preparation and post-synthetic modifications
have not been studied in depth,[3,4] particularly in the case of
red-emitting dye 1 (Scheme 1) with promising spectral prop-
erties. Recently we reported the synthesis and properties of
the red-emitting rhodamine dye KK114 (compound 2a in
Scheme 1),[5] which turned out to be a bright and photosta-
ble fluorescent marker in various optical microscopy and
nanoscopy techniques.[6] In the conjugated state, the residue
of KK114 is negatively charged (due to the presence of the
two sulfonic acid groups). The negatively charged dye conju-
gates do not penetrate through the outer plasma membrane
of living cells,[6c] and this feature limits the application scope
of this dye. An additional negative charge can also strongly
influence the properties of lipids labeled with compound
2a.[5a] It was therefore necessary to further improve dye
KK114 in such a way that its net electrical charge in conju-
gates could be varied, for example, from +1 to À3. To
Red-emitting fluorescent dyes are widely used as makers in
optical microscopy, biological imaging and sensing, especial-
ly in recent years.[1] Good imaging performance in optical
microscopy and modern super-resolution techniques[2] de-
mands high absorption coefficients and large fluorescence
quantum yields of the markers, their resistance against pho-
tobleaching, and low background emission caused by unspe-
[a] Dr. K. Kolmakov, Dr. C. A. Wurm, Prof. S. Jakobs, Dr. V. N. Belov,
Prof. S. W. Hell
Department of NanoBiophotonics
Max Planck Institute for Biophysical Chemistry
Am Fassberg 11, 37077 Gçttingen (Germany)
Fax : (+49) 551 2012505
E-mail: cwurm@gwdg.de
vbelov@gwdg.de
shell@gwdg.de
[b] Dr. C. A. Wurm, Prof. S. Jakobs
Georg-August University of Gçttingen
Medical School, Department of Neurology
37073 Gçttingen (Germany)
[c] Dipl.-Ing. R. Hennig, Dr. E. Rapp
Max Planck Institute for Dynamics of Complex Technical Systems
Sandtorstrasse 1, 39106 Magdeburg (Germany)
Fax : (+49) 391 6110535
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201201168.
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

contrast due to unspecific binding, and the presence of two
diastereomers. The challenging part of this study was to
design new rhodamines without these shortcomings while
trying to keep or even surpass the valuable properties of the
best near-IR-emitting dyes. The most important parameters
that have to be maintained are the positions of absorption
and emission maxima (excitation with He–Ne laser at
633 nm or diode laser at 635–640 nm; sufficient optical cross
section for stimulated emission at 750–775 nm), high bright-
ness (product of the fluorescence quantum yield and the ab-
sorption coefficient), photostability under excitation and
stimulated emission depletion (STED) conditions (with exci-
tation light of 633 nm and powerful STED beam of 750–
775 nm, respectively), hydrophilic character (which hinders
the unspecific binding and provides high imaging contrast
due to the absence of the fluorescent background), and sta-
bility of an amino reactive derivative (NHS ester). The
chemical stability of the reactive markers in the solid state
and in aqueous solutions is a significant parameter, consider-
ing prolonged storage and/or shipping of the samples and
their stock solutions, and relatively long (normally 1 hour)
reaction times in aqueous buffers at pH 7–9 at room temper-
ature (typical conditions for the conjugation with proteins
and other biomolecules).
Scheme 1. Red-emitting rhodamine dye 1 as a scaffold with three variable
positions: A) methyl groups subjected to sulfonation (A1) or oxidation
with SeO2 followed by reduction with NaBH4 (A2); B) aromatic F-atoms
As regards the active ester of dye KK114 (2a in
Scheme 1), we found that its hydrolytic stability can be char-
acterized as medium (see Table 1). The NHS ester can be
isolated in the pure state by column chromatography over a
normal phase (SiO2) with eluents that contain acetonitrile
and water (due to the high polarity of this compound deco-
rated with two sulfonic acid residues). After direct lyophili-
zation of the fractions (without removal of acetonitrile), the
solid residue was analyzed by HPLC, and the area of the
peak which corresponds to the NHS ester was determined
to be about 85–90% (detection at 636 nm); the starting acid
was present in amounts of approximately 10–15% (HPLC
area). Polar solvents, especially water, solubilize a noticeable
amount of silica gel, which is difficult to remove completely
(by centrifugation and filtration). Silica gel, as an impurity
in “dried” solids, always retains some water, which is harm-
ful for hydrolytically unstable substances; as a result, the
content of the active ester in the solid material after freeze-
drying drops down to 60–80% and even further upon stor-
age under argon at À208C. Therefore, an important part of
this study was to improve the stability of the NHS esters,
maintaining their solubility in polar aprotic solvents (used
for the preparation of the stock solutions), to make the
markers easier to handle, taking aliquots, ship and store.
The parent compound—rhodamine 1[9]—is insoluble in
water (we could not even record its absorption spectrum in
pure water and had to add 10% (v/v) of methanol), but its
spectroscopic properties are very attractive (see Figure 1
and Table 2). Upon a closer look, the fluorescent dye 1 pos-
sess three variable positions (three functional groups; see
Scheme 1), which can be easily transformed in a combinato-
rial chemistry fashion. However, until our recent study, no
attempts to further redshift its absorption and emission
subjected to SN reactions with HSCH2CO2C2H5 (B1) or HS
ACHTNUGRTNE(NUGN CH2)2SO3Na
(B2); C) COOH group subjected to amidation with CH3NH(CH2)nCOOR
AHCTUNGTRENNUNG
(C1), or esterification with alcohols, e.g., allyl alcohol (C2). For a detailed
description of the reagents and conditions, see Scheme 2 and Scheme 3;
*) see ref. [5a]; **) see ref. [8].
pursue this goal, we used rhodamine dye 1 (Scheme 1) as a
scaffold for derivatization at various positions. This combi-
natorial approach is illustrated by Scheme 1, in which the
three variable positions in compound 1 are indicated as A,
B, and C. Proper chemical transformations enabled us to
obtain new red-emitting fluorescent dyes with high imaging
performance, variable net charge, various electrophoretic
mobilities, controlled polarity, desired solubility, and stable
N-hydroxysuccinimidyl (NHS) esters.
Results and Discussion
Motivation and possible synthesis strategies: Dye KK114, a
very bright and photostable near-IR fluorescent marker
(compound 2a in Scheme 1), has repeatedly demonstrated
its excellent imaging performance.[5,6] The disadvantages of
this compound are the lower photostability under confocal
conditions (compared to Atto647N; see below), the negative
charge in conjugates (therefore they are not cell-permeable)
and the relatively low chemical stability of the NHS ester.
Another widely used near-IR emitting dye is Atto647N.[7] Its
main drawbacks are the lower fluorescence quantum yield
in conjugates with proteins (relative to these of KK114), the
lipophilic character, which is reflected in a lower imaging
&
2
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Red-Emitting Rhodamines in Fluorescence Nanoscopy
FULL PAPER
well. In our previous studies, we used only one linker, the
N-methyl-b-alanine residue (taken as methyl or tert-butyl
ester). It would also be interesting to try a longer linker, for
example, N-methyl-4-aminobutyric acid, expecting better
stability of the NHS ester and, as a result, higher degrees of
labeling in aqueous media.
At both allylic sites (positions A in Scheme 1), a direct
oxidation to the corresponding diol (4a, Scheme 1) could be
a good alternative to sulfonation.[12a] Generally, alcohols
offer more opportunities for further modifications than any
other derivatives.[12b] Hydroxyl (or hydroxyalkyl) groups
alone substantially improve hydrophilic properties, increase
solubility in water and slightly increase the quantum yield of
fluorescence.[12a, 5b] Finally, as regards the general strategy of
the synthesis, direct modification of a chromophore-contain-
ing intermediate (e.g., rhodamine 1) is preferable. Conden-
sation is the crucial final step in the synthesis of rhoda-
mines; it proceeds under harsh conditions and always in the
presence of acids. The influence of new substituents and
functional groups on this reaction might be absolutely pro-
hibiting and/or require a long time to find proper condensa-
tion conditions and suitable protective groups for the hy-
droxyl.
Figure 1. Aqueous solutions of the fluorescent dyes (Scheme 1) exposed
to daylight. Left: compound 1; right: compound 2a (KK114). Note the
absence of the red fluorescence in the case of compound 2a. For spectral
and related data, see Table 2.
bands, to attach a convenient linker with reactive groups
(for conjugation with proteins and small molecules), and, fi-
nally, to improve its solubility of this compound in aqueous
solutions have been undertaken.[5a] We extended the direct
sulfonation at the allylic position of rhodamine dyes[10] to
tertiary amides of 1 and obtained dye KK114 (2a; see
Scheme 1 and Figure 1) with two sulfonic acid residues (and
the net charge of À1 in conjugates).[5a] The structure of this
dye has the unique combination of three structural features:
the tertiary amide bond, a benzene ring with four fluorine
atoms, and two sulfonic acid groups. However, we have not
explored any alternatives to the sulfonic acid groups—charg-
ed or uncharged polar groups that improve fluorescence
quantum yields and hinder the undesired aggregation of the
dye molecules in aqueous solutions.
Apart from the “conventional” carboxylic acid function
(position C, compound 1 in Scheme 1) and far less exploited
“allylic site” (position A), there is another variable position
in rhodamine 1, a fluorine atom, indicated as position B. At
the beginning of the work with amides of compound 1, we
observed the undesirable exchange of one fluorine atom to
methoxy group under mild basic conditions (with K2CO3) in
methanol.[5a,11] Later, we explored the substitution reaction
of one fluorine atom with ethyl thioglycolate
(HSCH2CO2Et) in the allyl ester of sulfonated rhodamine
3a.[8] After the substitution with thioglycolate had been per-
formed, it was important to learn that the spectral proper-
ties of the product 3b were almost unaffected. In fact, the
absorbance maximum in 3b shifted only by 1–2 nm to the
blue region (compared with the allyl ester of 3a),[8] and the
fluorescent quantum yield did not change much either.
Therefore, we considered the site with one fluorine atom
(position B in compound 1, Scheme 1) as an important tool
that could help to change and control the polarity and solu-
bility of the resulting dyes (in the case of using bifunctional
thiols with polar groups). It might be also important to
remove the most reactive fluorine atom from the fluorescent
dye, preventing any uncontrolled reactions with thiol groups
in cysteine residues of proteins or amino groups in biologi-
cally relevant substrates.
Hydroxyl-substituted red-emitting rhodamine dyes: Seleni-
um dioxide (SeO2) is one of the softest oxidants, known in
preparative organic chemistry.[13] In practice, oxidation of
“allylic” CH3 groups leads to the corresponding aldehydes
or mixtures with the alcohols. Subsequent reduction of the
aldehydes with sodium borohydride (NaBH4) makes the
separation of the two products unnecessary. For the oxida-
tion, 1,4-dioxane (often containing 5–10% water) is the sol-
vent of choice because it dissolves most organic substrates
and SeO2 as well. In our initial experiments with rhodamine
1 in anhydrous dioxane we observed a fast reaction (with
heating) that did produce the desired diol 4a (Scheme 1)
and the corresponding aldhehyde(s) with CH=O group(s) at
the allylic site(s), as established by the mass spectrometric
analysis. However, side products were also formed and the
isolation of the desired oxidized compounds proved difficult.
As a result, the isolated yields of 4a were poor. In aqueous
dioxane as the solvent (with up to 10% water), fewer un-
identified side products were detected, although the reaction
required a much longer time, even at reflux (at about
1008C). This reaction required a large excess of the oxidant.
Probably, for kinetic reasons, it is crucial to maintain high
concentrations of selenious acid (H2SeO3, the hydrated form
of SeO2) in the reaction solution.
Although rhodamine 1 is completely oxidized by SeO2 in
dry dioxane within 30 min at 60–808C, the full conversion of
the starting material and optimal yield of the oxidation reac-
tion in the presence of water are achieved after heating at
reflux temperature overnight. The reaction mixture that
contained diol 4a as a major product and variable amounts
of the mono- and di-aldehydes was subjected to an aqueous
workup, the dyes extracted, dissolved in ethanol and treated
with NaBH4 (to reduce the aldehydes to diol 4a). Luckily,
As regards the “conventional” position C at the carboxyl-
ic group (Scheme 1), we planned some improvements as
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

C. A. Wurm, V. N. Belov, S. W. Hell et al.
this operation did not further reduce the dye 4a to its color-
less leuco-form. Finally, diol 4a was isolated in a 35–50%
yield (several experiments), which can be considered as
quite satisfactory regarding that oxidation with SeO2 has not
been reported for rhodamine dyes so far.
We briefly checked if the same reaction sequence was ap-
plicable to the preparation of compounds 5-R straight from
the tertiary amides of the parent dye 1.[15] Unfortunately, the
method described above (oxidation followed by reduction;
see Scheme 2) did not work for compounds 5-R. We chose
aq. methanol upon further dilution with water, the solubility
of diol 4a in water was clearly more than 100 mmolLÀ1.
Even more importantly, as seen in Table 2, an increase in
the quantum yield of fluorescence was observed in aqueous
solutions: from 36% (compound 1) up to 53–55% (com-
pounds 2a,b, 6-H, and 7-H). However, the absorption maxi-
mum (621 nm) should be shifted further to the red spectral
region for a better match to the emission lines of commer-
cial red lasers (633 nm for a He–Ne laser and 635–640 nm
for a diode laser). Amidation of the free carboxyl group is
known to provide the required
bathochromic shift. Initially, we
used N-methyl-b-alanine esters
as amino components.[5] As
mentioned in the introduction,
the chemical stability of the dye
NHS esters is very important.
Expecting to improve it, we
used a longer linker, N-methyl-
4-aminobutyric acid. The 4-ami-
nobutyric acid residue is incor-
porated into many commercial-
ly available fluorescent dyes
which form relatively stable
NHS esters.[17] Earlier, we had
established that NHS esters of
the masked near-IR fluorescent
dyes with elongated linking
bridges demonstrate very good
stability as well.[8]
For the preparation of
methyl N-methyl-4-aminobuty-
rate (Scheme 2), the commer-
cial N-Z-4-aminobutyric acid
was subjected to the double
(N,O-) methylation with methyl
Scheme 2. Rhodamine dyes, modified with hydroxyl groups: a) SeO2, aq. 1,4-dioxane, reflux, overnight;
b) NaBH4, EtOH, 08C; c) HSACHTUNGTRNEUNG(CH2)2SO3H, Et3N, DMF, RT; d) 2-(1H-7-Azabenzotriazol-1-yl)-1,1,3,3-tetra-
methyl uronium hexafluorophosphate (HATU), Et3N, CH3CN, À5 to 08C; e) 0.05m KOH in aq. THF, RT;
iodide in the presence of silver
f) HATU, N-hydroxysuccinimide, Et3N, CH3CN; g) O-(N-Succinimidyl)-1,1,3,3-tetramethyluronium tetrafluo-
oxide,[16] and then benzyloxycar-
ACHTUNGTRENNUNGroACHTUNGTRENNUNGborate (TSTU), Et3N, CH3CN.
bonyl (Z) protection was
cleaved by hydrogenolysis over
palladium catalyst. A solution
ester 5-tBu[5a] as a model compound. The oxidation step was
relatively clean, but the reduction with NaBH4 caused
nearly complete discoloration of the reaction mixture and
led to a complex mixture of (leuco-) compounds. All at-
tempts to reoxidize the reaction mixture to one major prod-
uct with soft oxidants, like periodates or hydrogen peroxide,
failed. One can explain the selectivity of the oxidation and,
especially, of the reduction with the borohydride, assuming
that the starting compound 1, intermediates and diol 4a are
stabilized in their zwitterionic forms, which are less prone to
further reduction than the corresponding cationic forms
(e.g., amides 5-R, for which the zwitterionic form is obvious-
ly impossible).
of HCl in 2-propanol was added to the reaction mixture to
block the cyclization of the product to N-methylpyrrolidone.
Methyl N-methyl-4-aminobutyrate was isolated as a hydro-
chloride, which proved to be fairly stable.[14a,b]
For the reaction of this amino ester with the hydroxylated
dye 4a we used HATU, an active coupling reagent
(Scheme 2), and triethyl amine. In the presence of the
amino ester (a strong nucleophile), the free hydroxyl groups
did not react and thus require no protection. At low temper-
atures (À5 to 08C), despite noticeable cyclization of the
amino ester to N-methylpyrrolidone in a basic medium, the
coupling reaction was complete. The methyl ester 5-Me was
saponified, and the resulting dye, free acid 5-H, was isolated
in a pure state and explored. The properties of this and
other fluorescent dyes are given in Table 2, and the imaging
performance is discussed below. A dye with two hydroxyl
Next, the properties of the hydroxylated rhodamine dye,
diol 4a, were studied. In contrast to the parent compound 1,
which already precipitated from a 10 mmoL solution in 10%
&
4
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Red-Emitting Rhodamines in Fluorescence Nanoscopy
FULL PAPER
groups (5-H) demonstrates an intermediate value of the flu-
orescence quantum yield in aqueous solutions, and its solu-
bility in water is limited. The restricted solubility leads to
large deviations from the Beer–Lambert Law when the ab-
sorption spectra are measured in aqueous solutions.[14] These
non-linearities may be explained by the fact that at high
concentrations the dye 5-H sticks to the glass (quartz) walls
of the cuvette. The NHS ester of this dye (5-NHS) is easy to
prepare, very stable, yet rapidly reacts with amines and thus
provides sufficiently high degrees of labeling with antibod-
ies. The resulting images have high contrast with no agglom-
erates or fluorescent background (see Figure 3).
In a free state, dye 5-H exists as a zwitterion and in conju-
gates the dye residue is positively charged. The zwitterionic
character of this dye explains why this compound is only
slightly soluble in basic aqueous solutions (e.g., aqueous
NaHCO3, Et3N, or PBS buffer). The limited solubility in
water might be a problem in some applications, for example,
when high concentrations of the marker in aqueous solu-
tions are required. To improve the solubility in aqueous buf-
fers, we decided to modify dye 5-H with an additional hy-
drophilic anionic group. A fluorescent marker that bears
only one anion-providing group (except the carboxyl group)
would be of special interest. In conjugates or in the case of
the NHS ester, in which the carboxyl function is occupied,
the new dye will have a zero net charge. The resulting dye
will be the first one in this series to have hydrophilic groups
of two different types.
One fluorine atom in compound 1 and its derivatives (at
C-6’) can be selectively substituted with nucleophiles and it
therefore offers a very useful second variable position (B;
Scheme 1). From all available nucleophiles with additional
anionic groups, 2-mercaptoethyl sulfonic acid was chosen as
a reagent (see Scheme 2). First attempts to react it directly
(under basic conditions) with amide 5-H led to a mixture of
products. Luckily, in diol 4a the free carboxyl group reacted
cleanly, and only one fluorine atom was selectively substitut-
ed to form 4b (in the para-position to the carboxyl group, as
shown in Scheme 2). This result can be explained if we
assume that the aromatic nucleophilic substitution proceeds
in the “closed” (lactone) form of the fluorescent dye, in
which the carboxyl group forms in fact an ester with an
oxygen atom attached to the quaternary carbon (C-9=C-1’).
The ester group is known to be a stronger electron-with-
drawing substituent than an amide group or carboxylate
anion, and it directs the aromatic nucleophilic substitution
preferably into the para-position (to C-6’). ortho-Substitu-
tion (at C-4’) does not take place at all because this reaction
center is sterically hindered and has only one vicinal fluo-
rine atom with a strong ÀI effect, whereas the nucleophilic
substitution at C-6’ is additionally activated by two neigh-
boring F atoms. Due to the influence of the negatively
charged sulfonic acid group (field effect), the thiol group
proved to be not as reactive, as was in the case of alkyl mer-
captoacetates.[8] The lower reactivity of 2-mercaptoethyl sul-
fonic acid is an advantage because there is no need to con-
trol the reaction conditions very strictly and the reaction
was found to be selective and complete at room tempera-
ture in the presence of about four equivalents of the nucleo-
phile. The reaction product 4b can be extracted with di-
chloromethane in the course of aqueous workup, which sep-
arates the inorganic salts, the unreacted thiol, and the small
amount of the di-substituted derivative (with two
À
(CH2)2SO3 groups).
The linker (methyl N-methyl-4-aminobutyrate) was at-
tached to dye 4b according to the standard procedure (as
described above for dye 5-H) using HATU as a coupling re-
agent and Et3N as a base (Scheme 2). Saponification of 6-
Me with diluted aq. NaOH furnished the free acid 6-H. Iso-
lated as a trifluoroacetate (in the course of the RP chroma-
tography with CF3COOH in the eluent),[14a] the free acid is
only slightly soluble in water. However, the solubility is very
good when two equivalents of a base (Et3N, NaHCO3) are
added. Also importantly, the dye readily dissolves in PBS
buffer solutions at pH 7.4. Therefore, compound 6-H can be
characterized as “amphiphilic” because of its good solubility
in all organic solvents, except alkanes.
Luckily, the preparation and isolation of the correspond-
ing NHS ester (6-NHS) did not present much difficulty and
its stability proved to be surprisingly high. Particularly, the
neat compound does not develop noticeable changes (less
than 2% loss in the content of 6-NHS as established by
HPLC) after storage for one day at room temperature in
air. Also remarkably, this active ester stays almost un-
changed when left for one day in an aqueous PBS buffer
solution at pH 7.4 (with dye concentration of 100 mmoL).
On the other hand, the marker rapidly reacts with diluted
ammonia when added to this solution. In aqueous 0.1m solu-
tions of NaHCO3, this NHS ester is fairly stable for several
hours at room temperature, yet hydrolyzed completely when
left overnight. We can assume that at pH values in the range
of 7–8, compound 6-NHS still exists predominantly in its
zwitterionic form. The high stability of this NHS ester can
be due to zero net charge of its molecule and the absence of
the negatively charged groups in the close proximity to the
active ester residue, which never occurred before in this
series of dyes. Indeed, the NHS esters with sulfonic (2a,b)
or phosphoric acid groups (7) are negatively charged,
whereas those without anion-providing groups have a posi-
tive net charge (which requires a counter ion, like in com-
pound 1a,[5a] Figure 2).
The phosphorylated red emitting rhodamine dye: Our first
study on rhodamine 1 showed that the introduction of the
sulfonic acid groups drastically improves the solubility in
aqueous solutions, fluorescence quantum yield, photostabili-
ty, imaging contrast, ground-state depletion parameters and
other valuable dye properties.[5a] Importantly, the absorption
and emission maxima were only slightly shifted. Having ob-
tained diol 4a, we considered other modifications at OH
sites so that the dye would get polar and ionizable function-
al groups with acidic protons. Free hydroxyl groups offer a
variety of opportunities for derivatization (especially if the
carboxyl residue can be amidated with the functionally sub-
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

C. A. Wurm, V. N. Belov, S. W. Hell et al.
stituted amines, as described above, in the presence of the
unprotected hydroxyls).
It is clear that the dye modified in this way must be an
ester of a strong mineral acid. Esters of mineral acids obvi-
ously raise the hydrolytic stability concerns because labeling
with fluorescent markers is normally done in aqueous
NaHCO3 or PBS buffer solutions. Moreover, the final steps
of the synthesis involve saponification in alkaline solutions,
followed by the preparation of NHS esters (with N-hydroxy
succinimide) in basic media. Therefore, we had to exclude
nitric and sulfonic acid ester groups as candidates. Indeed,
esters of these acids are hydrolytically unstable and, in the
case of sulfonic acid, are even used as alkylating agents.
Phosphoric acid was thus chosen as an alternative functional
group. Indeed, this acid has a medium strength, an increased
number of acidic protons (relative to sulfuric acid), and
most importantly, its primary and secondary alkyl esters are
hydrolytically stable at pH 2–12.[18]
Scheme 3. Synthesis of rhodamine dyes with phosphorylated CH2OH
sites: a) 1H-tetrazole, THF, RT; b) tBuOOH, RT; c) [PdACHTUNGTRNEG(UN PPh3)4], Et3N,
HCOOH, THF, 408C; d) 0.05m KOH aq. THF, RT; e) TSTU, Et3N,
DMF, RT.
Finally, decoration with phosphoric acid residues might be
“biologically friendly” (e.g., make the dye less poisonous to
living cells). To the best of our knowledge, no attempts have
been undertaken so far to attach phosphoric acid residues
directly to fluorescent dyes.[19] In this respect, phosphoric
acid has an advantage over phosphonic acid,[1h,19b] because it
is a much stronger acid. For example, pKa and pKa values
for primary alkyl phosphates are 1.5–1.9 and 6.3–6.8, respec-
tively,[20a] whereas for primary alkyl phosphonates, the first
and second ionization constants were found to be 2.4–2.8
and 7.8–8.9, respectively.[20b] Therefore, at pH 8.0–8.3, which
is recommended for electrophoresis of DNA and RNA
probes, fluorescent dyes with primary phosphate groups will
exist in the fully ionized form and move as single bands
(peaks). Hydrolytic stability of primary phosphates is also a
benefit. The disadvantage of the primary phosphonates as
labels for biomolecules is that at pH 7–9 they are present as
mixtures of mono- and di-anions, which may form separate
bands in the course of capillary (gel)electrophoresis or
HPLC analysis.
In biochemistry, phosphorylation of hydroxyl-substituted
substrates (O-phosphorylation) is of special importance.[18]
One of the best ways of phosphorylation under mild condi-
tions involves the reaction of an alcohol with O,O’-diallyl-
N,N-diisopropylphosphoramidite, assisted by 1H-tetrazole,
and followed by a one-pot oxidation of the intermediate to
the phosphate.[21] In the final step, the allyl ester is cleaved
with a palladium catalyst.[21a] This sequence gives way to a
primary phosphoric acid ester with ROP(O)(OH)2 group(s).
We applied this phosphorylation protocol to rhodamine 5-
Me and obtained the corresponding bis-diallyl phosphate (7-
Me,All) with high yields as shown in Scheme 3. The excess
of the phosphorylating agent (diallyl N,N-diisopropylphos-
phoramidite) and the oxidant (tert-butyl hydroperoxide) did
not cause any complications.[14] The cleavage of all four allyl
methyl esters in this study. Neither primary phosphoric acid
ester groups nor fluorine atoms in the aromatic ring reacted
with alkali under these conditions. The final product of the
synthesis—dye 7-H (Scheme 3) with two primary phosphate
groups and a free carboxylic acid group—demonstrates high
extinction and a large fluorescence quantum yield, which is
as high as those of compounds 2a,b and 6-H. Importantly,
dye 7-H is perfectly soluble in aqueous PBS or NaHCO3 sol-
utions and moderately soluble in pure water. However, the
solubility of this dye in water containing an excess of tri-
fluoroacetic acid, which completely suppresses the ioniza-
tion, is very low. The absorption band is slightly shifted to
the red spectral region, relative to compound 6-H. As re-
gards the NHS ester of the phosphorylated dye 7-H, under
conventional conditions of HPLC analysis (0.1% TFA, pH
ꢀ2), it always gave two to three unresolved peaks of varia-
ble intensity.[14] This is probably due to the equilibrium be-
tween the protonated and deprotonated forms in the sub-
strate under acidic conditions at a pH that corresponds to
1
2
1
the pKa value of the phosphate group. We also noticed that
in the course of the chromatographic isolation of the NHS
ester (both over regular SiO2 and RP-SiO2), trifluoroacetic
acid strongly increases the adsorption of this compound so
that mobile phases with triethylamine were used for elu-
tion.[14] Despite the difficulties with its analysis, the NHS
ester of the phosphorylated dye (7-NHS) clearly consisted
of the required compound predominantly. In the ESI mass
spectrum, the peak with the required mass dominates
strongly; the material completely reacts with diluted aque-
ous ammonia (HPLC). An important feature is that the sol-
ution of 7-NHS in aqueous NaHCO3, which is relatively
stable alone, reacts with NH3 very quickly. Most importantly,
the amine-reactive compound 7-NHS provides sufficiently
high labeling degrees, when being coupled to antibodies.
The phosphorylated dye 7-H demonstrated good imaging
performance, especially in STED microscopy (see Figure 3).
groups with [PdACHTUNGTRENNUNG(PPh3)4] proceeded quite smoothly. The
methyl ester group in phosphorylated rhodamine 7-Me was
saponified in dilute aqueous alkaline solution to the free
acid 7-H under the conditions that were used for other
&
6
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Red-Emitting Rhodamines in Fluorescence Nanoscopy
FULL PAPER
Sulfonated dye with a longer linker and relative stability of
NHS esters: In an earlier study, comparing the stability of
the NHS esters of the dye with and without sulfonic acid
groups (compounds 1a and 2a in Figure 2), much better sta-
bility of the latter was noticed.[5a] We believe that, together
with other benefits, the SO3H groups in the dye stabilize the
NHS ester. This was also in line with the results of our pre-
vious study on certain carbopyronine dyes with an elongated
linker that contained one sulfonic acid residue.[5b] To evalu-
ate the influence of the linker length on the stability of NHS
esters, we obtained the analogue of dye KK114 (2a,
Scheme 1) with the linker elongated by one methylene
group (2b, Scheme 1) and explored the stability of its NHS
ester. The data is presented in Tables 1 and 2. In fact the
analogue without sulfonic acid group (compound 1c in
ref. [5b]) is rather unstable in water. In this regard, it is
worth mentioning that it was not even possible to isolate the
NHS ester of the “lipohilic” rhodamine without sulfonic
acid groups (1a, Figure 2) in satisfactory yields.[5a] In the
present study, we explored the “hydroxylated” analogue 5-H
of compound 1a (Scheme 2, Figure 2) with the same net
charge of +1, the NHS ester of which is fairly stable. One
concludes that the hydroxyl groups can stabilize active
esters in aqueous solutions. From the three dyes with sulfon-
ic acid groups discussed above (2a, 2b, and 6-H), dye 6-H is
clearly the best as regards the stability of the NHS ester. To
sum up, compound 6-H combines the following structural
features: three polar hydrophilic groups of two types, includ-
ing one SO3H moiety that provides a zero net charge to the
zwitterionic molecule of the marker and the dye residue in
conjugates, and the elongated linker bridge that generally
increases the stability of the active esters. As a result, dye 6-
H and its NHS ester, as zwitterions, are soluble in most or-
ganic solvents. High stability, purity, and reproducibly high
content of 6-NHS (>95% HPLC area) are advantageous
for taking aliquots, followed by evaporation of the solvent
in vacuo. An opportunity to take precise amounts of a
marker, in turn, is very important when controlled degrees
of labeling are required. Also importantly, inorganic salts
and silica gel are easy to remove from solutions of 6-NHS in
non-polar organic solvents.
Table 1. Stability of the NHS esters with sulfonic acid groups.[a]
Compound Number Number
of SO3H of CH2
Medium and exposure conditions
water aqueous PBS buffer, pH 7.4
groups
groups in RT
the linker 1 d
RT
1 d
+58C
1 d
Content of NHS ester
2a-NHS
2b-NHS
6-NHS
2
2
1
2
3
3
20
60
100
0
0
98
30
70
100
[a] For structures, see Scheme 1 and Figure 2. Content of NHS ester was
measured as HPLC area (%) in the mixture with the corresponding acid
after exposure.
NHS ester of the dye without the sulfonic acid groups (1a)
proved impossible to isolate.[5a] Dye 2b was synthesized
starting from the sulfonated rhodamine 3a,[8] as shown in
Scheme 1. Exploring dye 2b, which is spectrally identical to
dye 2a (KK114), we did observe better hydrolytic stability,
especially in the course of isolation. The latter involves
evaporation and freeze-drying of aqueous solutions, which
always caused a noticeable hydrolysis in case of 2a-NHS.
The hydrolytic stability of the two compounds (2a-NHS and
2b-NHS) in dilute aqueous solutions is similar, yet the
longer-chained homologue (2b-NHS) is more stable. The
stability of the NHS esters for three dyes (2a, 2b, and 6) in
water and PBS buffer are compared in Table 1. For the dyes
with two sulfonic acid groups, PBS buffer strongly acceler-
ates the hydrolysis, compared with pure water. Remarkably,
compound 6-NHS (Scheme 2) is very stable in PBS buffer
even at room temperature. It should be noted that all three
dyes have sulfonic acid groups, yet their positions and
number are different. We noticed that one sulfonic acid
group is enough to provide good solubility in water; it also
gives zero net charge to the molecule of an (active) ester or
to the zwitterionic dye residue in bioconjugates. Probably,
the better hydrolytic stabilities of the NHS esters correlate
with a zero charge of the molecule (at least in this series of
zwitterionic dyes). This is also true for the active ester 6-
NHS and the NHS ester of the carbopyronine dye with one
sulfonic acid group described in our previous study (com-
pound 22-NHS in ref. [5b]). Both of them demonstrated
very good hydrolytic stability, whereas the NHS ester of the
Spectral properties and imaging performance of modified
near-IR emitting dyes: Table 2 presents the most important
photophysical properties of ten dyes. Amidation of the car-
boxyl group (position C) produces a noticeable bathochro-
mic shift of 10–20 nm (cf. Figure 1). In the series of the five
dyes with linkers (1a, 2a, 5, 6, and 7 in Figure 2), all absorp-
tion maxima lie in the narrow region between 632 and
638 nm. Therefore, all these dyes may be efficiently brought
into the first excited state (S1) with a He–Ne laser (633 nm),
or with the red line of a diode laser (635–640 nm). As ex-
pected, the ionizable groups (sulfonic acid and primary
phosphate residues) increase the quantum yield of fluores-
cence and the solubility in water, relative to the non-substi-
tuted dyes 1, 1a, and 4a. The more intense fluorescence of
the sulfonated and phosphorylated dyes is due to the hydro-
philic ionizable groups preventing the aggregation of the
dye molecules and/or the formation of non-emitting dimers
in aqueous solutions.[22] The sulfonation increases the quan-
tum yield of fluorescence more than the hydroxylation
(compare the values for compounds 3a and 4a, as well as
2a and 5-H in the last column of Table 2). In water, com-
pounds 1, 1a, and the “hydroxylated” dyes (except 4b) dem-
onstrate deviations from the Beer–Lambert law. For exam-
ple, in Table 2 one can see that the “actual” extinction coef-
ficient is lower than anticipated for dyes 4a, 5, and 6. Such
deviations are typical for dyes with limited solubility and
are due to adsorption (or even precipitation) on the quartz
surface of a cuvette. Therefore, these dyes are very likely to
form aggregates in water. Nevertheless, for these hydroxACHTUNGTRENNUNGyl-
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

C. A. Wurm, V. N. Belov, S. W. Hell et al.
secondary antibodies and used for immunofluorescence la-
beling of tubulin. The following dyes were used: four new
dyes with linkers (2b, 5, 6, and 7), and two reference dyes,
KK114 (2a)[5] and Atto647N.[7] Several factors influence the
performance of a fluorophor in imaging applications: bright-
ness of the fluorescent label (proportional to the absorption
coefficient and the fluorescence quantum yield), sufficient
degree of labeling, photostability of the dye, solubility of the
fluorophor (which does not unspecifically bind with proteins
and can be completely removed after coupling), and so on.
It is typical that in conjugates with antibodies the quan-
tum yield of fluorescence decreases.[24] The fluorescence
quantum yields (Ffl.) in conjugates with antibodies often
depend on the degree of labeling (DOL):[25] higher quantum
yields are observed at lower DOL values (see values with
footnote [c] in Table 3). For the new dyes, in antibody conju-
gates, the fluorescence quantum yields were found to be
very high (up to 53% or even 64%; see data in Table 3).
These are, in fact, the highest quantum yields of fluores-
cence that have ever been reported for the near-IR emitting
fluorescent dye conjugates.
Figure 2. Derivatives of dye 1 with variable positions A, B, and C. For
spectral properties, see Table 2.
ACHTUNGTRENNUNGated dyes with linkers (5-H and 6-H), the quantum yields of
fluorescence in solutions are as good as that of the sulfonat-
ed and phosphorylated analogues (2a, 2b, and 7).[23] One
can see that substitution at the allylic sites (position A in
compound 1, Scheme 1 and Figure 2) practically does not
shift the absorption and emission maxima (Æ6 nm). The sul-
fonic acid group which substitutes one of four fluorine
atoms in the aromatic ring (position B in compound 1,
Figure 2) only slightly increases the fluorescence quantum
yield (compare dyes 4a and 4b, Table 2). This is also true
for the carbopyronine dye that had a sulfonic acid group at-
tached to the linker.[5b] Thus, two ionizable groups in the al-
lylic positions of the rhodamine scaffold have a crucial influ-
ence on the efficiency of emission.
Remarkably, the replacement of the fluorine with thiol
moiety (2-mercaptoethyl sulfonic acid) practically did not
shift the absorption and emission bands in the dyes (com-
pare dye 4a with 4b, and 5-H with 6-H in Table 2).
In conjugates with antibodies, the absorption and emission
bands of the fluorescent markers are further shifted to the
red region by approximately 5 nm, relative to the starting
dye in the form of an acid. This means that the small hypso-
chromic shift in dyes 5-H, 6-H, and 7-H (2–5 nm; see
Table 2) compared with dye KK114 (2a), is insignificant re-
garding their practical use as fluorescent markers. To evalu-
ate the imaging performance of the dyes in conventional
and super-resolution microscopy, six dyes were coupled to
Table 3. Conjugates of dyes with antibodies (Sheep anti-Mouse in all
cases) used for immunolabeling, imaging of tubulin filaments in fixed
PtK2 cells, as shown in Figure 3, and evaluation of bleaching rates, as
shown in Figure 4.
Conjugated dye[a]
2a
2b
5-H
6-H
7-H
2.5
64
100
Atto-647N
Degree of labeling[b]
Ffl. [%] in conjugates
DOLꢁFfl.
1
8
9
45
5
2.8
46
81
4.6
12
35
53[c]
25
15[c]
47
(normalized to 100%)
[a] For structures see Scheme 1 or Figure 2. [b] DOL: average amount of
the dye residues attached to one antibody molecule (Mꢀ150000).
[c] Dye 2a (KK114) had Ffl. =40% at DOL=2.2 and dye 5-H demon-
strated Ffl. =53% at DOL=0.7 (see footnote [f] to Table 2).
We found that all new dyes are suitable for immunofluo-
ACHUTNGERNrNUG esAHCTUNGTRNENcUGN ence assays. The confocal and STED nanoscopy images
were bright, demonstrated good signal-to-noise ratio, and
excellent optical resolution.[26] In this respect, the tested new
dyes seem advantageous compared to Atto647N (Figure 3).
Table 2. Properties of red-emitting rhodamine dyes (Figure 2).
Dye Polar groups
Linker Net
Solubility in PBS
NHS ester stability in
H2O or PBS buffer
Absorption lmax Emission lmax
e [MÀ1 cmÀ1
]
Ffl. [%]
(H2O)
charge[a] buffer, pH 7.4
[nm] (H2O)
[nm] (H2O)
ꢁ10À5 (H2O)
1
–
–
–
yes
–
–
–
+1
+1
À1
+1
0
+1
0
À3
À1
À1
insoluble
very low
excellent
low
–
615[b]
638
625
621
623
632
634
635
637
637
646
661
644
649
644
654
654
655
660
660
0.35[c]
0.73[c]
0.93
36
42
49
31
1a[5a]
3a
4a
4b
very low
–
–
2ꢁSO3H
2ꢁOH
2ꢁOH, SO3H
0.50[c]
0.95
good[d]
low
–
39
5-H 2ꢁOH
yes
good
excellent
good
moderate
good
0.87[c]
0.63[c]
0.75
0.92
0.90
45[e]
55
6-H 2ꢁOH, SO3H yes
7-H 2ꢁOPO(OH)2 yes
good[d]
good[d]
excellent
excellent
55[f]
53[f,e]
55
2a[5a] 2ꢁSO3H
yes
yes
2b
2ꢁSO3H
[a] In conjugates; in the free acid the charge is one unit more negative. [b] In water with 10% methanol. [c] Deviations from the Beer–Lambert law due
to the low polarity/solubility; extinction is lower than anticipated due to adsorption on a quartz surface. [d] Only slightly soluble in pure water, well-solu-
ble in basic media, including aqueous NaHCO3. [e] In conjugates with Goat-anti-Rabbit antibodies, compounds 2a and 5-H have Ffl. =40 and 53% at de-
grees of labeling (DOL) 2.2 and 0.7, respectively. [f] Lifetimes of the exited state for 2a and 7-H are 3.4 and 3.6 ns, respectively.
&
8
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Red-Emitting Rhodamines in Fluorescence Nanoscopy
FULL PAPER
Figure 3. Confocal (left) and STED (right) microscopy images of tubulin filaments in fixed PtK2 cells (raw data). The tubulin cytoskeleton was immuno-
labeled with the commercial dye Atto647N, and compounds 2a, 2b, 5-H, 6-H, or 7-H respectively; excitation with a 640 nm diode laser (ꢀ10 mW). De-
tection at 670Æ40 nm, pulsed STED at 760 nm with power of 200 mW at the objectives back aperture; pulse duration 300 ps at 76 MHz repetition rate).
The numbers (bottom, right) indicate the relative brightness (%) of the resulting image. The brightness of the individual images is normalized to the
brightest pixel in each image (to be able to evaluate contrast and background fluorescence), and the relative brightness of each image is indicated in its
lower right corner. The average number of fluorophores attached to one molecule of an antibody (DOL) is not matched for the different images (see
Table 3), and that influences the brightness of the images. Scale bar: 5 mm.
If we compare the experimental values for the relative
brightness in Figure 3 with the calculated values DOLꢀFfl.
(last line in Table 3), we may conclude that among the hy-
drophilic rhodamine dyes, the brightness of the images
(under confocal conditions) is only very roughly correlated
with the labeling degrees and fluorescence quantum yields
in conjugates. According to the data in Table 3, dyes 6-H
and 7-H are expected to provide the brightest images. How-
To evaluate and quantify the photostability of the new
dyes, bleaching curves under confocal and STED conditions
(Figure 4) were recorded. For this purpose, the tubulin cy-
toskeleton of fixed PtK2 cells was labeled using the antibod-
ies described in Table 3. The total fluorescence signal of the
same area in the immunolabeled cell samples over the
course of several scans was then compared. The photostabil-
ity of the reference dye Atto647N under confocal conditions
is the highest (under irradiation with 640 nm light). That is
not surprising, and can be explained by the absence of the
C=C double bond in the structure of Atto647N.[5b]
However, under STED conditions (under irradiation with
very strong 760 nm light), dyes 5-H, 6-H, and 2b were found
to be more photoresistant than Atto647N. Moreover, three
new dyes (5-H, 6-H, and 7-H) appear to be more photosta-
ble than KK114 (2a). These results indicate that, unexpect-
edly, the hydroxyl groups protect the fluorescent dyes from
photobleaching at least as well as sulfonic or phosphoric
acid residues, or even better.
ever, the highest brightness was detected for the phospho
ACHTUNGTNERrNUNG yl-
ACHTUNGTRENNUNG
dyes 2a and Atto647N give images that are darker than the
images obtained with other dyes.
Another important imaging parameter associated with
brightness is the contrast, or the absence of the fluorescent
background. In this respect, compounds 2a and 7-H are the
best because they demonstrate high contrast and very low
background (compare the STED images in Figure 3). The
imaging performance of the two hydroxylated dyes 5-H and
6-H is virtually the same, which is in a good agreement with
their properties in aqueous solutions (which are also similar;
see Table 2). The sulfonic acid group attached to position B
of the scaffold (see Scheme 1 and Figure 2) affects only solu-
bility and the stability of the NHS esters.
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
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

C. A. Wurm, V. N. Belov, S. W. Hell et al.
Figure 4. Photostability of dyes 2a, 2b, 5-H, 6-H, 7-H, and the commercial dye Atto647N in bioconjugates. Bleaching curves were measured by taking
consecutive confocal (left) and STED (right) microscopic images of tubulin filaments in fixed PtK2 cells. To this end, the tubulin cytoskeleton was immu-
nolabeled with the corresponding fluorophores; excitation with 640 nm diode laser (~10 mW). Detection at 670Æ40 nm; STED at 760 nm with 200 mW
at 76 MHz at the objectives back aperture; pulse duration 300 ps). The average number of fluorophores per antibody (DOL) is not matched (see
Table 3). This may influence the bleaching rates for different fluorophores.
cell biology. In addition to the available fluorescent fusion
proteins, several strategies such as SNAP-, CLIP- or Halo-
tags were established in recent years to label proteins in
living cells specifically with organic fluorophores.[27] Modi-
fied cellular enzymes are applicable for covalent addition of
a dye that in this case has to be bound with a certain small
molecule. However, only a limited number of dyes are cur-
rently available for these labeling strategies. The most im-
portant requirement for the dyes used in these labeling
Figure 5. Membrane-permeation experiments of the model rhodamine
strategies is that a sufficient amount of it needs to pass
dyes 5-Me (left) and 6-Me (right) (see Figure 2 for the structures of com-
through cellular membrane(s) to reach the target protein.
Dyes with positive or zero charged residues are the most
promising candidates for testing the cell permeability.[28]
Therefore we chose methyl esters 5-Me (with a positively
charged dye residue; q= +1) and 6-Me (with an uncharged
moiety; q=0) as model compounds imitating the corre-
sponding conjugates of dyes 5-H and 6-H. To explore the
permeability of the cell membrane, the following protocol
was used. The living mammalian cells (PtK2) were grown
overnight on cover slips, and were then subjected to 1 mg of
a dye in 1 mL of the growth medium for 10 min at room
temperature (the amount of dye used for the labeling of
SNAP-fusion proteins is generally below 5 mg in 1 mL).
After that, the cells were washed with a pure growth
medium (without a dye), and imaged using an epifluores-
cence microscope (Figure 5). We found that both dyes cross
the cellular membranes and are incorporated into cellular
structures, although to different extents. As visible in
Figure 5, the zwitterionic dye 6-Me can hardly be detected
in the cell, whereas the positively charged dye 5-Me enrich-
es the cellular structures, even if it is applied in high dilu-
tion. It may thus be useful for the new labeling strategies in
living cells. It is not a trivial result because the dye molecule
is large and relatively heavy (Mꢀ1000). In this respect, it is
important to note that compound 5-H gave bright and high
pounds 5-H and 6-H). Wide-field fluorescence images of living mammali-
an PtK2 cells after 10-minute incubation with aqueous buffer solution of
the dyes (1 mgmLÀ1). For comparison, the imaging conditions and color
tables are kept constant. Scale bar: 20 m.
contrast fluorescent images under confocal and STED con-
ditions (Figure 3), so that the method of STED nanoscopy is
expected to be applied for example in studying the SNAP-
tagged protein distribution using amides prepared from
compound 5-H. The neutral (zwitterionic with a zero net
charge) dye 6-Me was found to be less cell-permeable, as
can be seen in Figure 5 (at least at low concentrations).
Electrophoretic mobility of the sulfonated and phosphoryl-
AHCTUNGTREGaNNNU ted fluorescent dyes: To estimate the actual negative net
charges in dyes 6-H, 2b, and 7-H in aqueous solutions (at
pH 8), these compounds were analyzed by capillary gel elec-
trophoresis with laser-induced fluorescence detection (CGE-
LIF), utilizing a standard capillary DNA sequencer.[14a,29]
The electropherograms of the three dyes 6-H, 2b, and 7-H
are presented in Figure 6. Although the main emission
wavelength of the excitation source (argon laser (488 nm))
does not match the absorption bands of the dyes, these com-
pounds have sufficient absorption at 514 nm, and can be ex-
cited by this weaker line of the argon laser (514.5 nm). The
&
10
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Red-Emitting Rhodamines in Fluorescence Nanoscopy
FULL PAPER
readily cross the cellular membranes, even if applied in high
dilution. On the other hand, compound 6-Me, with a zero
net charge, proved to be far less cell-permeable. Important-
ly, aliphatic tertiary amides (dyes with a linker for conjuga-
tion) were prepared from the precursors with free carboxylic
acid group without protection of the hydroxyl. The dye with
two hydroxyl groups and an amide linker demonstrated very
high fluorescence quantum yield as a free acid and in conju-
gates with antibodies (at low degrees of labeling).
The hydroxymethyl residues in a rhodamine dye were
transformed into the primary phosphate groups. This study
is the first to describe a fluorescent dye with two primary
phosphate residues, which are expected to make the com-
pound more “bio-compatible”. The phosphorylated dye (7-
H in Figure 2) was shown to be an important alternative to
widely used sulfonated fluorescent dyes. High polarity and
the increased (relative to the sulfonic acid residue) number
of acidic protons in the phosphorus-containing dye results in
good solubility in neutral and basic aqueous solutions (pH>
7), accompanied by high fluorescence quantum yield. Far
better electrophoretic mobility (at pH 8) is the result of the
higher net charge (q=À4) of the phosphorylated dye.
Moreover, a third variable position was involved to im-
prove the water solubility of the hydroxyl substituted dye.
One of the four fluorine atoms in the aromatic ring was se-
lectively exchanged to the residue of 2-mercaptoethyl sul-
fonic acid—the commercially available bifunctional reagent.
This modification did not affect the spectral properties very
much and provided an “amphiphilic dye” (6-H, Figure 2),
which forms an exceptionally stable active NHS ester. Thus,
a convenient fluorescent marker was developed, which pro-
vides excellent imaging performance, is well-soluble in most
organic solvents, aqueous PBS buffer, or NaHCO3 solutions,
and is easy to purify and handle. The influence of three
polar substituents (hydroxyl, phosphoric, and sulfonic acid)
on the spectral properties and the imaging performance of
the near-IR-emitting rhodamine dye were explored. In
many respects, both hydroxyl and phosphoric acid ester
groups proved to be viable alternatives to “classical” sulfon-
ic acid groups. Special attention was paid to the stability and
solubility of the corresponding amino-reactive markers (the
NHS esters), which are very important and often decisive in
many biological applications. Finally, a simple method for
preparation of methyl N-methyl-4-aminobutyrate hydro-
chloride was used. This ester represents a versatile linker for
the preparation of (amino) reactive markers with improved
stability.
Figure 6. Electrophoretic mobility of sulfonated and phosphorylated fluo-
rescent dyes. From right to left: 6-H (green curve; q=À1), 2b (red curve;
q=À2), 7-H (blue curve, q=À4). See text for details.
molecular masses of the compounds 7-H, 2b, and 6-H, their
molecular shapes and thus their hydrodynamic radii are sim-
ilar. Therefore, their relative migration times depend almost
exclusively on their negative net charges. Comparing the mi-
gration times of the distinct peaks, one can clearly see that
the electrophoretic mobility of a dye is directly proportional
to its negative charge. Thus, the migration time of the phos-
phorylated dye 7-H (with two OP(OH)2 groups and net
charge q=À4) is two times shorter than that of 2b (the ana-
logue with two SO3H groups and q=À2), and the migration
rate of 2b is half that of 6-H (one SO3H group, q=À1).
Figure 6 proves that at pH 8, all the anion-providing groups,
particularly primary phosphate, are completely ionized.
Thus, it was shown that phosphorylation is indeed a new ef-
ficient tool to drastically increase the electrophoretic mobili-
ty of a dye. One small additional peak near the main peak
of 2b appears probably due to the presence of amide rota-
ACHTUNGTRENNUNGmers. The existence of these rotamers was confirmed by
NMR spectroscopy (see the Supporting Information), in
which certain signals are split.
Conclusion
Three variable positions in a scaffold of a rhodamine dye (1,
Scheme 1 and Figure 2) were utilized to obtain red-emitting
fluorescent dyes with new valuable properties. Particularly,
direct oxidation of the two methyl groups attached to the
double C=C bonds to the corresponding allylic alcohols was
explored. This transformation gives way to a more hydro-
philic compound with a considerably increased fluorescence
quantum yield (in aqueous solutions), and certain solubility
in water, although the starting material is insoluble. Hydrox-
ylation does not introduce an additional negative charge
and therefore does not inhibit cell-permeability. Derivatives
of compound 5-H (e.g., methyl ester 5-Me in Figure 2) with
two hydroxyl groups and positively charged dye residue
The new dyes demonstrated excellent imaging perform-
ance in conventional and super-resolution microscopy and
perfectly matched the required spectral parameters.
Acknowledgements
The authors are grateful to Dr. Mariano Bossi, Dr. Heiko Schill, Ma-
rianne Pulst, Nina Ohm and Jꢂrgen Bienert (MPI BPC Gçttingen), Rein-
hard Machinek, Dr. Holm Frauendorf and their co-workers (Institut fꢂr
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

C. A. Wurm, V. N. Belov, S. W. Hell et al.
Organische und Biomolekulare Chemie, Georg-August-Universitꢃt Gçt-
tingen) for recording spectra. We are indebted to Bundesministerium fꢂr
Bildung und Forschung (BMBF 513) for financial support in the program
Optische Technologien fꢂr Biowissenschaften und Gesundheit (FKZ
13N11066), and for support by the DFG-Research Center for Molecular
Physiology of the Brain. We appreciate the assistance of E. Rothermel
with the cell culture and J. Jethwa for proof-reading the manuscript.
2364–2371; d) C. Kukat, C. A. Wurm, H. Spꢆhr, M. Falkenberg,
N.-G. Larsson, S. Jakobs, Proc. Natl. Acad. Sci. USA 2011, 108,
13534–13539; e) C. A. Wurm, D. Neumann, M. A. Lauterbach, B.
Harke, A. Egner, S. W. Hell, S. Jakobs, Proc. Natl. Acad. Sci. USA
2011, 108, 13546–13551; f) Y. Hua, R. Sinha, C. S. Thiel, R. Schmidt,
J. Hꢂve, H. Martens, S. W. Hell, A. Egner, J. Klingauf, Nat. Neuro-
sci. 2011, 14, 833–839.
[7] Atto-Tec GmbH, Siegen, Germany. For the structure of Atto647N,
see ref. [5]. This dye is a carbopyronine with a positively charged
dye residue
[8] K. Kolmakov, C. A. Wurm, M. V. Sednev, M. L. Bossi, V. N. Belov,
S. W. Hell, Photochem. Photobiol. Sci. 2012, 11, 522–532.
[9] L. G. Lee, R. J. Graham, W. E. Werner, E. Swartzman, L. Lu, App-
tera Corp. USA, US Pat. 6372907, 2002.
[10] a) F. Mao, W.-Y. Leung, R. P. Haugland, Molecular Probes, Inc.,
Eugene, Oregon, USA, US Pat. 6130101, 2000; b) A. Zilles, J.
Arden-Jacob, K.-H. Drexhage, N. U. Kemnitzer, M. Hammers-
Schneider, Atto-Tec GmbH, Siegen, Germany, WO 2005/003086,
2005.
[11] See: a) Ref. [10a]; b) B. Agnew, K. R. Gee, T. G. Nyberg, Invitrogen,
USA, US Pat. 2007/0249014, 2007; c) S. M. Polyakova, V. N. Belov,
S. F. Yan, C. Eggeling, C. Ringemann, G. Schwarzmann, A. de Mei-
jere, S. W. Hell, Eur. J. Org. Chem. 2009, 5162–5177.
[12] a) For an example of improving the solubility in water by replacing
ethyl substituents in N,N,N’,N’-tetraethyl-3,6-diaminoxanthene frag-
ment by 2-hydroxyethyl groups, see: R. Bandichhor, A. D. Petrescu,
A. Vespa, A. B. Kier, F. Schroeder, K. Burgess, Bioconjugate Chem.
2006, 17, 1219–1225 (however, unprotected hydroxyls may not be
used in the coupling reactions involving the 2-carboxy group of the
rhodamine phenyl ring); b) except perhaps amino groups; but these
are prohibitive in the allylic positions, because in close proximity to
the fluorophore, they may quench the emission of light due to the
photoinduced electron transfer (PET) from the lone electron pair of
the nitrogen atom to the half-occupied HOMO of the fluorophore
in the excited state.
[13] N. Rabjohn, Selenium Dioxide Oxidation, Organic Reactions, (Ed.:
S. E. Denmark, Wiley Online Library, New York, 1976, pp. 261–
416,.
[14] a) See the Supporting Information for details; b) K. Kolmakov, S.
Nizamov, V. N. Belov, C. Wurm, K. Willig, S. W. Hell, H. Schill, L.
Kastrup, G. Donnert, J. Bꢂckers, Max-Planck Innovation, PCT/EP
2011/005231, 2011.
[1] For some recent examples, see: a) M. Sauer, K.-T. Han, R. Mꢂller,
S. Nord, A. Schulz, S. Seeger, J. Wolfrum, J. Arden-Jacob, G.
Deltau, N. J. Marx, C. Zander, K.-H. Drexhage, J. Fluoresc. 1995, 5,
247–261; b) J. Arden-Jacob, J. Frantzeskos, N. U. Kemnitzer, A.
Zilles, K.-H. Drexhage, Spectrochim. Acta 2001, 57A, 2271–2283;
c) C. Jung, B. K. Mꢂller, D. C. Lamb, F. Nolde, K. Mꢂllen, C.
Brꢃuchle, J. Am. Chem. Soc. 2006, 128, 5283–5291; d) K. Peneva, G.
Mihov, F. Nolde, S. Rocha, J. Hotta, K. Braeckmans, J. Hofkens, H.
Uji-i, A. Hermann, K. Mꢂllen, Angew. Chem. 2008, 120, 3420–3423;
Angew. Chem. Int. Ed. 2008, 47, 3372–3375; e) S. L. Niu, G. Ulrich,
R. Ziessel, A. Kiss, P.-Y. Renard, A. Romieu, Org. Lett. 2009, 11,
2049–2052; f) M. Fu, Y. Xiao, X. Qian, D. Zhao, Y. Xu, Chem.
Commun. 2008, 1780–1782; g) Z. Lu, N. Liu, S. J. Lord, S. D. Bunge,
W. E. Moerner, R. J. Twieg, Chem. Mater. 2009, 21, 797–810; h) F.
Bruyneel, L. DꢄAuria, O. Payen, P. J. Courtoy, J. Marchand-Brynaert,
ChemBioChem 2010, 11, 1451–1457; i) Y. Koide, Y. Urano, K. Ha-
naoka, T. Terai, T. Nagano, ACS Chem. Biol. 2011, 6, 600–608; j) L.
Yuan, W. Lin, Y. Yang, H. Chen, J. Am. Chem. Soc. 2012, 134,
1200–1211; k) M. P. Shandura, V. P. Yakubovskyi, A. O. Gerasov,
O. D. Kachkovsky, Y. M. Poronik, Y. P. Kovtun, Eur. J. Org. Chem.
2012, 1825–1834.
[2] For reviews on modern optical nanoscopy methods, see: a) S. W.
Hell, in Single Molecule Spectroscopy in Chemistry, Physics and Bi-
ology (Eds.: A. Grꢃslund, R. Rigler, J. Widengren), Springer, Berlin,
2010, pp. 365–398; b) D. Puliti, D. Warther, C. Orange, A. Specht,
M. Goeldner, Bioorg. Med. Chem. 2011, 19, 1023–1029; c) S. van de
Linde, S. Wolter, M. Sauer, Austr. J. Chem. 2011, 64, 503–511; d) T.
Mꢂller, C. Schumann, A. Kraegeloh, ChemPhysChem 2012, 13,
1986–2000; for the most recent references, see: e) G. T. Dempsey,
J. C. Vaughan, K. H. Chen, M. Bates, X. Zhuang, Nat. Methods
2011, 8, 1027–1040; f) M. Bates, G. T. Dempsey, K. H. Chen, X.
Zhuang, ChemPhysChem. 2012, 13, 99–107.
[3] For applications and spectral range of rhodamines as fluorescent
dyes, see: a) M. Beija, C. A. M. Afonso, J. M. G. Martinho, Chem.
Soc. Rev. 2009, 38, 2410–2433; b) M. S. T. GonÅalves, Chem. Rev.
2009, 109, 190–212; c) H. N. Kim, M. H. Lee, H. J. Kim, J. S. Kim, J.
Yoon, Chem. Soc. Rev. 2008, 37, 1465–1472; d) L. D. Patsenker,
A. L. Tataretz, Y. A. Povrozin, E. A. Terpetsching, Bioanal. Rev.
2011, 3, 115–137; e) R. P. Haugland, A Guide to Fluorescent Probes
and labeling Technologies, Invitrogen, Carlsbad, 2010, pp. 15–23,
35–41, 47–48, 74–77.
[4] For recent reviews on rhodamine synthesis and derivatization, see:
a) S. Murphree in Progress in Heterocyclic Chemistry, Vol. 22, (Eds.:
G. Gribble, J. A. Joule), Elsevier, Amsterdam, 2011, pp. 21–58;
b) L. M. Wysocki, L. D. Lavis, Curr. Opin. Chem. Biol. 2011, 15,
752–759; for the most recent references, see: c) T. Peng, D. Yang,
Org. Lett. 2010, 12, 496–499; d) A. Romieu, T. Bruckdorfer, G.
Clavꢅ, V. Grandclaude, C. Massif, P.-Y. Renard, Org. Biomol. Chem.
2011, 9, 5337–5342; e) J. B. Grimm, L. D. Lavis, Org. Lett. 2011, 13,
6354–6357.
[15] The final product, a red emitting dye, needs to have a linker for con-
jugation with the target molecules. The presence of a tertiary amide
(compounds 2a,b) or an ester group (compound 3b) additionally
redshift the absorption and emission bands to 5–10 nm.
[16] See, for example, the available structures of fluorescent dyes pro-
duced by Atto-Tec GmbH, Siegen (Germany), www.atto-tec.de.
[17] R. K. Olsen, J. Org. Chem. 1970, 35, 1912–1915.
[18] a) F. H. Westheimer, Science 1987, 235, 1173–1178; b) F. H. West-
heimer, S. Huang, F. Covitz, J. Am. Chem. Soc. 1988, 110, 181–185.
[19] See: a) J. Beechem, K. R. Gee, D. C. Hagen, I. D. Johnson, H.-C.
Kang, C. Pastula, Invitrogen Corp., Carlsbad, CA, USA, US pat.
7282339, 2007; b) E. Lukhatov, EPOCH Biosciences, Inc., Washing-
ton DC, USA, WO 2009/046165, 2009.
[20] a) F. Wold, C. E. Ballou, J. Biol. Chem. 1957, 227, 301–312; b) L. D.
Freedman, G. O. Doak, Chem. Rev. 1957, 57, 479–523.
[21] a) Y. Kagoshima, M. Mori, E. Suzuki, N. Kobayashi, T. Shibayama,
M. Kubota, Y. Kamai, T. Konosu, Chem. Pharm. Bull. 2010, 58,
794–804; b) S. L. Beaucage, Oligodeoxyribonucleotides Synthesis
Phosphoramidite Approach in Methods in Molecular Biology: Proto-
cols for Oligonucleotides and Analogs (Ed.: S. Agrawal), Humana
Press Inc., Totowa, 1993, pp. 33–61.
[22] O. Valdes-Aguilera, D. C. Neckers, Acc. Chem. Res. 1989, 22, 171–
177.
[23] That means that the polarity and solubility in water do not directly
correlate with the fluorescence quantum yield in solutions. For an
example of the hydrophilic rhodamine derivative with four 2-hy-
[5] a) K. Kolmakov, V. N. Belov, J. Bierwagen, C. Ringemann, V.
Mꢂller, C. Eggeling, S. W. Hell, Chem. Eur. J. 2010, 16, 158–166;
b) K. Kolmakov, V. N. Belov, C. A. Wurm, B. Harke, M. Leuteneg-
ger, C. Eggeling, S. W. Hell, Eur. J. Org. Chem. 2010, 3593–3610.
[6] For leading references on applications of red-emitting dye KK114,
see: a) R. Schmidt, C. A. Wurm, A. Punge, A. Egner, S. Jakobs,
S. W. Hell, Nano Lett. 2009, 9, 2508–2510; b) J. Bꢂckers, D. Wild-
anger, G. Vicidomini, L. Kastrup, S. W. Hell, Opt. Express 2011, 19,
3130–3143; c) P. A. Pellett, X. Sun, T. J. Gould, J. E. Rothman, M.-
Q. Xu, I. R. CorrÞa Jr., J. Bewersdorf, Biomed. Opt. Exp. 2011, 2,
&
12
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Red-Emitting Rhodamines in Fluorescence Nanoscopy
FULL PAPER
droxyethyl groups replacing ethyl substituents in N,N,N’,N’-tetraeth-
yl-3,6-diaminoxanthene fragment, see reference [12a].
[24] See: V. P. Boyarskiy, V. N. Belov, R. Medda, B. Hein, M. Bossi, S. W.
Hell, Chem. Eur. J. 2008, 14, 1784–1792.
[25] For the protocols of the labeling reaction between proteins and
NHS esters of the fluorescent dyes and estimation of DOL values,
see: www.abberior.com.
[26] C. A. Wurm, K. Kolmakov, F. Gçttfert, S. Berning, M. L. Bossi, S.
Jakobs, G. Donnert, V. N. Belov, S. W. Hell, Journal of Optical
Nanoscopy, unpublished results.
[27] H. M. OꢄHare, K. Johnsson, A. Gautier, Curr. Opin. Struct. Biol.
2007, 17, 488–494.
[28] The outer part of the plasma membrane is negatively charged, and,
due to electrostatic repulsion, the negatively charged ions as a rule
do not penetrate through it (if there is no specific transport mecha-
nism for them).
[29] Recorded electropherogram raw data were processed and analyzed
using glyXtool V3.3 software: R. Hennig, U. Reichl, E. Rapp, Gly-
coconjugate J. 2011, 28, 331.
Received: April 5, 2012
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
13
&
ÞÞ
These are not the final page numbers!

C. A. Wurm, V. N. Belov, S. W. Hell et al.
A study in scarlet: Rhodamines that
absorb at 621–637 nm and emit at 644–
660 nm were prepared and used in flu-
orescence microscopy and nanoscopy.
They are photostable, feature variable
polarities and net charges, high stabili-
ties of N-hydroxysuccinimidyl (NHS)
esters, large fluorescence quantum
yields in aqueous solutions, and anti-
body conjugates.
Fluorescence
K. Kolmakov, C. A. Wurm,*
R. Hennig, E. Rapp, S. Jakobs,
V. N. Belov,* S. W. Hell* . . . . &&&& — &&&&
Red-Emitting Rhodamines with
Hydroxylated, Sulfonated, and Phos-
phorylated Dye Residues and Their
Use in Fluorescence Nanoscopy
“A Study in Scarlet”…… by Sir Arthur Conan
Doyle was a detective novel introducing Sherlock
Holmes (who was fond of chemistry) and the magni-
fying glass as an investigative contraption. Color,
chemistry, and contraption all relate to the picture,
which shows one of the best red-emitting rhodamine
dyes described by C. A. Wurm, V. N. Belov, S. W.
Hell et al. in their Full Paper on pages && ff. The
tubular network in a cell was stained with the dye
and imaged in a modern light microscope, a refine-
ment of the magnifying glass, as well as by STED
microscopy.
&
14
&
www.chemeurj.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!