1. Introduction
Reactive oxygen species (ROS)—principally superoxide (•O
2-), radical hydroxyl ion (•OH), and nonradical hydrogen peroxide (H
2O
2)—are present in nearly all aerobic cell types [
1]. When reactive radicals such as superoxide and nitric oxide hybridize, they form peroxynitrite (ONOO
-), which is a highly reactive oxidizing agent, more toxic than the parent species. Peroxynitrite (ONOO
-) and nitric oxide (•NO) are types of reactive nitrogen species (RNS), which, together with ROS, may transiently exceed the capacity of the body to remove them—a condition known as oxidative stress [
2]. Oxidative stress is associated with numerous adverse health effects [
1,
2,
3]. At the cellular level, these ill effects include irreversible oxidation of DNA, lipids, and amino-acid side chains (e.g., protein carbonylation) [
1]. However, ROS/RNS species play a central role in essential physiological functions, such as the activation of macrophages to deploy ROS in fighting infection [
4]. This ROS-mediated “respiratory burst” is a key component of the inflammatory response necessary for the regulation of cell growth and survival.
Despite the short-term benefit of this immune response, chronic inflammation is characteristic of diseases such as cancer, atherosclerosis, rheumatoid arthritis, Parkinson’s disease, and Alzheimer’s [
1,
2,
4,
5]. Chronic inflammation involves a state of prolonged oxidative stress in which the normal mechanisms of inflammatory resolution are impaired or do not function, such as the recruitment of anti-inflammatory M2 macrophages [
6]. Chronically elevated levels of ROS, then, may have far-reaching detrimental effects in the body [
7].
Hydrogen peroxide is one of the primary redox signals in the body [
8]. It is a non-radical product of the superoxide dismutase SOD-catalyzed dismutation of superoxide, which is produced continuously in mitochondria and elsewhere [
9]. The most prevalent sources of endogenous superoxide are the electron transport chain of mitochondria, which generates superoxide radicals and occasionally hydrogen peroxide. Superoxide is generated and released by the NADPH oxidase (NOX) complexes as shown in
Figure 1 [
3]. Studies have shown that dual oxidase (DUOX) enzymes and the isoform NOX4 can produce H
2O
2 directly [
10].
Much of the oxidative damage caused by hydrogen peroxide occurs indirectly, e.g., through the Fenton reaction in the presence of ferrous iron and the subsequent production of hydroxyl radicals [
11]. Elevated H
2O
2 levels are a hallmark of inflammation and oxidative stress [
2]. Since H
2O
2 is the most stable ROS in vivo, H
2O
2 is a suitable choice of biomarker in an assay of inflammation [
12].
Current methods to detect and quantify oxidative stress, including ROS and RNS and their reactive intermediates, include spectroscopic methods, fluorescent-dependent methods [
8,
13], the use of chemiluminescent probes [
14,
15], spectrophotometric methods [
16], chromatographic methods [
17], and electrochemical sensors [
18]. Each of these methods suffer from limitations such as the quenching of fluorescence when fluorescent probes or dyes are used and their inability to penetrate deeper layers of tissue. More recently, Vaneev et al. [
19] reported in vitro and in vivo electrochemical measurement of ROS using an intracellularly placed platinized nanoelectrode. Although in vitro work appeared promising, the detection was somewhat invasive, having to pierce the cellular membrane, possibly compromising membrane integrity. This method also appeared to be impractical in the case of ROS detection in tissues, where the thickness of the tissue sample would provide inaccurate results.
The Amplex Red reagent is commonly used as an extracellular indicator of H
2O
2 production. Amplex Red is oxidized by horseradish peroxidase (HRP) in the presence of H
2O
2 to form resorufin, a fluorogenic molecule [
20]. This assay is highly specific for H
2O
2 and, because H
2O
2: resorufin stoichiometry is 1:1, may be used to quantify extracellular H
2O
2 production [
20]. However, similar to all fluorescent molecules, resorufin is subject to photobleaching with the subsequent attenuation of fluorescence. This tendency presents a practical problem in the design of an assay of inflammation.
The use of chronoamperometry in the detection of H
2O
2 and other ROS/RNS released by a small population of RAW 264.7 cells in a Pt MEA-containing microfluidic chamber has been demonstrated [
21]. However, the intricacy of such a design may be of limited utility in a medical laboratory.
On the other hand, the ready availability of ceramic Pt MEA probes may permit the development of a more rapid scalable assay of extracellular H
2O
2. Furthermore, the ease of modifying the microelectrode surface allows for a wide range of multiplexed applications demonstrated from our group: Hossain et al. report the fabrication of a multiplexed glutamate/gamma-aminobutyric acid (GABA) Pt MEA probe via the addition of glutamate oxidase (GlOX) and GABase to separate microelectrodes within the same probe; this probe was used to detect ex vivo release of GABA in hippocampal rat brain slices [
22]. Scoggin et al. report the use of GlOX-modified Pt MEA probes in vitro to detect glutamate uptake in astrocytes versus glioma cells [
23].
In our study, we employ a similar approach to Vaneev [
19] and Hossain and Scoggin [
19,
22,
23], using a Pt MEA probe in an extracellular detection strategy. The immediate goal of this project is the non-invasive post hoc detection of stable concentrations of H
2O
2 released by a large population of macrophages and accurately detectable above baseline levels. Hence, real-time detection of H
2O
2 production by a small population of cells, as in Amatore et al. [
21], was not attempted. Instead, H
2O
2 release was induced as follows: 1 × 10
5 RAW 264.7 cells were incubated with varying concentrations of LPS for 6–48 h before testing for 4–8 min with the probe, to capture the electrochemical signals over different time periods and durations of probing. Sies et al. [
8] reported the normal intracellular physiological H
2O
2 concentration as 1–10 nM in human liver, with extracellular concentrations being 100-fold. As our study was based on cell monolayer culture, our detection range planned to cover the oxidative eustress and distress ranges of the RAW 264.7 cells, with the expectation that the level of cellular complexity would be much lower in this study, enabling adequate detection.
To assess the feasibility of the Pt MEA as an H2O2 assay, the following hypotheses were investigated: (1) the probe will sense H2O2 over a physiologically relevant linear range (1–10 µM), (2) RAW 264.7 cells will produce H2O2 in response to LPS doping, (3) the probe will detect significantly more H2O2 in culture medium in LPS-doped samples than in untreated controls, (4) LPS-stimulated H2O2 production will be greater at the physiologically relevant temperature of 37 °C than at 4 °C, (5) dexamethasone, an anti-inflammatory steroid, will attenuate H2O2 production in LPS-doped samples, and (6) LPS doping and the operation of the probe will not adversely affect cell viability.
3. Results and Discussion
3.1. Calibration of the Probe in the Presence of H2O2
Calibration curves are shown in
Figure 5; amperometry was performed in 10 mL of continuously stirred DMEM (+0.7 V voltage step versus Ag/AgCl; recording frequency: 10 Hz). H
2O
2 was aliquoted into the stirred media at 60 s intervals, as shown in
Table 1, over the physiologically relevant H
2O
2 range of 1–10 µM. Probe calibration in this range ensured that H
2O
2 levels for monolayer cells before as well as after LPS-induced oxidative stress were capable of reliable detection by this probe, as supported by limited, but similar studies showing linearity of the standard curves at such low extracellular concentrations and periods of measurement [
26,
27].
In the example calibration shown in
Figure 5, noise is seen at points of H
2O
2 addition (e.g., at 1980s in
Figure 5a) and due to disturbances in the media (e.g., at 2030–2040s). However, this is a typical characteristic feature of the probe calibration curve and is expected.
Care was taken to minimize background noise during testing. The analysis of experimental data was automated in MATLAB; briefly, random noise was filtered via the averaging of detected current at fifty evenly spaced time points across the last minute of testing. Additionally, the use of cell-free controls in each experiment helped ensure that noise of a longer duration would not artifactually increase the current and hamper accurate measurement.
3.2. Optimization of LPS and H2O2 Doping Protocol
To test the ability of the probe to detect H2O2 released by LPS-doped activated macrophages, 1 × 105 cells/well were seeded in 24-well plates and incubated for 24 h. When the cells were 50% confluent, they were doped with 200 ng/mL or 500 ng/mL LPS according to Method 1 (see Method 1), incubated for 24 h, and tested. The following controls were included in this experiment: For negative controls, there were media with cells (media/cells) only (no doping of LPS). For positive controls, there were (a) media/cells with H2O2 (2 min), 2 µL of stock H2O2 was aliquoted two minutes before testing into wells containing media and cells (final concentration in well: 4 µM) and (b) media/cells with H2O2 (24 h), 2 µL of stock H2O2 was aliquoted 24 h before testing (i.e., at the time of LPS addition) to wells containing media and cells. This control was included to ensure that a known concentration of exogenous H2O2 would remain stable enough in solution to be detected by the probe. The difference in signal between both positive controls would be used to assess the stability of extracellular H2O2 released by LPS-doped cells.
To assess the efficacy of this protocol, the following experiment was performed: RAW 264.7 cells were seeded in 24-well plates, as above, incubated for 24 h, and doped with 200 ng/mL or 500 ng/mL LPS. The resulting data showed an enhanced H
2O
2-generating response in LPS-doped cells, with significantly elevated H
2O
2 production in both 200 ng/mL (
p = 0.039) and 500 ng/mL-treated (
p = 0.021) cells (
Figure 6). H
2O
2 produced was not significantly different between the two groups.
As shown in
Figure 6, H
2O
2 showed apparent persistence in the 24-h positive control: currents generated (i.e., H
2O
2 detected) at E1 in the H
2O
2 (2 min) and H
2O
2 (24 h) groups were not significantly different (
p = 0.99998). The normalized H
2O
2 concentration in both positive controls was ~6 µM—consistent with the expected ~4 µM increase above average H
2O
2 concentration in the cells-only negative control (1.88 ± 0.30 µM). Given the possibility of macrophages metabolizing external H
2O
2 to concentrations below the LOD of the probe [
27], the similar signals obtained in both H
2O
2-positive controls indicate a robust ability to detect H
2O
2 incubated for long periods with live cells. This result extended to LPS-doped cells as well. Taken together, these two results validate the hypothesis that stable extracellular concentrations of LPS-induced H
2O
2 may be detected in RAW cells via the probe.
3.3. LPS Dosage Concentration for H2O2 Detection
Figure 7 is to show the optimal LPS concentration for extracellular H
2O
2 production; therefore, only two selected concentrations of LPS were included in the next experiment. RAW 264.7 cells were seeded in 24-well plates as above and doped with LPS in the following (final) concentrations: 200 ng/mL, 500 ng/mL, 800 ng/mL, and 1000 ng/mL. The cells were incubated for 12 or 24 h before testing. H
2O
2 production in RAW 264.7 cells showed no significant LPS dose dependence (
Figure 8), implying that the range of LPS concentrations assayed, 200 ng/mL
–1 µg/mL, is saturating yet sublethal. Extracellular H
2O
2 production was more pronounced at the two lower concentrations of LPS (
Figure 7). The apparent stability of extracellular H
2O
2 was further supported by this experiment: between 12 h and 24 h, only the highest dosage of LPS (1 µg/mL) showed any statistically significant reduction in H
2O
2 detected by the probe, whereas the lower concentrations remained relatively stable over this period. Given the toxicity of LPS and the potential concomitant reduction in cell viability [
28], the lowest dose, 200 ng/mL, was selected as the best concentration for further study.
3.4. Incubation Time Study
Figure 8 summarizes time studies that were performed to discern the effect of incubation time on H
2O
2 production. The optimal LPS concentration of 200 ng/mL was used. RAW 264.7 cells were seeded in 24-well plates as described previously, then doped with 200 ng/mL LPS and incubated for 6, 12, 24, or 48 h.
At each of the four time points, H2O2 was significantly elevated in LPS-treated cells versus cell-only controls (p < 0.001). Somewhat unexpectedly, the LPS-stimulated H2O2 detected was not statistically different across the four time points. This prolonged stability indicated that 200 ng/mL LPS produced a stable, saturating concentration of extracellular H2O2. While comparing these incubation timepoints, LPS stimulation of the cells at all these timepoints exhibits statistically similar [H2O2] release. Thus, selecting an optimal LPS incubation time was somewhat arbitrary; therefore, one or more of these incubation timepoints were chosen for subsequent experiments.
3.5. Incubation Temperature Study
The inflammatory response depends on metabolically functioning cells, which must initiate a host of synthetic and transport processes. In order to determine that the H2O2 measured is due to cellular processes, LPS-doped cells were incubated at 4 °C. The functioning of mammalian cells in vitro depends strongly on incubation temperature, it was expected that H2O2 production in LPS-doped RAW cells incubated at 4 °C would be near zero.
To assess the effect of reduced incubation temperature on H
2O
2 production, RAW 264.7 cells were seeded in 24-well plates as described above. All plates were initially incubated at 37 °C, 5% CO
2 to ensure the cells could adhere and proliferate. After doping the cells with 200 ng/mL LPS, cells were incubated at either 4 °C or 37 °C for 48 h. Subsequent testing with the ROS probe revealed the expected lack of extracellular H
2O
2 produced by 4 °C-incubated cells as shown in
Figure 9.
H2O2 was significantly elevated when LPS-doped cells were incubated at 37 °C, whereas little or no H2O2 were detected for cells incubated at 4 °C, thus supporting the metabolic energy dependence of H2O2 production in LPS-activated macrophages.
3.6. Effect of Dexamethasone on H2O2 Production
Figure 10 shows the effect of an anti-inflammatory drug on H
2O
2 production. It was expected that the anti-inflammatory glucocorticoid dexamethasone (Dex) would counter the action of LPS, thus reducing cumulative H
2O
2 production in LPS-doped RAW 264.7 cells [
29]. To assess the effect of treating cells with Dex in conjunction with LPS, cells were treated with 200 ng/mL LPS and either 200 nM or 400 nM Dex, then incubated for 6, 12, 24, or 48 h.
At each of the four time points, cells treated with LPS and Dex (200 nM and 400 nM) produced significantly more H2O2 than cell-only controls (p < 0.001). The H2O2 production effect was mitigated by Dex at three time points: at 6, 12, and 24 h after treatment, LPS + Dex groups produced significantly less H2O2 than the LPS-only positive control (6 h and 12 h: p < 0.05; 24 h: p < 0.001). Between 6 and 24 h, these data appear to show the anti-inflammatory action of Dex being gradually outpaced by LPS-stimulated H2O2 production. Yet, 48 h after treatment, the anti-inflammatory, ROS-suppressive effect was strongest (p < 0.001). Unexpectedly, Dex alone stimulated significant H2O2 production (p < 0.05) at 48 h. This seeming paradox should be investigated further. Despite the imperfect trend in the data, these results support that the probe may be used as an extracellular assay not just of H2O2, but for inflammation in general.
3.7. Viability Assay of LPS-Doped and ROS-Probed Cells
In order to assess the viability of cells after LPS stimulation, an MTT assay kit (Promokine) was used. Cells were cultured in 96-well plates and subjected to one of the following conditions: (1) negative control, (2) 200 ng/mL LPS, (3) 500 ng/mL LPS, (4) exposure to the ROS probe (+0.7 V against Ag/AgCl) for 5 min, (5) exposure to the ROS probe (+0.3 V against Ag/AgCl) for 5 min.
With both 200 ng/mL and 500 ng/mL LPS, RAW 264.7 cells exhibited viability slightly greater than 100% (
p > 0.05): LPS appears to encourage proliferation at these doses as shown in
Figure 11. This result agrees with a phenomenon observed in the inverted microscope: after incubation with LPS in 24-well plates, these cells often appeared more confluent than those of the negative control (data not shown).
Tukey–Kramer analysis of these results indicated that neither LPS doping nor use of the probe significantly reduced cell viability (
p > 0.05). At both 200 ng/mL and 500 ng/mL, RAW 264.7 cells exhibited viability slightly greater than 100% (
p > 0.05), LPS appears to enhance cell proliferation at these doses, supporting evidence in previous studies [
9,
30] and demonstrating that LPS can stimulate immune cell proliferation. This result agrees with a phenomenon observed in the inverted microscope: after incubation with LPS in 24-well plates, these cells often appeared more confluent than those of the cell-only control. This effect might be attributed to the greater number of cells accounting for some of the increase in cumulative H
2O
2 seen in LPS-doped groups.
The reduction in viability at both voltage steps was just shy of statistical significance. However, in this assay, the probe was used in 100 µL, rather than 1 mL, of media (i.e., a 96-well plate, rather than a 24-well plate). It is possible that the closer confines of the 96-well plate caused additional stress to the cells when the probe was introduced and operated.