Brought to you by Portland Press Ltd.
Published on behalf of the International Federation for Cell Biology
Cancer Cell death Cell cycle Cytoskeleton Exo/endocytosis Differentiation Division Organelles Signalling Stem cells Trafficking
Cell Biology International (2008) 32, 1486–1496 (Printed in Great Britain)
Simultaneous use of electrochemistry and chemiluminescence to detect reactive oxygen species produced by human neutrophils
Sergey Shleevab*, Jonas Wetteröc, Karl‑Eric Magnussond and Tautgirdas Ruzgasa
aBiomedical Laboratory Science, Health and Society, Malmö University, 20506 Malmö, Sweden
bLaboratory of Chemical Enzymology, Institute of Biochemistry, 119071 Moscow, Russia
cRheumatology/AIR, Department of Clinical and Experimental Medicine, Linköping University, 58185 Linköping, Sweden
dMedical Microbiology, Department of Clinical and Experimental Medicine, Linköping University, 58185 Linköping, Sweden


Abstract

A novel approach for the simultaneous optical and electrochemical detection of biologically produced reactive oxygen species has been developed and applied. The set-up consists of a luminol-dependent chemiluminescence assay combined with two amperometric biosensors sensitive to superoxide anion radicals (O2) and hydrogen peroxide (H2O2), respectively. The method permits direct, real-time in vitro determination of both extra- and intracellular O2 and H2O2 produced by human neutrophil granulocytes. The rate of O2 production by stimulated neutrophils was calculated to about 10−17mol s−1 per single cell. With inhibited NADPH oxidase, a distinct extracellular release of H2O2 instead of O2 was obtained from stimulated neutrophils with the rate of about 3·10−18mol s−1 per single cell. When the H2O2 release was discontinued, fast H2O2 utilisation was observed. Direct interaction with and possibly attachment of neutrophils to redox protein-modified gold electrodes, resulted in a spontaneous respiratory burst in the population of cells closely associated to the electrode surface. Hence, further stimulation of human neutrophils with a potent receptor agonist (fMLF) did not significantly increase the O2 sensitive amperometric response. By contrast, the H2O2 sensitive biosensor, based on an HRP-modified graphite electrode, was able to reflect the bulk concentration of H2O2, produced by stimulated neutrophils and would be very useful in modestly equipped biomedical research laboratories. In summary, the system would also be appropriate for assessment of several other metabolites in different cell types, and tissues of varying complexity, with only minor electrode modifications.


Keywords: Superoxide anion radical, Hydrogen peroxide, Neutrophil, Biosensor, Luminol-dependent chemiluminescence, NADPH oxidase.

*Corresponding author. Biomedical Laboratory Science, Health and Society, Malmö University, 20506 Malmö, Sweden. Tel.: +46 40 665 7414; fax: +46 40 665 8100.


1 Introduction

Reactive oxygen species (ROS), e.g. the superoxide anion radical (O2), hydroxyl radical (OH), peroxyl radical (ROO) and hydrogen peroxide (H2O2), as well as their adducts with nitric oxide (e.g. peroxynitrite, ONO2), are essential in the human defence against infections. They also have important non-bactericidal functions in a number of cells and tissues, viz. in signal transduction, proliferation, thrombosis, inflammation and cancer (Finkel, 1999; Lander, 1997; Stief, 2000). Human polymorphonuclear neutrophilic granulocytes (neutrophils) are the most abundant type of white blood cells and form an integral part of the human immune system. In circulation neutrophils are spheres with an average diameter and volume of 8μm and 300am3, respectively (Bainton, 1977; Simchowitz et al., 1993), whereas in peripheral blood smears their diameters increases up to 15–20μm (e.g. Bamberg and Johnson, 2002). Neutrophils respond to infection with a respiratory “burst” during which O2 is reduced to O2 by the NADPH oxidase. Superoxide may be released both to the phagosome and to the extracellular compartment (Babior, 1999). The short-lived ROS intermediate O2 is potentially hostile to the host since it may exceed certain “antioxidant” levels locally in tissue (Finkel and Holbrook, 2000; White et al., 1994).

It is self-evident that reliable methods for ROS detection should have wide applicability. Such measurements are, however, seriously confounded by the evanescent nature of ROS and the multiple cellular mechanisms evolved to maintain these substances at low concentrations (Tarpey and Fridovich, 2001). Thus, techniques more particular than those presently available are required for the analysis of ROS. Ideally, a technique designed to measure cellular production of ROS should be very specific, sensitive, affordable, easy to handle and standardise, and also bio-inert (Dahlgren and Karlsson, 1999). No technique hitherto developed has satisfied all these criteria.

One popular concept for O2 detection is based on the chemiluminescence (CL) principle (Tarpey and Fridovich, 2001). Several compounds have been used to enhance CL response, including luminol (LumH2), isoluminol, lucigenin, and coelenterazine (Dahlgren and Karlsson, 1999; Faulkner and Fridovich, 1993; Lucas and Solano, 1992). LumH2 is frequently used because it can access intracellular sites of O2 generation, shows minimal toxicity and has a very high sensitivity (Faulkner and Fridovich, 1993; Yao et al., 2002). In aqueous alkaline solutions and in the presence of an oxidant, LumH2 undergoes oxidation with emission of light (Roswell and White, 1978). H2O2 is the most useful oxidant, but it requires a catalyst, such as transition metal ions, hemin, or peroxidase. Instead of a catalyst, an electrode with positive applied potential can be used to produce electro-generated CL from the LumH2-H2O2 system (Epstein and Kuwana, 1967; Kuwana, 1963). Depending on the catalyst used, the pH range for efficient LumH2 CL generally falls between 7 and 11 (Cormier and Prichard, 1968). Despite its popularity, LumH2 is actually not ideally suited for specific O2 detection within living cells for several reasons. Firstly, O2 does not react directly with LumH2 (Faulkner and Fridovich, 1993). Secondly, LumH2 has an intrinsic ability to generate O2, when its univalently oxidised form is autooxidised. In the case of activated neutrophils, univalent oxidants can be myeloperoxidase (MPO) or extracellular horseradish peroxidase (HRP) in vitro, plus H2O2 (Allred et al., 1980). Thirdly, the quantum yield of O2 is limited at neutral (physiological) pH (Faulkner and Fridovich, 1993). Moreover, H2O2 is also able to trigger LumH2 based luminescence (Lucas and Solano, 1992) and ONO2, a product of the chemical reaction of O2 and NO, reacts directly with LumH2 to produce CL (Radi et al., 1993). Still, in the present study LumH2-dependent CL have been used as a very sensitive but not very selective method to detect total intra- and extracellular ROS production by neutrophils, taking into account the advantages and disadvantages of this compound.

Because ROS are very difficult to detect due to their high reactivity and short life time (Amatore et al., 2000; Finkel and Holbrook, 2000; White et al., 1994), it is appealing to exploit the advantages of electrochemical biosensors, e.g. real-time detection with high sensitivity and selectivity. Although several electrochemical biosensors for the separate measurement of O2 and H2O2 have been presented (Alpeeva et al., 2005; Campanella et al., 2000, 1997; Emreguel, 2005; Ferapontova et al., 2001; Ge and Lisdat, 2002; Lindgren et al., 2000; Lisdat et al., 1999; McNeil et al., 1995; McNeil and Manning, 2002; Scheller et al., 1999; Shleev et al., 2006), only a limited number of studies describe their simultaneous use (Krylov et al., 2006; Lvovich and Scheeline, 1997; Shipovskov et al., 2004). Furthermore, to our best knowledge simultaneous optical (using LumH2) and electrochemical (using biosensors) monitoring of O2 and H2O2 production by human neutrophil granulocytes has not been performed. Thus, the objective of the present work was to investigate ROS production in isolated human neutrophils using combined electrochemical and optical detection.

2 Materials and methods

2.1 Chemicals and buffers

Na2HPO4, KH2PO4, and NaCl were obtained from Merck GmbH (Darmstadt, Germany). Xanthine, fMLF-peptide (fMLF), diphenyleneiodonium (DPI), LumH2 (5-amino-2,3-dihydro-1,4-phtalazinedione), potassium superoxide (KO2) and dimethylsulphoxide (DMSO) were from Sigma (St. Louis, MO, USA). 3,3′-Dithiobis(sulfosuccinimidylpropionate) (DTSSP) was from Pierce Biotechnology (Chester, UK). 11-Mercapto-1-undecanol (MU) and 11-mercaptoundecanoic acid (MUA) were obtained from Aldrich (Steinheim, Germany). Absolute ethanol (99.7%) was from Solveco Chemicals AB (Täby, Sweden). All aqueous solutions were prepared using deionised water (18MΩ) purified with a Milli-Q system (Millipore, Milford, CT, USA). The main buffers were phosphate-buffered saline (PBS; 10mM sodium hydrogen phosphate, 10mM potassium dihydrogen phosphate and 0.15M sodium chloride, pH 7.3), and Krebs–Ringer phosphate buffer supplemented with 10mM glucose, 1.5mM MgSO4, and 1.1mM CaCl2 (pH 7.3; KRG). KRG was also prepared without the addition of CaCl2 to avoid aggregation of neutrophils.

2.2 Proteins and enzymes

Pseudomonas aeruginosa azurin (MW 14kDa), cytochrome c (cyt c, MW 12.4kDa) from horse heart, and xanthine oxidase (XOD, Grade III, 1–2units/mg, MW 300.0kDa) from bovine milk were purchased from Sigma. Superoxide dismutase (SOD, 2500–7000units/mg, MW 31.2kDa) from bovine erythrocytes, horseradish peroxidase (HRP, Type VI, 300units/mg, MW 40kDa, RZ 2.7), and catalase (30,000units/mg, MW 250kDa) from human erythrocytes were obtained from Boehringer Mannheim GmbH (Mannheim, Germany).

2.3 Cells

Peripheral human polymorphonuclear neutrophil granulocytes (neutrophils) were isolated from heparinised (10U/ml) whole blood immediately following venipuncture of apparently healthy non-medicated volunteers (Boyum, 1968; Ferrante and Thong, 1980). Whole blood was put on top of one part of Lymphoprep (Axis-Shield PoC AS, Oslo, Norway) layered over four parts of Polymorphprep (Axis-Shield PoC AS) and centrifuged for 40min at 480×g at room temperature. The neutrophil fraction was collected and washed for 10min at 480×g in room temperature PBS, and erythrocytes were eliminated by a short hypotonic lysis in ice-cold distilled water. Cells were washed twice at 200×g at 4°C in Ca2+-free KRG. The isolated neutrophils were counted in a Coulter Counter ZM Channelyser 256 (Coulter-Electronics Ltd., Luton, UK), showed excellent viability, <0.1 platelet contaminations per neutrophil and were kept on melting ice until experiments were performed.

2.4 Preparation of the biosensors

2.4.1 Preparation of H2O2 sensitive HRP-modified graphite electrode

An HRP-modified spectrographic graphite electrode (HRP-SPGE) was used for electrochemical detection of hydrogen peroxide (Alpeeva et al., 2005; Ruzgas et al., 1996). Working electrodes with an outside diameter of 3.05mm were prepared from rods of solid spectroscopic graphite type RW001 (Ringsdorff Werke GmbH, Bonn, Germany). The end of the graphite rod was polished on wet fine emery paper (Tufback Durite, P1200) and then carefully washed with deionised water. The enzyme was adsorbed onto the polished surface by placing 10μl aliquots of a 2mg/ml HRP solution on the dry electrode surface. Non-adsorbed enzyme was removed by repeated flushing in deionised water for at least 1min. The HRP-SPGE electrode was then stored at room temperature for at least 6h in PBS for stabilisation before use. Calibration of the electrodes was performed using standard solutions of H2O2, as well as superoxide production by reaction of KO2 with water (Shipovskov et al., 2004).

2.4.2 Preparation and electrochemical characterisation of the O2 sensitive gold electrodes

Three types of O2 sensitive biosensors were prepared using 2 different redox proteins, azurin (Shleev et al., 2006) and cyt c (Ge and Lisdat, 2002; Manning et al., 1998), immobilised on thiol-modified gold disk electrodes from Bioanalytical Systems (model MF-2014, West Lafayette, IN, USA) with an area of about 2.5·10−6m2. The bare gold electrode surface was first polished in DP-Suspension, and was then polished in alumina FF slurry (0.25 and 0.1μm, respectively, Struers, Copenhagen, Denmark), rinsed in Millipore water with 10min sonications in between. The electrodes were cycled 30 times in 0.5M H2SO4, and kept in concentrated H2SO4.

Prior to bio-modification gold electrodes were rinsed thoroughly with Millipore water. The first type of biosensor, based on azurin adsorbed on the DTSSP-modified gold electrode (azurin-DTSSP electrode), was prepared in accordance with our previous studies (Shleev et al., 2006). The second type of O2 sensitive gold electrode (Cyt-DTSSP) was prepared according to Manning and co-authors (Manning et al., 1998). In the third case, a gold electrode with cyt c adsorbed on a long-chain thiol layer (cyt-MU electrode) was used (Ge and Lisdat, 2002).

After preparation, the quality of the electrodes was controlled by cyclic voltammetry. Cyclic voltammograms (CVs) of the protein-modified electrodes were recorded using a one channel three-electrode potentiostat (BAS CV-50W Electrochemical Analyser with BAS CV-50W software v 2.1, Bioanalytical Systems, West Lafayette, IN, USA) and one single-compartment 20ml electrochemical cell. The reference electrode was an Hg|Hg2Cl2|KClsat (240mV vs. NHE), and a platinum wire served as a counter electrode.

The calibration of the electrodes was performed using chemically or enzymatically produced superoxide as described elsewhere (Ge and Lisdat, 2002; Shleev et al., 2006; Tammeveski et al., 1998). Basic electrochemical characteristics of the biosensors used in the present study are listed in Table 1. The response time of the superoxide sensitive electrodes is less than 1ms, as briefly described below. Thus, it is in principle possible to perform real-time monitoring of superoxide radicals using all three types of biosensors described.


Table 1.

Basic electrochemical characteristics of the biosensors used in present study

Type of biosensorAnalyteSensitivity (A m−2 M−1)Stability (h)Preparation time (h)Reference
HRP-SPGEH2O25.0 · 102>8≈4([Alpeeva et al., 2005] and [Ruzgas et al., 1996])
Azurin-DTSSPO2radical dot6.0 · 102≈5≈20(Shleev et al., 2006)
Cyt-DTSSPO2radical dot0.5 · 1026–8≈20(Tammeveski et al., 1998)
Cyt-MUO2radical dot2.5 · 102≥8≈30(Ge and Lisdat, 2002)


The approximate response time (), which is related to recharging of the electrode, can be calculated from Eq. (1) (Bard and Faulkner, 1980).

(1) where is the resistance of the solution, ca. 200Ω in our studies, and is the double layer capacitance of a protein monolayer based electrode, ca. 10−5F cm−2.

An approximate response time (), which is related the diffusion of superoxide to the electrode surface, can be calculated from Eq. (2).

(2) where δ is the distance between the electrode surface and superoxide (1micron in our studies) and D is the diffusion coefficient of superoxide ions in the solution (2·10−5cm2 s−1).

2.5 Detection of ROS

2.5.1 Luminometric detection of ROS (LumH2-dependent CL)

The intra- and extracellular neutrophil generation of ROS was studied by LumH2-dependent CL in a six channel Biolumat (LB 9505 C, Berthold Co., Wildbaden, Germany). Measurements were performed in polypropylene tubes with neutrophils (1·106 up to 5·106cells/ml) at 37°C in KRG with or without HRP (4U/ml) and LumH2 (56.4μM). Intracellular ROS were also detected in the presence of scavengers, SOD (200U/ml) and catalase (2000U/ml), as well as an inhibitor of the NADPH oxidase, DPI (5.0μM). In some experiments neutrophils were given 5min for pre-warming and stabilisation before being stimulated with 0.1μM fMLF.

2.5.2 Electrochemical detection of O2 and H2O2

The extracellular generation of O2 and H2O2 in isolated neutrophils was studied using either a multi-channel custom-built potentiostat (Institute of Biochemistry, Vilnius, Lithuania; http://www.tpa.lt/5FP/LIFE/life_intellisens.html), or a one channel CV-50W potentiostat. Initially, two working electrodes, O2 and H2O2 sensitive biosensors, were used simultaneously in one Biolumat channel. Different potentials were applied to the HRP-SPGE and O2 sensitive gold working electrodes, −50 and +250mV, respectively, and thus two chronoamperometric curves were obtained simultaneously. Finally, only one biosensor was used, either an O2 or a H2O2 sensitive electrode, as described below. In vitro electrochemical measurements were performed using a silver wire in PBS as a combined reference and counter electrode. All measurements described were performed at least in triplicate using freshly prepared O2 and H2O2 sensitive electrodes.

3 Results

The H2O2 and O2 sensitive biosensors were combined with the LumH2 CL assay to estimate the ROS production by human neutrophils. It was found that simultaneous use of both H2O2 and O2 sensitive electrodes in one Biolumat channel with accompanying CL detection resulted in significant interference between the electrochemical and optical detections, as well as in minor cross talk between two electrochemical signals. For example, LumH2 induced a substantial background current from the HRP-SPGE because of enzymatic and, possibly, electrochemical reactions of LumH2 on the electrode surface. Thus, to avoid interference and cross talk, electrochemical and optical monitoring were performed in separate Biolumat channels and only one biosensor, either a H2O2 or a O2 sensitive electrode, was used for each experimental run (Fig. 1). In order to normalise the results, the electrochemical and spectral measurements were started simultaneously along with the subsequent addition of cells and chemical compounds.


Fig. 1

Detection scheme of the optical-electrochemical set-up to monitor ROS production by human neutrophils. a – computer for the detection of O2 or H2O2 signals; b – potentiostat CV-50W; c – combined reference and counter electrode (Ag wire); d – working electrode (O2-sensitive gold electrode or H2O2-sensitive graphite electrode); e – electrochemical cell (first channel of Biolumat); f – optical compartment (channel No. 6 of Biolumat); g – optical signal from the LumH2-dependent CL compartment; h – channel No. 6 of Biolumat; i – computer for the detection of CL, 1–6 – corresponding channels of the Biolumat.


3.1 Chemiluminescence detection of ROS in combination with the H2O2 sensitive biosensor

First, the HRP-SPGE and LumH2 were combined. Initially, both electrochemical and CL signals were stabilised and baselines were established. Afterwards, aliquots carrying 1·106 up to 5·106 neutrophils per ml were added to the electrochemical and optical compartments (Fig. 1) and distinct signals were obtained. A typical CL signal and a current response recorded with the Biolumat and HRP-SPGE are shown in Fig. 2A,B, respectively. One part of the immediate response upon addition of neutrophils (as indicated by arrows in all figures) shown by the CL assay can be explained by the neutrophil transfer to the 37°C milieu after storage on melting ice (Fig. 2A). For comparison, modest mechanical disturbance of the electrochemical system, and possibly cells attaching to the electrode surface, also contributed significantly to the huge current step obtained directly after addition of cells (Fig. 2B). However, a minor contribution to the current step corresponding to ROS production could depend on the long stabilisation time, and/or the reductive direction of the current compared to the oxidative one that is always obtained with O2 sensitive biosensors (Fig. 2B). While a low steady-state level of the CL signal could be seen after stabilisation (corresponding to 3·106–9·106 counts per min), the electrochemical response was just at the background level. The data scattering in the CL assay is probably the result of different durations between neutrophil isolation and the ROS measurements.


Fig. 2

Response curves for ROS from neutrophils following stimulation with 0.1μM fMLF in the presence (solid line) and absence (dotted line) of 4U/ml extracellular HRP. (A) LumH2-dependent CL signal (KRG buffer; 56.4μM LumH2). (B) Amperometric detection of extracellular H2O2 using the HRP-modified SPGE (KRG buffer). Arrows indicate the addition of neutrophils, fMLF, and catalase into cuvettes equipped with the HRP-modified SPGE-based biosensor or containing the luminol reagent. Final concentration of cells in the cuvette: 106 cells/ml; 37°C; cpm=counts per minute.


For the following 300s, the cells warm up and the signals from both methods stabilise (Fig. 2). Addition of a potent receptor agonist (fMLF) resulted in a marked CL signal, which reached its maximum at about 500s, i.e. &007E;100s after stimulation. The peak response was ca. 30 times higher (2.4·108–5.5·108cpm), when compared to the signal without stimuli (Fig. 2A). While the CL signal decreased rapidly after it had reached maximum, the accumulation of H2O2 could still be observed, however, with a significantly reduced rate of accumulation (cf. Fig. 2A,B). The extracellular concentration of H2O2 at 507s, i.e. at the maximum CL, coincides perfectly with a small peak on the chronoamperometric curve. Here the H2O2 concentration was calculated to be close to 0.6μM (Fig. 2). H2O2 continued to accumulate and the highest level of extracellular H2O2 produced by fMLF-stimulated neutrophils was found to be ca. 3.0μM. Addition of catalase, a specific H2O2 utilising enzyme catalysing peroxide disproportionation, resulted in the complete loss of the electrochemical signal from the H2O2-SPGE. Indeed, it rapidly reached the background level, confirming the selectivity of the biosensor. Both the CL signals and the extracellular concentration of H2O2 were dependent on the concentration of neutrophils from 1·106 to 5·106 cells per ml.

It is interesting to compare the results from two independent methods, electrochemistry and CL, in the presence and absence of the extracellular catalyst HRP. The data (Fig. 2) displays a drop in both the electrochemical and CL signal in the absence of the enzyme with similar temporal patterns of both curves (cf. solid and dotted lines in Fig. 2). In the absence of HRP the CL signal decreased to 1/7 (i.e. by 86%) of its value in the presence of the enzyme, whereas the total electrochemical signal decreased by only 25%. A decrease in the CL signal without extracellular HRP could be expected since peroxidases are typical catalysts required for the light emission reactions at neutral pH (Cormier and Prichard, 1968).

Inhibition of the NADPH oxidase by DPI resulted in dramatic changes in both the patterns and amplitudes of the optical and electrochemical signals (Fig. 3). First, both signals dropped to near background levels immediately following DPI addition. Second, the fMLF-stimulated responses from Biolumat and the H2O2 sensitive biosensor were &007E;70 and two times lower than without inhibition of the enzyme after cell stimulation. Incidentally, the rate of H2O2 accumulation during the initial inhibition stage decreased by 400%, e.g. from 0.12nA s−1 for intact cells to 0.03nA s−1 for DPI-inhibited neutrophils. Third, an additional fMLF-triggered secondary broad peak (Fig. 3A, at 950s) appeared in the CL response, while no further accumulation of extracellular H2O2 was observed in the electrochemical system. In contrast to the experiments with intact neutrophils where further accumulation of H2O2 was observed, no H2O2 was detected in the solution at 500s after the stimulation in the presence of DPI (Fig. 3B). Finally, the peaks of the optical and electrochemical curves did not coincide in time after the inhibition of NADPH oxidase contrary to findings for intact neutrophils (cf. Fig. 3A,B).


Fig. 3

Response curves for ROS from neutrophils following stimulation with 0.1μM fMLF after the addition of 5μM DPI. (A) LumH2-dependent CL signal (KRG buffer; 56.4μM LumH2; 4U/ml HRP). (B) Amperometric detection of extracellular H2O2 using the HRP-modified SPGE (KRG buffer; 4U/ml HRP). Arrows indicate the addition of neutrophils, DPI, and fMLF into cuvettes equipped with the HRP-modified SPGE-based biosensor or containing the luminol reagent. Final concentration of cells in the cuvette: 106 cells/ml; 37°C.


3.2 Chemiluminescence detection of ROS in combination with O2 sensitive biosensors

The basic parameters of the azurin-DTSSP, cyt-MU, and cyt-DTSSP-modified electrodes are listed in Table 1. The sensitivity of the biosensors towards O2 was approximately 6.0·102, 2.5·102, and 0.5·102A m−2 M−1, respectively. Thus, macroscale electrodes would generally be able to sense the typically very small extracellular amount of O2, which may be released in vivo. To compare the different assays, all experiments were synchronised as far as possible. Acknowledging the complexity of studies on radicals in biological systems, the use of several alternative sensing chemistries is important to avoid misinterpretation due to possible artefacts. It should be emphasised also that the three different O2 sensitive biosensors showed qualitatively similar results (not shown). The azurin-DTSSP-modified gold electrode is probably the most sensitive biosensor for O2 detection reported so far and, hence, only the data obtained with this particular electrode are presented and discussed in detail below.

From the beginning of the measurements both electrochemical (O2 sensitive biosensors) and optical (ROS sensitive CL assay) signals were stabilised. After the stabilisation the neutrophils (1–5·106 cells per ml) were added to the electrochemical and CL compartments. Typical azurin-DTSSP electrode current and CL responses are shown in Fig. 4. The data from the CL assay was similar to the previous experiments (compare Figs. 2A with 4A,B). The addition of catalase to the luminometer compartment resulted in a significant drop of the CL signal (Fig. 4B).


Fig. 4

Response curves for ROS from neutrophils following stimulation with 0.1μM fMLF. (A, B) LumH2-dependent CL signal (KRG buffer; 56.4μM LumH2; 4U/ml HRP). (C, D) Amperometric detection of extracellular O2 using the Azurin-DTSSP-immobilised gold electrode (KRG buffer only). Arrows indicate the addition of neutrophils, fMLF, and catalase into cuvettes equipped with the azurin-DTSSP-based biosensor or containing the luminol reagent. Final concentration of cells in the cuvette: 106 cells/ml; 37°C.


Fig. 5

Response curves for ROS from neutrophils following stimulation with 0.1μM fMLF. (A) LumH2-dependent CL signal (KRG buffer; 56.4μM LumH2; 4U/ml HRP). (B) Amperometric detection of extracellular O2 using the azurin-DTSSP-immobilised gold electrode (KRG buffer; 4U/ml HRP). Arrows indicate the addition of neutrophils, fMLF, and SOD into cuvettes equipped with the azurin-DTSSP-based biosensor or containing the luminol reagent. Final concentration of cells in the cuvette: 106 cells/ml; 37°C.



Contrary to the H2O2 sensitive electrode, the azurin-based biosensors displayed pronounced current responses (0.5–2.0nA), corresponding to an extracellular O2 concentration of about 0.5–2.0μM, prior to the stimulation of the neutrophils (Figs. 4C and 6B). After 300s of equilibration the signals from either assay were stable and fMLF was added simultaneously to both channels. Surprisingly, the agonist increased the biosensor response only moderately, i.e. by 0.1nA (corresponding to ca. 0.1μM of O2) in the case of the azurin-DTSSP electrode (Fig. 4D). Subsequent addition of catalase resulted in a sharp decrease of the CL signal (Fig. 4B) and a very small amperometric increase, respectively (Fig. 4D).


Fig. 6

Response curves for ROS from unstimulated neutrophils. (A) LumH2-dependent CL signal (KRG buffer; 56.4μM LumH2; 4U/ml HRP). (B) Amperometric detection of extracellular O2 using the Azurin-DTSSP-immobilised gold electrode (KRG buffer; 4U/ml HRP). 1, bold lines – cells from melting ice directly; 2, solid lines – cells pre-incubated for 5min at 37°C; 3, dotted lines – cells pre-incubated for 5min at 37°C with 5μM DPI. Arrows indicate the addition of neutrophils into cuvettes equipped with the azurin-DTSSP-gold biosensor or containing the luminol reagent. Final concentration of cells in the cuvette: 106 cells/ml; 37°C.


In another experiment, HRP was present in the electrochemical cell (Fig. 5B), but this did not affect the electrochemical signal. By contrast, addition of SOD instead of catalase resulted in a sharp decrease of both the electrochemical (&007E;0.75nA, Fig. 5B) and the optical signal (&007E;1·108cpm, Fig. 5A). However, the current from the azurin-DTSSP electrode did not reach the background level, which is contrary to what was observed with the HRP-based H2O2 sensitive biosensor after the addition of catalase (cf. Figs. 2B and 5B).

The step of the current from the O2 sensitive biosensors directly after neutrophil addition, the increase in the reductive current after fMLF stimulation, and the SOD depleted current were all highly dependent on the amount of neutrophils in the solution, ranging from 1·106 to 3·106 cells per ml. A further increase of the cell density did not affect the O2 sensitive electrochemical signals (data not shown).

To clarify the nature of the initial current step after cell injection, additional experiments were performed with unstimulated neutrophils. Cell transfer from 4 to 37°C resulted in an altered CL signal (Fig. 6A), suggesting a mainly temperature-dependent production of ROS. Inhibition of neutrophils by DPI caused an almost complete disappearance of both the optical and electrochemical signals, confirming the cellular origin of ROS (Fig. 6).

4 Discussion

Both the LumH2-dependent CL signal and the current from the ROS sensitive biosensors depended quantitatively on the cell density and addition of different scavengers, i.e. SOD and catalase, which is in good agreement with previously published data (Allred et al., 1980; Dahlgren et al., 1985; Dahlgren and Karlsson, 1999; McNeil et al., 1989, 1992, 1995). To evaluate the complex experiments several aspects have to be considered. Firstly, the LumH2-dependent CL assay reflects both intra- and extracellular production of ROS, whereas the electrochemical signal corresponds specifically to the O2 and H2O2 released from neutrophils. When using HRP as a catalyst, O2 is the primary extracellular oxygen species in the peroxidase-catalysed excitation of LumH2, and thus H2O2 released from the cells does not significantly affect the CL (Dahlgren and Karlsson, 1999; Lock et al., 1988; Lundqvist and Dahlgren, 1996). Secondly, SOD and catalase can only scavenge extracellular metabolites. However, these scavengers might still affect intracellular responses since the removal of ROS from the medium, could promote the expulsion of H2O2 and O2 from the cytoplasm to maintain equilibrium (Halliwell and Gutteridge, 1990). Thirdly, only O2 in close proximity to the electrode contributes to the amperometric response due to the fast spontaneous dismutation to H2O2 at neutral pH. A rough calculation based on the diffusion coefficient of O2 in water (1.8·10−9m2 s−1) and the short life time of O2 (less than 0.1s at physiological conditions) indicates that the critical distance does not exceed 15μm. Coincidentally, the extracellular H2O2 concentration can be precisely measured via electrochemical detection in bulk solution since it is quite stable compared to O2.

Together these findings lead to the following. First, since there was no significant effect of LumH2 itself on neutrophils in our studies, on inhibition of O2 release (Faldt et al., 1999) or on LumH2-induced CL (Allred et al., 1980), in the absence of extracellular HRP the LumH2-dependent CL signal corresponds to a first approximation to intracellular production of ROS. Therefore, the difference between the two curves presented in Fig. 2A reflects the extracellular release of O2 from neutrophils. Moreover, the maximum of the LumH2-dependent CL signal almost coincided with the first maximum on the chronoamperometric curve from the H2O2 sensitive biosensor (cf. Figs. 2A,B). In accordance with chemical stoichiometry, one mole of H2O2 is produced from 2moles of O2 through fast spontaneous dismutation. After fMLF stimulation a respiratory “burst” of human neutrophils results in a very rapid, but rather short-lived (&007E;100s duration) extracellular release of O2, &007E;10−17mol s−1 from a single neutrophil, as calculated from the maximal concentration of O2, 1.2μM from 0.6μM H2O2; Fig. 2B, produced in 100s by 106 cells in 1ml volume. Thus, in one second a single cell releases &007E;107 molecules of O2 upon stimulation. This is a quite reasonable value taking into account the reported turnover number of NADPH oxidase in human neutrophils of 12–25s−1 (Glass et al., 1986).

Second, inhibiting NADPH oxidase with DPI drastically decreased both intra- and extracellular ROS production from either fMLF-stimulated or temperature-stressed neutrophils (Fig. 3). In fact, the total CL signal, as calculated from the total area under the graph, decreased by more than a factor of 150 compared to the signal from unimpaired cells. In contrast, the total amount of H2O2, as determined from the total area of the electrochemical signal, produced by the same number of cells was found to be only 10 times smaller than without inhibition. Moreover, while the maximum value of the optical signal was found to be 70 times lower upon addition of DPI, (cf. Figs. 2A and 3A), the corresponding peaks on the chronoamperometric curves differed only by a factor of two (cf. Figs. 2B and 3B). This discrepancy of more than one order of magnitude between the electrochemical and optical data cannot be explained by a difference in extra- and intracellular production of ROS. The experiments showed, however, a substantial extracellular release of H2O2 instead of O2 from neutrophils with DPI-inhibited NADPH oxidase; in the initial phase it was &007E;0.3·10−17mole s−1 of H2O2 by a single cell, based on the H2O2 concentration of &007E;0.4μM produced in 150s by 106 neutrophils in 1ml. Thus, our observations confirm a satisfactory selectivity of the LumH2-dependent CL assay towards extracellular O2 in the presence of HRP. Given the stoichiometry of the O2 dismutation reaction in non-inhibited cells releasing O2, the extracellular production of H2O2 by stimulated neutrophils with inhibited NADPH oxidase is halved as compared to the usual respiratory “burst”. Thus, in the presence of DPI, neutrophils still respond to fMLF, but by releasing H2O2 instead of O2. Previously published data support H2O2 generation by neutrophils with inhibited NADPH oxidase, when stimulated with six-formylpterin (Yamashita et al., 2001). This bifurcating behaviour of neutrophils may be important physiologically, and could possibly be exploited for medical applications.

As mentioned earlier, pre-incubation of cells at 37°C before injection into the Biolumat channels resulted in a significant drop in the CL signal, reflecting the production of ROS by stressed neutrophils during warming after storage on melting ice. The rate and total amount of ROS produced by the stressed neutrophils were, however, almost negligible compared to those of stimulated cells (Figs. 2A and 6A). At the same time the O2 sensitive electrochemical signals seemed insensitive to pre-warming, whereas DPI inhibition resulted in an almost complete disappearance of both optical and electrochemical signals, confirming their O2 origin (Fig. 6). The current from the O2 sensitive biosensor corresponds specifically to O2 being released from neutrophils in very close proximity to the electrodes, i.e. neutrophils directly adhering to the biosensors. This proposition also helps us explain the unexpected results when O2 sensitive electrodes were used. Hence, while the dominant population of neutrophils, i.e. non-adsorbed, react as expected on the specific fMLF-stimulation by a respiratory “burst”, the cells on the electrode surface are already in some sense stimulated by the electrode surface or the adsorbed redox proteins, thus giving a seemingly insignificant response to fMLF (Figs. 4 and 5). Moreover, SOD inhibited the electrochemical signal significantly, but not back to the background level, possibly due to steric factors prohibiting the enzyme to penetrate between the cells and the electrode (Fig. 5B). On the other hand, addition of catalase increased the signal (Fig. 4D), most probably by removal of H2O2 from the bulk medium and promoting the generation of O2 in the cytoplasm. Yet, the current increase was limited by the fixed number of neutrophils on the electrode surface. Using the area of the gold electrodes (2.5·10−6m2) and the size of a single neutrophil (0.2nm2), the approximate number of cells attached to the electrode surface was calculated to be 1.25·104. Taking into account the known life time of O2 (ca. 0.1s), the rate of O2 production by stimulated neutrophils (10−17mole s−1 of O2 per single neutrophil), and the small confined volume (≈10−11m3) limited by several hundred nm distance between the electrode and cells, the local steady-state concentration of O2, which arises from the equilibrium between O2 production and spontaneous dismutation, can be estimated to reach several μM in a few hundred sec after injection of neutrophils. Interestingly, closely related values for the steady-state concentration of O2 were indeed obtained from the O2 sensitive biosensors, i.e. about 1μM. Electrochemical detection of O2 from living cells in bulk solution is accordingly only possible in a very small volume, where cells are separated somehow from the electrode surface. This approach has recently been realised by Chang et al. (2005a, b), where cells were attached to a glass spacer or a COSTAR® membrane insert, a few hundred μm away from the surface of an O2 sensitive biosensor. An excellent correlation between the rate of the current increase from O2 sensitive gold electrodes and the number of cells seeded was obtained (Chang et al., 2005a, b), whereas in the present work and some other studies (McNeil et al., 1992, 1995), the response curve was not linear, but showed a saturation behaviour at higher neutrophil loadings, reflecting the attachment of the latter to the electrode surface.

To meet all the criteria for sensitive and selective measurements of ROS produced by living cells, the design of an electrochemical system should be recognised as the most promising device for real-time determination of steady-state O2 concentrations without interference from cellular metabolism and associated regulatory pathways (Campanella et al., 2000; Chang et al., 2005a, b; Manning et al., 1998; McNeil and Manning, 2002). However, placing the biosensors in a compartment with a spectral system (Chang et al., 2005b) requiring a dye probe to produce the optical signal, limits the applicability. Therefore, we performed the electrochemical and optical assays in two separate compartments along with simultaneous real-time monitoring of both signals. Our methodology enables continuous monitoring of both O2 and H2O2 produced by human neutrophil granulocytes. By suitable modifications the system can be even more versatile, i.e. by employing available biosensors for glutamate and nitric oxide detection (Chang et al., 2005a; O'Neill et al., 2004) and by using more specific and/or sensitive CL compounds, like isoluminol which is highly selective with respect to extracellular ROS only (Dahlgren and Karlsson, 1999; Lundqvist and Dahlgren, 1996). Hence, it would be possible to extend our approach for simultaneous, direct, quantitative real-time measurements of ROS (e.g. O2 and H2O2), nitrogenous species (e.g. nitric oxide), and other messengers (e.g. glutamate) produced by different cells types or even more complex biological systems. The experimental set-up of the present approach (Fig. 1) could easily be expanded with additional biosensors or other CL compounds, provided the six channel Biolumat is combined with a sensitive multi-channel potentiostat together with suitable software for multi-parameter data analysis.

In summary, the present work demonstrates a reasonably simple, moderately expensive, and extraordinarily powerful methodology to study the production of intra- and extracellular ROS from human neutrophils. It is also potentially useful in the study of other metabolites from different living cells or even tissues. The methodology might be a very useful tool in the research on the mechanisms of drug action and drug discovery processes, and would certainly contribute to the research of oxidative stress therapy.

Acknowledgements

The authors thank the European Community (contract QLK3-CT-2001-00244), the Swedish Research Council, the Swedish Society for Medicine, and the Goljes Minne, Magn Bergvall, Lars Hiertas Minne, and Nanna Svartz research foundations for their financial support. We also thank Dr. Zoltan Blum (Biomedical Lab Science, Health and Society, Malmö University) for critical reading and helpful suggestions.

References

Allred CD, Margetts, J, Hill, HR. Luminol-induced neutrophil chemiluminescence. Biochim Biophys Acta 1980:631:380-5
Medline   1st Citation   2nd   3rd  

Alpeeva IS, Niculescu-Nistor, M, Leon, JC, Csoeregi, E, Sakharov, IY. Palm tree peroxidase-based biosensor with unique characteristics for hydrogen peroxide monitoring. Biosens Bioelectron 2005:21:742-8
Crossref   Medline   1st Citation   2nd   3rd  

Amatore C, Arbault, S, Bruce, D, de Oliveira, P, Erard, M, Vuillaume, M. Analysis of individual biochemical events based on artificial synapses using ultramicroelectrodes: cellular oxidative burst. Faraday Discuss 2000:116:356
1st Citation  

Babior BM. The participation of the hemes of flavocytochrome b245 in the electron transfer process in NADPH oxidase. Blood 1999:93:4449
Medline   1st Citation  

Bainton DF. Differentiation of human neutrophilic granulocytes: normal and abnormal. Prog Clin Biol Res 1977:13:1-27
1st Citation  

Bamberg R, Johnson, J. Segmented neutrophil size and platelet morphology in HIV/AIDS patients. Clin Lab Sci 2002:15:18-22
Medline   1st Citation  

Bard AJ, Faulkner, LR. Electrochemical methods: fundamentals and applications. 1980:
Medline   1st Citation  

Boyum A. Isolation of mononuclear cells and granulocytes from human blood. Isolation of mononuclear cells by one centrifugation, and of granulocytes by combining centrifugation and sedimentation at 1g. Scand J Clin Lab Invest Suppl 1968:97:77-89
Medline   1st Citation  

Campanella L, Favero, G, Tomassetti, M. A modified amperometric electrode for the determination of free radicals. Sens Actuat 1997:B44:559-65
1st Citation  

Campanella L, Favero, G, Persi, L, Tomassetti, M. New biosensor for superoxide radical used to evidence molecules of biomedical and pharmaceutical interest having radical scavenging properties. J Pharm Biomed Anal 2000:23:69-76
Crossref   Medline   1st Citation   2nd  

Chang S-C, Pereira-Rodrigues, N, Henderson, JR, Cole, A, Bedioui, F, McNeil, CJ. An electrochemical sensor array system for the direct, simultaneous in vitro monitoring of nitric oxide and superoxide production by cultured cells. Biosens Bioelectron 2005:21:917-22
Crossref   Medline   1st Citation   2nd   3rd   4th  

Chang S-C, Rodrigues, NP, Zurgil, N, Henderson, JR, Bedioui, F, McNeil, CJ. Simultaneous intra- and extracellular superoxide monitoring using an integrated optical and electrochemical sensor system. Biochem Biophys Res Commun 2005:327:979-84
Crossref   Medline   1st Citation   2nd   3rd   4th  

Cormier MJ, Prichard, PM. An investigation of the mechanism of the luminescent peroxidation of luminol by stopped flow techniques. J Biol Chem 1968:243:4706-14
Medline   1st Citation   2nd  

Dahlgren C, Karlsson, A. Respiratory burst in human neutrophils. J Immunol Methods 1999:232:3-14
Crossref   Medline   1st Citation   2nd   3rd   4th   5th  

Dahlgren C, Aniansson, H, Magnusson, KE. Pattern of formylmethionyl-leucyl-phenylalanine-induced luminol- and lucigenin-dependent chemiluminescence in human neutrophils. Infect Immun 1985:47:326-8
Medline   1st Citation  

Emreguel E. Development of a new biosensor for superoxide radicals. Anal Bioanal Chem 2005:383:947-54
Crossref   Medline   1st Citation  

Epstein B, Kuwana, T. Electro-oxidation of phthalhydrazides. J Electroanal Chem 1967:15:389-97
Crossref   1st Citation  

Faldt J, Ridell, M, Karlsson, A, Dahlgren, C. The phagocyte chemiluminescence paradox: luminol can act as an inhibitor of neutrophil NADPH-oxidase activity. Luminescence 1999:14:153-60
Crossref   Medline   1st Citation  

Faulkner K, Fridovich, I. Luminol and lucigenin as detectors for superoxide radical. Free Rad Biol Med 1993:15:447-51
Crossref   Medline   1st Citation   2nd   3rd   4th  

Ferapontova EE, Grigorenko, VG, Egorov, AM, Borchers, T, Ruzgas, T, Gorton, L. Mediatorless biosensor for H2O2 based on recombinant forms of horseradish peroxidase directly adsorbed on polycrystalline gold. Biosens Bioelectron 2001:16:147-57
Crossref   Medline   1st Citation  

Ferrante A, Thong, YH. Optimal conditions for simultaneous purification of mononuclear and polymorphonuclear leucocytes from human blood by the Hypaque–Ficoll method. J Immunol Methods 1980:36:109-17
Crossref   Medline   1st Citation  

Finkel T. Signal transduction by reactive oxygen species in non-phagocytic cells. J. Leukocyte Biol 1999:65:337-40
1st Citation  

Finkel T, Holbrook, NJ. Oxidants, oxidative stress and the biology of ageing. Nature 2000:408:239-47
Crossref   Medline   1st Citation   2nd  

Ge B, Lisdat, F. Superoxide sensor based on cytochrome c immobilized on mixed-thiol SAM with a new calibration method. Anal Chim Acta 2002:454:53-64
Crossref   1st Citation   2nd   3rd   4th   5th  

Glass GA, DeLisle, DM, DeTogni, P, Gabig, TG, Magee, BH, Markert, M. The respiratory burst oxidase of human neutrophils. Further studies of the purified enzyme. J Biol Chem 1986:261:13247-51
Medline   1st Citation  

Halliwell B, Gutteridge, JMC. Role of free radicals and catalytic metal ions in human disease: an overview. Methods Enzymol 1990:186:1-85
Medline   1st Citation  

Krylov AV, Adamzig, H, Walter, AD, Loechel, B, Kurth, E, Pulz, O. Parallel generation and detection of superoxide and hydrogen peroxide in a fluidic chip. Sens Actuat 2006:B119:118-26
1st Citation  

Kuwana T. Electrooxidation followed by light emission. J Electroanal Chem 1963:6:164-7
Crossref   1st Citation  

Lander HM. An essential role for free radicals and derived species in signal transduction. FASEB J 1997:11:118-24
Medline   1st Citation  

Lindgren A, Ruzgas, T, Gorton, L, Csoregi, E, Bautista Ardila, G, Sakharov, IY. Biosensors based on novel peroxidases with improved properties in direct and mediated electron transfer. Biosens Bioelectron 2000:15:491-7
Crossref   Medline   1st Citation  

Lisdat F, Ge, B, Ehrentreich-Foerster, E, Reszka, R, Scheller, FW. Superoxide dismutase activity measurement using cytochrome c-modified electrode. Anal Chem 1999:71:1359-65
Crossref   1st Citation  

Lock R, Johansson, A, Orselius, K, Dahlgren, C. Analysis of horseradish peroxidase-amplified chemiluminescence produced by human neutrophils reveals a role for the superoxide anion in the light emitting reaction. Anal Biochem 1988:173:450-5
Crossref   Medline   1st Citation  

Lucas M, Solano, F. Coelenterazine is a superoxide anion-sensitive chemiluminescent probe: its usefulness in the assay of respiratory burst in neutrophils. Anal Biochem 1992:206:273-7
Crossref   Medline   1st Citation   2nd  

Lundqvist H, Dahlgren, C. Isoluminol-enhanced chemiluminescence: a sensitive method to study the release of superoxide anion from human neutrophils. Free Rad Biol Med 1996:20:785-92
Crossref   Medline   1st Citation   2nd  

Lvovich V, Scheeline, A. Amperometric sensors for simultaneous superoxide and hydrogen peroxide detection. Anal Chem 1997:69:454-62
Crossref   1st Citation  

Manning P, McNeil, CJ, Cooper, JM, Hillhouse, EW. Direct, real-time sensing of free radical production by activated human glioblastoma cells. Free Rad Biol Med 1998:24:1304-9
Crossref   Medline   1st Citation   2nd   3rd  

McNeil CJ, Manning, P. Sensor-based measurements of the role and interactions of free radicals in cellular systems. Rev Mol Biotechnol 2002:82:443-55
Crossref   1st Citation   2nd  

McNeil CJ, Smith, KA, Bellavite, P, Bannister, JV. Application of the electrochemistry of cytochrome c to the measurement of superoxide radical production. Free Rad Res Commun 1989:7:89-96
Crossref   1st Citation  

McNeil CJ, Greenough, KR, Weeks, PA, Self, CH, Cooper, JM. Electrochemical sensors for direct reagentless measurement of superoxide production by human neutrophils. Free Rad Res Commun 1992:17:399-406
Crossref   1st Citation   2nd  

McNeil CJ, Athey, D, Ho, WO. Direct electron transfer bioelectronic interfaces: application to clinical analysis. Biosens Bioelectron 1995:10:75-83
Crossref   Medline   1st Citation   2nd   3rd  

O'Neill RD, Chang, S-C, Lowry, JP, McNeil, CJ. Comparisons of platinum, gold, palladium and glassy carbon as electrode materials in the design of biosensors for glutamate. Biosens Bioelectron 2004:19:1521-8
Crossref   Medline   1st Citation  

Radi R, Cosgrove, TP, Beckman, JS, Freeman, BA. Peroxynitrite-induced luminol chemiluminescence. Biochem J 1993:290:51-7
Medline   1st Citation  

Roswell DF, White, EH. The chemiluminescence of luminol and related hydrazides. Methods Enzymol 1978:57:409-23
1st Citation  

Ruzgas T, Csoeregi, E, Emmneus, J, Gorton, L, Marko-Varga, G. Peroxidase-modified electrodes: fundamentals and application. Anal Chim Acta 1996:330:123-38
Crossref   1st Citation   2nd  

Scheller W, Jin, W, Ehrentreich-Forster, E, Ge, B, Lisdat, F, Buttemeier, R. Cytochrome c-based superoxide sensor for in vivo application. Electroanalysis 1999:11:703-6
Crossref   1st Citation  

Shipovskov S, Ferapontova, EE, Gazaryan, I, Ruzgas, T. Recombinant horseradish peroxidase- and cytochrome c-based two-electrode system for detection of superoxide radicals. Bioelectrochemistry 2004:63:277-80
Crossref   Medline   1st Citation   2nd  

Shleev S, Wetterö, J, Magnusson, K-E, Ruzgas, T. Electrochemical characterization and application of azurin-modified gold electrodes for detection of superoxide. Biosens Bioelectron 2006:22:213-9
Crossref   Medline   1st Citation   2nd   3rd   4th   5th  

Simchowitz L, Textor, JA, Cragoe, EJ. Cell volume regulation in human neutrophils: 2-(Aminomethyl)phenols as chloride channel inhibitors. Am J Physiol 1993:265:C143-55
Medline   1st Citation  

Stief TW. The blood fibrinolysis/deep-sea analogy. A hypothesis on the cell signals singlet oxygen/photons as natural antithrombotics. Thrombosis Res 2000:99:1-20
Crossref   1st Citation  

Tammeveski K, Tenno, TT, Mashirin, AA, Hillhouse, EW, Manning, P, McNeil, CJ. Superoxide electrode based on covalently immobilized cytochrome c: modeling studies. Free Rad Biol Med 1998:25:973-8
Crossref   Medline   1st Citation   2nd  

Tarpey MM, Fridovich, I. Methods of detection of vascular reactive species. Nitric oxide, superoxide, hydrogen peroxide, and peroxynitrite. Circulation Res 2001:89:224-36
Crossref   Medline   1st Citation   2nd  

White CR, Brock, TA, Chang, LY, Crapo, J, Briscoe, P, Ku, D. Superoxide and peroxynitrite in atherosclerosis. Proc Natl Acad Sci USA 1994:91:1044-8
Crossref   Medline   1st Citation   2nd  

Yamashita K, Arai, T, Fukuda, K, Mori, H, Ishii, H, Nishioka, M. 6-Formylpterin intracellularly generates hydrogen peroxide and restores the impaired bactericidal activity of human neutrophils. Biochem Biophys Res Commun 2001:289:85-90
Crossref   Medline   1st Citation  

Yao D, Vlessidis, AG, Evmiridis, NP, Zhou, Y, Xu, S, Zhou, H. Novel chemiluminescence method for detection of superoxide anions and its application to dry-cured meat. Anal Chim Acta 2002:467:145-53
Crossref   1st Citation  


Received 10 January 2008/23 July 2008; accepted 12 August 2008

doi:10.1016/j.cellbi.2008.08.016


ISSN Print: 1065-6995
ISSN Electronic: 1095-8355
Published by Portland Press Limited on behalf of the International Federation for Cell Biology (IFCB)