Live-cell imaging with water-soluble aminophenoxazinone dyes synthesised through laccase biocatalysis

Article abstract of DOI:10.1002/cbic.201000145

Aminophenoxazinone dyes with variable water solubilities were assayed for the first time in a live-cell imaging application. Among a library of ten sulfonylated chromophores, one compound gave excellent results as an endocytic marker, showing a precise su

Full text of DOI:10.1002/cbic.201000145

DOI: 10.1002/cbic.201000145
Live-Cell Imaging with Water-Soluble Aminophenox-
azinone Dyes Synthesised through Laccase Biocatalysis
Frꢀdꢀric Bruyneel,[a] Ludovic D’Auria,[b] Olivier Payen,[a] Pierre. J. Courtoy,[b] and
Jacqueline Marchand-Brynaert*[a]
Aminophenoxazinone dyes with variable water solubilities
were assayed for the first time in a live-cell imaging applica-
tion. Among a library of ten sulfonylated chromophores, one
compound gave excellent results as an endocytic marker,
showing a precise subcellular distribution. The compound was
compared to four commercial vital tracers, including Lucifer
Yellow. The first laccase-mediated regioselective synthesis of a
diphosphorylated 2-aminophenoxazinone dye was also de-
scribed. This compound, water-soluble at 10À2 m, displayed
modest fluorescence properties and the ability to complex
Mg2+ and Ca2+ cations, therefore giving fluorescence quench-
ing.
Introduction
Fluorescent organic dyes are widely used as markers in the
field of molecular biology and biomedical research.[1,2] Such
nonradioactive probes allow the labelling of biomolecules that
are of interest for imaging cell structure and function.[3,4]
Indeed, the subcellular distribution, physical association and
specific interactions of biological molecules at, and inside the
cell are of central importance. All of these aspects are particu-
larly relevant to the study of the growing number of human
disorders that are related to a defect in the endo-lysosomal
system.[5] Moreover, metabolic mapping in living cells encom-
passes a large set of cellular events, among which endocytic
pathways remain an important research field.[6,7] In practice,
live-cell imaging with endocytic markers is usually explored
among the first potential applications whenever one desires to
develop a new class of fluorophores for molecular biology. On
the other hand, whereas numerous available synthetic fluores-
cent dyes exhibit high quantum yields, narrow emission bands
up to the near-infrared and good photostability, the severe
drawback of poor water solubility decreases the utility of most
of them.[1] Commonly, improved water solubility is obtained
through the insertion of ionisable hydrophilic groups into the
fluorophore core structure, or by dye conjugation to biopoly-
mers such as carbohydrates.[8,9] In the former strategy, func-
tions such as sulfonic, and/or ammonium groups are the most
popular ones, and have been widely used in the case of ben-
zo[a]phenoxazine and xanthene dyes, thus leading to several
families of water-soluble fluorophores endowed with high
quantum yields.[10,11] The classical synthetic pathway towards
benzo[a]phenoxazine dyes involves a nitrosative condensation
that is poorly efficient with sulfonated building blocks.[12] This
only improves solubility at the last stage of the chemical
synthesis, and thus considerably limits the diversity of novel
probe libraries.
vented by low water solubility and the absence of a general
methodology for the production of diversely substituted phe-
noxazinone scaffolds.[14] Recently, our group reported a novel
synthetic approach, based on the use of laccase as a biocata-
lyst, which showed promising results for the “green” produc-
tion of phenoxazinone dyes featuring variable water solubili-
ty.[15] From a pool of diverse benzenesulfonyl derivatives as
substrates, the last synthetic step corresponds to a cyclising
oxidative dimerisation conveniently performed at room tem-
perature, in aqueous solution, and with air as oxidant.
Herein, we report an extension of this methodology by the
production of the first phosphorylated aminophenoxazinone
dye. This water-soluble compound, along with the previously
prepared series of sulfonylated phenoxazinone dyes (see
Scheme 1 for general core structures) were further studied for
their optical properties and screened for their potential in fluo-
rescence live-cell imaging. The best candidate (compound 8c)
was finally characterised as an endocytic tracer of the degrada-
tive pathway, by reference to known fluorescent probes target-
ed to endosomes, lysosomes or the Golgi complex, by using
double-labelling experiments.
[a] Dr. F. Bruyneel, Dr. O. Payen, Prof. Dr. J. Marchand-Brynaert
Institute of Condensed Matter and Nanosciences (IMCN)
Organic and Medicinal Chemistry Unit (CHOM)
Universitꢀ catholique de Louvain
Bꢁtiment Lavoisier, Place Louis Pasteur 1, 1348 Louvain-la-Neuve (Belgium)
Fax: (+32)10-474168
E-mail: jacqueline.marchand@uclouvain.be
[b] L. D’Auria, Prof. Dr. P. J. Courtoy
Cell Biology unit (CELL) and Platform for Imaging Cells and Tissues (PICT)
de Duve Institute (ICP-UCL-7541), Universitꢀ catholique de Louvain
75, avenue Hippocrate, 1200 Bruxelles (Belgium)
Aminophenoxazinone dyes and pigments have been known
for several decades, most of them being of natural sources.[13]
Their potential use as fluorescent probes has been mainly pre-
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cbic.201000145.
ChemBioChem 2010, 11, 1451 – 1457
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1451

J. Marchand-Brynaert et al.
intermediate was produced in situ and then treated with
2.4 equivalents of diethyl chlorophosphate (step b) to give
compound 3 in 70% yield after chromatography. Next the
methyl ether function was cleaved using an excess of boron
reagent in dichloromethane (step c). After usual work-up and
precipitation in diethyl ether, pure benzoxazole compound 4
was obtained in 78% yield. Interestingly, spontaneous intramo-
lecular cyclisation was observed here, in the same way as pre-
viously described for the sulfonated analogue.[15a] This original
2-tert-butyl-4-phosphonic acid benzoxazole was hydrolysed in
a 12n HCl solution at reflux (step d) to give the precursor 5 as
a white powder in 70% yield, after concentration and precipi-
tation.
Scheme 1. General structures (I) and (II) of studied fluorescent dyes. FG=
functional group responsible for the solubility properties.
Results and Discussion
Chemistry
From a synthetic point of view, the Michaelis–Arbuzov reaction
was the traditional method employed when phosphorylated
aromatic dyes had to be produced.[16] However, this strategy
implies the regioselective synthesis of aromatic halides as start-
ing materials. The directed-metallation reaction has proven to
be an alternative method of choice to insert selectively an oxi-
dised phosphorous group in organometallic chemistry.[17]
Based on a synthetic approach previously used by our group
for selective aromatic sulfonation, we have taken advantage of
the insertion of a lithiated intermediate into a PÀCl bond,
which leads to the corresponding phosphorylated precursor in
good yield and complete selectivity. Next, a laccase was used
as catalyst to promote the oxidative cyclisation of the sub-
strate into the first diphosphorylated 2-aminophenoxazinone
dye produced as a single regioisomer. The synthetic pathways
for the efficient synthesis of our phosphorylated building block
and its dimerisation are outlined in Scheme 2.
To date, there is no example of a laccase-mediated oxidative
coupling reaction that makes use of organic phosphorylated
compounds as substrates. Sulfonic and phosphonic groups
being considered as bioisosteres of the carboxylic function,[18]
the previous conditions of our laccase-catalysed synthesis in
water were applied to compound 5. In the presence of a cata-
lytic amount of laccase in an aqueous buffer at pH 6, substrate
5 was converted at room temperature (step e) into the corre-
sponding 1,9-disubstituted phenoxazinone dye 6 as a >95%
pure compound with an excellent yield of 90% after isolation.
Briefly, the reaction was monitored by HPLC until complete
conversion of 5; next the laccase was inactivated by adding
40% (v/v) of methanol, and concentrated in vacuo before
freeze-drying of the remaining solution. The residual red
powder was fully characterised by HRMS and IR, NMR (1H and
13C) and UV–visible spectroscopy. The novel bis-phosphorylat-
ed dye 6 is highly water-soluble: solutions of 6 up to a concen-
tration of 10À2 m can be prepared in PBS or pure water. In
methanol, the highest concentration reached is about 5ꢂ
10À4 m. Syntheses and structural elucidations of the related bis-
sulfonylated dye 7 and derivatives 8a–g and 9 (see Table 1)
were previously reported.[15b]
Table 1. UV/Vis absorption data of dyes 6–9.[a]
Compound Functional group[b] Core[b] eꢂ103 [mÀ1 cmÀ1
]
lmax [nm][c]
6
7
PO3H
SO3H
SO2NH2
SO2NH-cyclohexyl
SO2NH(CH2)3N(Me)2
SO2NH-phenyl
SO2NH(CH2)2NH2
SO2NHCH2CO2Me
SO2NH(CH2)2CH2OH
SO2NH-cyclohexyl
I
I
I
I
I
I
I
I
I
8.80
9.80
4.04
10.80
7.20
0.48
6.02
5.90
12.60
3.95
241, 438
240, 418, 441
240, 436, 451
236, 424, 456
239, 422, 442
240, 430
240, 423, 442
238, 422, 442
240, 420, 440
238, 420, 430
8a
8b
8c
8d
8e
8 f
8g
9
Scheme 2. Synthesis of aminophenoxazinone 6. Reagents and conditions:
a) 1.1 equiv tBuCOCl, 3 equiv NaHCO3, H2O/EtOAc, 258C, 4 h; b) 2.4 equiv
nBuLi, 2.4 equiv TMEDA, THF, 2.4 equiv (EtO)2POCl, À108C, 2 h, then over-
night at 258C; c) 3 equiv BBr3, CH2Cl2, 08C, 2 h, then 258C, 18 h; d) HCl
(12n), 1008C, 24 h; e) MeOH/NH4OAc (0.2m, pH 6), laccase 100 ULÀ1, air,
258C, 24 h.
II
[a] Measured in PBS at pH value of 7.4. [b] See Figure 1. [c] The under-
lined value corresponds to the given e (molar coefficient extinction).
Briefly, commercially available o-anisidine (1) was quantita-
tively protected (step a) as N-pivaloyl derivative (2). The crude
intermediate was submitted to regioselective lithiation be-
cause of the slight difference in directing ability in favour of
the N-pivaloyl function over the O-methyl ether. By using
2.4 equivalents of nBuLi in the presence of tetramethylethyl-
enediamine (TMEDA) as activating agent, the lithiated aromatic
Absorption and fluorescence properties
The photophysical properties of the aminophenoxazinone
dyes featuring variable water solubilities were evaluated under
simulated physiological conditions, that is, in phosphate-buf-
fered-saline solution (PBS; 0.15m) at pH 7.4. The UV–visible
1452
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1451 – 1457

Live-Cell Imaging with Aminophenoxazinone Dyes
spectra of compounds 6–9 showed absorptions between 420
and 450 nm with e values ranging from 480 to 12600mÀ1 cmÀ1
(Table 1 and Scheme 1). Data collections were performed at
concentrations ranging from 5ꢂ10À4 to 1ꢂ10À6 m to ensure
full solubility of all compounds. Rather low e values were ob-
tained for dyes 8a, 8d and 9, whereas the e values for 8c, 8e
and 8 f were slightly higher. The highest e values were record-
ed for the dyes 6, 7, 8b and 8g. Still, the absorbance of
these newly synthesised compounds was ten- to 100-fold
lower than for a classical benzo[a]phenoxazine dye (e=42.4ꢂ
103 mÀ1 cmÀ1).[10a]
Next, the fluorescence properties of the functionalised ami-
nophenoxazinone dyes 6–9 were evaluated at concentrations
ranging from 5ꢂ10À5 to 1ꢂ10À4 m by using rhodamine 6g
(Rho-6g) as reference (FF (quantum yield)=0.95, see the Sup-
porting Information); the results are summarised in Table 2.
Table 2. Fluorescence data of dyes 6–9.[a]
Figure 1. Absorption (solid line, left axis) and fluorescence (dashed line, right
axis) spectra of 5ꢂ10À5 m 8c in PBS at pH 7.4. A.U.=arbitrary unit of fluores-
cence.
Compound
lexc [nm]
lem [nm]
FF
Stokes’ shift [nm]
6
7
488
488
488
488
488
488
488
488
488
488
500–650
500–600
520–680
520–680
500–700
570–620
500–700
500–700
500–700
520–650
0.003
0.001
0.009
0.007
0.017
0.0004
0.007
0.014
0.005
0.002
110
~100
122
130
116
~100
146
146
~130
~100
8a
8b
8c
8d
8e
8 f
8g
9
theless, a first screen by live-cell imaging was undertaken
using CHO (Chinese Hamster Ovary) cells after 1 hour of incu-
bation at 378C with the entire series of phenoxazinone dyes at
1 mgmLÀ1, with the exception of compound 8d. The results of
this visual screen are summarised in Table 3. It immediately ap-
[a] Measured in PBS (0.15m) at pH value of 7.4.
Table 3. Preliminary screen by live cell imaging, (6–9, ~5ꢂ10À3 m).[a]
Dye concentrations were adjusted to yield an identical optical
density of 0.06 in PBS (0.15m). Emission spectra after excitation
at l=488 nm were recorded between 500 and 700 nm. All
dyes exhibited some fluorescence, albeit with a rather low in-
tensity. Clear emission maxima were observed around 560 nm,
while some dyes, such as 7, 8d, 8g and 9, emitted broadly
from 500 to 650 nm without clear maxima. Relative quantum
yields measured by reference to Rho-6g were in general rather
low, almost 100 times lower than that of Lucifer Yellow (LY;
FF =0.21, see the Supporting Information).[19] Two dyes, 8c
and 8 f, emerged as most promising based on this criterion,
with quantum yields close to 0.02. Dyes 6, 8a–c and 8e exhib-
ited a large Stokes’ shift ranging from 110 to almost 150 nm
depending on the functional groups inserted in the core struc-
tures (see Figure 1). Compounds bearing sulfonylated side
chains were always more efficient than the parent sulfonic acid
7, whereas compound 6, bearing free phosphonic acids, still
gave a quite significant Stokes’ shift. Hydrophobicity did not
seem to play a major role either. Whereas dyes 6 and 7 exhibit-
ed similar molar extinction coefficients, a puzzling difference in
emission spectra was observed.
Compound
Fluorescence
Location
Aggregate
Cell state[b]
6
7
null
low
low
high
medium
low
medium
medium
null
–
–
–
–
–
–
good
good
stressed
good
good
good
good
stressed
8a
8b
8c
8e
8 f
8g
9
intracellular
extracellular
intracellular
intracellular
intracellular
intracellular
extracellular
massive
–
small
–
–
massive
[a] Excitation at 488 nm, Emission filter at 526–532 nm. Transmission at
5%; detection in the green channel. [b] Stress was attributed to aggrega-
tion. Due to the variable water solubility of dyes, final concentrations of
ethanol in this preliminary screen ranged from 0.5 to 5% for 8a, b, e, f,
g, and 9, and were below 0.5% for 6, 7 and 8c.
peared that the significant hydrophobicity of compounds 8b
and 9 caused massive aggregation, therefore preventing any
application to biological systems. Overall, the cell did not
appear to be appreciably altered, despite the rather high final
ethanol concentration in this screen.
Although it was the first time that such heterocyclic chromo-
phores became available for testing in biological media, it
might seem brave to apply dyes 6–9 in fluorescence labelling
experiments, given their rather limited quantum yield. Never-
The candidate selected for further imaging was the water-
soluble compound 8c, which also offers the highest quantum
yield. Furthermore, its intracellular distribution in well-defined
spots readily pointed to its potential as an endocytic tracer.
ChemBioChem 2010, 11, 1451 – 1457
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1453

J. Marchand-Brynaert et al.
The puzzling result observed with compound 6 seemed, a
priori, quite disappointing. Such an absence of fluorescence,
when even compound 7 could be detected in the experiments,
led us to consider that the data obtained with compound 6
were the result of an artefact. Bearing in mind the well-known
ability of phosphonate groups to chelate divalent cations, we
envisaged a quenching of fluorescence due to the presence of
Mg2+ and Ca2+ ions in the culture medium of CHO cells. Such
a loss of fluorescence due to the chelation of Mg2+ and/or
Ca2+ was indeed confirmed independently (Supporting Infor-
mation). Whereas compound 6 (4.6 mm) fluoresced in pure
water and PBS (0.15m NaCl, 0.8 mm PO42À), the quenching was
complete in the presence of PBS doped with Ca2+/Mg2+
(3 mm each) and similarly in aqueous solutions of each cation
(3 mm).
Subcellular localisation of compound 8c
Several endocytic pathways, all based on pinching-off vesicles
from the plasma membrane invaginations, allow bulk or selec-
tive internalisation of extracellular, membrane impermeant
compounds.[5] Bulk (i.e., fluid-phase or nonspecific adsorptive)
endocytosis allows us simultaneously to explore the various
internalisation portals and intracellular endocytic routes.[7] The
endo-lysosomal system comprises early, sorting, late and recy-
cling endosomes, secondary lysosomes and the trans-Golgi
network. Although early endosomes are primarily involved in
sorting internalised compounds and membrane, recycling en-
dosomes ensure the homeostasis of the plasma membrane.[22]
Lysosomes are usually the final destination where internalised
macromolecules are to be degraded. A large variety of fluores-
cent probes are available to analyse this complex system.[20–23]
Some of these, such as LY, are widely used as tracers for fluid-
phase endocytosis.[24] Other fluorescent dyes have shown in-
trinsic properties that permit the selective labelling of well-
defined intracellular compartments, such as the recycling path-
way (transferrin),[23] lysosomes[24] or the Golgi complex
(BODIPY-ceramide).[25]
Figure 2. Evidence for endocytic uptake of compound 8c. A), B), E), F) Con-
trol or C), D) ATP-depleted CHO cells were incubated for 45 min with 1.5 mm
of compound 8c at 378C in A), B) control or C), D) ATP-depleting medium;
or E), F) at 48C in the presence of 10 mm CellTracker, for global cell imaging.
The figure shows dark-field fluorescent images (left), merged with transmis-
sion images (right). In the upper panel, control cells showed bright intracel-
lular punctate labelling, mainly perinuclear (arrowheads). The boxed area at
A) is enlarged in the inset. As shown by the central and lower panels, ATP-
depletion (C, D) or incubation at 48C (E, F) does not lead to green intracellu-
lar labelling. In (F), the integrity of the plasma membrane is demonstrated
by cytosolic labelling with CellTracker Red. These experiments demonstrate
that 8c accumulates in cells by a temperature and energy-dependent mech-
anism that is consistent with endocytosis. N=nucleus, scale bar=10 mm.
brane-permeant in their nonprotonated form and become
sequestrated upon protonation in acidified, closed compart-
ments. As shown in Figure 3, compound 8c fully segregated
from recycling endosomes and the Golgi complex, and was
exclusively targeted to late endosomes/lysosomes.
Finally, to test whether compound 8c qualified as a bulk en-
docytic tracer, its endocytic trafficking was compared with that
To rule out diffusion as a mechanism contributing to intra-
cellular labelling by compound 8c, two classical endocytic con-
trols were carried out: energy depletion and low-temperature
block. As illustrated in Figure 2, cells incubated with 1.5 mm of
compound 8c for 45 min in serum-free medium showed multi-
ple intracellular dots, with preferential clustering around the
nucleus. In contrast, ATP-depleted cells exposed to com-
pound 8c at 378C, or untreated cells incubated with this com-
pound at 48C showed no fluorescence. The integrity of the
plasma membrane was verified by simultaneous labelling with
CellTracker Red, a fluorescent membrane-permeant probe that
becomes membrane-impermeant after conjugation with gluta-
thione (see the Supporting Information for chemical struc-
tures).
Figure 3. Internalised compound 8c is targeted to late endosomes/lyso-
somes. CHO cells were incubated with 8c at 378C for 45 min together with
A) (in red) 0.7 mm Alexa 568-transferrin (tracer of recycling endosomes),
B) 5 mm BODIPY-ceramide (a vital stain of the Golgi complex) or C) 50 nm Ly-
soTracker (to label late endosomes/lysosomes by acidotropic sequestration).
Note that internalised 8c fully segregates from recycling endosomes and
the Golgi complex and is exclusively targeted to late compartments of the
degradative pathway. N=nucleus, scale bar=10 mm.
We next analysed the fate of internalised compound 8c, by
reference to the vital tracers of recycling endosomes (Alexa
568-transferrin), the Golgi complex (BODIPY-ceramide, a meta-
bolic precursor of BODIPY-sphingomyelin), and LysoTracker
Red, which, like chloroquine and acridine orange, are mem-
1454
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1451 – 1457

Live-Cell Imaging with Aminophenoxazinone Dyes
of LY, as illustrated in Figure 4. Interestingly, although com-
pound 8c is a cationic species, due to protonation of its terti-
ary amino groups, and the bisulfonate compound LY is an
anionic species, both endocytic tracers labelled similar struc-
tures and apparently progressed towards lysosomes at a com-
parable rate.
an endocytic tracer could be ascribed to the benefit of mem-
brane adsorption, due to charge complementarity with the cell
surface glycocalyx. While the other sulfonylated water-soluble
chromophores 7–9 suffer from much lower quantum yields,
they might eventually reveal interesting properties yet to be
explored. Finally, dyes 8e–g feature at least one free reactive
function for further conjugation to biomolecules and grafting
to materials, which opens the route towards selective labelling
experiments.[26]
Experimental Section
1
General methods: H, 13C and 31P NMR spectra were recorded with
a Bruker Avance 500 or a Bruker Avance 300 spectrometer. Spectra
were obtained in [D6]DMSO, CDCl3, D2O or CD3OD. Chemical shifts
are reported in ppm relative to tetramethylsilane.[27] 13C NMR spec-
tra were obtained with broadband proton decoupling (D1=5 s) or
DEPTQ. Low-resolution mass spectra were acquired on a Thermo
Finnigan LCQ spectrometer, either in positive- or negative-mode
ESI. HRMS analyses by using ESI or FAB methods were performed
at the University of Mons Hainaut (Belgium) or at the University of
Oxford (UK). Melting points were recorded with an Electrothermal
apparatus calibrated with benzoic acid. Infrared spectra were re-
corded with a Shimadzu FTIR 84005 spectrometer using KBr plates.
The HPLC system consisted of a Waters Alliance 2699 separation
module, and a Waters 2998 photodiode array detector; analytical
separations were performed on a Waters XterraMSC18 column
(4.6 mmꢂ100 mm, 5 mm; Waters, Milford, MA, USA) equipped with
a conventional precolumn. Detection was performed at 220/254/
280 nm with a control at 440/460 nm and on-line UV/Vis absorb-
ance scans were performed. Flow rate was 1 mLminÀ1. Initial condi-
tions were ACN 100% for 5 min, then a hyperbolic gradient to
75:25 of solvents A (water/HCO2H, 100:0.1 v/v) and B (ACN/HCO2H,
100:0.1 v/v) over 5 min, then a hyperbolic gradient to 25:75 of A/B
in 10 min. Then, the cycle was terminated by returning to the ini-
tial conditions over 5 min. The injection volume was 10 mL.
Figure 4. Despite bearing opposite net charges, compound 8c and Lucifer
Yellow are expected to follow late endosomes/lysosomes by default. CHO
cells were incubated at 378C for 45 min with 50 nm LysoTracker Red togeth-
er with 1.5 mm compound 8c (left) or 2.2 mm LY (right), and then briefly
washed and studied by live confocal microscopy. Optical sections at the
level of the nucleus (N) are shown. The boxed areas, enlarged tenfold below,
suggest that pleiomorphic late endosomes, filled by endocytosis of the
membrane-impermeant tracers, 8c or LY, dock with lysosomes concomitant-
ly labelled by acidotropic sequestration of the membrane-permeant probe,
LysoTracker Red. Scale bars: upper panel 10 mm, lower panel 1 mm.
Materials: Lucifer Yellow (LY) was purchased from Sigma. Other
commercial fluorescent probes, namely Tfn-Alexa 568, ceramide-
BODIPY-589, Cell-Tracker-Red and LysoTracker-Red were obtained
from Invitrogen. Laccase from Trametes versicolor was purchased as
a brownish powder form Bioscreen e.K. (Uebach-Palenberg, Germa-
ny) under the commercial trademark Oxizym LA (batch no. LA
2000/001).
Conclusions
We here provide the first report on a series of water-soluble
aminophenoxazinone dyes that have been evaluated for vital
imaging in cell biology. These heterocyclic compounds were
produced by using an enzymatic synthetic strategy, as an alter-
native to the nitrosative or chemical oxidative coupling of phe-
nolic precursors. Thus, the first 1,9-bis-phosphorylated amino-
phenoxazinone dye (6) has been synthesised in excellent yield
by taking advantage of laccase-mediated oxidation in water.
The fluorescence quenching observed when 6 was mixed with
Mg2+ and/or Ca2+ brought about a heavy handicap in the cell
labelling application, but could lead to its development as a
sensor for variations in divalent cation concentrations in the
high range. Among the first generation of tested dyes, namely
the sulfonylated compounds 7–9, one dye, 8c, appears partic-
ularly promising as a presumably bulk endocytic tracer with
targeting by default to lysosomes. Despite having a tenfold
lower quantum yield than Lucifer Yellow, compound 8c gener-
ated intracellular signals of comparable intensity. However, it
remains to be tested whether this useful “compensation” for
Fluorescence measurements: UV/Vis spectra were recorded on a
UV/Vis–NIR Varian-CARY spectrophotometer (l given in nm). Emis-
sion fluorescence spectra were recorded on a Fluorolog-3 spectro-
fluorometer (Jobin Yvon Horiba). The excitation spectra (emission
at 500–700 nm, window of 526–534 nm used for FF) and emission
spectra (excitation at 488 nm) of aminophenoxazinone dyes were
determined in PBS (pH 7.4) over a range of concentrations around
10À5 m (optical density of 0.06). The e values were also obtained in
phosphate-buffered solutions. The fluorescence intensities in pure
water were measured from the corrected emission curve by using
the area function of the fluorometer. Relative fluorescence quan-
tum yields were obtained by using Rho-6g (FF =0.95) as the refer-
ence.
Quenching of fluorescence: Fluorescence experiments of dyes 6,
7, 8a, 8e and 8g in the presence of Mg2+ (3 mm) and/or Ca2+
(3.6 mm) either in PBS or pure water were realised in triplicate in a
96-well plate. Fluorescence was recorded by using a HP Fluoro-
count mounted with an excitation filter at 488 nm and an emission
ChemBioChem 2010, 11, 1451 – 1457
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1455

J. Marchand-Brynaert et al.
filter at 530 nm. Blanks were recorded with and without dyes in
PBS or water. A quenching phenomenon was observed only for
dye 6.
(75 MHz; CDCl3): d=16.3 (d, 4J(C-P)=6.8 Hz, PO(CH2CH3)2), 27.6
(C(CH3)3), 39.8 (C(CH3)3), 56.3 (OCH3), 63.6 (d, 3J(C-P)=5.3 Hz, PO-
(CH2CH3)2), 117.7 (d, 4J(C-P)=2.6 Hz, C4), 124.2 (d, 1J(C-P)=182.
2
3
0 Hz, C1), 124.4 (d, J(C-P)=6.8 Hz, C6), 127.2 (d, J(C-P)=16.8 Hz,
C5), 129.4 (d, 2J(C-P)=6.1 Hz, C2), 155.4 (d, 3J(C-P)=16.2 Hz, C3),
176.6 ppm (CO amide); 31P NMR (121 MHz; CD3OD): d=18.55 ppm
(s, PO(OEt)2); MS (ESI) m/z (%): 344.18 (100) [M+H]+; HRMS (ESI):
calcd for C16H26NO5NaP: 366.1446, found: 366.1434.
Cell culture and labelling: CHO cells were propagated in DMEM/
F-12 supplemented by 10% foetal calf serum and antibiotics.[28] For
experiments, cells were seeded at ~20000 cm2 on Lab-Tek cham-
bers (Nunc, Roskilde, Denmark) and grown until ~90% confluency
(2 days). For the primary screening, probes were dissolved in etha-
nol (20 mgmLÀ1), then diluted to 1 mgmLÀ1 in PBS. For the analysis
of the subcellular localisation of dye 8c, this probe was dissolved
from a stock solution in ethanol (200 mgmLÀ1) into DMEM contain-
ing defatted BSA (1 mgmLÀ1) to prevent adsorption to the culture
vessel, to a final concentration of 1 mgmLÀ1 (1.5 mm). For compa-
rative studies with LY, both compounds were added to 0.5% etha-
nol, (final concentration).
2-tert-Butylbenzoxazole-4-phosphonic acid (4): The glassware was
flame-dried under argon. A solution of BBr3 in dichloromethane
(1m, 18.63 mL, 3.0 equiv) was added dropwise to a solution of
compound 3 (2.132 g; 6.21 mmol) in anhydrous dichloromethane
(70 mL) at 08C. The mixture was stirred at 08C for 1 h, then at
room temperature for 18 h. The mixture was cooled to 08C, and
water was added dropwise to quench the residual BBr3. The solu-
tion was evaporated, and the residue was filtered on RP-18 silica
gel with methanol/water (1:1), then methanol. After evaporation,
both residues were dissolved several times in diethyl ether; each
time, the formed white solid was filtered and dried. The title com-
pound was isolated with an overall yield of 77.9% (1.234 g). M.p.
Live-cell imaging by confocal microscopy: Extensively washed
cells were examined under a Zeiss LSM510 confocal microscope
equipped with an incubation chamber set at 378C (Zeiss; Incuba-
tor XL/LSM) and a 63ꢂ immersion oil, NA 1.4 objective. Images
were recorded sequentially in the green, and then the red chan-
nels. For commercial probes, the laser was set at 2–5% (detector
gain, 700; amplifier offset, À0.5). For aminophenoxazinone probes
the laser was set at 4–8% (detector gain, 750; amplifier offset,
À0.5).
1
204–2078C; H NMR (300 MHz; CD3OD): d=1.51 (s, 9H; tBu), 7.41
(td, 3J(H-H)=7.8 Hz, 4J(H-P)=3.2 Hz, 1H; H6), 7.74 (d, 3J(H-H)=
7.7 Hz, 1H; H7), 7.79 ppm (dd, 3J(H-H)=7.50 Hz, 4J(H-P)=14.3 Hz,
1H; H5); 13C NMR (75 MHz; CD3OD): d=28.7 (tBu), 35.4 (C-(CH3)3),
115.0 (C7), 124.4 (d, 1J(C-P)=185.8 Hz, C4), 125.2 (d, 3J(C-P)=
2
2
14.6 Hz, C6), 128.8 (d, J(C-P)=7.8 Hz, C5), 143.2 (d, J(C-P)=6.7 Hz,
C9), 152.5 (d, 3J(C-P)=15.2 Hz, C8), 175.7 ppm (C2); 31P NMR
(121 MHz; CD3OD): d=11.02 ppm (s, PO(OH)2); MS (ESI): m/z (%):
256.18 (100) [M+H]+; HRMS (ESI): calcd for C11H15NO4P: 256.0739,
found 256.0736.
Organic synthesis
N-(2-Methoxy-phenyl)-2,2-dimethyl-propionamide
(2):
NaHCO3
(11.17 g, 3.0 equiv) and pivaloyl chloride (6 mL, 1.1 equiv) were
added to a solution of o-anisidine 1 (5.46 g, 5.00 mL, 44.33 mmol)
in ethyl acetate/water (130:150 mL). The solution was stirred at
room temperature for 6 h. The layers were separated, and the or-
ganic layer was washed with aqueous 1n HCl. The organic layer
was dried over MgSO4 and evaporated. The title compound was
obtained in a quantitative yield (10.86 g). 1H NMR (300 MHz, CDCl3):
2-Amino-3-hydroxy-benzenephosphoric acid (5): A solution of 4
(0.410 g; 1.61 mmol) in HCl (12n, 20 mL) was heated at 1008C for
24 h. The reaction mixture was cooled down to room temperature
and evaporated. The residue was added with water, and the re-
maining solid was dissolved in methanol/diethyl ether. After evapo-
ration under vacuum, the title compound was obtained as a white
solid (0.21 g) with a yield of 69%. M.p. 202–2068C; 1H NMR
(300 MHz; [D6]DMSO): d=6.42 (td, 3J(H-H)=7.6 Hz, 4J(H-P)=
4.08 Hz, 1H; H5), 6.74 (dd, 3J(H-H)=7.41 Hz, 4J(H-H)=1.2 Hz, 1H;
H4), 6.88 ppm (ddd, 3J(H-H)=7.6 Hz, 4J(H-H)=1.2 Hz, 3J(H-P)=
12.1 Hz, 1H; H6); 13C NMR (75 MHz; [D6]DMSO): d=115.1 (d, 1J(C-
3
d=1.31 (s, 9H; tBu), 3.87 (s, 3H; OCH3), 6.85 (dd, J(H-H)=7.9 Hz,
3
4
4J(H-H)=1.8 Hz, 1H; H6), 6.94 (td, J(H-H)=7.6 Hz, J(H-H)=1.8 Hz,
3
4
1H; H4), 7.01 (td, J(H-H)=7.9 Hz, J(H-H)=1.8 Hz, 1H; H5), 8.13 (s,
3
4
1H; NH amide), 8.41 ppm (dd, J(H-H)=7.6 Hz, J(H-H)=1.8 Hz, 1H;
H3); 13C NMR (75 MHz; [D]CDCl3): d=27.6 (C(CH3)3), 40.0 (C(CH3)3),
55.8 (OCH3), 109.8 (C6), 119.5 (C3), 121.1 (C4), 123.4 (C5), 127.9 (C2),
148.0 (C1), 176.5 ppm (CO amide).
3
P)=177 Hz, C1), 116.1 (d, J(C-P)=14.7 Hz, C5), 116.2 (C4), 122.7 (d,
2J(C-P)=6.9 Hz, C6), 138 (d, J(C-P)=8.5 Hz, C2), 144.6 ppm (d, J(C-
P)=16.5 Hz, C3); 31P NMR (121 MHz; [D6]DMSO): d=15.87 ppm (s,
PO(OH)2); MS (ESI) m/z (%): 190.14 (100) [M+H]+; HRMS (ESI): calcd
for C6H9NO4P: 190.0269, found 190.0273.
1-Diethoxyphosphoryl-2-(2,2-dimethyl-propionylamino)-3-methoxy-
benzene (3): The glassware was flame-dried under argon. TMEDA
(6.95 mL, 2.4 equiv) and a solution of nBuLi in hexane (2.5m,
18.5 mL, 2.4 equiv) were added dropwise to a solution of 2 (4 g,
19.30 mmol) in anhydrous THF (40 mL) at À108C. The addition
lasted for 5 min. The mixture was stirred at À108C for 15 min, then
at room temperature for 3 h. At À108C, diethylchlorophosphate
(6.72 mL, 2.4 equiv) was added dropwise to the solution. The reac-
tion mixture was stirred at À108C for 2 h, then warmed to room
temperature over 17 h. An aqueous solution of 1 n HCl was added
and the mixture was extracted with ethyl acetate. The organic
phase was washed with water, dried over MgSO4 and evaporated.
The red oil (7.92 g) was purified by column chromatography on
flash silica gel (eluent: cyclohexane/ethyl acetate 1:1, then ethyl
acetate); two fractions were obtained. The first one was an orange
oil corresponding to the starting compound 2 (0.72 g; recovery=
18.2%; conversion=81.8%). The second fraction was a yellow solid
corresponding to the title compound 3 (4.56 g; yield=68.9%).
1H NMR (300 MHz; CD3OD): d=1.30 (m, 6H; PO(CH2CH3)2), 1.32 (s,
9H; tBu), 3.83 (s, 3H; OCH3), 3.99–4.12 (m, 4H; PO(CH2CH3)2), 7.31–
7.36 (m, 1H; H5), 7.41–7.46 ppm (m, 2H; H4 and H6); 13C NMR
2
3
2-Amino-3-oxo-3H-phenoxazine-1,9-diphosphonic acid (6): A solution
of 5 (0.2 g, 1.06 mmol) in MeOH (25 mL) was diluted to a final
volume of 500 mL with water (220 mL) and ammonium acetate
buffer (250 mL, 0.2 mm, pH 6). The reaction was started by adding
a solution of laccase (5 mL, 104 ULÀ1), and the stirred mixture was
kept for 24 h at 258C under air. The reaction was stopped by
adding 40% (v/v) of methanol and concentrated under vacuum to
a volume of 300 mL. The crude mixture was freeze-dried to give a
red powder (0.2 g) with a yield of 90%. The isolated product was
obtained with traces of acetate, which were removed after repeat-
ed freeze-drying. HPLC: tR =7.78 min; m.p.>2808C (decomp.);
1H NMR (300 MHz; [D6]DMSO/D2O, 1:1): d=6.37 (s, 1H; H4), 7.41
(ddd, 3J(H-H)=7.0, 8.0 Hz, 4J(H-P)=3.6 Hz, 1H; H7), 7.49 (d, 3J(H-
H)=7.9 Hz, 1H; H6), 7.63 ppm (dd, 3J(H-H)=7.0 Hz, 3J(H-P)=
11.7 Hz, 1H; H8); 13C NMR (125 MHz, [D6]DMSO/D2O, 1:1) d=103.7
1
2
(C4), 103.9 (d, J(C-P)=171.2 Hz, C1), 117.2 (C6), 128.0 (d, J(C-P)=
1456
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 1451 – 1457

Live-Cell Imaging with Aminophenoxazinone Dyes
6.8 Hz, C8), 128.8 (d, 3J(C-P)=15.0 Hz, C7), 133.2 (d, 2J(C-P)=
5.38 Hz, C14), 138.1 (d, 1J(C-P)=170.2 Hz, C9), 141.5 (d, 3J(C-P)=
11.8 Hz, C13), 146.2 (d, 2J(C-P)=5.6 Hz, C2), 148.0 (d, 2J(C-P)=
4.5 Hz, C11), 149.6 (d, 3J(C-P)=12.8 Hz, C12), 181.3 ppm (d, 3J(C-
P)=15 Hz, C3); 31P NMR (121 MHz; [D6]DMSO/D2O): d=9.87,
7.18 ppm (s, PO(OH)2); IR (KBr) n˜ =960–1200, 1120–1220
(RPO(OH)2), 1530–1700 (iminoquinone), 3030–3400 (NH2) cmÀ1; UV/
Vis (phosphate buffer, pH 7.4): lmax (e)=241, 452 nm (8800); MS
(ESI) m/z (%): 370.96 (100) [MÀH]À, 392.97 (40) [MÀ2H+Na],
290.95 (55), 273.06 (30), 220.20 (60), 78.84 (30); HRMS (ESI, Àve
mode) calcd for C12H9N2O8P2: 370.9834, found: 370.9854.
[10] a) J. Jose, Y. Ueno, K. Burgess, Chem. Eur. J. 2009, 15, 418–423; b) V.
Frade, M. S. T. Goncalves, J. Moura, Tetrahedron Lett. 2006, 47, 8567–
8570.
[11] V. P. Boyarskiy, V. Belov, R. Medda, B. Hein, M. Bossi, S. W. Hell, Chem. Eur.
J. 2008, 14, 1784–1792.
[12] a) N.-H. Ho, R. Weissleder, C.-H. Tung, Tetrahedron 2006, 62, 578–585;
b) J. Jose, K. Burgess, Tetrahedron 2006, 62, 11021–11037.
[13] G. W. K. Cavill, P. S. Clezy, J. R. Tetaz, R. L. Werner, Tetrahedron 1959, 5,
275–280.
[14] a) G. L. Newton, J. R. Milligan, Radiat. Phys. Chem. 2006, 75, 473–478;
b) A. Butenandt, Justus Liebigs Ann. Chem. 1957, 602, 72–79.
[15] a) F. Bruyneel, E. Enaud, L. Billottet, S. Vanhulle, J. Marchand-Brynaert,
Eur. J. Org. Chem. 2008, 72–79; b) F. Bruyneel, O. Payen, A. Rescigno, B.
Tinant, J. Marchand-Brynaert, Chem. Eur. J. 2009, 15, 8283–8295.
[16] P. Savignac, B. Iorga, Modern Phosphonate Chemistry, CRC, Boca Raton,
2003, p. 530.
Acknowledgements
[17] I. Bonnaventure, A. B. Charette, J. Org. Chem. 2008, 73, 6330–6340.
[18] A. Macchiarulo, R. Pellicciari, J. Mol. Graph. Modell. 2007, 26, 728–739.
[19] W. W. Stewart, Cell 1978, 14, 741–759.
[20] a) J. M. Oh, S. J. Choi, S. T. Kim, J. H. Choy, Bioconjugate Chem. 2006, 17,
1411–1417; b) X. Zhang, Y. Jin, M. R. Plummer, S. Pooyan, S. Gunaseelan,
P. J. Sinko, Mol. Pharmaceutics 2009, 6, 836–848.
[21] a) I. P. Chang, K. C. Hwang, C.-S. Chiang, J. Am. Chem. Soc. 2008, 130,
15476–15481; b) A. Vertut-Doi, S. Onishi, J. Bolard, Antimicrob. Agents
Chemother. 1994, 38, 2373–2379.
[22] M. Marsh, Endocytosis, Frontiers in Molecular Biology, Oxford University
Press, 2001, p. 283.
This research was supported in part by the European Commis-
sion, Sixth Framework Program (NMP2-CT2004-505899, SOPHIED)
and in part by the UCL, Belgium Concerted Research Programme
ARC 08/13-009. J.M.-B. is a senior research associate of FRS
(FNRS, Belgium). Investigations at the CELL Unit were supported
by FNRS, Rꢀgion wallonne, Rꢀgion bruxelloise, Loterie nationale,
Concerted Research Actions and Interuniversity Attraction Poles.
[23] M. Mettlen, A. Platek, P. Van Der Smissen, S. Carpentier, M. Amyere, L.
Lanzetti, P. de Diesbach, D. Tyteca, P. J. Courtoy, Traffic 2006, 7, 589–
603.
Keywords: biocatalysis
fluorescent probes · lysosomes
· dyes/pigments · endocytosis ·
[24] P. de Diesbach, F. N’Kuli, C. Berens, E. Sonveaux, M. Monsigny, A. C.
Roche, P. J. Courtoy, Nucleic Acids Res. 2002, 30, 1512–1521.
[25] N. G. Lipsky, R. E. Pagano, Science 1985, 228, 745–747.
[26] a) J. Yin, A. J. Lin, P. D. Buckett, M. Wessling-Resnick, D. E. Golan, C. T.
Walsh, Chem. Biol. 2005, 12, 999–1006; b) K. D. Bellve, D. Leonard, C.
Standley, L. M. Lifshitz, R. A. Tuft, A. Hayakawa, S. Corvera, K. E. Fogarty,
J. Biol. Chem. 2006, 281, 16139–16146.
[27] H. E. Gottlieb, V. Kotlyar, A. Nudelman, J. Org. Chem. 1997, 62, 7512–
7515.
[28] E. Wiame, D. Tyteca, N. Pierrot, F. Collard, M. Amyere, G. Noꢃl, J. Des-
medt, M. C. Nassogne, M. Vikkula, J. N. Octave, M. F. Vincent, P. J. Cour-
toy, E. Boltshauser, E. Van Schaftingen, Biochem. J. 2010, 425, 127–136.
[1] L. D. Lavis, R. T. Raines, ACS Chem. Biol. 2008, 3, 142–155.
[2] M. S. T. Goncalves, Chem. Rev. 2009, 109, 190–212.
[3] R. Reents, M. Wagner, S. Schlummer, J. Kuhlmann, H. Waldmann, Chem-
BioChem 2005, 6, 86–94.
[4] X. Nan, A. M. Tonary, A. Stolow, X. S. Xie, J. P. Pezacki, ChemBioChem
2006, 7, 1895–1897.
[5] a) S. D. Conner, S. L. Schmid, Nature 2003, 422, 37–44; b) S. Mayor, R. E.
Pagano, Nat. Rev. Mol. Cell Biol. 2007, 8, 603–612.
[6] H.-L. Lee, E. A. Dubikovskaya, H. Hwang, A. N. Semyonov, H. Wang, L. R.
Jones, R. J. Twieg, W. E. Moerner, P. A. Wender, J. Am. Chem. Soc. 2008,
130, 9364–9370.
[7] P. Cupers, A. Veithen, A. Kiss, P. Baudhuin, P. J. Courtoy, J. Cell Biol. 1994,
127, 725–735.
[8] R. Bianchini, G. Catelani, R. Cecconi, F. D’Andrea, L. Guazzelli, J. Isaad, M.
Rolla, Eur. J. Org. Chem. 2008, 444–454.
Received: March 6, 2010
[9] S. Niu, G. Ulrich, R. Ziessel, A. Kiss, P. Renard, A. Romieu, Org. Lett. 2009,
11, 2049–2052.
Published online on June 8, 2010
ChemBioChem 2010, 11, 1451 – 1457
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
1457