Quenching and generation of singlet oxygen by hydroethidine and related chromophores

Article abstract of DOI:10.1016/j.cplett.2009.05.032

We have found that the oxidative probe hydroethidine (HE) is an efficient ¹O₂ quencher with k_q = 1.1 × 10⁹ M^-1 s^-1 in CH₃CN. The structurally related MitoSOX Red had k_q

Full text of DOI:10.1016/j.cplett.2009.05.032

Chemical Physics Letters 475 (2009) 116–119
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
Quenching and generation of singlet oxygen by hydroethidine
and related chromophores
P.J. Bilski * 1, B. Karriker, C.F. Chignell 2
,
Laboratory of Pharmacology, ETP, NIEHS, NIH, RTP, NC 27709, United States
a r t i c l e i n f o
a b s t r a c t
We have found that the oxidative probe hydroethidine (HE) is an efficient 1O2 quencher with
Article history:
Received 9 February 2009
In final form 14 May 2009
Available online 18 May 2009
9
À1 À1
9
À1 À1
k
q
= 1.1 Â 10 M
s
in CH
3
CN. The structurally related MitoSOX Red had k
q
= 2.1 Â 10 M
s
in
2 3
D O–CH CN containing DNA. Ethidium cation, the main oxidation product, and three model compounds,
5
-ethyl-5,6-dihydro-3,8-dinitro-6-phenyl-6-phenanthridinol, 3-amino-2-naphthenol, and 2-aminophe-
1
1
nol, were lesser
O
2
quenchers. Another product, 2-OH-ethidium cation, quenched
2
O with
8
À1 À1
k
q
= 5.5 Â 10 M
s
in a D O–CH CN mixture. Interestingly, both ethidium and 2-OH-ethidium cations
2
3
1
1
are weak
tively. Thus,
O
2
1
photosensitizers with quantum yields of
O
2
production in CH
3
CN of 0.06 and 0.03, respec-
O can be involved in photochemical experiments with MitoSOX Red and HE.
2
Ó 2009 Published by Elsevier B.V.
1
. Introduction
Hydroethidine (HE) and the mitochondrial probe MitoSOX Red
other deactivation modes of 1
O
2
. We have reported that 2-OH-E
and E were generated during the photosensitized oxidation of HE
1
[12], which may involve
O
2
since the HE chromophore appears
1
share the same core chromophore. Both are redox probes widely
used to investigate oxidative stress by detecting the presence of
reactive oxygen species (ROS) in cellular systems. They are useful
because the oxidation products of the HE chromophore fluoresce
strongly when intercalated into DNA, which makes them easy to
detect in cells. For a long time, the ethidium cation (E) was be-
lieved to be a sole fluorescent oxidation product. Lately, however,
a new (hydroxylated) product, the 2-OH-ethodium cation (2-OH-
E), was reported from oxidation by the superoxide radical anion
to be a very efficient O quencher. There is indeed a contribution
2
to the formation of the fluorescent products from photochemically
induced oxidation processes. However, the mechanism is compli-
cated depending on the photosensitizer used. In this study, we only
1
address the primary step of
2
O quenching, whose solid kinetics
provides an excellent characterization of the quenchers, often sug-
gesting that they may be also good substrates for O
our present investigation on the purely kinetic aspects of the
1
2
. We focused
1
O
2
quenching, leaving chemical oxidation mechanisms for future
studies.
[
1,2], which is claimed to be highly specific for in vitro superoxide
1
1
detection [3].
2 2
When oxidized, a O quencher may often become a O photo-
A strong red fluorescence emitted upon the excitation of DNA-
intercalated E and 2-OH-E is used to quantify oxidative stress in
sensitizer, and such a paradigm is at least partially true for the
main oxidation products of HE. Interestingly, we found that one
1
the intracellular milieu. Such stress is often initiated photochemi-
of them, 2-OH-E, also retains a very high
O
2
quenching ability.
1
1
cally with singlet oxygen ( O
2
) involvement. Oxidative stress in-
In the present work we have investigated the quenching of O
2
1
duced by
O
2
is likely to occur in some mammalian tissues such
by HE and related chromophores (see Scheme 1) in different sol-
as the skin and the eye [4,5]. Many other biological systems,
including plants [6–8] and marine habitat [9] where light and
vents, some containing DNA. We also tested E and 2-OH-E for
O photo-production.
2
1
photosensitizers are prevalent, may also be prone to (photo)oxida-
1
tive stress involving O
2
.
2
. Experimental
Previous studies investigated the formation of superoxide-in-
1
duced products including
O
2
-dependent controls and concluded
does not oxidize HE [10,11]. This suggests the absence
of any interaction, even though the reports did not investigate
Acetonitrile, Rose Bengal, perinaphthenone, deuterium chloride,
-ethyl-5,6-dihydro-3,8-dinitro-6-phenyl-6-phenanthridinol
1
that
O
2
5
(
EDDPP), ethidium bromide, 3-amino-2-naphthenol, and 2-amino-
phenol and calf thymus DNA type I were purchased from Sigma Al-
drich Chemical Co. (Milwaukee, WI, USA) and were used as received.
Deuterium oxide was from Cambridge Isotope Laboratories (Ando-
ver, MA, USA). MitoSOX Red was from Invitrogen Corporation
(Carlsbad, CA, USA). Hydroethidine was purchased from Tyger Sci-
*
Corresponding author.
E-mail address: bilski@niehs.nih.gov (P.J. Bilski).
NIEHS contract to SRA-International.
1
2
This Letter is dedicated to the memory of Dr. Colin Chignell.
0
009-2614/$ - see front matter Ó 2009 Published by Elsevier B.V.
doi:10.1016/j.cplett.2009.05.032

P.J. Bilski et al. / Chemical Physics Letters 475 (2009) 116–119
117
NH2
NH2
NO2
N
N
N
NH2
NH2
NO2
(
CH2)6
P+(C6H5)3
OH
5
-Ethyl-5,6-dihydro-3,8-dinitro-
Hydroethidine
MitoSOX Red
OH
-6-phenyl-6-phenanthridinol
HE
EDDPP
NH2
NH2
OH
N+
N+
NH2
NH2
NH2
3
-Amino-2-naphthenol
OH
Ethidium cation
2-OH-Ethidium cation
NH2
E
2
-OH-E
2-Aminophenol
Scheme 1.
entific Inc. (Ewing, NJ, USA). 2-OH-ethidium was synthesized and
purified as described [13]. De-ionized water was used to prepare
aqueous solutions. The samples were prepared in a suprasil fluores-
1O
was produced by single pulse excitation (duration, 20 ns) of
2
RB [16] at 532 nm, or of perinaphthenone [17] at 355 nm. Photo-
sensitizers were dissolved in air saturated solvent mixtures in a
1
cence cuvette (0.5 cm pathlength) and were air-equilibrated.
0.5 cm pathlength suprasil fluorescence cuvette. O
2
lifetimes were
1
1
Singlet oxygen lifetime was measured using a time-resolved O
2
2
calculated from monoexponential decays of O phosphorescence.
1
spectrometer as described previously [14,15]. Briefly, the instru-
ment utilized a Surelite II laser (Continuum, Santa Clara, CA) for
2
At higher quencher concentrations, when the O signals were
weak, some decays could deviate from a perfect exponential decay
mode. However, their best approximations were still exponential.
Deviations from exponential decays may resulted in a lower corre-
lation coefficient for the lines used to extract the rate constants
from the experimental points, which accounts for the experimental
excitation and a germanium diode (Model 403 HS, Applied Detec-
1
tor Corporation, Fresno, California) for
2
O detection. An HP
5
4111D Digitizing Oscilloscope (Hewlett Packard, Colorado
Springs, CO) interfaced with a PC was used for data acquisition.
1
All solutions for O
2
quenching experiments were freshly prepared
q
error of k determination (Table 1). At higher quencher concentra-
1
before experiments, which were performed at room temperature.
tions, a decrease in the
O
2
signal amplitude was also observed.
Table 1
Rate constants for 1
O
quenching by HE and related compounds.
Solvent
2
Compound
Rate constant
Photosensitizer DNA concentration
À1 À1
8
(
%, vol:vol)
MeCN
0% MeCN 50% D
M
s
 10
6.2 ± 0.2
21.2 ± 1
5.2 ± 0.9
MitoSOX Red
Hydroethidine
RB
5
2
O
O
P; DNA, 1 mg/ml
P
1
MeCN
0% MeCN 80% D
10.9 ± 0.4
10.1 ± 0.3
RB
RB
2
2
9.3 ± 0.3
RB; DNA, 1.6 mg/ml
Propylene carbonate
Ethanol
Methanol
8.4 ± 0.2
3.1 ± 0.1
3.3 ± 0.3
31 ± 2.0
2.1 ± 0.3
RB
RB
RB
P
5
5
0% MeCN 50% D
0% MeCN 50% D
2
O
2
O, 500 mM DCl
P
Ethidium Bromide
MeCN
0% MeCN 70% D
0.254 ± 0.008
0.23 ± 0.01
P
3
2
O
O
P; DNA, 1.7 mg/ml
P
0.66 ± 0.05
2
-OH-Ethidium
MeCN
0% MeCN 50% D
3.4 ± 0.2
RB
RB
5
2
5.54 ± 0.06
0.062 ± 0.005
0.67 ± 0.03
*
EDDPP
MeCN
MeCN
MeCN
RB
RB
3
-Amino-2-naphthenol
-Aminophenol
2
0.53 ± 0.01
1.01 ± 0.03
RB
RB
2
2
0% MeCN, 80% D O
*
EDDPP, 5-ethyl-5,6-dihydro-3,8-dinitro-6-phenyl-6-phenanthridinol; RB, Rose Bengal; P, perinaphthenone.

1
18
P.J. Bilski et al. / Chemical Physics Letters 475 (2009) 116–119
This decrease is most likely caused by the interaction of the photo-
-0.40
1
sensitizer triplet with the quenchers, which does not preclude O
2
1
quenching experiments. At high quencher concentrations, all
O
2
-0.45
-0.50
-0.55
-0.60
photosensitizers are likely to be quenched in their triplet state.
1
However, the time-resolved detection of
2
O phosphorescence al-
lows the observation of the selective interaction of the quencher
1
1
2 2
with O , as long as the O phosphorescence is strong enough to
be registered. The observed rate constants (the inverse of the ob-
served phosphorescence lifetimes) from the samples containing
different concentrations of quencher were plotted vs. quencher
concentrations yielding a linear plot of the Stern–Volmer equation,
1
2
k
obs = k
o
q
+ k [Q].
Singlet oxygen phosphorescence spectra were measured as de-
3
1
-0.65
scribed previously [18], using a steady-state O
featuring an optimized optical system as in our pulse O
photometer [14]. Briefly, the apparatus utilized a germanium diode
Model 403 HS, Applied Detector Corporation, Fresno, California), a
2
laser spectrometer
1
2
spectro-
-0.70
(
-5
-4
-4
0
.0
5.0x10
1.0x10
1.5x10
light chopper with a monochromator in conjunction with an effi-
cient optical system, and a digital lock-in amplifier (Stanford Re-
search, Sunnyvale, CA) for signal detection. Samples were excited
from a 500 W Hg lamp operating at 300 W through a 546 nm inter-
Time (s)
Fig. 1. Examples of 1
mixtures containing 1.7 mg/mL DNA and increasing concentrations of
ethidium bromide (0 M, 250 M, 500 M). Perinaphthenone was used as a O2
2
O phosphorescence decay [1–3] in 30:70 (%, vol:vol), MeCN:
2
D O
ference filter in combination with a KG3 heat-removing glass filter.
l
l
l
1
1
The
O
2
phosphorescence spectra were recorded during one ca.
photosensitizer excited at 355 nm by a single laser pulse.
3
0 s scan over a range of 1200–1350 nm and were normalized to
the same number of absorbed photons at the excitation wave-
length. Absorption spectra were acquired using an HP Diode Array
Spectrophotometer model 8452A (Hewlett Packard Co., Palo Alto,
HE
CA). The quantum yields of 1
O
2
production by ethidium bromide
4
8
6
4
.0x10
.0x10
.0x10
and 2-OH-ethidium cation were calculated using RB as a reference.
1
4
4
4
1 +DNA
3
. Results and discussion
EB
Hydroethidine and the mitochondrial probe MitoSOX Red are
widely used cellular fluorescence probes to investigate oxidative
2
stress [3,19–21], which can be induced by photochemical means
1
2.0x10
involving
O
2
[22]. Thus, the primary focus of this work is the
O with HE, which is also an active chromophore
2
2 + DNA
1
interaction of
-
5
-4
-4
-4
-4
in MitoSOX Red. We have included in our investigations the main
oxidation products ethidium (E) and 2-OH-ethidium (2-OH-E)
cations, and three structurally related compounds that we used
0.0
5.0x10 1.0x10 1.5x10 2.0x10 2.5x10
Compounds (M)
Fig. 2. Quenching of 1
Observed first order rate constants for O2 quenching as a function of EB and HE
concentrations. HE solutions contained 20:80 (%, vol:vol), MeCN:D O HE solutions,
without DNA [1] and in the presence of 1.6 mg/mL DNA (1 + DNA). Ethidium
bromide solutions contained 30:70 (%, vol:vol), MeCN:D O mixture, without DNA
2] and in the presence of 1.7 mg/mL DNA (2 + DNA). Rose Bengal was used as a O
2
as models (Scheme 1). The main oxidation products of HE turned
2 2
O by hydroethidine and ethidium bromide in D O solution.
1
1
out to be weak O
2
photosensitizers (vide infra).
2
1
3
.1. O
2
quenching
2
1
[
photosensitizer excited at 532 nm by a single laser pulse. Corresponding rate
constants are shown in the Table 1.
The most accurate and virtually artifact-free way to investigate
1O
1
2
interaction with diverse molecules is time-resolved IR O
2
1
2
phosphorescence. Fig. 1 shows examples of O decay in solutions
containing increasing concentrations of ethidium bromide. When
Different solvents may affect the 1O2 quenching [23]. Several
solvents and solvent combinations were used for the quenching
experiments with HE, showing that the quenching was generally
more efficient in polar and protic milieus. In aprotic but polar pro-
pylene carbonate, the quenching was comparable to that in MeCN
plotted vs. quencher concentration, the observed rate constants
1
for the O
2
decays yield straight lines whose slopes are the quench-
) values. Fig. 2 shows an example of how the
ing rate constant (k
q
quenching rate constant was obtained in a MeCN:EtOH, 2:8 vol:vol
solvent mixture containing HE and DNA in comparison to E bro-
1
mide. All the k
q
values for O
2
quenching by HE and related com-
(Table 1). Even though the lowest k values were in methanol and
q
1
pounds are compiled in Table 1.
ethanol, those values still make HE a very efficient O2 quencher,
Because cellular imaging experiments with the HE chromo-
phore involve its interaction with DNA, we examined how the
presence of DNA may affect the quenching. The presence of DNA
comparable to the azide anion [24]. The kq were generally higher
in solvent mixtures containing D O and depended slightly on the
2
9
À1 À1
photosensitizer used. The highest value, k = 2.1 Â 10 M
q
s ,
did not significantly decrease the k
q
values for HE (Fig. 2), while
was observed for MitoSOX Red (Table 1), making it a superb mito-
chondrial quencher, close to the diffusion limit, and comparable in
effectiveness to the carotenes. Because different solvents model
well the aqueous and organic cellular compartments, our data sug-
we observed a slight increase for MitoSOX Red (Table 1). DNA itself
did not show any quenching (data not shown). These results show
that both HE and MitoSOX Red can quench 1
2
O very efficiently
1
regardless of the presence or absence of DNA, which suggests that
the same thing may happen in the cellular environment.
gest that the HE moiety is a very efficient O2 quencher indepen-
dent of cellular localization.

P.J. Bilski et al. / Chemical Physics Letters 475 (2009) 116–119
119
1
interaction with O
2
as compared to the ethidium cation that can-
5
5
4
4
4
4
1
1
8
6
4
2
.2x10
.0x10
.0x10
.0x10
.0x10
.0x10
1
not deprotonate at this pH. Since the increased O
2
quenching may
9
-1 -1
k = (3.1 +/- 0.2) x 10 M s
q
2
also relate to the –OH and –NH moieties in ortho configuration, we
1
included in our investigation two simpler aromatic compounds
bearing both groups (Scheme 1). The quenching by 3-amino-2-
naphthenol and 2-aminophenol was ca. one order of magnitude
lower, suggesting that the HE core is equally essential for efficient
1
2
O quenching. This observation emphasizes the uniqueness of the
1
+ DCl
molecular structure of the HE moiety in terms of its efficient inter-
1
2
action with O .
8
-1 -1
k = (2.1 +/- 0.3) x 10 M s
q
1
3.2. O
2
photosensitization
-
5
-4
-4
0
.0
6.0x10
1.2x10
1.8x10
Oxidation transforms HE from a potent 1O
2
quencher into a
1
Hydroethidine (M)
weak O
can photosensitize O
2
photosensitizer, as both E and, to lesser extend, 2-OH-E
1
2
. Fig. 4 shows the steady-state spectrum of
Fig. 3. Quenching of 1
Plots of observed first order rate constants for
concentration in 50:50 (%, vol:vol) MeCN:D O mixtures in the absence [1] of DCl
O
by neutral and protonated forms of HE in D
O solution.
1
2
2
O
2
phosphorescence generated in MeCN by E in comparison to
1
2
O quenching as a function of HE
the RB standard. The quantum yields are 0.06% for E and 0.03%
for 2-OH-E. While these quantum yields are low, they may be suf-
ficient to induce a mild phototoxicity during prolonged fluores-
cence imaging, which, to our knowledge, was largely unknown
and rarely considered before.
2
q
and in the presence (1 + DCl) of 0.5 M DCl. Calculated rate constants, k , are shown
in the figure.
1
Rose Bengal
5
4
3
2
1
0
4. Conclusions
We have shown that the HE moiety is a very efficient 1
quencher, and that its main oxidation products, ethidium and 2-
O
2
1
OH-ethidium cations, are weak
O
2
photosensitizers. Our findings
1
show that O
2
can be involved in photochemical experiments with
MitoSOX Red and HA, also bringing to attention the possibility of
weak photosensitization that may be exerted by ethidium bromide
during DNA fluorescence imaging.
Ethidium Bromide
2
Acknowledgment
This research was supported by the Intramural Research
Program of the NIH, NIEHS.
1200
1230
1260
1290
1320
1350
λ (nm)
Fig. 4. Production of 1
Phosphorescence spectra, 1, of Rose Bengal (50
100 M). The compounds were excited using an interference filter transmitting at
46 nm.
O
by ethidium bromide compared to Rose Bengal in MeCN.
M) and 2, ethidium bromide
References
2
l
(
5
l
[1] H.T. Zhao, S. Kalivendi, H. Zhang, J. Joseph, K. Nithipatikom, J. Vasquez-Vivar, B.
Kalyanaraman, Free Radic. Biol. Med. 34 (2003) 1359.
[
2] H.T. Zhao, J. Joseph, H.M. Fales, E.A. Sokoloski, R.L. Levine, J. Vasquez-Vivar, B.
Kalyanaraman, Proc. Natl. Acad. Sci. USA 102 (2005) 5727.
The efficacy of 1
O
quenching depends greatly on the quench-
[3] J. Zielonka, J. Vasquez-Vivar, B. Kalyanaraman, Nat. Protoc. 3 (2008) 8.
2
[
[
[
4] S. Beatty, H.H. Koh, D. Henson, M. Boulton, Surv. Ophthalmol. 45 (2000) 115.
5] K. Scharffetter-Kochanek et al., Exp. Gerontol. 35 (2000) 307.
6] I.M. Moller, P.E. Jensen, A. Hansson, Annu. Rev. Plant Biol. 58 (2007) 459.
er’s molecular electron density. We exploited protonation as a sim-
ple way to decrease electron density for the HE chromophore. We
also used a model compound in which the two –NH
2
groups in the
[7] M.E. Daub, M. Ehrenshaft, Annu. Rev. Phytopathol. 38 (2000) 461.
[
[
8] J.P. Knox, A.D. Dodge, Phytochemistry 24 (1985) 889.
9] S.L.H. Sandvik, P. Bilski, J.D. Pakulski, C.F. Chignell, R.B. Coffin, Mar. Chem. 69
HE molecule are oxidized to –NO groups that are electron-with-
2
drawing (EDDPP, Scheme 1). Both the protonation and the change
(
2000) 139.
in substituent should affect the electron density distribution in the
[10] A.V. Bizyukin, L.G. Korkina, B.T. Velichkovskii, Bull. Exp. Biol. Med. 119 (1995)
347.
1
core HE chromophore and resultant O
2
quenching. When we mea-
in 50% MeCN, 50% D O solu-
tion of HE in the absence and presence of deuterium chloride
Fig. 3), we found that HE protonation did indeed exert a profound
effect on the quenching. The expected lower electron density in
protonated HA and in EDDPP lowered the 1
quenching efficacy
by about an order of magnitude, making the k values similar to
[
[
11] V.P. Bindokas, J. Jordan, C.C. Lee, R.J. Miller, J. Neurosci. 16 (1996) 1324.
12] P. Bilski, B. Karriker, E.F. DeRose, C.F. Chignell, Free Radic. Biol. Med. 39 (2005)
S07.
1
sured quenching rate constant of O
2
2
[
[
13] J. Zielonka, H.T. Zhao, Y.K. Xu, B. Kalyanaraman, Free Radic. Biol. Med. 39
(
(
2005) 853.
14] P. Bilski, C.F. Chignell, J. Biochem. Biophys. Meth. 33 (1996) 73.
O
2
[15] P. Bilski, M.Y. Li, M. Ehrenshaft, M.E. Daub, C.F. Chignell, Photochem. Photobiol.
1 (2000) 129.
7
q
[
[
16] P. Bilski, R. Dabestani, C.F. Chignell, J. Phys. Chem. 95 (1991) 5784.
17] R. Schmidt, C. Tanielian, R. Dunsbach, C. Wolff, J. Photochem. Photobiol. a-
Chem. 79 (1994) 11.
those observed for E and 2-OH-E (Table 1). These data show that
the positive electric charge acquired by the HE either by proton-
ation or during oxidation is a crucial factor in determining the effi-
[18] P. Bilski, B.Z. Zhao, C.F. Chignell, Chem. Phys. Lett. 458 (2008) 157.
[19] K.M. Robinson, M.S. Janes, J.S. Beckman, Nat. Protoc. 3 (2008) 941.
1
cacy of interaction with O
2
.
[
[
20] K.M. Robinson et al., Proc. Natl. Acad. Sci. USA 103 (2006) 15038.
21] J. Zielonka et al., Free Radic. Biol. Med. 44 (2008) 835.
While the above conclusion is generally true, one partial excep-
tion is 2-OH-E which, despite its expected cationic nature, we
[22] A. Krieger-Liszkay, C. Fufezan, A. Trebst, Photosynth. Res. 98 (2008) 551.
[23] P. Bilski, R.N. Holt, C.F. Chignell, J. Photochem. Photobiol. a-Chem. 109 (1997)
1
found to be quite an efficient O
2
quencher, especially when com-
243.
pared to E (Table 1). However, at neutral pH, 2-OH-E can be par-
tially in a zwitter-ionic form, which can lead to a stronger
[
24] M.Y. Li, C.S. Cline, E.B. Koker, H.H. Carmichael, C.F. Chignell, P. Bilski,
Photochem. Photobiol. 74 (2001) 760.