A fluorescent probe for fast and quantitative detection of hydrogen sulfide in blood

Article abstract of DOI:10.1002/anie.201104236

The dose makes the poison: Hydrogen sulfide is an important gasotransmitter for which rapid detection agents are needed. A hydrogen sulfide probe, which allows for fast (within seconds), selective, and quantitative detection in buffer solution, serum, and whole blood is designed, synthesized, and used for detection of hydrogen sulfide (see picture). Copyright

Full text of DOI:10.1002/anie.201104236

Communications
DOI: 10.1002/anie.201104236
Fluorescent Probes
A Fluorescent Probe for Fast and Quantitative Detection of Hydrogen
Sulfide in Blood**
Hanjing Peng, Yunfeng Cheng, Chaofeng Dai, Adrienne L. King, Benjamin L. Predmore,
David J. Lefer, and Binghe Wang*
Hydrogen sulfide (H2S), well-known for its unpleasant rotten
egg smell, was traditionally considered as a toxic gas.
However, recent studies have demonstrated that hydrogen
sulfide is an endogenously produced gaseous signaling com-
pound (gasotransmitter) with importance on a par with that of
the other two known endogenous gasotransmitters, nitric
oxide (NO)[1] and carbon monoxide (CO).[2] H2S has been
recognized for mediating a wide range of physiological
effects. Studies have shown that H2S can have an effect on
the cardiovascular system[3] by acting as a K-ATP channel
opener.[4] Several studies have shown the protective roles of
H2S, in situations such as myocardial ischemia, most likely
through a combination of its antioxidant and anti-apoptotic
signaling effects.[5] Further studies also showed that H2S may
be of therapeutic benefit for the treatment of ischemia-
induced heart failure.[6,7] It is also a modulator in the central
nervous system,[8–10] respiratory system, gastrointestinal
system, and endocrine system.[11] It seems that hydrogen
sulfide exhibits almost all the beneficial effects of NO without
generating the toxic reactive oxygen species (ROS). In
addition, it also acts as an antioxidant or scavenger of ROS.
Furthermore, research has indicated that the hydrogen sulfide
level is related to diseases such as the Down syndrome[12] and
Alzheimerꢀs disease.[13] Therefore, recent years have seen a
steady increase in interest in understanding the physiological
and pathological functions of hydrogen sulfide.[11,14,15] One
significant limiting factor in studying hydrogen sulfide is the
lack of sensors and agents that allow for its rapid and accurate
detection. Methods using colorimetric,[16–18] electrochemical
analysis,[19–21] and gas chromatography have been
reported.[22,23] However, catabolism of hydrogen sulfide is
known to be fast, which could result in continuous fluctuation
in its concentration, leading to difficulties in the accurate
analysis of this important molecule. Current methods do not
allow for fast, accurate, and real-time determination. Differ-
ent endogenous sulfide concentrations have been reported
with most publications suggesting the sulfide concentration in
blood is in the range of 10–100 mm.[24–29] There have been other
studies suggesting much lower sulfide concentrations.[30,31]
Therefore, new methods are needed for the efficient detection
of sulfide in biological systems.
With the idea of developing a new method that will be
useful for rapid assay of hydrogen sulfide concentrations
under physiological conditions, we undertook the effort of
searching for a selective chemosensing agent for this impor-
tant gasotransmitter. For easy use in a biology laboratory, the
chemosensing agent should 1) act fast (within seconds) under
mild conditions, 2) be chemically stable for long-term storage,
3) be sensitive for detection under near physiological con-
ditions, 4) show a linear concentration–signal relationship
within physiologically relevant hydrogen sulfide concentra-
tion ranges for easy quantitation, 5) show minimal or no
interference by other anions in the blood, and 6) be functional
in aqueous solutions and blood plasma. Herein, we report the
development of a fluorescent chemoprobe and its application
in the determination of hydrogen sulfide in aqueous solution,
serum, and whole blood.
Fluorescence is one of the most sensitive detection
methods. Thus we were interested in selecting a fluorophore,
which has a high quantum yield, emits at a long wavelength,
and responds to hydrosulfide by fluorescent property changes.
Dansyl is a commonly used fluorophore, and well-known for
its strong fluorescence and long emission wavelength. We
were interested in designing a sulfide-sensitive agent using
this fluorophore by taking advantage of the known unique
reduction of an azido group by hydrogen sulfide.[32] We
reasoned that the reduction of an azido group attached to a
strongly electron-withdrawing group would occur at an
accelerated rate. Because of the difference in electronegativ-
ity of the azido and amino groups and the added degree of
rotational freedom for the azido group, reduction of sulfonyl
azide into sulfonamide should trigger a change in the
electronic properties and thus the fluorescent properties of
the dansyl moiety.[33] Therefore, we synthesized dansyl azide
(DNS-Az, 2, Scheme 1).[34]
[*] H. Peng, Y. Cheng, Dr. C. Dai, Prof. B. Wang
Department of Chemistry, and Center for Biotechnology
and Drug Design, Georgia State University
P.O. Box 4098, Atlanta, GA 30302-4098 (USA)
E-mail: wang@gsu.edu
Homepage: http://chemistry.gsu.edu/Wang.php
A. L. King, Dr. B. L. Predmore, Prof. D. J. Lefer
Department of Surgery, Division of Cardiothoracic Surgery
Emory University School of Medicine, Carlyle Fraser Heart Center
550 Peachtree Street NE, Atlanta GA 30308-2247 (USA)
[**] We gratefully acknowledge financial support of this work by the
University Fellowship Program and the Molecular Basis of Disease
Program at Georgia State University through fellowships to H.P.
and Y.C., the support of the NIH through a grant (GM084933) to
B.W. and grants (5R01HL092141-03 and 5R01HL093579-02) to D.J.L
and to A.L.K (FIRST Fellowship Program of the NIH,
3K12GM000680-10S1) as well as funds from the Carlyle Fraser
Heart Center of Emory University Hospital Midtown.
DNS-Az (2) by itself is nonfluorescent. However, upon
addition of hydrogen sulfide the DNS-Az solution showed a
strong fluorescence enhancement, as expected (Na2S was
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201104236.
9672
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9672 –9675

that of sulfide (Figure 2b, for details see Figure S5 in the
Supporting Information). The response of DNS-Az (2) to
other reducing agents, such as thiophenol, benzyl mercaptan,
Scheme 1. DNS-Az (2) as a fluorescent probe for sulfide.
used as a hydrogen sulfide source in all experiments). The
addition of 25 mm hydrogen sulfide led to a 40-fold fluores-
cence enhancement in 20 mm sodium phosphate buffer
(pH 7.5) with 0.5% Tween-20 (Buffer/Tween). The detection
limit was as low as 1 mm with a signal-to-noise ratio (S/N) of
3:1. HPLC, MS, and NMR analysis confirmed that the
fluorescence increase was due to the formation of dansyl
amide (3, Scheme 1, see the Supporting Information for
spectral data).
Figure 2. Hydrogen sulfide concentration-dependent fluorescence
intensity changes determined using 96 well plates: DNS-Az 200 mm,
Na2S 0–100 mm in commercial bovine serum and buffer/Tween (excita-
tion filter 340 nm, emission filter 535 nm).
To study the selectivity of this chemoprobe for sulfide, the
fluorescent properties of DNS-Az (2) in the presence of
various anions were examined in buffer/Tween. No compa-
rable response was observed from other anions (Figure 1).
Because the detection is based on the reducing properties of
sulfide, other possible reducing anions such as iodide,
bromide, fluoride, bisulfite, and thiosulfate were also tested.
Overall 18 anions were screened and no obvious response was
observed for most of the anions at a concentration of 1 mm,
which is a 40-fold higher concentration than that of sulfide.
and cysteine was also tested. Benzyl mercaptan was the only
compound that induced strong responses (about 1/5 of that of
sulfide, see Figure S11 in the Supporting Information), which
could pose an interference problem. However, this finding
should not be a practical issue because benzyl mercaptan is
rarely found in biological systems. We also found that DNS-
Az was recalcitrant to the possible displacement reaction
resulting from attack by an amino group. DNS-Az showed
very limited response to glycine and lysine at concentrations
as high as 50 mm (see Figure S12 in the Supporting Informa-
tion).
The same results were observed for ClÀ at 150 mm, HCO3
À
and citrate at 100 mm, which are equal to or higher than the
physiological concentrations of these anions. None of the
other anions should have a concentration higher than 1 mm
under normal physiological conditions. Among all the anions,
2À
only HSO3À, S2O42À, and S2O5 led to some fluorescent
intensity increases. However, the extent of the fluorescence
increase was far smaller than that caused by sulfide even when
the concentrations of those anions were four-fold higher than
A linear relationship is always important for easy and
accurate analysis. Thus, the dependence of fluorescent
changes on hydrogen sulfide concentration was studied
using both a fluorometer and
a microplate reader. DNS-Az
(2) reacts with sulfide essen-
tially quantitatively even in
aqueous solution. The fluores-
cence intensity showed a repro-
ducible linear relationship in
buffer/Tween against hydrogen
sulfide (Figure 2). When the
sulfide concentration is higher
than that of DNS-Az (2), the
plot was found to reach a
plateau (see Figure S9 in the
Supporting
Information),
which means that the stoichi-
ometry of this reaction was 1:1.
Thus far, all the selectivity
and linearity studies suggest
that DNS-Az (2) can be used
for the determination of sulfide
Figure 1. a) Fluorescence intensity changes of DNS-Az (2) in solution after addition of sulfide and other
anions. b) Fluorescence spectra of DNS-Az (2) upon addition of various anions (DNS-Az 100 mm, ClÀ
150 mm, HCO3À, citrate 100 mm, HSO3À, S2O42À, and S2O52À 100 mm as well as HSÀ 25 mm and all other
anions at 1 mm in 20 mm sodium phosphate buffer (pH 7.5) with 0.5% Tween-20, lex =340 nm). Anions
tested: ClÀ, BrÀ, IÀ, FÀ, OHÀ, OAcÀ, CNÀ, N3À, NO2À, HCO3À, HSO3À, SO42À, S2O32À, S2O42À, S2O5
,
2À
HPO42À, and citrate.
Angew. Chem. Int. Ed. 2011, 50, 9672 –9675
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 9673

Communications
concentrations in a biological sample. For a further test, DNS-
Az (2) was evaluated in commercially available bovine serum.
Upon addition of sulfide, the solution of 2 in serum showed
very significant fluorescent intensity increases. Though
bovine serum showed background fluorescence (Figure 3
Figure 4. Determination of the hydrogen sulfide concentration in
blood using C57BL6/J mice (see Table 1) and 96 well plates (DNS-Az
200 mm, excitation filter 360 nm, emission filter 528 nm). The zero
point was obtained by trapping sulfide with ZnCl2 and the calibration
curve was obtained by using an internal standard method. (Hydrogen
sulfide concentrations 0, x, x+10, x+50, and x+100 mm).
Figure 3. Reaction time profile of DNS-Az (2) and hydrogen sulfide
(DNS-Az 100 mm, H2S 30 mm in bovine serum).
and Figure S8 in the Supporting Information at 0 s), it was
negligible compared to the strong fluorescence of dansyl
amide (3) obtained from the reaction (Figure 3). It should be
noted that the reaction went very fast in bovine serum
(complete within seconds, Figure 3 and Figure S8 in the
Supporting Information). This is very important considering
the fast metabolism and volatile nature of hydrogen sulfide in
biological systems. This unprecedented fast response could
provide the possibility of quantitative detection without any
pretreatment of samples. The reaction time profile was
studied in different solvent systems (see Figures S6 and S7
in the Supporting Information).
Table 1: Measurement of hydrogen sulfide concentrations in mouse
blood.
Mouse
c(H2S) [mm]
1
25.0
2
30.2
3
26.1
4
48.4
5
28.9
Average
31.9Æ9.4
An excellent linear relationship was also obtained in
bovine serum (Figure 2). The standard curve covers the range
of reported endogenous levels of hydrogen sulfide, indicating
that this probe is very suitable for the detection of sulfide in
biological samples. Due to changes in the microenvironment
and viscosity of the medium, fluorescence intensity and
emission wavelength vary when different media are used. A
standard addition procedure (internal spiking) was used for
accurate measurement of samples when the medium was not
readily available for generation of a calibration curve. Over-
all, the results suggest that the anions and biological
substrates normally encountered in the blood do not pose a
problem in the quantitative detection of hydrogen sulfide in a
biological sample.
Encouraged by the promising results obtained in buffer
and bovine serum, we applied this fluorescent chemoprobe to
the determination of hydrogen sulfide concentrations in
blood using the C57BL6/J mouse model (Figure 4 and
Figure S3 in the Supporting Information), which we have
been using in other related studies. A standard addition
procedure was used in the experiment. Five mice were used in
this study (Table 1). The average sulfide concentration in
these five mice was (31.9 Æ 9.4) mm, very similar to previous
reported values in mouse plasma (34.1 mm).[27] This has
confirmed that our fluorescent chemoprobe can indeed be
used in the detection of hydrogen sulfide in real biological
samples.
In conclusion, a novel reduction-sensitive fluorescence
chemoprobe was developed for hydrogen sulfide detection in
aqueous solutions, including blood serum and whole blood.
The probe was found to be very selective for sulfide among
18 anions tested and other common reducing species, with a
detection limit of 1 mm in buffer/Tween and 5 mm in bovine
serum with a signal-to-noise ratio of 3:1. The linear relation-
ship obtained in bovine serum covers the reported endoge-
nous concentration range of hydrogen sulfide. The probe was
used in the detection of hydrogen sulfide in blood using the
C57BL6/J mouse model. The result (31.9 Æ 9.4 mm) was very
close to the previously reported serum concentration of
hydrogen sulfide. In addition, this agent is characterized by
simplicity and ease in measurements, and compatibility with
96 well plates and microplate reader, which are readily
accessible in biology labs. Hydrogen sulfide concentrations
in biological systems are tightly regulated and can experience
rapid changes. The unprecedented fast response by DNS-Az
(2) to sulfide allows it to be used for the detection of transient
changes in hydrogen sulfide levels without sample pretreat-
ment. The probe, DNS-Az (2), is simple in structure, very easy
to synthesize, stable, and amenable to long-term storage. A
new research field has emerged in the past 15 years because of
the newly recognized significance of hydrogen sulfide as an
endogenous gasotransmitter. The molecular mechanism of
cellular actions of sulfide remains to be understood, and novel
H2S-releasing drugs need to be developed. We feel that this
9674 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 9672 –9675

[16] W. Lei, P. K. Dasgupta, Anal. Chim. Acta 1989, 226, 165.
[17] M. N. Hughes, M. N. Centelles, K. P. Moore, Free Radical Biol.
Med. 2009, 47, 1346.
[18] D. Jimꢁnez, R. Martinez-Manez, F. Sancenon, J. V. Ros-Lis, A.
Benito, J. Soto, J. Am. Chem. Soc. 2003, 125, 9000.
[19] J. E. Doeller, T. S. Isbell, G. Benavides, J. Koenitzer, H. Patel,
R. P. Patel, J. R. Lancaster, V. M. Darley-Usmar, D. W. Kraus,
Anal. Biochem. 2005, 341, 40.
fast, selective, efficient, and low-cost detection method for
sulfide will be very useful in the booming research field of
hydrogen sulfide.
Received: June 20, 2011
Revised: July 29, 2011
Published online: August 31, 2011
[20] D. G. Searcy, M. A. Peterson, Anal. Biochem. 2004, 324, 269.
[21] C. J. Richardson, E. A. M. Magee, J. H. Cummings, Clin. Chim.
Acta 2000, 293, 115.
[22] J. Radford-Knoery, G. A. Cutter, Anal. Chem. 1993, 65, 976.
[23] P. R. Bꢁrubꢁ, P. D. Parkinson, E. R. Hall, J. Chromatogr. A 1999,
830, 485.
[24] R. Hyspler, A. Ticha, M. Indrova, Z. Zadak, L. Hysplerova, J.
Gasparic, J. Churacek, J. Chromatogr. B 2002, 770, 255.
[25] Y.-H. Chen, W.-Z. Yao, B. Geng, Y.-L. Ding, M. Lu, M.-W. Zhao,
C.-S. Tang, Chest 2005, 128, 3205.
[26] H. Jiang, H. Wu, Z. Li, B. Geng, C. Tang, J. First Mil. Med. Univ.
2005, 25, 951.
[27] L. Li, M. Bhatia, Y. Z. Zhu, R. D. Ramnath, Z. J. Wang, F. B. M.
Anuar, M. Whiteman, M. Salto-Tellez, P. K. Moore, FASEB J.
2005, 19, 1196.
[28] L. R. Goodwin, D. Francom, F. P. Dieken, J. D. Taylor, M. W.
Warenycia, R. J. Reiffenstein, G. Dowling, J. Anal. Toxicol. 1989,
13, 105.
[29] J. C. Savage, D. H. Gould, J. Chromatogr. Biomed. 1990, 526,
540.
[30] N. L. Whitfield, E. L. Kreimier, F. C. Verdial, N. Skovgaard,
K. R. Olson, Am. J. Physiol. Regul. Integr. Comp. Physiol. 2008,
294, R1930.
Keywords: analytical methods · azides · fluorescent probes ·
hydrogen sulfide
.
[1] E. Culotta, D. E. Koshland, Science 1992, 258, 1862.
[2] T. Morita, M. A. Perrella, M. E. Lee, S. Kourembanas, Proc.
Natl. Acad. Sci. USA 1995, 92, 1475.
[3] D. J. Lefer, Proc. Natl. Acad. Sci. USA 2007, 104, 17907.
[4] W. M. Zhao, J. Zhang, Y. J. Lu, R. Wang, EMBO J. 2001, 20,
6008.
[5] J. W. Calvert, S. Jha, S. Gundewar, J. W. Elrod, A. Ramachan-
dran, C. B. Pattillo, C. G. Kevil, D. J. Lefer, Circ. Res. 2009, 105,
365.
[6] J. W. Calvert, M. Elston, C. K. Nicholson, S. Gundewar, S. Jha,
J. W. Elrod, A. Ramachandran, D. J. Lefer, Circulation 2010, 122,
11.
[7] J. W. Elrod, J. W. Calvert, J. Morrison, J. E. Doeller, D. W. Kraus,
L. Tao, X. Y. Jiao, R. Scalia, L. Kiss, C. Szabo, H. Kimura, C. W.
Chow, D. J. Lefer, Proc. Natl. Acad. Sci. USA 2007, 104, 15560.
[8] K. Abe, H. Kimura, J. Neurosci. 1996, 16, 1066.
[9] D. Boehning, S. H. Snyder, Annu. Rev. Neurosci. 2003, 26, 105.
[10] H. Kimura, Mol. Neurobiol. 2002, 26, 13.
[31] J. Furne, A. Saeed, M. D. Levitt, Am. J. Physiol. Regul. Integr.
Comp. Physiol. 2008, 295, R1479.
[32] F. Kazemi, A. R. Kiasat, S. Sayyahi, Phosphorus Sulfur Silicon
Relat. Elem. 2004, 179, 1813.
[33] A. P. de Silva, H. Q. N. Gunaratne, T. Gunnlaugsson, A. J. M.
Huxley, C. P. McCoy, J. T. Rademacher, T. E. Rice, Chem. Rev.
1997, 97, 1515.
[11] A. Martelli, L. Testai, M. C. Breschi, C. Blandizzi, A. Virdis, S.
Teddei, V. Calderone, Med. Res. Rev. 2011, DOI: 10.1002/
med.20234.
[12] P. Kamoun, M.-C. Belardinelli, A. Chabli, K. Lallouchi, B.
Chadefaux-Vekemans, Am. J. Med. Genet. Part A 2003, 116, 310.
[13] K. Eto, T. Asada, K. Arima, T. Makifuchi, H. Kimura, Biochem.
Biophys. Res. Commun. 2002, 293, 1485.
[34] A. Takadate, K. Tahara, H. Fujino, S. Goya, YakuGaku Zasshi
1986, 106, 36.
[14] P. K. Moore, M. Bhatia, S. Moochhala, Trends Pharmacol. Sci.
2003, 24, 609.
[15] R. Wang, FASEB J. 2002, 16, 1792.
Angew. Chem. Int. Ed. 2011, 50, 9672 –9675
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9675