Pyrene-based fluorescent nitric oxide cheletropic traps (FNOCTs) for the detection of nitric oxide in cell cultures and tissues

Article abstract of DOI:10.1002/chem.201000029

The synthesis and the structural and spectroscopic characterization of nonfluorescent, pyrene-based cyclic o-quinodimethanes are reported. These compounds react efficiently with nitric oxide (NO) in a formal cheletropic manner, by which the fluorescent aromatic pyrene system is regenerated. The NO trapping capabilities and kinetics of the fluorescent nitric oxide cheletropic traps (FNOCTs) are assessed in THF and buffered aqueous solution by ESR, UV/Vis, and fluorescence spectroscopy, by employing NO solutions and NO released from N-diazeniumdiolates (NONOates). Prototypal biological applications include the quantitation of NO production from cultured rat alveolar macrophages and the endothelium of porcine aorta, which demonstrate a sensitivity for NO detection in the nanomolar range.

Full text of DOI:10.1002/chem.201000029

FULL PAPER
DOI: 10.1002/chem.201000029
Pyrene-Based Fluorescent Nitric Oxide Cheletropic Traps (FNOCTs) for the
Detection of Nitric Oxide in Cell Cultures and Tissues
Peter M. Dꢀppe,[a] Peter M. Talbierski,[a] Frank S. Hornig,[a] Ursula Rauen,[b]
Hans-Gert Korth,[a] Timo Wille,[b] Roland Boese,[c] Thorsten Omlor,[c]
Herbert de Groot,[b] and Reiner Sustmann*[a]
Abstract: The synthesis and the struc-
tural and spectroscopic characteriza-
tion of nonfluorescent, pyrene-based
cyclic o-quinodimethanes are reported.
These compounds react efficiently with
nitric oxide (NO) in a formal chele-
tropic manner, by which the fluores-
cent aromatic pyrene system is regener-
ated. The NO trapping capabilities and
kinetics of the fluorescent nitric oxide
cheletropic traps (FNOCTs) are as-
sessed in THF and buffered aqueous
solution by ESR, UV/Vis, and fluores-
cence spectroscopy, by employing NO
solutions and NO released from N-dia-
zeniumdiolates (NONOates). Prototy-
pal biological applications include the
quantitation of NO production from
cultured rat alveolar macrophages and
the endothelium of porcine aorta,
which demonstrate a sensitivity for NO
detection in the nanomolar range.
Keywords: bioinorganic chemistry ·
cheletropic traps
spectroscopy nitrogen oxides
quinodimethanes
·
fluorescence
·
·
Introduction
An important aspect of NO biochemistry is its detection
under physiological conditions. To this end, a variety of
methods have been developed in the past 20 years.[4] An
ideal procedure would be the direct detection of NO as a
radical with the capability of temporal and spatial resolution
without significantly interfering with the biological function
of the living system. However, none of the existing methods
seems to be ideally suited under the diversity of chemical
and physiological conditions. Several methods are disruptive,
that is, they destroy the cellular system while quantifying
NO; other procedures require conversion of NO to other,
more reactive nitrogen species prior to detection, and thus
are lacking true specificity for NO. Methods that specifically
trap the NO radical and convert it to fluorescent entities,
such as the fluorescent nitric oxide cheletropic trap
(FNOCT) method described herein, allow quantitation of
NO production with temporal and spatial resolution, but
might affect the physiological function of NO. On the other
hand, electrochemical sensor technologies are well suited
for long-term monitoring of NO with negligible disturbance
of biological function, but lack spatial resolution, and quan-
titation as well as intracellular NO detection are difficult to
perform. Thus, the various methods for NO detection
should be regarded, and used, as complementary techniques.
o-Quinodimethanes are useful reactive intermediates in
the synthesis of carbon backbones.[5] Due to their reactivity,
o-quinodimethanes generally escape isolation in pure form.
The importance of nitric oxide in biology and medicine[1]
was recognized by the award of the 1998 Nobel prize in
physiology and medicine to Furchgott, Ignarro, and
Murad.[2] Many physiological functions of NO, for example,
blood pressure regulation or defense against pathogens by
the immune system, have been uncovered and thoroughly
investigated. Meanwhile, seven volumes of Methods in En-
zymology have been specifically devoted to NO.[3]
[a] Dipl.-Chem. P. M. Dꢀppe, Dr. P. M. Talbierski, Dr. F. S. Hornig,
Dr. H.-G. Korth, Prof. Dr. R. Sustmann
Institut fꢀr Organische Chemie
Universitꢁt Duisburg-Essen
Campus Essen, Universitꢁtsstr. 5, 45141 Essen (Germany)
Fax : (+49)201-183-4259
E-mail: reiner.sustmann@uni-duisburg-essen.de
[b] Prof. Dr. U. Rauen, Dr. T. Wille, Prof. Dr. H. de Groot
Institut fꢀr Physiologische Chemie
Universitꢁtsklinikum Essen
Hufelandstr. 55, 45122 Essen (Germany)
[c] Prof. Dr. R. Boese, Dr. T. Omlor
Institut fꢀr Anorganische Chemie
Universitꢁt Duisburg-Essen
Campus Essen, Universitꢁtsstr. 5, 45141 Essen (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201000029.
Chem. Eur. J. 2010, 16, 11121 – 11132
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11121

More than a decade ago, we discovered that tetramethyl-o-
quinodimethane (1), although being short lived and only
present in solution at very low concentration (ca. 10ꢀ4 m),
readily reacts with NO to form a stable tetramethyl-substi-
tuted isoindolyl nitroxide (TMIO) radical 2 (Scheme 1).[6]
Scheme 2. Phenanthrene-based fluorescent nitric oxide cheletropic traps
(FNOCTs).
Scheme 1. Prototypal nitric oxide cheletropic traps (NOCTs).
quenchers.[13] This conversion may be achieved either by un-
specific thermal degradation of the nitroxide (provided the
aromatic fluorophore is preserved in such a process) or by
reduction to the corresponding hydroxylamines 7 by addi-
tion of suitable reductants. Biological reductants such as as-
corbate, glutathione, a-tocopherol, or various carbohydrates
are ubiquitously present in living systems. Accordingly, we
found that in living cells the reduction of the nitroxides 6a–
c to the hydroxylamines 7a–c readily takes place without
the necessity to add external reductants.
The maximum excitation and emission wavelengths of the
parent system 5a are too short for biological applications,[12]
therefore, a dimethylamino group was introduced at the 2-
position of the phenanthrene ring (5b), which shifted the ex-
citation and emission maxima in the reduced NO adduct 7b
to longer wavelengths (lex =380, lem =460 nm).[14] Several
applications of FNOCT 5b have been reported.[15] However,
the disadvantage of a pH dependence of the fluorescence of
5b prompted us to prepare the corresponding methoxy-sub-
stituted FNOCT (5c), which indeed proved to be superi-
or.[16]
A further improvement of the FNOCTs with respect to
fluorescence wavelength and intensity was expected for
pyrene as a fluorophoric group. Herein, we report on the
preparation, structural and spectroscopic properties, and
NO trapping characteristics of the pyrene-based FNOCTs
8a–c (Scheme 3). The substituents R were varied to check
whether the NO trapping capabilities are affected by sub-
stituents at the positions where NO addition takes place.[17]
In addition, the X-ray structure of o-quinodimethane 8b is
Compound 1 was generated photochemically from the cor-
responding substituted indanone 3. The reaction was con-
ceived for the detection of NO in biological systems by ESR
spectroscopy. With regard to the formal NO trapping reac-
tion, o-quinodimethanes such as 1 were christened “nitric
oxide cheletropic traps” (NOCTs).
Because compound 1 is rather unstable and insoluble in
water, which is a requirement for applications in biological
systems, research was continued with the goal to obtain
more stable o-quinodimethanes with improved water solu-
bility and stability.[7] Initially, the success was rather limited,
since in most cases the photolysis of derivatives of indanone
3 provided mainly the corresponding benzocyclobutenes,[8]
which turned out not to be ideally suited as depot com-
pounds for the o-quinodimethanes due to a slow rate of
electrocyclic ring opening.[9]
Jones and Pomfret were the first to show that polycyclic
o-quinodimethanes, in which the terminal double bonds are
part of a ring system, inherit a certain stability that allows
isolation, although the compounds were still very reactive.[10]
A modification, compound 4, was sufficiently water soluble
and proved to be highly reactive towards NO, but the pro-
duced nitroxide radical was thermally too unstable to relia-
bly allow NO quantification by means of ESR spectrosco-
py.[11]
As a consequence of this disadvantage, we changed the
methodology of NO detection in biological systems from
ESR to fluorescence spectroscopy. We took advantage of
the fact that on reaction of the o-quinodimethane with NO
the aromatic system is regenerated (compare reaction 1!2
in Scheme 1). The parent structure 4, therefore, was modi-
fied in such a way that a fluorescent aromatic system, in this
case a phenanthrene unit, is formed on reaction with NO.[12]
The related NO scavengers 5 (Scheme 2) were named “fluo-
rescent nitric oxide cheletropic traps” (FNOCTs). However,
before fluorescence can be detected, the initially formed
paramagnetic nitroxides 6 have to be converted to a diamag-
netic product because nitroxide radicals are fluorescence
Scheme 3. Pyrene-based FNOCTs of this study.
11122
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11121 – 11132

Fluorescent Nitric Oxide Cheletropic Traps
FULL PAPER
reported. Compound 8a turned out to be the most sensitive
FNOCT produced so far. NO trapping studies on 8a in cell
culture systems as well as in a tissue model (pig aorta) were
carried out to establish the detection limits for NO and to
work out a protocol for the application of 8a in biological
systems.
the reaction of 9 with 3-pentanone required two additional
steps to afford the Diels–Alder dimer 12 of 10c. Because
the dimer and monomer exist in equilibrium at elevated
temperature, Diels–Alder trapping of the monomeric cyclo-
pentadienone was possible at 1208C by employing norbor-
nene-(E)-2,3-diethyldicarboxylate as
a solvent, thereby
giving 13c in 43% yield.
1
The H NMR spectra of the raw reaction products were
dominated (80–90%) by the signals of a single stereoisomer
showing two characteristic multiplet resonances at d=0.30
to 0.39 and ꢀ0.32 to ꢀ0.60 ppm (see the Experimental Sec-
tion). The high-field shift of these resonances was attributed
to a shielding of the hydrogen atoms of the norbornane
methylene bridge by the aromatic pyrene system. This im-
plied an anti arrangement of the CO and CH2 bridges, as in-
dicated in Scheme 4.[21] Thus, out of the four possible diaste-
reomers of ketones 13a–c from addition of dienes 10 in the
endo or exo orientation to the endo or exo side, respectively,
of the norbornene system (see the Supporting Information,
Scheme S1),[22] only the products from addition to the exo
side of the norbornene system are formed. The preferred
addition of the cyclopentadienones 10 in an endo orientation
has previously been proven by single-crystal X-ray crystal-
lography for 13c[20] and is also found here to be true for 13b
(see below).
Results and Discussion
Synthesis: FNOCTs 8 were synthesized as outlined in
Scheme 4. Pyrene-4,5-dione (9) was obtained in 45% yield
by oxidation of pyrene with sodium periodate in the pres-
Photolysis of 13a–c in THF in a Rayonet reactor with
300 nm light gave solutions of the o-quinodimethanes 8a–c,
which were characterized by high-resolution mass spectrom-
1
etry and H and 13C NMR spectroscopy. No second isomer
of compounds 8a–c could be detected in the NMR spectra.
This fact is in line with the formation of only the two possi-
ble exo-norbornene Diels–Alder adducts 13, since both give
rise to the same exo-fused o-quinodimethane 8 (Scheme 3)
on photodecarbonylation. The UV/Vis spectra show the typ-
ical broad band of o-quinoid compounds with a lmax (loge)
between 400 and 450 nm [8a: lmax (loge): 443 (3.56), 8b: 408
(3.86), 8c: 412 (3.12) nm; Figure 1]. Compound 8b was iso-
lated in pure crystalline form, and single crystals suitable for
X-ray structure analysis could be obtained. Attempts to
grow single crystals of 8a and 8c were unsuccessful.
Scheme 4. Preparation of pyrene-based FNOCTs 8a–c. [a] 12: Diels–
Alder dimer of 10c;[20] see the Supporting Information, Figure S1, for
structure.
ence of rutheniumACHTUNGTRENNUNG
(III) chloride monohydrate as catalyst.[18]
Reaction of 9 with 1,3-diphenylacetone gave the diphenylcy-
clopentadienone 10a[19] in 45% yield, which on Diels–Alder
reaction with norbornene-(E)-2,3-dicarboxylic acid afforded
the bridged ketone 13a (ꢁ33%). Reaction of 9 with 3-oxo-
dimethylglutarate did not produce the desired cyclone 10b,
but gave only the monocondensation product, hydroxycyclo-
pentenone 11. Thus, a second elimination of water to form
10b did not take place under these conditions. However,
water could be eliminated from 11 by boiling in acidic etha-
nol solution to give in situ 10b, which in the presence of a
tenfold excess of norbornene-(E)-2,3-diethyldicarboxylate
provided the Diels–Alder adduct 13b (35%). Details of the
preparation and characterization of the methyl-substituted
derivative 13c have been reported.[20] Briefly, in this case
Figure 1. UV/Vis spectra of 8a (c) and 8b (a) (50 mm each) in THF.
Chem. Eur. J. 2010, 16, 11121 – 11132
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11123

R. Sustmann et al.
X-ray structures of 8b and 13b: Single crystals of the major
isomers of 8b and 13b were obtained by diffusion crystalli-
zation in dichloromethane/diethyl ether. Figure 2 displays
The experimental structures of 13b and 13c are very well
reproduced by DFT calculations on the B3LYP/6-31G(d)
level (here, the diethyl ester groups at the norbornane ring
were modeled by dimethyl ester groups), thus confirming
the weak, bowl-shaped deformation of the pyrene moiety
(Supporting Information, Scheme S2). It is worth noticing
that the computations predict that the anti-(exo–endo) struc-
tures of 13b and 13c (compare Figure 2) are higher in
energy by DE0 =2.9 and 4.0 kcalmolꢀ1, respectively, than the
lowest-energy syn-(exo–endo) isomers (Supporting Informa-
tion, Schemes S1 and S2), in accordance with a kinetic con-
trol of the Diels–Alder addition.
The X-ray structure of 8b shows two molecules of 8b in
the asymmetric unit of the monoclinic crystals, together with
one molecule of diethyl ether. The two molecules of 8b
assume slightly different conformations of the terminal ester
groups at the norbornane scaffold. The central o-quinoid
ring is slightly twisted against the pyrene ring system (dihe-
dral angle C55-C66-C71-C64, ca. 198), that is, the molecule
shows a screwed deformation along the horizontal axis bi-
ꢀ
ꢀ
secting the C59 C60 and C65 C72 bonds, so that the two
inner ester groups are located somewhat above and below
the pyrene plane, respectively. Nevertheless, the p system of
pyrene and the two o-quinoid C=C double bonds of 8b can
be considered to be still fully conjugated.[23] In both con-
formers of 8b, the C=C double-bond lengths C14=C21 and
C5=C16 differ slightly (1.354 and 1.348 ꢃ, respectively). The
bond between the two exo double bonds (C16=C21) is
1.474 ꢃ, that is, about 0.1 ꢃ longer than in the bridged ke-
tones 13b and 13c (C17=C18, 1.372 ꢃ; the related bond
length in pyrene is 1.338 ꢃ[24]). The bond lengths C1=C17
and C6=C18 in 13b are 1.515 ꢃ (13c: 1.522 ꢃ) as expected
for C=C single bonds. Details of the X-ray analysis of 8b
and 13b are given in the Supporting Information and can be
obtained from The Cambridge Crystallographic Data
Centre.[25]
The B3LYP/6-31G(d)-computed gas-phase structure of 8b
was found to be in excellent agreement with the experimen-
tal X-ray structure. In particular, the screwed deformation
of the pyrene–o-quinoid system was nicely reproduced, with
a computed torsion angle of the C55-C66-C71-C64 fragment
of 228, compared with the experimental torsion angle of 198.
A virtually identical structure was found for the methyl-sub-
stituted derivative 8c (dihedral angle ca. 278; Supporting In-
formation, Scheme S3).
Figure 2. ORTEP plots of the X-ray structures of 8b (bottom; only one
of the two conformers in the unit cell is shown) and 13b (top; anti-(exo–
endo) diastereomer).
ORTEP plots of these structures determined at ꢀ708C. Sim-
ilar to what has previously been found for 13c, the key fea-
ture of ketone 13b is that the CO and CH2 bridges are ar-
ranged in an anti position, with the methylene bridge adopt-
ing an endo arrangement. This confirms the preferred endo
reaction mode of the cyclopentadienone to the exo side of
the norbornene system. The X-ray structure further proves
that the strong high-field shift of the H NMR signals of the
hydrogen atoms of the methylene bridge (C26) (see above)
is due to the shielding by the aromatic pyrene system. The
Reaction with nitric oxide—ESR spectroscopy: To ascertain
that FNOCTs 8a–c are capable of scavenging the NO radi-
cal, their reaction with NO (Scheme 5) was studied by ESR
spectroscopy. After mixing of solutions of 8a–c in THF with
an equimolar amount of a freshly prepared, deoxygenated
solution of NO in THF, ESR signals of nitroxides 14a–c
were observed (Figure 3). The spectra show the expected
hyperfine splitting (hfs) pattern as previously observed for
nitroxides 6 (Scheme 2), characterized by a nitrogen-14 trip-
1
distance of the endo hydrogen at C26 to the C17 C18
pyrene bond is only about 2.06 ꢃ.
let hfs of a
(14N)=2.3–2.6 mT and g factors close to 2.0060,
ACHTUNGTRENNUNG
ꢀ
both typical for bridgehead nitroxides[26] (Scheme 5). The
11124
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11121 – 11132

Fluorescent Nitric Oxide Cheletropic Traps
FULL PAPER
The two resolved hydrogen hfs patterns have previously
been attributed to the two bridgehead g-hydrogen atoms
and one of the four e-hydrogen atoms that is in a zigzag ar-
rangement with the radical center.[12,14,16] Thus, the hfs sug-
gested that nitroxides 14 assume an anti-endo structure, that
is, NO has reacted at the more accessible side opposite to
the CH2 bridge of the norbornane moiety. Nitroxide radicals
14 thus should have the same anti-(exo–endo) configuration
as their parent ketones 13 (compare Figure 2). This interpre-
tation is fully confirmed by the hfs data obtained from
B3LYP/6-31G(d) DFT calculations of the anti-(exo–endo)
isomers of 14b and 14c (Figure 4; computed structures and
Scheme 5. Reaction of FNOCTs 8 with NO and reduction of nitroxides
14 to fluorescent hydroxylamines 15.
Figure 4. B3LYP/6-31G(d)-optimized structure of nitroxide anti-(exo–
endo)-14c and computed 14N and 1H ESR hyperfine splittings [mT] (hy-
drogen atoms denoted by open circles). The hyperfine splittings of the
pyrene hydrogen atoms are smaller than 0.001 mT; the a values of the
bridgehead methyl groups are the averages over the three hydrogen
atoms.
Figure 3. ESR spectra of nitroxides 14a–c from reaction of 8a–c with NO
in THF at room temperature. a) 14a, 25 min, b) 14b, 28 min, c) 14c, 23 h
after mixing of the reactants.
spectra were analyzed by least-squares fitting of simulated
spectra to the experimental spectra. The simulations confirm
that the additional splitting of the three nitrogen lines that
are visible in the spectra of 14a and 14b are due to two
equivalent hydrogen atoms with a(1H)=0.07–0.08 mT and
one hydrogen with a(1H)=0.046–0.052 mT (Table 1). For
14c, these splittings are covered by a small septet splitting
of about 0.030 mT from the two bridgehead methyl groups.
The small splittings listed in the two right-hand columns of
Table 1 are estimates from best-fitted simulations of expand-
ed lines recorded at low modulation using the quantum
chemical computed data (see below) as starting values.
hfs of the other stereoisomers are provided as Supporting
Information, Scheme S4). Notably, it is the exo-oriented hy-
drogen (He1) at the distant positions of the norbornane
system and not the one at the methylene bridge that gives
1
rise to the second largest H hfs.
Interestingly, the computed hfs data indicate that the
oxygen of the pyramidal NO group in the anti-(exo–endo)
isomers of nitroxides 14 is preferably oriented toward the
norbornane moiety. The inverted orientation toward the
pyrene unit was calculated to be higher in energy by 3–
4 kcalmolꢀ1 (Supporting Information, Scheme S4); thus, the
rapid inversion equilibrium[27] lies largely on the side as dis-
played in Figure 4.
We observed that the rate of reaction with NO varied
strongly among o-quinodimethanes 8a–c. The time evolu-
tion of the ESR signal of 8a was fast (see rate constant,
below), attaining the maximum signal intensity within 4 min
after mixing, that is, within the shortest period in which the
first spectrum could be fully recorded, whereas for 8b and
8c it took 28 min and 23 h, respectively, to produce a similar
signal intensity. This observation is important with respect
to the feasibility of FNOCTs 8 to monitor NO production in
biological systems (see below).
Table 1. ESR hyperfine splittings [mT] and g-factors of nitroxide radicals
14.[a]
Radical
g
a
(14N)
a
R
a
(1He1)
a
G
(1Hother
aACHTUNGERNTUNG )
factor
(1H)
(1H)
14a
14b
14c
2.00618 2.286 0.077
2.00596 2.532 0.071
2.00605 2.297 0.084
0.050
0.046
0.052
0.021
0.022
0.034
0.011 (3H)
0.010 (3H)
0.030 (6H)
0.011 (3H)
[a] In THF, T=228C.
Chem. Eur. J. 2010, 16, 11121 – 11132
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11125

R. Sustmann et al.
Reaction with nitric oxide—fluorescence spectroscopy: As
noted above, on reaction of FNOCTs 8 with NO the aro-
matic pyrene system will be regenerated, which should give
rise to similar fluorescence emissions to those observed for
the parent ketones 13.[12,14,16] To avoid the self-quenching of
the fluorescence of nitroxide radicals 14, the NO trapping
experiments were carried out in the presence of ascorbic
acid or Trolox (a water-soluble analogue of a-tocopherol).
These agents rapidly reduce radicals 14 to the corresponding
hydroxylamines 15 (Scheme 5), by this means achieving a
full fluorescence response. Due to their low solubility in
water, the fluorescence properties of 15a–c were first estab-
lished in THF. The excitation and emission spectra of 15a
are shown in Figure 5, and the spectra of 13a are shown for
conditions, that is, in buffered aqueous solution at pH 7.3–
7.4, were further pursued for FNOCT 8a only.
Reaction of 8a with NO in aqueous solution: The problem
of the low water solubility of FNOCT 8a was circumvented
by preparing fresh 5 mm solutions in deoxygenated DMSO
and taking aliquots of this solution to attain the required
concentrations in buffered solution. In cell culture experi-
ments, the final concentration of DMSO did not cause any
toxicity (data not shown).
As commonly known for polycyclic aromatic compounds,
FNOCTs 8 tend to aggregate in aqueous solution, which
leads to partial quenching of their fluorescence and to the
appearance of broad excimer fluorescence bands in the 450–
550 nm range.[17] We observed that aggregation can be sup-
pressed by addition of small amounts of ionic or nonionic
surfactants. In fact, by using ketone 10c, this property is the
basis of a recently developed, versatile method for the deter-
mination of the critical micelle concentration (cmc) of sur-
factants.[20] In nonbiological applications, maximal and re-
producible fluorescence response was achieved by addition
of 0.4% sodium dodecyl sulfate (SDS). Because such con-
centrations of SDS, however, lead to the destruction of cell
membranes, the nonionic surfactant Pluronic F127 was
tested for use in cell culture experiments. We found that
0.02% Pluronic F127 proved to be the best for both the pre-
vention of aggregation of 8a and low background fluores-
cence, although inferior to that of 0.4% SDS. Thus, 0.02%
Pluronic F127 was added to all biological experiments re-
ported below. The NO trapping capabilities of 8a in cell-
free systems were established with both 0.4% SDS and
0.02% Pluronic F127.
Figure 5. Fluorescence excitation (a) and emission spectra (c)
(a.u.=arbitrary units) of 1) 13a (10 mm) and 2) 15a in THF at 238C.
comparison. Spectra of 13b, 13c, 15b, and 15c are given in
the Supporting Information (Figure S2). We note that
FNOCTs 8 are only very weakly fluorescent; their fluores-
cence emission is generally covered by some background
fluorescence from small fractions of unreacted ketones 13
and/or decomposition products (see below).
Another important property for biological applications of
FNOCTs is their long-term stability under physiological
conditions. The good long-term stability of FNOCT 8a is
evident from the lower traces in Figure 6, which show an in-
crease of the background fluorescence by only about 3% of
By using fluorescein as a standard, the fluorescence quan-
tum yield of 13a was determined to be QF =0.47. From com-
parison of the fluorescence intensities (see Figure 5), the
quantum yield of 15a was estimated to be QF =0.54. These
are the highest quantum yields for all precursor ketones and
NO trapping products of our FNOCTs produced so far.
The time evolution of the emission bands of hydroxyla-
mines 15 was similar to what had been observed for the
growth of the ESR signals of nitroxides 14, that is, a con-
stant maximum emission of 15a was achieved within 2 min
after addition of NO, but for 15b and 15c it took several
hours or even a day, respectively. Similar time dependencies
were also observed in aqueous solution. Thus, the replace-
ment of the two bridgehead phenyl groups by ester or
methyl groups led to a significant decrease of the rate of
NO trapping. This sluggishness unfortunately rendered 8b
and 8c largely unsuitable for studies in biological systems, in
which the sequence of events is often the question. For that
reason, the NO trapping capabilities under physiological
Figure 6. Time dependence of the fluorescence emission of 15a at lem
=
380 nm (lexc =350 nm) from reaction of FNOCT 8a (5 mm) with a saturat-
ed solution of NO (50 mm) (*) and PAPA/NO (12.5 mm) (*) in deoxy-
genated phosphate buffer of pH 7.3 (0.2% SDS; 50 mm ascorbate) at
238C. Data are corrected for the initial background fluorescence of 8a
(5.2 a.u.). The time evolution of the background fluorescence of 8a in the
absence (5) and presence (ꢄ) of 100 mm DTPA under otherwise identical
conditions is also shown.
11126
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11121 – 11132

Fluorescent Nitric Oxide Cheletropic Traps
FULL PAPER
the maximum value achieved with excess NO within 2.5 h at
238C. No difference was found in air-saturated solution. We
further found that the addition of 100 mm of the heavy-metal
chelator diethylene triamine pentaacetic acid (DTPA) has a
beneficial effect. In the presence of DTPA, the background
fluorescence increased by only about 1.5% under the same
conditions (even less, ꢁ1%, when Trolox was used instead
of ascorbate at 378C, see below). This observation suggests
that part of the slow increase of the background fluores-
cence of 8a in phosphate buffer might derive from rearoma-
tization processes induced by heavy-metal ions, preferably
iron. Therefore, 100 mm DTPA was added in all further ex-
periments.
The NO trapping capabilities of 8a in aqueous buffer of
pH 7.3 were established in oxygen-free solution by reaction
with a saturated solution of NO as well as by reaction with
(Z)-1-[N-(3-ammoniopropyl)-N-(n-propyl)amino]diazen-1-
ium-1,2-diolate (PAPA/NO) (Figure 6). In deoxygenated so-
lution, a constant fluorescence level of 15a was reached
within 4 min after addition of NO solution, similar to the ex-
periments in THF.
Figure 7. Time evolution of the fluorescence intensity of 15a at lem
=
400 nm (lexc =340 nm) from reaction of 8a (10 mm) with varying concen-
trations (0–40 mm) of the NO donor MAHMA/NO in bicarbonate-free,
air-saturated Krebs–Henseleit buffer of pH 7.4 (0.02% Pluronic F-127,
1.1% DMSO, 100 mm Trolox, 100 mm DTPA) at 378C. Numbers denote
the concentration of MAHMA/NO after addition; the arrow marks the
point of addition of MAHMA/NO to the buffer. Solid lines are least-
squares fits to a first-order rate law. Similar experiments in phosphate
buffer in the presence of 0.4% SDS are displayed in Figure S3 in the
Supporting Information.
When PAPA/NO was used as NO source, the fluorescence
intensity of 15a increased exponentially with a rate constant
of k=(1.53ꢂ0.03)ꢄ10ꢀ4 sꢀ1 (half-life=(76ꢂ1) min), in ex-
cellent agreement with the reported half-life of PAPA/NO
of 77 min at 228C and pH 7.4,[28,29] to finally approaching the
same maximum value, thus indicating complete conversion
of 8a.
The second-order reaction of 8a with NO in oxygen-free
solution was measured by following the decay of the o-qui-
noid absorption band at lmax =443 nm (compare Figure 1)
under first-order conditions at a 10- to 100-fold excess of
NO. In phosphate buffer of pH 7.3 a trapping rate of
in the Supporting Information (Figure S3, 0–3 mm MAHMA/
NO in the presence of 0.4% SDS). At all concentrations,
the growth of the fluorescence followed a first-order rate
law. In phosphate buffer and in the presence of 0.4% SDS,
a rate constant of NO release of k=(11.4ꢂ1.6)ꢄ10ꢀ3 sꢀ1
(half-life=(61ꢂ8) s) was determined, in excellent agree-
ment with the literature value for decomposition of
MAHMA/NO in neat buffer. In the presence of 0.02% Plur-
onic F127 as a solubilizing agent, the rate of NO release was
about half as fast: k=(6.3ꢂ0.8)ꢄ10ꢀ3 sꢀ1; half-life=(110ꢂ
14) s.[32] The fluorescence intensity of 15a approached a con-
stant value for [MAHMA/NO]ꢃ20 mm, which indicates a
complete reaction of 8a at a fourfold excess of NO
(Figure 8).
kACHTUNGTRENNUNG
value is close to that of 5b under similar conditions.[14]
In the presence of oxygen, the NO trapping reaction com-
petes with the autoxidation of NO to give finally nitrite. In
accordance with this finding, in previous studies a reduced,
up to two-thirds, fluorescence level of FNOCT–NO adducts
compared with oxygen-free conditions has generally been
observed for equimolar concentrations of the reactants
when the experiments were carried out in the open air.[30]
The detection limit for NO under aerobic conditions was
probed with (Z)-1-{N-methyl-N-[6-(N-methylammonio)hex-
yl]amino}diazen-1-ium-1,2-diolate (MAHMA/NO) as
a
source for NO in the presence of both 0.4% SDS and
0.02% Pluronic F127 in phosphate and Krebs–Henseleit
buffer, respectively. In these experiments, Trolox was used
as reducing agent because it gave rise to a slower increase in
background fluorescence than ascorbic acid, which turned
out to be beneficial for the detection of low amounts of NO.
MAHMA/NO is characterized by the release of two equiva-
lents of NO with a half-life of about 1 min at 378C in buf-
fered solution of pH 7.4.[29,31] The time profiles of the fluo-
rescence of 15a from 10 mm 8a as a function of the
MAHMA/NO concentration are shown in Figure 7 (0.0–
40 mm MAHMA/NO in the presence of Pluronic F127) and
Figure 8. Dependence of the maximum fluorescence intensity of 15a on
the concentration of MAHMA/NO from the data of Figure 7. Data are
meansꢂstandard deviation of three independent measurements. The
solid line is an empirical least-squares fit to a single exponential. Inset:
Expanded region at low MAHMA/NO concentration. The dashed line is
a linear fit (r2 =0.9948) through the data points in the 0–1.0 mm range.
Chem. Eur. J. 2010, 16, 11121 – 11132
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11127

R. Sustmann et al.
For MAHMA/NO concentrations ꢄ1 mm (ꢄ2 mm NO),
the fluorescence intensity was strictly linearly dependent on
the NO concentration (Figure 8, inset), thus indicating com-
plete trapping of NO at a fivefold excess of the FNOCT in
this concentration range. Note, however, that the excess of
FNOCT required for quantitative trapping of NO, and vice
versa, in the presence of a given concentration of oxygen de-
pends on the absolute concentration of NO, since the com-
peting autoxidation process is of second order in NO con-
centration. For instance, at 4 mm NO an approximate eight-
fold excess of 8a is required for quantitative trapping (Sup-
porting Information, Figure S4). From Figure 8 it can be in-
ferred that at 10 mm equimolar concentrations of FNOCT 8a
and NO the final fluorescence intensity is about 65% of its
theoretical maximum, whereas at 4 mm equimolar concentra-
tions under the conditions of Figure S4 (Supporting Infor-
mation) only about 33% of the released NO is being
trapped.
At higher concentrations of 8a, the fluorescence intensity
tends to decrease (see the Supporting Information, Fig-
ure S4). This fact most likely arises from inhibition of the
NO trapping reaction due to formation of aggregates of 8a,
because at higher SDS concentrations this effect can be fully
reversed (data not shown). Thus, the concentration at which
FNOCT 8a should preferably be applied in cell-free studies
with 0.4% SDS as solubilizing agent should exceed the ex-
pected NO level by a factor of five, but should not exceed
50 mm.
phages with lipopolysaccharide (LPS) in cell culture
medium for 18 h at 378C. For control, a fraction of the cells
from the same batch were left unstimulated, but treated oth-
erwise identically. Activated cells were then incubated in
Krebs–Henseleit buffer in the presence of d-glucose
(10 mm), DTPA (100 mm), Trolox (100 mm), Pluronic F127
(0.02%), 8a (10 mm), and l-arginine (0.5 mm). Because basal
NO production from macrophages may also be stimulated
by the stress exerted by the isolation of the cells, nonactivat-
ed cells were incubated in the presence of the NO synthase
(NOS) inhibitor NMA (NG-monomethyl-l-arginine=N5-
[imino(methylamino)methyl]-l-ornithine) (0.1 mm) instead
of l-arginine to suppress any NO production from intracell-
ularly available l-arginine. After 1 h, samples of the super-
natant were taken, which, compared with the same solution
incubated without cells, showed a fluorescence increase at
lexc =340 nm, lem =400 nm of (33.8ꢂ8.2) a.u. for activated
macrophages and of (1.4ꢂ0.3) a.u. (=4% of the intensity
from activated cells) for nonactivated macrophages, thus in-
dicating that NO formed by the cells was trapped by 8a.
Previously, on application of FNOCT 5c we observed that
part of 5c (and thus also part of the related NO adduct 7c)
was cell-bound at the end of the experiment,[16] that is, the
sensor had been taken up by the cells and/or incorporated
into cellular membranes. Therefore, we also assessed the
cell-bound fraction of NO adduct 15a by lysis of the cells
with SDS at the end of the incubation period. Fluorescence
of the cell lysates mirrored the fluorescence in the superna-
tant, with (14.2ꢂ3.4) a.u. for activated cells and (1.4ꢂ
0.3) a.u. (=10% of the fluorescence from activated cells)
for nonactivated, NMA-treated cells. To quantify the frac-
tion of cell-associated FNOCT 8a, 100 mm MAHMA/NO
was then added to the supernatant and lysate of the cell cul-
tures to assure complete conversion of residual 8a to hy-
droxylamine 15a. These controls showed that about 70% of
the applied FNOCT 8a had remained dissolved in the super-
natant whereas about 30% had been cell-associated at the
end of the cellular experiment (Table 2). For comparison, re-
lated data for the phenanthrene-based FNOCTs 5b and 5c
are also listed in Table 2. Notably, the dimethylamino-substi-
tuted derivative 5b appeared to be more strongly bound to
the cells than the other two FNOCTs.
The previously prepared FNOCTs all exhibited a high
specificity for the NO radical. Other reactive oxygen and ni-
Cꢀ
trogen species, such as H2O2, superoxide (O2 ), hydroperox-
ꢀ
C
ides, HO , or peroxynitrite (ONOO ), did not produce fluo-
rescence even at nonphysiological high levels. This was also
found to be true for FNOCT 8a (see the Supporting Infor-
mation, Figure S5). The above-mentioned observation that
in the presence of molecular oxygen the NO-mediated fluo-
rescence intensity is always reduced compared with oxygen-
free conditions also proves that the autoxidation products of
NO, namely, NO2, N2O3, and N2O4, do not produce fluores-
cence significantly exceeding the background level on reac-
tion with 8a.
Detection and quantitation of biologically produced nitric
oxide: The aim of our efforts is the development of a sensi-
tive probe for NO that can be applied in biological systems.
Therefore, FNOCT 8a was first tested for potential cell tox-
icity. Satisfyingly, no toxicity against isolated hepatocytes,
liver endothelial cells, and alveolar macrophages could be
detected when 8a was added at a final concentration of
10 mm to Krebs–Henseleit buffer in which the cells were in-
cubated (data not shown). At a final concentration of 50 mm
of 8a, an increasing toxicity was observed when the expo-
sure periods exceeded 2 h, most likely due to the marked
cellular accumulation of the compound (see below).
Table 2. Fraction of cell association of FNOCTs.[a]
FNOCT
Cell-associated FNOCT [%]
8a
5b
5c
29.9ꢂ4.9
47.9ꢂ3.3
24.9ꢂ5.1
[a] Rat alveolar macrophages were incubated for 1 h with the FNOCTs
(10 mm each) in Krebs–Henseleit buffer (pH 7.4). Thereafter, the super-
natant was removed, cells were lysed with 0.4% SDS, and the FNOCT in
the supernatant as well as in the lysate was assessed fluorimetrically after
reaction with excess NO. Cell-bound FNOCT is given as fluorescence of
hydroxylamines 15a, 7b, and 7c, respectively, as the percentage of the
total hydroxylamine fluorescence (lysate+supernatant), corrected for
background fluorescence and volume after complete conversion of the
FNOCTs.
As a standard test system for the detection of biologically
produced NO we used isolated rat alveolar macrophages.
NO production was stimulated by incubation of the macro-
11128
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11121 – 11132

Fluorescent Nitric Oxide Cheletropic Traps
FULL PAPER
According to the data of Table 2, a sizable fraction of the
FNOCTs remains associated with the cells, and thus in sub-
sequent experiments devoted to establishing the sensitivity
of 8a in biological systems, the fluorescence from both the
supernatant and the cell lysate was assessed. It should be
noted, however, that the amount of NO trapped in the su-
pernatants always correctly reflected the differences of total
NO production across the experimental sets; thus, determi-
nation of NO in the supernatant alone can be used for the
establishment of relative NO production. In Figure 9, the re-
ties in the cellular supernatant and in the cell lysates were
compared with data from calibration curves generated with
MAHMA/NO. These experiments indicated that (2.1ꢂ
1.2) nmol NO was produced by 105 cells over a period of
30 min [=(0.7ꢂ0.4) nmolminꢀ1 (106 cells)ꢀ1].
As FNOCT 8a turned out to be very sensitive for NO de-
tection, we then checked whether it may also be used for
time-resolved monitoring of NO production. To this end,
compound 8a (10 mm) was added to preactivated and nonac-
tivated alveolar macrophages and the time evolution of fluo-
rescence was monitored by means of a microplate reader.
As shown in Figure 10, from the preactivated cells an almost
Figure 9. Fluorescence detection of NO release (60 min) from stimulated
rat alveolar macrophages by reaction with FNOCTs 5b (lem =460, lexc
=
380 nm), 5c (lem =380, lexc =320 nm), and 8a (lem =400, lexc =340 nm)
(10 mm each) under standard conditions (0.02% Pluronic F127, 100 mm
DTPA, 100 mm Trolox, 10 mm glucose) in the presence of l-arginine
(black bars) and of unstimulated controls additionally supplemented with
NMA (gray bars) in 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid
(HEPES)-buffered, bicarbonate-free Krebs–Henseleit buffer (pH 7.4) at
378C.
Figure 10. Fluorescence monitoring of NO production from preactivated
(18 h activation with LPS) (circles) and nonactivated (diamonds) rat al-
veolar macrophages by reaction with FNOCT 8a (10 mm (open symbols)
and 50 mm (closed symbols)]. For preactivated cells, l-arginine (0.5 mm)
was added as a substrate for NO synthase. For nonactivated controls, the
NO synthase inhibitor NMA (0.1 mm) was added. Experiments were per-
formed in HEPES-buffered, bicarbonate-free Krebs–Henseleit buffer.
Data are averages from two microplate-well readings each. Dashed lines
are linear least-squares fits through the data points in the 0–45 min
region. The solid lines are just intended to guide the eye.
sults are displayed for the total (i.e., supernatant+lysate)
fluorescence increase of LPS-activated alveolar macrophag-
es (with l-arginine as the substrate for NO synthase) and
nonactivated cells (with NMA as the inhibitor of NO syn-
thase). The same protocol was carried out with the phenan-
threne-based FNOCTs 5b and 5c. Comparison of the fluo-
rescence intensities from experiments in the presence of l-
arginine with those in the presence of NMA reveals that 8a
is more sensitive and less susceptible to nonspecific conver-
sion to fluorescent products than FNOCTs 5b and 5c em-
ployed in previous biological studies (Figure 9). It can easily
be seen that FNOCT 8a would be able to detect far lower
amounts of NO than that released by the average 9ꢄ104
macrophages of these experiments.
The relative fluorescence recorded in the experiments
with nonactivated cells, in which basal NO formation was
additionally inhibited, reflects the reduced stability of
FNOCTs 5b and 5c in terms of NO-independent regenera-
tion of the fluorescent aromatic system. The data displayed
in Figure 9 thus demonstrate that 8a should be the molecule
of choice for cell experiments.
linear increase of the fluorescence signal was monitored for
the first 40–50 min of the experiment, whereas the low fluo-
rescence emission from the nonactivated controls remained
essentially constant. Increasing the concentration of 8a to
50 mm resulted in a somewhat increased fluorescence intensi-
ty as a result of an initial, slight (ꢁ25%) increase of the
rate of NO trapping. At both concentrations, the growth of
the fluorescence signal was somewhat diminished for reac-
tion times ꢃ50 min. From Figure 10 it can be inferred that
an increase of the concentration of 8a above 10 mm is of
little advantage in terms of NO detection, since the back-
ground signal is also increased.
It is known that the amount of NO released by the induci-
ble NO synthase (iNOS) of macrophages generally is far
higher than the amount of NO released by the constitutive,
calcium-dependent NO synthase of endothelial cells
(eNOS). The NO release by the latter is usually only detect-
ed indirectly, for example, by the effects of NO synthase in-
hibitors on NO-mediated responses like vasodilation. As
FNOCT 8a showed very good sensitivity for NO detection
in the cell-free experiments as well as in the experiments
For the quantitation of NO production from alveolar mac-
rophages, experiments similar to those shown in Figure 9
were performed, but with an incubation time of only 30 min
to avoid oversaturation of FNOCT 8a. Fluorescence intensi-
Chem. Eur. J. 2010, 16, 11121 – 11132
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11129

R. Sustmann et al.
2 mW; modulation amplitude: 0.03 mT; sweep range: 7 mT; sweep time:
15 min. Spectra simulation was performed with the WinSim program.[33]
with macrophages, we assessed whether it may detect re-
lease of NO by endothelial cells. To this end, small segments
of porcine aorta were longitudinally sliced open and the re-
sulting sheet was fixed in plane by means of a custom-de-
signed press-on chamber (see the Supporting Information,
Figure S6) with the endothelial lining looking upwards. This
provided a chamber with a 2.2 cm2 endothelial cell surface
that was filled with Hanksꢅ balanced salt solution (400 mL)
containing d-glucose (10 mm), 8a (50 mm), DTPA (100 mm),
Trolox (100 mm), l-arginine (0.5 mm), or, for controls, NMA
(0.1 mm) and Pluronic F127 (0.02%). The calcium iono-
phore ionomycin (10 mm) was added (in the presence of l-ar-
ginine) to part of the aortic segments. After 30 min of incu-
bation at 378C, a sample of the supernatant was taken and
its fluorescence determined and compared with that of stan-
dard samples generated with MAHMA/NO incubations in
cell-free solutions. From these experiments an NO produc-
tion of (0.34ꢂ0.11) mm in ionomycin-stimulated aortic seg-
ments and of (0.35ꢂ0.27) mm in the arginine-supplemented
incubations in the absence of ionomycin, as compared with
the fluorescence of NMA-treated controls, was determined.
Thus, an equal amount of about 0.14 nmol NO was formed
by the 2.2 cm2 ionomycin-treated as well as by the ionomy-
cin-untreated endothelium. In both cases the signal could be
inhibited by NMA, so there is no doubt that the fluores-
cence response truly indicated NO production. The NO for-
mation already observed in the absence of ionomycin can
reasonably be attributed to stimulation by the stress exerted
on the endothelium by the harvesting and preparation of
the tissue samples. In any case, these experiments underline
the sensitivity of FNOCT 8a for NO detection in biological
samples.
Crystallography: X-ray diffraction data for 8b and 13b were collected
with a Siemens SMART three-axis goniometer with APEX II area detec-
tor system at 203(2) K. Unit cell dimensions were determined in high 2q
angles using the XSCANS program. Three standard reflections measured
after every 97 reflections exhibited no significant loss in intensity. The
structures were solved by direct methods and refined by least-squares
techniques on F2 by using the SHELXTL program. Compound 8b crys-
tallizes in the monoclinic space group P21/c with z’=2 and a=12.979(14),
b=27.62(3), c=18.394(14) ꢃ, g=108.22(2)8, and R1=0.0558. 13b crys-
¯
tallizes in the triclinic space group P1, a=7.9014(5), b=12.7820(8), c=
17.3751(11) ꢃ, a=70.855(3), b=84.788(3), g=80.709(3)8 with z’=1 and
R1=0.0549.
Computational methods: DFT calculations were performed with the
Gaussian 03 suite of quantum chemical programs.[34] Geometries were
fully optimized on the (U)B3LYP level using the 6-31G(d) basis sets.
NO trapping from NO solution, PAPA/NO, and MAHMA/NO: These
experiments are described in the Supporting Information.
Cell isolation and culture: Alveolar macrophages were isolated from
male Wistar rats (350–380 g) by bronchoalveolar lavage as described pre-
viously.[35] The cells were seeded onto two four-well cell culture plates
(Nunc) and cultured in Dulbeccoꢅs modified Eagleꢅs medium (DMEM)
supplemented with l-glutamine (2 mm) and gentamicin (50 mgmLꢀ1).
One hour after seeding, cells were washed with warm Hanksꢅ balanced
salt solution (HBSS; 137 mm NaCl, 5.4 mm KCl, 1.0 mm CaCl2, 0.5 mm
MgCl2, 0.4 mm KH2PO4, 0.3 mm Na2HPO4, 25 mm HEPES, pH 7.4) and
supplied with fresh medium (DMEM) supplemented with l-glutamine
(2 mm), gentamicin (50 mgmLꢀ1), and 10% heat-inactivated fetal calf
serum. Rat hepatocytes and liver endothelial cells were isolated and cul-
tured as described previously.[36]
Assessment of FNOCT toxicity: The potential toxicity of FNOCT 8a was
assessed in cultured hepatocytes, liver endothelial cells, and alveolar mac-
rophages. To this end, cultured cells were incubated in Krebs–Henseleit
buffer containing d-glucose (10 mm), Pluronic F127 (0.02%), and
FNOCT 8a (10 or 50 mm) or solvent (DMSO) for 2–6 h at 378C. Cell
injury was assessed by the release of lactate dehydrogenase[36] (hepato-
cytes, liver endothelial cells) or by trypan blue uptake[37] (alveolar macro-
phages).
NO trapping from alveolar macrophages: Four to five hours after cell iso-
lation and 18 h prior to the experiments, cells in half of the wells were
primed for NO production by the addition of LPS (0.5 mgmLꢀ1) for 18 h.
For experiments, cells were washed with HBSS, counted, and supplied
with modified Krebs–Henseleit buffer (115 mm NaCl, 25 mm NaHCO3,
5.9 mm KCl, 1.2 mm MgCl2, 1.2 mm NaH2PO4, 1.2 mm Na2SO4, 2.5 mm
CaCl2, 20 mm HEPES, pH 7.4) supplemented with d-glucose (10 mm),
DTPA (100 mm), Trolox (100 mm), and FNOCTs 8a, 5b, or 5c (10 mm
each). The cell density in the experiments was (4.8ꢂ1.3)ꢄ104 cellscmꢀ2
[(9.0ꢂ2.5)ꢄ104 cellswellꢀ1]. l-Arginine (0.5 mm) was added to the wells
with activated cells, and NMA (0.1 mm) was added to the other wells.
Cells were incubated at 378C in a humidified atmosphere containing
95% air/5% CO2 for 30 min (one series with 8a) or 1 h (all other series).
After this incubation, the supernatant was removed and fluorescence de-
termined with the microplate reader at lexc =340 nm, lem =400 nm (gain
2000) at the same instrument settings as in the cell-free experiments. For
5b, lexc =380 nm, lem =460 (gain 2000) and for 5c lexc =320 nm, lem =380
(gain 1500) were used. With FNOCT 8a, cells were also assessed continu-
ously in the plate reader in one series. After removal of the supernatant,
cells were lysed in HBSS containing SDS to a final concentration of
0.4% (w/v), DTPA (100 mm), and Trolox (100 mm). Cell lysates were cen-
trifuged to remove cell debris, and supernatants were then assessed in
the same way as the samples of the supernatant Krebs–Henseleit buffer.
Blanks of all solutions with all additives including FNOCTs 8a, 5b, and
5c were incubated (in the absence of cells) for the same time periods and
measured at the same time as samples. Blank values were subtracted
from all sample values. Cell blanks were incubated in the absence of
FNOCT. However, cellular autofluorescence proved to be negligible in
the experimental setting used here. The fluorescence values obtained are
Conclusion
The pyrene-based FNOCT 8a represents a promising and
versatile tool for the monitoring of NO production in cell
cultures and tissues as well as in cell-free systems.
Experimental Section
Materials and instrumentation: The preparation of compounds 9, 12, and
13c and the X-ray structure of 13c have been reported previously.[20] Pro-
cedures for the preparation of all other compounds and spectroscopic
data are given in the Supporting Information. IR spectrometer: Bio-Rad
FTS 135; UV/Vis spectrophotometer: Cary 300 Bio (Varian); ESR: X-
band spectrometer ER-420 (Bruker); fluorescence spectrometry: FL
3095 J&M GmbH (software FL 3095-Fluoroscan); fluorescence micro-
plate reader: BMG Labtech FLUOstar Optima; mass spectrometer: BI-
OTOF III (Bruker); 1H and 13C NMR spectrometer: Bruker DRX 500.
The undeuterated solvent was used as an internal standard. The assign-
ment of 1H NMR signals was supported by 2D experiments. Melting
point: apparatus 9100 (Electrothermal), melting points are uncorrected.
ESR measurements: X-band ESR spectra were recorded in argon-flush-
ed, septum-capped 4 mm outer diameter quartz tubes by injecting a satu-
rated NO solution in THF (30 mL) into 1 mm FNOCT in oxygen-free
THF (100 mL). Microwave frequency: 9.48 GHz; microwave power:
11130
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11121 – 11132

Fluorescent Nitric Oxide Cheletropic Traps
FULL PAPER
either given as counts, arbitrary units (corrected for the fluorescence of
blanks, for volume, and cell density), or were corrected for blanks, con-
verted to NO concentrations based on calibration curves similar to the
one shown in Figure 8, and then corrected for volumes and cell density.
After the fluorescence of cellular supernatants and cellular lysates was
determined, MAHMA/NO was added in large excess (100 mm) to all sam-
ples to obtain maximal FNOCT fluorescence; these values were used to
determine FNOCT distribution (supernatant versus cell-associated).
[4] a) Biology of Nitric Oxide (Ed.: S. Moncada), Portland Press,
London, 2000; b) Nitric Oxide Protocols (Ed.: M. A. Titheradge),
Humana Press, Totowa, 1998; c) Nitric Oxide Protocols (Ed.: A.
Hassid), Humana Press, Totowa, 2004; d) Methods in Nitric Oxide
Research (Eds.: M. Feelisch, J. S. Stamler), Wiley, New York, 1996.
[5] a) C. W. G. Fishwick, D. W. Jones in The Chemistry of Quinoid Com-
pounds, Vol. 2 (Eds.: S. Patai, Z. Rappoport), Wiley, New York,
1988, pp. 403–453; b) N. Martin, C. Seoane, M. Hanack, Org. Prep.
Proced. 1991, 23, 237–272; J. L. Segura, N. Martin, Chem. Rev. 1999,
99, 3199–3246.
[6] H. G. Korth, K. U. Ingold, R. Sustmann, H. De Groot, H. Sies,
Angew. Chem. 1992, 104, 915–917; Angew. Chem. Int. Ed. Engl.
1992, 31, 891–893.
[7] H.-G. Korth, R. Sustmann, P. Lommes, T. Paul, A. Ernst, H.
de Groot, L. Hughes, K. U. Ingold, J. Am. Chem. Soc. 1994, 116,
2767–2777.
NO trapping from segments of porcine aortae: Porcine aortae were ob-
tained from the local abattoir within 20–30 min of death by exsanguina-
tion after electrical stunning and transported to the laboratory within
30 min. The aortae were transported in HBSS + d-glucose (10 mm) at
room temperature. In the laboratory, aortae were freed from surrounding
tissue and extensively rinsed with sterile HBSS + glucose (room temper-
ature). Thereafter, the thoracic aorta was opened longitudinally, the pos-
terior quarter of the circumference excised, and the aorta cut into seg-
ments of length approximately 2 cm. Each aortic segment was placed, en-
dothelial surface upwards, on to a cork plate. A ring of stainless steel
(inner diameter 1.67 cm, height 1.5 cm) with a sharp edge at the bottom
was pressed on to an aortic segment and fixed with two cannulae to form
a fluid-tight chamber on the endothelial surface with an area of 2.2 cm2.
This chamber was filled with HBSS (400 mL) containing d-glucose
(10 mm) and l-arginine (0.5 mm) or, for control, NMA (0.1 mm) and seg-
ments were preincubated at 378C for 30 min. Thereafter, the supernatant
was discarded and replaced by fresh HBSS (400 mL) containing d-glucose
(10 mm), DTPA (100 mm), Trolox (100 mm), FNOCT 8a (50 mm), 0.02%
Pluronic F127, and l-arginine (0.5 mm) or, for control, NMA (0.1 mm).
Aortic segments were incubated at 378C for 30 min, then the supernatant
(300 mL) was sampled and fluorescence assessed in a fluorescence micro-
plate reader at lexc =340 nm, lem =400 nm. Calibration curves were ob-
tained with MAHMA/NO in HBSS containing the same additives as for
the aortic measurements.
[8] T. Paul, M. A. Hassan, H.-G. Korth, R. Sustmann, D. V. Avila, J.
Org. Chem. 1996, 61, 6835–6848.
[9] a) T. Paul, R. Boese, I. Steller, H. Bandmann, G. Gescheidt, H.-G.
Korth, R. Sustmann, Eur. J. Org. Chem. 1999, 551–563; b) The ring
opening of benzocyclobutene to o-quinodimethane is 13.1 kcalmolꢀ1
endothermic: W. R. Roth, B. P. Scholz, Chem. Ber. 1981, 114, 3741–
3750; W. R. Roth, V. Rekowski, S. Bçrner, M. Quast, Liebigs Ann.
Chem. 1996, 409–430.
[10] D. W. Jones, A. Pomfret, J. Chem. Soc. Perkin Trans. 1 1991, 263–
267.
[11] I. Ioannidis, M. Bꢁtz, T. Paul, H.-G. Korth, R. Sustmann, H.
de Groot, Biochem. J. 1996, 318, 789–795.
[12] M. Bꢁtz, H.-G. Korth, R. Sustmann, Angew. Chem. 1997, 109, 1555–
1557; Angew. Chem. Int. Ed. Engl. 1997, 36, 1501–1503.
[13] S. A. Green, D. J. Simpson, G. Zhou, P. S. Ho, N. V. Blough, J. Am.
Chem. Soc. 1990, 112, 7337–7346.
[14] P. Meineke, U. Rauen, H. de Groot, H.-G. Korth, R. Sustmann,
Chem. Eur. J. 1999, 5, 1738–1747.
[15] P. Meineke, U. Rauen, H. de Groot, H.-G. Korth, R. Sustmann,
Biol. Chem. 2000, 381, 575–582.
[16] F. S. Hornig, H.-G. Korth, U. Rauen, H. de Groot, R. Sustmann,
Helv. Chim. Acta 2006, 89, 2281–2296.
[17] F. S. Hornig, Dissertation, Universitꢁt Duisburg–Essen (Germany),
2007.
Acknowledgements
The excellent technical assistance of N. Boschenkov is gratefully ac-
knowledged.
[18] J. Hu, D. Zhang, F. W. Harris, J. Org. Chem. 2005, 70, 707–708.
[19] D. W. Jones, J. Chem. Soc. Perkin Trans. 1 1977, 980–987.
[20] A. Mohr, P. Talbiersky, H.-G. Korth, R. Sustmann, R. Boese, B.
Blꢁser, H. Rehage, J. Phys. Chem. B 2007, 111, 12985–12992.
[21] Only one other isomer (10–20%) of 13a–c could be detected in the
NMR spectra of the raw reaction products. According to the chemi-
cal shifts of the methylene bridge protons (d=1.7–1.8 and 0.9–
1.0 ppm), these minor isomers were attributed to the syn-(exo–exo)
diastereomer (see the Supporting Information, Scheme S1), by com-
parison with literature data on decahydro-1,4-exo–endo-5,8-dimetha-
nonaphthalen-9-one (1.74 and 0.94 ppm) and related compounds: T.
Svensson, S. Winstein, J. Am. Chem. Soc. 1972, 94, 2336–2347. No
attempts were made to isolate these isomers in the pure form.
[22] We use notations like “exo–endo” etc. to indicate Diels–Alder addi-
tion to the exo side of the norbornene in the endo orientation of the
diene. The additional syn–anti notation refers to the relative orienta-
tion of the two bridges.
[1] a) Nitric Oxide (Eds.: B. Tota, B. Trimmer), Elsevier, Amsterdam,
2007; b) Pathophysiology and Clinical Applications of Nitric Oxide
(Ed.: G. M. Rubanyi), Harwood Academic Publishers, Amsterdam,
1999; c) S. Lamas, E. Cardenas, Nitric Oxide, Cell Signaling, and
Gene Expression, CRC Press, Boca Raton, 2006; d) A. R. Butler, R.
Nicholson, Life, Death and Nitric Oxide, RSC, Cambridge, 2003;
e) Nitric Oxide: Biology and Pathobiology (Ed.: L. J. Ignarro), Aca-
demic Press, San Diego, 2000; f) Nitric Oxide. Principles and Actions
(Ed.: J. Lancaster, Jr.), Academic Press, San Diego, 1996.
[2] a) R. F. Furchgott, Angew. Chem. 1999, 111, 1990–2000; Angew.
Chem. Int. Ed. 1999, 38, 1870–1880; b) L. J. Ignarro, Angew. Chem.
1999, 111, 2002–2013; Angew. Chem. Int. Ed. 1999, 38, 1882–1892;
c) F. Murad, Angew. Chem. 1999, 111, 1976–1989; Angew. Chem.
Int. Ed. 1999, 38, 1856–1868.
[23] Only two X-ray structures of o-quinodimethanes have been reported
so far: in the monocyclic 7,7,8,8-tetraaryl-o-quinodimethane report-
ed by Iwashita et al. (S. Iwashita, E. Ohta, H. Higuchi, H. Kawai, K.
Fujiwara, K. Ono, M. Takenaka, T. Suzuki, Chem. Commun. 2004,
2076–2077) the conjugated o-quinoid system is strongly distorted
due to the steric interaction of the four phenyl groups. Very recently,
Ghereg et al. (D. Ghereg, S. E.-C. El Kettani, M. Lazraq, H. Ranai-
vonjatovo, W. W. Schoeller, J. Escudie, H. Gornitzka, Chem.
Commun. 2009, 4821–4823) reported on a planar o-quinodimethane
stabilized by germyloxyl rings at the terminal positions.
[24] R. Allmann, Z. Kristallogr. Kristallgeom. Kristallphys. Kristallchem.
1970, 132, 129–151.
[3] a) Special issue on: “Nitric Oxide, Part A: Sources and Detection of
NO; NO Synthase”: Methods Enzymol. 1996, 268; b) Special issue
on: “Nitric Oxide, Part B: Physiological and Pathological Process-
es”: Methods Enzymol. 1996, 269; c) Special issue on: “Nitric Oxide,
Part C: Biological and Antioxidant Activities”: Methods Enzymol.
1999, 301; d) Special issue on: “Nitric Oxide, Part D: Oxide Detec-
tion, Mitochondria and Cell Functions, and Peroxynitrite Reac-
tions”: Methods Enzymol. 2002, 359; e) Special issue on: “Nitric
Oxide, Part E”: Methods Enzymol. 2005, 396; f) Special issue on:
“Nitric Oxide, Part F”: Methods Enzymol. 2008, 440; g) Special
issue on: “Nitric Oxide, Part G: Oxidative and Nitrosative Stress in
Redox Regulation of Cell Signaling”: Methods Enzymol. 2008, 441.
Chem. Eur. J. 2010, 16, 11121 – 11132
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11131

R. Sustmann et al.
[25] CCDC-757257 (8b) and 757258 (13b) contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[26] I. Morishima, K. Yoshikawa, K. Bekki, M. Kohne, K. Arita, J. Am.
Chem. Soc. 1973, 95, 5815–5817.
T. A. Hatton, Macromolecules 1994, 27, 2414–2425 (Pluronic F127)
and ref. [20] (SDS).
[33] a) WinSim, D. R. Duling, Laboratory of Molecular Biophysics, Re-
search Triangle Park, North Carolina, 1994; b) D. R. Duling, J.
Magn. Reson. Ser. B 1994, 104, 105–110.
[34] Gaussian 03, Revision E.01, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[27] The inversion barrier of cyclic nitroxides is very low, generally less
than 2 kcalmolꢀ1; see: L. Horny, F. Mariotti, M. Quack, Chimia
´
2008, 62, 256–259; I. Komaromi, J. M. J. Tronchet, J. Phys. Chem.
1995, 99, 10213–10220.
[28] J. A. Hrabie, J. R. Klose, D. A. Wink, L. K. Keefer, J. Org. Chem.
1993, 58, 1472–1476.
[29] L. K. Keefer, R. W. Nims, K. M. Davies, D. A. Wink in Methods in
Enzymology, Vol. 268 (Ed.: L. Packer), Academic Press, San Diego,
1996, pp. 281–293.
[30] V. Stenert, Dissertation, Universitꢁt Essen (Germany), 2003.
[31] a) D. L. Mooradian, T. C. Hutsell, L. K. Keefer, J. Cardiovasc. Phar-
macol. 1995, 25, 674–678; b) K. M. Davies, D. A. Wink, J. E. Saave-
dra, L. K. Keefer, J. Am. Chem. Soc. 2001, 123, 5473–5481.
[32] Because the decomposition of N-diazeniumdiolates is a proton-cata-
lyzed process that can be enhanced by incorporation in micelles of
anionic surfactants like SDS (see: A. Mohr, T. Pozo Vila, H.-G.
Korth, H. Rehage, R. Sustmann, ChemPhysChem 2008, 9, 2397–
2405), the lower rate constant for MAHMA/NO decomposition in
the presence of the nonionic surfactant Pluronic F127 suggests a
slightly diminished local proton concentration in the vicinity of the
NONOate group of the NO donor by association with the surfac-
tant. This phenomenon is currently under investigation. Note that
the applied concentrations of Pluronic and SDS are above the cmc
values of these surfactants, see: P. Alexandridis, J. F. Holzwarth,
[35] J.-F. Wang, P. Komarov, H. de Groot, Arch. Biochem. Biophys. 1993,
304, 189–196.
[36] U. Rauen, B. Polzar, H. Stephan, H. G. Mannherz, H. de Groot,
FASEB J. 1999, 13, 155–168.
[37] U. Rauen, M. Hanßen, W. Lauchart, H. D. Becker, H. de Groot,
Transplantation 1993, 55, 469–473.
Received: January 6, 2010
Published online: August 2, 2010
11132
www.chemeurj.org
ꢂ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11121 – 11132