Visualization of mitochondrial membrane potential and reactive oxygen species via double staining.

Abstract

Quantitative and qualitative analysis of both generated reactive oxygen species (ROS) and mitochondrial membrane potential cannot be detected simultaneously. We here introduce a simple, new double staining method. We have successfully used this for several years utilizing cerium for ROS detection and JC-1 staining to assess the mitochondrial membrane potential. The resultant signals on laser confocal images can be localized in the same cells and can easily quantify them. We used a confocal microscope along with our new, combined staining method to both visualize mitochondrial membrane potential (DeltaPsim) and imaged ROS. These were quantified by JC-1 staining and by cerium ions with reflectance in a method modified in our laboratory. To test this double labeling technique we used PC 12 cells subjected to 1 h hypoxia and 24h re-oxygenization. We are able to produce a quantitative analysis of red/green signals of JC-1 that reflected the energy state of the cells. Cerium reflectance correlates with the amount of ROS release in the same cells. Significant differences have been calculated after hypoxia and re-oxygenation in both modality of the cell staining. The red/green ratio was 18.2+/-9.3 (n=30) in normoxic cells versus 1.65+/-0.9 (n=30) in the hypoxia/re-oxygenation group (p<0.05). In the same randomly selected cells the average cerium reflectance signal intensity was 2.5+/-1.2 (n=30) in the control group while 5.8+/-3.1 (n=30) in the hypoxia/re-oxygenation group (p<0.05). This assay, by characterizing hypoxic injury and re-oxygenization induced ROS production, offers a qualitative and quantitative method to detect the consequences of oxidative stress in experimental conditions and to detect different cell protective strategies.

Full text

Page 1

Neuroscience Letters 399 (2006) 206–209
Visualization of mitochondrial membrane potential and reactive
oxygen species via double staining
G´ eza Szil´ agyia,∗, L´ aszl´ o Simona, P´ eter Koskaa, G´ eza Telekb, Zolt´ an Nagya
aNational Stroke Center, Department of Vascular Neurology, Semmelweis University, Budapest, Hungary
bIIIrd Department of Surgery, Semmelweis University, Budapest, Hungary
Received 19 November 2005; received in revised form 30 January 2006; accepted 31 January 2006
Abstract
Quantitativeandqualitativeanalysisofbothgeneratedreactiveoxygenspecies(ROS)andmitrochondrialmembranepotentialcannotbedetected
simultaneously. We here introduce a simple, new double staining method. We have successfully used this for several years utilizing cerium for
ROS detection and JC-1 staining to assess the mitochondrial membrane potential. The resultant signals on laser confocal images can be localized
in the same cells and can easily quantify them. We used a confocal microscope along with our new, combined staining method to both visualize
mitochondrial membrane potential (?Ψm) and imaged ROS. These were quantified by JC-1 staining and by cerium ions with reflectance in a
method modified in our laboratory. To test this double labeling technique we used PC 12 cells subjected to1h hypoxia and 24h re-oxygenization.
We are able to produce a quantitative analysis of red/green signals of JC-1 that reflected the energy state of the cells. Cerium reflectance correlates
with the amount of ROS release in the same cells. Significant differences have been calculated after hypoxia and re-oxygenation in both modality
of the cell staining. The red/green ratio was 18.2±9.3 (n=30) in normoxic cells versus 1.65±0.9 (n=30) in the hypoxia/re-oxygenation group
(p<0.05). In the same randomly selected cells the average cerium reflectance signal intensity was 2.5±1.2 (n=30) in the control group while
5.8±3.1 (n=30) in the hypoxia/re-oxygenation group (p<0.05). This assay, by characterizing hypoxic injury and re-oxygenization induced ROS
production, offers a qualitative and quantitative method to detect the consequences of oxidative stress in experimental conditions and to detect
different cell protective strategies.
© 2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Mitochondrial membrane potential; Reactive oxygen species; JC-1; Fluorescence; Visualization; Confocal microscopy
Reactiveoxygenspeciesplayapivotalroleindifferentneurode-
generative diseases and in ischemic/hypoxic conditions. The
reactive oxygen species (ROS) produced by mitochondria is
an important factor in cell necrosis and/or apoptosis. By the
detection and quantification of ROS we are able to identify an
energy crisis of the effected cells and evaluate the efficacy of
a range of possible drug intervention strategies. ROS detection
is technically difficult, since they react quickly and readily with
othermolecules.Chemiluminescence’sassays[3],fluorescence-
based analysis, enzymatic assays [1], and electron spin trapping
methods[14]arefrequentlyused,buttheyfailtoprovideagood
visualization in individual cells. A recently developed method
of ROS analysis [9], offers a high standard visualization by use
∗Correspondenceat:NationalInstituteofPsychiatryandNeurology,National
Stroke Center and Department of Vascular Neurology, Semmelweis University,
H˝ uv¨ osv¨ olgyi ´ ut 116, H-1021 Budapest, Hungary. Tel.: +36 1 391 5331;
fax: +36 1 391 5440.
E-mail address: szilagyi@radi.sote.hu (G. Szil´ agyi).
of a confocal microscope. Utilizing another fluorescence mito-
chondrial staining method could provide a greater amount of
information, which reflects the membrane potential of mito-
chondria [12]. This combined staining method provides both
qualitative and quantitative data on the cellular level.
To test our new, combined staining method, we used a PC
12 cell culture. PC 12 cells were grown in Dulbecco’s modi-
fied Eagle’s medium (DMEM) with 10% calf serum, 5% horse
serum, 2mM l-glutamine, and antibiotics. Nerve growth factor
(NGF) was also a component in the tissue culture medium. The
cells were grown on glass (d=12mm) on collagen.
In vitro hypoxia was produced by Argon gas using the
method published by Kusumoto et al. [6]. Following 1h of
oxygen deprivation the cultures (n=6) were returned to the
normal condition for 24h (re-oxygenization) [5]. Control cul-
tures (n=6) were maintained in the incubator. A blood-gas
analyzer (ABL Radiometer, Copenhagen) was used to mea-
sure oxygen pressure in the medium fluid. The partial oxygen
pressure was 154.65±1.35mmHg (n=6) in normoxic, and it
0304-3940/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.neulet.2006.01.071

Page 2

G. Szil´ agyi et al. / Neuroscience Letters 399 (2006) 206–209
207
was 51.96±2.53mmHg (n=6) in hypoxic conditions (aver-
age±S.E.M.).
Cultures were stained with propidium iodide (PI) (1.5?g/ml
insaline)for2mintolabeldeadcells.Viabilitywascalculatedon
randomly selected visual field using a fluorescence microscope
with 490nm excitation and 520nm barrier filters. Cell-death
was expressed as a mean of percentage of PI labeled cells in 3
separate cultures in 12 samples.
The new, combined staining procedure here follows. DMEM
was removed and 300?l of JC-1 in a concentration of 10?g/ml
was dissolved in physiological saline added to the cultures for
10min. Afterwards, the cell cultures were rinsed with saline.
Thenthecellswerestainedwith300?lof20mmol/lCeCl3solu-
tion(dissolvedinlactatedRinger)for2min.TheCeCl3solution
was removed, and then the cell culture was rinsed with saline
again. Next, the cells were fixed in a 0.25% buffered glutaralde-
hyde solution for 2min. The fixed cells on glass were covered
with the Vectashield mounting medium for fluorescence study
(Vector Laboratories, Inc. Burlingame, CA) and put on glass
slides.
Intracellulardistributionofthedyeswasassessedbyconfocal
microscope. (BIO-RAD MRC 1024 confocal system, Bio-Rad
Corp., Hertfordshire, England) on a Nikon OPTIPHOT inverted
microscope (Donsanto Corp., Nattick, Massachusetts).
For simultaneous visualization – of the green fluorescence
from JC-1 monomer and the red fluorescence from JC-1 J-
aggregate – a long pass filter system was used (excitation
lines 488 and 568nm, T1 trichroic and T2a 560 DRLP mir-
ror, 522 DF 35 and 585 LP emission filter set). For detection
of cerium labeled cells, single channel detection was utilized in
the reflectance mode (excitation line 488nm of Krypton-Argon
laser,B1blockandT1trichroicmirror,Bluereflectionfilterset).
Fixed exposure time and laser intensities were utilized during
all acquisitions.
Forquantitativeanalysishigh-resolution(820×)imageswere
taken.Thequantitativeanalysisofreflectanceofceriumcomplex
and red/green fluorescence signal was detected and the built-in
evaluation software was used for calculation (LaserSharp Pro-
cessing Bio-Rad Corp., Hertfordshire, England). The average
intensities of reflectance (reflecting cerium precipitates), the
ratio of green (JC-1 monomers) and red fluorescence (JC-1 J-
aggregates) were measured within randomly selected cells [10].
In both the experimental and control group there were 30 indi-
vidual cells were measured in six identical cultures. All data
was statistically evaluated using the Mann–Whitney test and the
Kolgomorov-Smirnov test. The intensity curve was made by
SPSS 12.0 (LEAD Technologies, US) statistical program. Data
was considered to be statistically significant if p<0.05.
The aggregated form of JC-1 molecules (red fluorescence)
accumulated in functional mitochondria. On the other hand, the
green fluorescence signal from JC-1 monomers is evenly dis-
tributed in the cytoplasm (Fig. 1A and B). In the control groups,
the red/green ratio was 18.2±9.3, while in the hypoxia/re-
oxygenization group this ratio was as low as 1.65±0.9. This
difference was significant (p<0.05). The cerium reflectance
appeared as fine red signals in the cells. The distribution and
location of ROS was evaluated in the same cells where JC-1
signalsweredetected(Fig.1CandD).Theaveragesignalinten-
sity was 2.5±1.2 (n=30) in the control group while it was as
high as 5.8±3.1 (n=30) in the hypoxia/re-oxygenation group
(p<0.05).
There was an inverse relationship between ROS signal and
JC-1 ratio. A low JC-1 ratio means that there will be a low
amount of the aggregated form of JC-1 in the mitochondria
and this correlates with a high amount of ROS (Fig. 2). Our
double-labeling data correlated well with the PI staining which
representsthecellinjurylevel.ThepercentageofthePIpositive
cells after hypoxia/re-oxygenization was about double as it was
inthenormoxiagroup(43.5±3.2%versus18.7±2.3%(n=12,
p<0.05)).
We used a combination of non-toxic JC-1 staining to detect
mitochondrial membrane integrity and cerium to visualize ROS
Fig. 1. High power confocal micrographs represent cerium reflectance and JC-1 fluorescence in PC 12 cells. (A, C) images visualized identical cells from normoxic
control. (A) demonstrates the JC-1 overlay image of 522 and 585nm fluorescence. The red dominance reflects the aggregated form of JC-1, the consequence of
well-preserved mitochondrial membrane potential. (C) Cerium reflectance red signal is contrasted by green signal of JC-1 monomers for better visualization. (B, D)
shows identical PC 12 cells from hypoxia/re-oxygenated group. In (B) there is an overlay image of JC-1. The green dominance reflects the collapsed mitochondrial
membrane potential. In (D) the high intensity of red signal correspond to the high ROS production. (A, B, C, D ×420).

Page 3

208
G. Szil´ agyi et al. / Neuroscience Letters 399 (2006) 206–209
Fig. 2. The relationship between the level of cerium reflectance and red/green
ratio of JC-1. The coherent value could describe with an inverse curve (F=238,
p<0.01, ANOVA). The linear regression model was significant too (F=51,
p<0.01, ANOVA).
releasedafterhypoxia/re-oxygenization.Thecombinedstaining
technique resulted in well-defined fluorescence signals which
could be quantified.
Mitochondrial dysfunction is a final common denominator
in several chronic neurodegenerative diseases and in ischemic
brain injury. In ischemic lesions, ROS are the most detrimental
factor for cells. Our combined staining method can characterize
the cell lesion in a quantitative manner.
JC-1 is a non-toxic fluorescence probe to monitor membrane
potential. This dye has no affect on living cells, including their
respiration [8]. JC-1 monomers accumulate selectively in mito-
chondria and subsequently aggregate as a consequence of the
membrane potential. This reaction is pH independent within the
physiological range. The electrochemical gradient is responsi-
ble for this J aggregation [12]. At the low level transmembrane
potential low JC-1 concentration can be detected in an aggre-
gated form, which emits red fluorescence. In contrast JC-1 at
high concentrations forms aggregates. The red/green ratio is an
indicator of ?Ψ, which is independent of dye-loading, light-
scattering, and other optical factors. During re-oxygenation, the
?Ψ falls down significantly as this phenomenon has been doc-
umented by others [10].
With the cerium method, we could visualize ROS at the cel-
lular level. This semi-quantitative method is based on a change
in the level of cerium reaction products resulting in oxidases
and phosphatases [9]. It has been demonstrated that the highest
reflectance intensities are accompanied by cerium perhydroxide
(CeIII[OH]2OOHorCeIV[OH]3OOH)[4,15].Theceriumreacts
not only with the ROS, but it also reacts with phosphate deriva-
tives. These cerium phosphate derivatives do not have a relevant
level of reflectance using a confocal laser scanning microscope
[16]. The quantitative analysis of pixel-by-pixel measurement –
of dye distribution present in the cells subjected to hypoxia/re-
oxygenation and in control cells – showed a significant increase
in cerium signals present in the cells after their hypoxic insult
as compared to controls.
In vivo, ROS are produced by enzymatic (i.e., NADPH oxi-
dase, cytochrome P450dependent oxygenase, xantine oxidase)
and by nonenzymatic routes. The mitochondrial electron trans-
portchaincontainsseveralredoxcentersthatmayleakelectrons
tooxygen(i.e.,complexI,III),constitutingtheprimarysourceof
ROSinmosttissues.MitochondrialROSproductionisincreased
by a high level of reduced electron carriers and by a large ?µH+
in mitochondrial membrane.
The reduction of membrane potential is accompanied by a
decrease in H2O2 production in isolated mitochondria [13].
On the other hand, in cells after hypoxia/re-oxygenation the
mitochondrial membrane potential falls and ROS production
increases. This could occur by complex I inhibition [2] or
oxidant-mediated disruption of membrane integrity [7,17].
In our experiment we were able to simultaneously measure
these two modalities in the same cells. This relationship has
an inverse feature described with this equation: y=7.41x−0.47,
where y is the cerium signal and x is the JC-1 ratio.
This novel double staining method is easily reproducible.
Furthermore, it is a good tool for screening neuroprotective,
ROS blocking, drug-candidate molecules [11].
Acknowledgements
ThisstudywassupportedbytheHungarianNationalScience
Foundation (OTKA) 2001 T-037887 and ETT 096/2003 grants.
References
[1] Y. Cao, P. Guo, Y. Xu, B. Zhao, Simultaneous detection of NO and ROS
by ESR in biological systems, Methods Enzymol. 396 (2005) 77–83.
[2] C. Chinopoulos, L. Tretter, V. Adam-Vizi, Depolarization of in situ mito-
chondria due to hydrogen peroxide induced oxidative stress in nerve
terminals: inhibition of a-ketoglutatare dehydrogenase, J. Neurochem.
73 (1999) 220–228.
[3] T.J. Guzik, K.M. Channon, Measurement of vascular reactive oxygen
species production by chemiluminescence, Methods Mol. Med. 108
(2005) 73–89.
[4] K.J. Halbhuber, C. Scheven, G. Jirikowski, H. Feuerstein, U. Ott,
Reflectance enzyme histochemistry (REH): visualization of cerium-based
and DAB primary reaction products of phosphatases and oxidases in
cryostat sections by confocal laser scanning microscopy, Histochem.
Cell Biol. 105 (1996) 239–249.
[5] K. Kovecs, K. Komjati, T. Marton, J. Skopal, P. Sandor, Z. Nagy, Hyper-
capnia stimulates prostaglandin E(2) but not prostaglandin I(2) release
in endothelial cells cultured from microvessels of human fetal brain,
Brain Res. Bull. 54 (2001) 387–390.
[6] M. Kusumoto, E. Dux, W. Paschen, K.A. Hossmann, Susceptibility of
hippocampal and cortical neurons to argon-mediated in vitro ischemia,
Journal of Neurochemistry 67 (1996) 1613–1621.
[7] J. Levraut, H. Iwase, Z.H. Shao, T.L. Vanden Hoek, P.T. Schumacker,
Cell death during ischemia: relationship to mitochondrial depolarization
and ROS generation, Am. J. Physiol. Heart Circ. Physiol. 284 (2003)
549–558.
[8] M. Reers, T.W. Smith, L.B. Chen, J-aggregate formation of a carbo-
cyanine as a quantitative fluorescent indicator of membrane potential,
Biochemistry 30 (1991) 4480–4486.
[9] J.M. Robinson, B.E. Batten, Localization of cerium-based reaction prod-
ucts by scanning laser reflectance confocal microscopy, J. Histochem.
Cytochem. 38 (1990) 315–318.
[10] M.N. Sharikabad, K.M. Ostbye, O. Brors, Increased, [Mg2+]o reduces
Ca2+ influx and disruption of mitochondrial membrane potential dur-
ing reoxygenation, Am. J. Physiol. Heart Circ. Physiol. 281 (2001)
2113–2123.
[11] L. Simon, G. Szilagyi, Z. Bori, G. Telek, K. Magyar, Z. Nagy, Low
dose (-)deprenyl is cytoprotective: it maintains mitochondrial membrane

Page 4

G. Szil´ agyi et al. / Neuroscience Letters 399 (2006) 206–209
209
potential and eliminates oxygen radicals, Life Sci. 78 (2005) 225–
231.
[12] S.T. Smiley, M. Reers, C. Mottola-Hartshorn, M. Lin, A. Chen,
T.W. Smith, G.D. Steele Jr., L.B. Chen, Intracellular heterogene-
ity in mitochondrial membrane potentials revealed by a J-aggregate-
forming lipophilic cation JC-1, Proc. Natl. Acad. Sci. U.S.A. 88 (1991)
3671–3675.
[13] A.A. Starkov, G. Fiskum, Regulation of brain mitochondrial H2O2pro-
duction by membrane potential and NAD(P)H redox state, J. Neurochem.
86 (2003) 1101–1107.
[14] M.M. Tarpey, D.A. Wink, M.B. Grisham, Methods for detection of
reactive metabolites of oxygen and nitrogen: in vitro and in vivo con-
siderations, Am. J. Physiol. Regul. Integr. Comp. Physiol. 286 (2004)
431–444.
[15] G. Telek, J. Regoly-Merei, G.C. Kovacs, L. Simon, Z. Nagy, J. Hamar,
F. Jakab, The first histological demonstration of pancreatic oxidative
stress in human acute pancreatitis, Hepatogastroenterology 48 (2001)
1252–1258.
[16] G. Telek, J.Y. Scoazec, J. Chariot, R. Ducroc, G. Feldmann, C.
Roz, Cerium-based histochemical demonstration of oxidative stress in
taurocholate-induced acute pancreatitis in rats, J. Histochem. Cytochem.
47 (1999) 1201–1212.
[17] J.F. Turrens, Mitochondrial formation of reactive oxygen species, J.
Physiol. 552 (2003) 335–344.