This study demonstrates the use of flow cytometry to detect reactive oxygen species (ROS) production resulting from activation of the FcγR. This method can be used to assess changes in the antimicrobial and redox signaling function of phagocytes in response to immune complexes, opsonized microorganisms, or direct FcγR cross-linking.
The oxidative or respiratory burst is used to describe the rapid consumption of oxygen and generation of reactive oxygen species (ROS) by phagocytes in response to various immune stimuli. ROS generated during immune activation exerts potent antimicrobial activity primarily through the ability of ROS to damage DNA and proteins, causing death of microorganisms. Being able to measure ROS production reproducibly and with ease is necessary in order to assess the contribution of various pathways and molecules to this mechanism of host defense. In this paper, we demonstrate the use of fluorescent probes and flow cytometry to detect ROS production. Although widely used, fluorescent measurement of ROS is notoriously problematic, especially with regards to measurement of ROS induced by specific and not mitogenic stimuli. We present a detailed methodology to detect ROS generated as a result of specific FcγR stimulation beginning with macrophage generation, priming, staining, FcγR cross-linking, and ending with flow cytometric analysis.
Reactive oxygen species (ROS) are reactive molecules or free radicals that are by-products of aerobic respiration (reviewed in 1). These include the superoxide anion, peroxide, hydrogen peroxide, hydroxyl radical, and hydroxyl ions, among others. Under normal physiologic conditions, ROS are produced mainly by the mitochondria and nicotinamide adenine dinucleotide phosphate (NADPH) oxidases and are rapidly detoxified by various enzymes and proteins such as superoxide dismutase and glutathione. An exaggerated production of ROS or a defect in the ability to remove ROS can result in oxidative stress, whereby reactive oxygen species promote the damage of proteins, lipids, and DNA leading to cellular stress or death and pathological disease states. However, it is currently appreciated that ROS can also act as signaling molecules (redox signaling), and ROS-mediated modification of various molecules and pathway intermediates can influence cellular metabolism, proliferation, survival, inflammatory signaling, and aging2. In phagocytic cells, ROS plays an essential role in providing antimicrobial activity during the so-called “respiratory burst”1,3,4,5,6. During the response of phagocytes to external stimuli, components of the NADPH oxidase complex (p40phox, p47phox, p67phox) translocate from the cytosol to the phagosomal membrane containing the gp91phox and p22phox subunits, and together with the actions of Rac1/2, form a fully functional NADPH oxidase enzyme complex. The assembled NADPH oxidase then utilizes NADPH to reduce oxygen to superoxide within the phagosomal vacuole. Superoxide anions can directly cause damage or be dismutated into hydrogen peroxide. Both superoxide and hydrogen peroxide can react with other molecules to generate highly reactive hydroxyl radicals. Damage is mediated by reaction of these ROS with iron-sulfur clusters on proteins or by causing base oxidation of DNA, ultimately leading to restricted microbial metabolism or death of the microbe5. The importance of the NADPH oxidase enzyme complex and ROS produced during the respiratory burst is illustrated clinically in patients with Chronic Granulomatous Disease (CGD)7,8,9,10. Individuals with CGD possess mutations in gp91phox, resulting in a lack of ROS production and susceptibility to recurrent infections with bacteria and fungi which are not usually a concern with immunocompetent individuals. Therefore, whether studying oxidative stress, redox signaling, or host defense, being able to measure ROS production in real-time is a useful endeavor.
Multiple assays have been utilized to measure ROS production or the results of oxidative stress11,12,13. Among these, one of the most widely used is the fluorescent probe 2’,7’ dichlorodihydrofluorescein diacetate (DCFH2-DA)14. This molecule is colorless and lipophilic. Diffusion of DCFH2-DA across the cell membrane allows it to be acted upon by intracellular esterases, which deacetylates it into DCFH2, rendering it cell impermeable. The actions of multiple types of ROS (hydrogen peroxide, peroxynitrite, hydroxyl radicals, nitric oxide, and peroxy radicals) on DCFH2 oxidize it into DCF which is fluorescent (reported Ex/Em: 485-500 nm/515-530 nm) and can be detected using a flow cytometer equipped with a standard filter set for fluorescein (FL1 channel). Superoxide does not strongly react with DCFH2 but can react with another probe dihydroethidium (DHE) to yield the fluorescent product 2-hydroxyethidium (as well as other fluorescent superoxide-independent oxidation products)15. The fluorescent products of DHE oxidation can be detected using an excitation wavelength of 518 nm and an emission wavelength of 605 nm (FL2 channel). Although relatively simple to use, utilization of these probes for detection of ROS requires knowledge of their limitations and careful incorporation of staining procedures and controls into the specific assay being performed in order to have valid experimental results and conclusions. The following protocol demonstrates the use of a commercially available kit employing these 2 probes designed to measure ROS by flow cytometry. We stain primed bone marrow-derived macrophages with these probes and induce ROS production through FcγR cross-linking. We present representative data obtained using this protocol and stress appropriate precautions that must be undertaken for successful experimentation.
The protocol for animal handling was approved by the Institutional Animal Care and Use committee (IACUC) of University of Central Florida.
1. Generation of bone marrow derived macrophages (BMDMs)
2. Harvesting, seeding and priming of BMDMs
3. Reagent and material preparation for ROS measurement
4. Assay conditions and controls
5. Cell preparation
6. Performing the assay
7. Flow cytometry data analysis and anticipated results
8. Cell surface staining in combination with flow cytometric analysis of ROS production (Optional)
NOTE: This step provides a protocol for staining macrophages with a cell-surface marker prior to stimulation of the FcγR and ROS measurement. This may be useful in assessing ROS production in mixed cell populations. It is important to choose an antibody for macrophage surface marker conjugated to an appropriate fluor that does not interfere with the fluorescence from the oxidative stress or superoxide detection reagents. In this protocol, an antibody for mouse F4/80 conjugated to Alexa fluor 647 is used.
Using the protocol outlined within, we present representative data demonstrating flow cytometric detection of ROS production resulting from stimulation of WT C57BL/6J BMDMs through the FcγR. As expected, we observe minimal changes in FL1 or FL2 fluorescence above background levels in unstimulated cells (Figure 3A, compare “stained, unstimulated” vs “unstained, unstimulated” dot plots). We observe a marked increase in FL1 and FL2 fluorescence when cells are stimulated with FcγR cross-linking agent (Figure 3A, compare “stained and stimulated via Fc cross-linking” samples vs “stained, unstimulated” dot plots). Lastly, when cells were treated with ROS inhibitor prior to FcγR cross-linking, this increased fluorescence is brought back to basal levels (Figure 3A, compare “stained and treated with ROS inhibitor and stimulated via Fc cross-linking” vs “stained and stimulated via Fc cross-linking” dot plots). This is also evident when data is presented as a histogram for each channel (Figure 3B) or when data is presented as a percentage of cells positive for either the green or orange ROS probes (Figure 3C). A similar trend is also apparent when data is presented as MFI, although the reduction in orange fluorescence with pre-treatment with ROS inhibitor is not captured as well when presented as MFI versus as a percentage (Figure 3C). We also present the results of 3 independent experiments performed on different days (Figure 3, Experiment 1, 2, and 3). The average values and corresponding standard error of the mean are indicated in the graphs (Figure 3C).
We also present unsuccessful experimentation, where sub-optimal ROS production as a result of FcγR stimulation was observed (Figure 4). A minimal increase in FL1 and FL2 fluorescence was detected when comparing “stained and stimulated via Fc cross-linking” samples vs “stained, unstimulated” samples (Figure 4A,B,C). This is presented alongside a successful experiment to highlight the large differences between the expected percentages or MFI increases and the observed values in the unsuccessful experiment.
The current protocol utilizes a 24 h priming step. When comparing a 24 h versus a 48 h priming time, we observed no marked difference in the percentage of cells positive for the green, oxidative stress reagent (Figure 5A, top histograms and Figure 5B green probe, % positive). However, increasing the priming time to 48 h did increase the percentage of cells positive for the orange fluorescence (Figure 5A, lower histograms and Figure 5B orange probe, % positive). This was similarly reflected when data was presented as MFI. This suggests that for optimal detection of all ROS species, a 48 h priming time may be more ideal.
Given that, due to the cost or the time needed for experimentation, use of a kit to perform this assay may not be an option. For this reason, we also tested similar components to those provided in the kit and purchased these from standard vendors (Thermofisher, EMD Millipore, Cayman). We find that using individually procured components and the same experimental protocol for cell loading and FcγR stimulation, we can recapitulate many of the same findings we observed using the kit (Figure 6A,B). However, although increases in fluorescence were apparent with stimulation, a higher level of ROS production was observed using the kit. This may indicate that use of individually procured components may be feasible but would need to be further optimized for this specific assay.
Lastly, we demonstrate that it is also possible to combine cell-surface staining with these ROS probes. We use a known macrophage marker, F4/80, conjugated to Alexa 647 and perform cell surface staining prior to treatment with ROS inhibitor, ROS inducer, or specific stimuli to induce ROS production. We demonstrate in Figure 7B that macrophages respond as expected when treated with FcγR crosslinking agent and FcγR crosslinking agent + ROS inhibitor (Figure 7B, orange vs green dot plots). Furthermore, we can observe increased orange or green fluorescence specifically generated by the F4/80 labeled cells upon treatment with FcγR crosslinking agent and reduced with pre-treatment with ROS inhibitor (Figure 7B, F4/80 vs green and F4/80 vs orange dot plots).
Figure 1: Flow cytometric assessment of appropriate generation of BMDMs. Wild type BMDMs were generated and were left either unstained or stained with FITC anti-mouse F4/80 or Alexa 647 anti-mouse CD11b. FSC (x-axis) vs SSC (y-axis) plots were generated and macrophages (BMDMs) were gated to exclude dead cells and debris. Using a plot of FSC-H(x-axis) vs FSC-A (y-axis), a singlet gate was generated. Gating on singlets, histograms were generated to show cells stained with either FITC F4/80 or APC CD11b in comparison to the isotype stained control. A) Correct BMDM differentiation with more than 95% of cells staining positive for CD11b or F4/80. B) Incorrect BMDM differentiation where less than 95% of cells are staining positive for F4/80 and 2 peaks are present for CD11b. Please click here to view a larger version of this figure.
Figure 2. Performing compensation when using green and orange ROS probes. Compensation will require unstained untreated cells, cells stained with green ROS probe and treated with ROS inducer, and cells stained with orange ROS probe and treated with ROS inducer. Dot plots for unstained untreated cells are used to determine quadrant gates. Data for cells singly stained with either green ROS probe or orange ROS probe and treated with ROS inducer are shown prior to, and after, compensation was applied. The compensation matrix is then applied to all subsequent experimental samples. Please click here to view a larger version of this figure.
Figure 3. Measurement of ROS in response to specific FcγR stimulation using green and orange ROS probes and assessment of assay reproducibility. Wild-type bone marrow-derived macrophages (BMDMs) were generated, primed, and were left either unstained or stained with a cocktail of green and orange ROS probes. Stained BMDMs were either left untreated, stimulated through their FcγRs using murine anti-BSA IgG1 + BSA for 30 min, or treated with ROS inhibitor prior to stimulation via FcγR cross-linking. A) Dot plots, showing an increase in the percentages of cells in the upper left, upper right, and lower right quadrants upon specific FcγR stimulation, which is reduced in the presence of ROS inhibitor. B) Histograms of each fluorescence channel, showing the marker gate to determine cells positive for each probe. C) Presentation of the data as a percentage of cells positive for each probe or as an increase in MFI. Three independent, representative experiments are presented. The mean for the 3 experiments and standard errors of the mean are shown as lines within the graphs. Fc XL, Fc cross-linking; NAC, N-acetyl-L-cysteine. Please click here to view a larger version of this figure.
Figure 4. Examples of successful and suboptimal FcγR stimulation. Wild-type bone marrow-derived macrophages (BMDMs) were generated, primed, and were left either unstained or stained with a cocktail of green and orange ROS probes. Stained BMDMs were either left untreated, stimulated through their FcγRs using murine anti-BSA IgG1 + BSA for 30 min, or treated with ROS inhibitor prior to stimulation via FcγR cross-linking. Representative results for successful and suboptimal stimulation are shown as A) dot plots, B) histograms, or C) the percentage of cells positive for each probe or as an increase in MFI. Successful stimulation shows increased fluorescence in the upper left, upper right and lower right quadrants of the FL1 vs FL2 plot and increased MFI and percentage of positive cells stained with each probe upon FcγR stimulation. Suboptimal stimulation shows minimal increase in MFI or percentage of positive cells. Fc XL, Fc cross-linking; NAC, N-acetyl-L-cysteine. Please click here to view a larger version of this figure.
Figure 5. Effect of priming time on ROS generation upon FcγR stimulation. BMDMs were generated, primed for either 24 or 48 h, and were left either unstained or stained with a cocktail of green and orange ROS probes. Stained BMDMs were either left untreated, stimulated through their FcγRs using murine anti-BSA IgG1 + BSA for 30 min, or treated with ROS inhibitor prior to stimulation via FcγR cross-linking. A) Histograms for the fluorescence induced by each probe upon stimulation of macrophages primed for either 24 or 48 h. B) Percentage of cells positive for each probe upon stimulation of macrophages primed for either 24 or 48 h. Priming macrophages for 48 h resulted in an increase in percent of cells positive for orange fluorescence (or an increase in the MFI for the FL2 channel) compared to priming the macrophages for 24 h. The mean for the 3 experiments and standard errors of the mean are shown as lines within the graphs. Fc XL, Fc cross-linking; NAC, N-acetyl-L-cysteine. Please click here to view a larger version of this figure.
Figure 6. Flow cytometric ROS measurement upon FcγR cross-linking using reagents from different vendors. BMDMs were generated, primed for 24 h, and were left either unstained or stained with a cocktail of oxidative stress and superoxide detection probes. Stained BMDMs were either left untreated, stimulated with ROS inducer, stimulated through their FcγRs using murine anti-BSA IgG1 + BSA for 30 min, or treated with ROS inhibitor prior to stimulation via FcγR cross-linking. Probes, ROS inducer and ROS inhibitor were used either from the kit or were purchased separately from different vendors and used at a similar concentration. A) Dot plots showing a side-by-side comparison of results obtained using kit and non-kit components. B) Histograms showing a side-by-side comparison of results obtained using kit and non-kit components. Fc XL, Fc cross-linking; NAC, N-acetyl-L-cysteine. Please click here to view a larger version of this figure.
Figure 7. Combining cell-surface staining with flow cytometric measurement of ROS production upon FcγR cross-linking. A) Dot plots for the various fluorescent channels using unstained or singly stained compensation controls. Wild-type BMDMs were generated, primed for 24 h, and were left either unstained, stained with Alexa 647 anti-mouse F4/80 only, stained with green ROS probe only and treated with ROS inducer, or stained with orange ROS probe only and treated with ROS inducer. Dot plots for unstained untreated cells are used to determine quadrant gates. Plots demonstrate expected results if channels are correctly compensated. The compensation matrix was then applied to all subsequent experimental samples. B) Wild-type BMDMs were generated, primed for 24 h and stained with Alexa 647 anti-mouse F4/80. Afterward, BMDMs were either left unstimulated, stimulated through their FcγRs using murine anti-BSA IgG1 + BSA for 30 min, or treated with ROS inhibitor prior to stimulation via FcγR cross-linking. Dot plots demonstrate that green or orange fluorescence specifically produced by F4/80 cells can be detected. Please click here to view a larger version of this figure.
DCFH2-DA and DHE-based detection of ROS is a widely-used technique14,15. Ease of use and the adaptability of these ROS probes for kinetic microplate formats, fluorescence microscopy or flow cytometric analysis has contributed to their popularity. However, in our studies of FcγR-mediated macrophage functions, there did not seem to be a standard protocol for performing this assay for flow cytometric analysis of FcγR cross-linked cells. Given the reactive nature of the analytes being assessed, we have found that timing is of crucial importance in making sure that reproducibility is achieved. Although we have also performed kinetic microplate assays, there can be a wide variability in the intensity of fluorescence from experiment to experiment, making it difficult to aggregate biological replicates for statistical analysis. Flow cytometric use of these probes allows for quantification of “ROS positive” cells which resolves some of these issues but is not without complications. This is especially true when using a flow cytometer equipped with an autosampler. In this case, analysis of a large number of samples will impact the amount of time the cells are exposed to various stimuli. To mitigate this, we have incorporated the lag time in between sample analysis into the protocol to make sure that the analysis is performed on cells stimulated for a similar amount of time.
Another important factor in flow cytometric ROS analysis is the inclusion of appropriate controls. It cannot be overstated that every experiment must have (at least) all the experimental and compensation controls we list in this protocol. These are important to ensure the validity of each experiment as well as to be able to justifiably exclude samples if needed. Some examples of this include if macrophages did not differentiate/mature appropriately and exhibited much lower ROS production even with inducer treatment. Another case could be inappropriate activation of macrophages even prior to FcγR cross-linking, resulting in a high basal ROS signal. Use of ROS inhibitors in conjunction with specific FcγR cross-linking showing disappearance of fluorescent signals is another important control we have often found lacking in other studies performing similar assays. The ROS inhibitor used in this study, N-acetyl-L-cysteine, is regarded as a universal ROS inhibitor. However, if one is interested in ROS produced by specific enzymes, other specific ROS inhibitors can be included in the assay as well. Some examples include mefanamic acid (cyclooxygenase-dependent ROS inhibitor), apocynin (NADPH oxidase-dependent ROS inhibitor), or allopurinol (a xanthine oxidase-dependent ROS inhibitor).
It is additionally crucial to understand what is actually being measured in this readout. As previously mentioned, DCFH2-DA measures multiple ROS species, and so, fluorescence resulting from the green probe cannot be used to discriminate one ROS species from another14. Likewise, the orange probe (likely DHE), although often cited as superoxide specific, can produce both superoxide-specific and superoxide-independent oxidation products, which cannot be distinguished without the use of additional techniques such as HPLC15. However, for the purposes of measuring “total ROS” or “induced ROS” during the respiratory burst, use of these probes alongside the appropriate controls, may be acceptable. Many new fluorescent based ROS sensors have emerged or are being developed11,13. Some, such as the redox-sensitive fluorescent proteins, have the advantage of a dynamic measurement of ROS production due to their reversible oxidation. However, these would require genetic manipulation of cells, which may not always be feasible or desired. In many cases, fluorescent probe-based ROS detection is still a valid and useful tool, as long as the experiment is standardized, well-controlled, and the conclusions derived from such experiments are not overstated.
The benefits to using this commercially available ROS kit is that all necessary reagents including ROS inducers, scavengers, and titrated probes are included and “ready-to-use”. If the period of experimentation is anticipated to be brief (completed within a week), this kit can provide a more cost-effective method for performing flow cytometry-based ROS detection without the need to buy or optimize each specific component individually. We additionally demonstrate using this kit with a specific agonist (FcγR stimulation) and demonstrate reproducibility across experiments; we assess effects of priming times on ROS generation; we compare using similar probes, inhibitors, and inducers from different companies; we provide examples of unsuccessful experimentation; and lastly, we combine cell-surface staining with use of these ROS probes. This would potentially allow for simultaneous detection of ROS from different cell types within a mixed population, requiring less sample and reagents. For example, this might be particularly useful when differentiating between macrophage and neutrophil-derived ROS, the main cell types capable of responding to FcγR ligation. For multiple experiments which have to be performed with long intervening periods in between, use of a kit might not be as ideal. Multiple aliquots of the probes are not provided and once reconstituted, the manufacturer only recommends that reagents be used within a week (as ROS probes are highly reactive). If it is anticipated that this period is incompatible with experimentation, we recommend that individual reagents rather than a kit be purchased separately and reconstituted only during the day of experimentation. As we demonstrate here, further optimization of the individually purchased components may additionally be needed prior to the assay. Overall, we hope that this work provides a useful resource for researchers using flow cytometry to reproducibly measure ROS generation.
The authors have nothing to disclose.
The authors would like to thank other members of the Tigno-Aranjuez Lab including Madelyn H. Miller, Omar Cardona, Andjie Jeudy, and Roopin Singh for their help in laboratory upkeep and mouse colony maintenance. Support for this research was provided by grant R00 HL122365 and Start-up funds to J.T.T-A.
Anti-BSA IgG1 | Innovative Research | IBSA9E2C2 | |
Alexa Fluor 647 Rat IgG2b, κ Isotype Ctrl Antibody | BioLegend | 400626 | |
Anti-mouse CD16/32 | BioLegend | 101302 | |
Anti-mouse F4/80 antibody conjugated to Alexa Fluor 647 | BD Biosciences | 565853 | |
Anti-mouse F4/80 antibody conjugated to FITC | BioLegend | 123108 | |
Anti-mouse/human CD11b antibodyconjugated to Alexa Fluor 647 | BioLegend | 101218 | |
beta-mercaptoethanol (BME) | Sigma | M3148-100ml | |
Bovine Serum Albumin (BSA) FractionV | Fisher | BP1600-100 | |
C57BL/6J | Jackson labs | Stock No.000664 | |
CM-H2DCFDA | Molecular Probes | C6827 | Can be a substitute for oxidative stress detection reagent in the Enzo kit |
Dihydroethidium (DHE) | Molecular Probes | D11347 | Can be a substitute for superoxide detection reagent in the Enzo kit |
DMEM 1x | Corning | 10-013-CV | |
DMEM no phenol red | Gibco | 31053-028 | |
DMF Anhydrous | Acros Organics | 61094-1000 | |
Fetal Bovine Serum (FBS) | VWR | 97068-085 | |
FITC Rat IgG2a, κ Isotype Ctrl Antibody | BioLegend | 400506 | |
HEPES (1M) | Gibco | 15630-080 | |
L glutamine | Gibco | 25030-081 | |
LADMAC cells | ATCC | CRL-2420 | |
MEM | Corning | 10-010-CV | |
mouse IFN-g | GoldBio | 1360-06-100 | |
N-Acetyl-L-cysteine | EMD Milipore | 106425 | Can be a substitute for ROS inhibitor/scavenger in the Enzo kit |
Novocyte flow cytometer with autosampler | Acea | 2060R | |
Pyocyanin (ROS inducer) | Cayman chemical | 10009594 | Can be a substitute for inducer in the Enzo kit |
ROS-ID total ROS/superoxide detection kit | ENZO | ENZ-51010 | |
Sodium pyruvate (100mM) | Gibco | 11360-070 | |
Trypsin-EDTA (0.25%) | Gibco | 25200-056 |