Selective detection of superoxide anion radicals generated
from macrophages by using a novel fluorescent probe
Jing Jing Gao
1
, Ke Hua Xu
1
, Bo Tang
1
, Ling Ling Yin
1
, Gui Wen Yang
2
and Li Guo An
2
1 College of Chemistry, Chemical Engineering and Materials Science, Shandong Normal University, Jinan China
2 College of Life Science, Shandong Normal University, Jinan, China
Reactive oxygen species (ROS) such as the superoxide
anion radical (O
2
–
Æ), hydroxyl radical (HOÆ) and hydro-
gen peroxide (H
2
O
2
) are important mediators in
various pathological diseases [1]. Free radicals, in par-
ticular O
2
–
Æ, have been found extensively in myocardial

superoxide dismutase (SOD)-inhibitable Nitro Blue
tetrazolium [8], chemiluminescence [9,10] and fluor-
escence [11]. Among these methods, fluorescence
detection appears to be particularly attractive because
it is able make superoxide anion radicals in living cells
‘visible’ in situ using confocal laser scanning micro-
scopy. Although several types of fluorescent probe for
detecting and imaging O
2
–
Æ have been described [11–
13], those with high selectivity, sensitivity and practi-
cality are rare. For example, hydroethidine is the most
commonly used fluorescent probe for O
2
–
Æ [14], but its
selectivity could be improved [15]. In addition, owing
to the short half-life of the superoxide anion radical,
there is an exigent need for researchers to develop fast-
response probes to trap O
2
–
Æ in order to investigate its
mode of production, metabolism and trafficking.
Therefore, considering the design-applicable probes, it
is important that the reaction rate and selectivity are
improved to avoid potential side reactions from other
ROS under conditions that guarantee high sensiti-
vity. If novel fluorescence probes that overcome these

used to image O
2
–
Æ in living cells. Better selectivity for O
2
–
Æ over competing
cellular reactive oxygen species and some biological compounds illustrates
the advantages of our method. Under optimal conditions, the linear calib-
ration range for superoxide anion radicals was 5.03 · 10
)9
)3.33 · 10
)6
m.
The detection limit was 1.68 · 10
)9
m. Fluorescence images of probe-
stained macrophages stimulated with 4b-phorbol 12-myristate 13-acetate
were obtained successfully using a confocal laser scanning microscope.
Abbreviations
DBZTC, 2-chloro-1,3-dibenzothiazolinecyclohexene; ROS, reactive oxygen species; SOD, superoxide dismutase.
FEBS Journal 274 (2007) 1725–1733 ª 2007 The Authors Journal compilation ª 2007 FEBS 1725
problems were available, they would contribute greatly
to the elucidation of the roles of O
2
–
Æ in living cells.
In this study, a fast-reaction fluorescent probe for
O
2

and biological compounds, especially H
2
O
2
. Because
the level of H
2
O
2
is high compared with O
2
–
Æ, due to
accumulation in biological systems, H
2
O
2
may be the
most competitive ROS in the quantitative detection of
O
2
–
Æ. In this study, DBZTC showed a selective
response to O
2
–
Æ that was > 500-fold greater than the
response to H
2
O

(Hepes, 0.02 m, pH 7.4). As indicated in Fig. 1,
DBZTC showed low blank fluorescence, although the
addition of different concentrations of XA ⁄ XO trig-
gered prompt increases in fluorescence (k
ex ⁄ em
¼
485 ⁄ 559 nm). The XA ⁄ XO system was used as the
main source of O
2
–
. It has been reported that on the
catalysis of XO at pH 7.40, XA is oxidized to uric acid
[16]. XA + 2O
2
+H
2
O ¼ uric acid + 2O
2
–
Æ +2H
+
(single electron transfer). XA can also be oxidized by
O
2
with double electrons XA + O
2
+H
2
O ¼ uric
acid + H

addition of different concentrations XA ⁄ XO (final concentration:
0 ⁄ 0, 2 ⁄ 2, 3 ⁄ 3, 4 ⁄ 4, 6 ⁄ 6, 10 ⁄ 10, 15 ⁄ 15 and 20 ⁄ 20 l
M ⁄ mU) to a
solution of DBZTC.
Selective detection of superoxide anion radicals J. J. Gao et al.
1726 FEBS Journal 274 (2007) 1725–1733 ª 2007 The Authors Journal compilation ª 2007 FEBS
between fluorescent intensity and O
2
–
Æ concentration in
the range 5.03 · 10
)9
)6.66 · 10
)6
m. The regression
equation was F ¼ 3815.8 [O
2
–
Æ](lm) +2579.1 (Fig. 2).
The detection limit was 1.68 · 10
)9
m. This result
showed that DBZTC could detect O
2
–
Æ both qualita-
tively and quantitatively.
Reaction conditions
To determine the optimum reaction conditions for
the analysis of O

precision and sensitivity of the analytical method. The
relative fluorescence intensity increased as the DBZTC
concentration increased (< 9 lm), remained constant
at DBZTC concentrations of 9–18 lm, then decreased
(Fig. 4). A suitable concentration of DBZTC was
advantageous, whereas superfluous DBZTC could
quench fluorescence. Therefore, 10 lm of DBZTC was
used throughout.
Effects of other ROS and biological compounds
To assess the selectivity of the method, the effect of
other ROS and biological compounds on the determin-
ation of 3.33 lm O
2
–
Æ was examined individually. An
error of ± 5.0% in the relative fluorescence intensity
was considered tolerable. Little or no interference was
encountered with (tolerance ratio in mol): NaClO, tert-
butyl hydroperoxide, H
2
O
2
(500);
1
O
2
, glutathione,
1,4-hydroquinone (100); V
C
(50); HOÆ (5); ONOO

(ONOO
–
) and nitric oxide (NOÆ) were obtained from 3-
morpholinosydnonimine hydrochloride (SIN-1), and
3-(aminopropyl)-1-hydroxy-3-isopropyl-2-oxo-1-triazene,
respectively. The results are summarized in Fig. 5.
DBZTC appears to be a highly selective fluorescent
probe for O
2
–
Æ.
Effect of SOD on fluorescence intensity of the reaction
of DBZTC with superoxide
In order to further confirm that the changes in fluores-
cence of the probe solution were caused by O
2
–
Æ, SOD,
a scavenger of O
2
–
Æ, was used in the reactive system.
After reaction of SOD (150 U) with XA ⁄ XO
(10 lm ⁄ 10 mU) in Hepes buffer (20 mm) had been car-
ried out for 30 min at 37 °C, DBZTC was added and
the reaction was immediately diluted with doubly dis-
tilled water. The mixture was equilibrated and kept for
5 min before measurement. As can be seen from
Fig. 6, fluorescence intensity was markedly suppressed
by addition of SOD. However, when SOD was

indicated that a larger conjugated structure of the
DBZTC oxide was formed. From product analysis and
the fluorescence properties, we propose that the mode
Fig. 4. Effect of the fluorescent probe concentration: XA (10 lM),
XO (10 mU), Hepes (20 m
M).
Fig. 5. Selectivity of DBZTC for O
2
–
Æ at pH 7.40 (20 mM Hepes).
Fluorescence response of DBZTC to O
2
–
Æ, other ROS and biological
compounds. Bars represent the final integrated fluorescence
response (F) over the initial integrated emission (F
0
). Initial spectra
were acquired in a 10 l
M solution of DBZTC. Light grey bars repre-
sent the addition of an excess of the appropriate other ROS and
biological compounds (1.66 m
M for H
2
O
2
, NaClO, and tert-butyl
hydroperoxide, 0.33 m
M for
1

(Scheme 2). Upon treatment with O
2
–
Æ, DBZTC, an
almost nonfluorescent compound, was oxidized at
pH 7.40 by deleting hydrogen [19] to yield DBZTC
oxide, which has better rigidity and a larger conjugated
system (free electron pairs on S also conjugating with
phenyl ring).
In order to validate the molar ratio of the reaction
between the probe and O
2
–
Æ, we synthesized pure
DBZTC oxide and tested the fluorescence intensity of
the oxide at different concentrations. A linear relation-
ship between DBZTC oxide concentration and the fluor-
escence intensity was obtained (Fig. 7). The linear
equation was as followed: F ¼ 4268 [DBZTC oxide] +
288. Comparing this with the linear correlation between
the fluorescent intensity and the concentration of
O
2
–
Æ (F ¼ 3815.8 [O
2
–
Æ] + 2579.1), we may deduce that
DBZTC reacts with O
2

2
–
Æ in cells
We used RAW264.7 macrophages to investigate the
potential of DBZTC to detect O
2
–
Æ in living cells. Macr-
ophages were seeded onto a glass slide and the concen-
tration adjusted to 1 · 10
6
cellsÆmL
)1
with Dulbecco’s
modified Eagle’s medium containing 10% fetal bovine
serum, 1% penicillin and 1% streptomycin. Cells were
loaded with DBZTC (10 lm, Hepes, pH 7.40) by
incubation at 37 °C for 15 min and showed negligible
intracellular background fluorescence (Fig. 9A). Macr-
ophages, which were loaded with DBZTC (10 lm),
were stimulated with 4b-phorbol 12 myristate 13 acet-
ate (2 ngÆmL
)1
)at37°C for 12 h; an obvious increase
Fig. 7. Linear correlation between the fluorescence intensity and
the concentration of DBZTC oxide.
Scheme 2. A mechanism for the reaction of DBZTC with O
2
–
Æ.

incubated with DBZTC (10 l
M)at37°C for 15 min. (B) Fluorescence image of probe-stained RAW264.7 macrophages stimulated with
4b-phorbol 12 myristate 13 acetate (2 ngÆmL
)1
)at37°C for 12 h. (C) Bright-field image of probe-stained RAW264.7 macrophages stimulated
with 4b-phorbol 12 myristate 13 acetate (2 ngÆmL
)1
)at37°C for 12 h to confirm viability. (D) Fluorescence image of cells incubated with
100 l
M Tiron at 37 °C for 1 h after 4b-phorbol 12 myristate 13 acetate stimulation for 12 h, followed by loading with probe at 37 °C for
15 min (scale bar ¼ 15 l
M).
Selective detection of superoxide anion radicals J. J. Gao et al.
1730 FEBS Journal 274 (2007) 1725–1733 ª 2007 The Authors Journal compilation ª 2007 FEBS
In order to further investigate the feasibility of the
proposed method in biological systems, determination
of O
2
–
Æ in cell extracts was performed. The fluorescence
intensity of cell extracts treated with a Tiron solution
was used as a blank value. The detected O
2
–
Æ mean
content of 4b-phorbol 12 myristate 13 acetate-stimula-
ted cell extracts was 0.92 · 0.02 lm, based on a corre-
lation between the fluorescence intensity of the extracts
and the regression equation. The average recovery test
was carried out using the standard addition method,

from a 500-fold molar excess of H
2
O
2
). Furthermore,
DBZTC was confirmed as a cell-permeable probe and
was able to respond to micromolar changes in O
2
–
Æ con-
centration within living cells. Overall results establish
the potential value of the probe for facilitating investi-
gations of the generation, metabolism, and mechanisms
of superoxide-mediated cellular homeostasis and injury.
Experimental procedures
Apparatus
1
H NMR spectra were recorded on a Bruker Avance 300.
Elemental analysis was performed on a Perkin–Elmer Series
II CHNS ⁄ O analyzer. IR spectra were recorded on PE-983
IR spectrometer (KBr discs cm
)1
, Perkin–Elmer, Norwalk,
CT). All pH measurements were made using a pH-3c digital
pH meter (Shanghai Lei Ci Device Works, Shanghai,
China) with a combined glass–calomel electrode. Fluorimet-
ric spectra were obtained with a FLS-920 Edinburgh fluor-
escence spectrometer with a xenon lamp and 1.0 cm quartz
cells. Confocal fluorescence imaging was captured using a
Zeiss LSM 510 META scanning microscope containing an

aqueous solution), H
2
O
2
(30% aqueous solution), sodium
hypochlorite (NaOCl, 5% aqueous solution), dimethyl-
sulfoxide, 3-morpholinosydnonimine hydrochloride, 3-(ami-
nopropyl)-1-hydroxy-3-isopropyl-2-oxo-1-triazene, phorbol
12-myristate 13-acetate, and RPMI-1640 medium were from
Sigma. 1,4-Hydroquinone was from Fluka. 4,5-Dihydroxy-
1,3-benzenedisulfonic acid disodium salt (Tiron) was from
Shanghai Reagent Co. Ltd (Shanghai, China). All chemi-
cals were of analytical reagent grade, and double-distilled
water was used throughout. RAW264.7 cells were from
American Type Culture Collection (Manassas, VA).
Synthesis and properties of DBZTC
2-chloro-1-formyl-3-hydroxymethylenecyclohexene [21]
A solution of 40.0 mL of dimethylformamide in 40.0 mL of
methylene chloride was chilled in an ice bath. A solution of
37.0 mL of phosphorus oxychloride in 35.0 mL of methy-
lene chloride was added dropwise with stirring. Then 10.0 g
of cyclohexanone was added to the mixture. The solution
was refluxed for 3 h, cooled, poured onto 200 g of ice, and
Table 1. Determination of superoxide anion radicals in cell extracts (n ¼ 6). DBZTC (0.01 mM), Tiron solution (1 lM), 4b-phorbol 12 myristate
13 acetate (PMA; 2.0 ngÆL
)1
), XA (1.00 mM,30lL) ⁄ XO (1 U, 30 lL), Hepes (0.10 M, pH 7.40).
Sample
O
2

DBZTC
We added 2.50 g (0.02 mmol) of o-aminobenzenothiol,
1.73 g (0.01 mmol) of 2-chloro-1-formyl-3-hydroxymethyl-
enecyclohexene and 50.0 mL of a mixture of 1-butanol and
benzene (7 : 3 v ⁄ v) to a 100 mL flask. The mixed solution
was heated under reflux for 3 h. After the solvent was
removed under reduced pressure, a yellow crude solid was
obtained. This was recrystallized from benzene and dried
under a vacuum (2.087 g, 54% yield).
1
H NMR (300 MHz,
DMSO-d
6
,25°C, TMS): d 8.10–7.90 (m, 2H, benzol-H),
7.75 (d, 2H, benzol-H), 7.47–7.39 (m, 2H, benzol-H), 7,24
(d, 2H, benzol-H), 4.10 (m, 2H, N-H), 3.12 (m, 2H,
methine-H), 2.3 (m, 3H, cyclohexene-H), 1.15 (m, 4H,
cyclohexene-H). IR (KBr pellet): (cm
)1
) 3245 (N–H), 3110
(C–H), 1592 (C ¼ C). Melting point: 95–97 °C. Elemental
analysis (%) calcd for C
20
H
19
N
2
S
2
Cl (found): C 62.10

N
2
S
2
Cl (found): C 62.75 (62.71),
H 3.92 (3.84), N 7.32 (7.37).
Determination of O
2
–
Æ
Into a 10 mL color comparison tube were added 1.00 mL
of DBZTC (1.00 · 10
)4
m), 0.10 mL of XA solution
(1.00 mm), 0.10 mL of XO (1.00 U) and 2.00 mL of Hepes
buffer (0.10 m) in turn. After diluted to 10.00 mL volume
with double-distilled water, the mixture was equilibrated
and was laid aside at 37 °C for 10 min before determin-
ation. The fluorescence intensity was measured at k
ex ⁄ em
¼
485 ⁄ 559 nm against a reagent blank at the same time. The
excitation and emission slit were set to 3.5 and 3.5 nm,
respectively.
Preparation and staining of RAW264.7
macrophages cultures
RAW264.7 macrophages were cultured in Dulbecco’s modi-
fied Eagle’s medium containing 10% fetal bovine serum,
1% penicillin, and 1% streptomycin at 37 °C (w/v) in a 5%
CO

)1
. Some cells were added to a
Tiron solution and incubated for 1 h as an O
2
–
. scavenger,
others were stimulated with 4b-phorbol 12 myristate 13
acetate at 37 °C for 12 h. All cells were then incubated with
DBZTC for 30 min at 37 °C and harvested by centrifuga-
tion in the cold, and washed twice with 0.9% NaCl solu-
tion. These cells were suspended again in a volume of
Hepes equal to that in which they had been grown, and dis-
rupted for 10 min in a VC 130PB ultrasonic disintegrator
(Sonics & Materials Inc., Newtown, CT, USA). During
sonic disruption, the temperature was maintained below
4 °C with circulating ice water. The broken cell suspension
was centrifuged at 1435 g for 5 min and the pellet discar-
ded. Cell extracts that had been added to a Tiron solution
were divided into six parts. The 4b-phorbol 12 myristate 13
acetate-stimulated cell suspension was divided into 12 parts
and XA ⁄ XO was added into six parts in turn.
Acknowledgements
This work was supported by Program for New Cen-
tury Excellent Talents in University (NCET-04–0651),
the National Natural Science Foundation of China
(Nos 20335030 and 20575036).
Selective detection of superoxide anion radicals J. J. Gao et al.
1732 FEBS Journal 274 (2007) 1725–1733 ª 2007 The Authors Journal compilation ª 2007 FEBS
References
1 Chu CT, Levinthal DJ, Kulich SM, Chalovich EM &

monitoring of oxygen free radical production using
chemiluminescence in vivo. J Cereb Blood Flow Metab
15, 929–940.
10 Peters O, Back T, Lindauer U, Busch C, Megow D,
Dreier J & Dirnagl U (1998) Increased formation of
reactive oxygen species after permanent and reversible
middle cerebral artery occlusion in the rat. J Cereb
Blood Flow Metab 18, 196–205.
11 Mu
¨
nzel T, Afanas’ev IB, Klescchyov AL & Harrison
DG (2002) Detection of superoxide in vascular tissue.
Arterioscler Thromb Vasc Biol 22, 1761–1768.
12 Hatsuo M, Kayko Y, Yoko N, Iho K, Leila H, Narit-
sugu U, Shoko Y, Masako F, Yuka F, Yuji Y et al.
(2005) A design of fluorescent for superoxide based on
a nonredox mechanism. J Am Chem Soc 127, 68–69.
13 Xu KH, Liu X, Tang B, Yang G, Yang Y & An LG
(2006) Design of a phosphinate-based fluorescent probe
for superoxide detection in mouse peritoneal
macrophages. Eur J Chem 13, 1411–1416.
14 Tarpey MM & Fridovich I (2001) Methods of detection
of vascular reactive species: nitric oxide, superoxide,
hydrogen peroxide, and peroxynitrite. Circ Res 89,
224–236.
15 Benov L, Sztejnberg L & Fridovich I (1998) Critical
evaluation of the use of hydroethidine as a measure of
superoxide anion radical. Free Radical Biol Med 25,
826–831.
16 Fridovich I (1982) Measuing the activity of superoxide