Ultraviolet irradiation increases green fluorescence of dihydrorhodamine (DHR) 123: false‐positive results for reactive oxygen species generation

Abstract Dihydrorhodamine (DHR) 123 is a fluorophore commonly used for measuring reactive oxygen species (ROS), often after exposing cells to ultraviolet (UV) irradiation or oxidative burst inducers such as Phorbol 12‐myristate 13‐acetate (PMA). However, the negative effects of UV irradiation on oxidation of DHR123 itself to green fluorescence rhodamine (R) 123 under different experimental conditions (e.g., different buffers, media, cells, ROS detection techniques) have not been fully appreciated. We determined the effect of UV on DHR123 fluorescence, using a cell‐free system, and A549 epithelial cells, NIH/3T3 fibroblast cells, Jurkat T cells, primary human T cells, HL‐60 neutrophils and primary human neutrophils. We found that UV irradiation rapidly increases green fluorescence of DHR123 in cell‐free solutions. The intensity of green fluorescence increases with increasing amounts of DHR123 and UV exposure. The fluorescence increase was greater in Roswell Park Memorial Institute medium (RPMI) than DMEM media. The presence of DMSO (0–1.25%, v/v) in RPMI further increases the fluorescence signal. Phosphate buffered solution (PBS) and Hanks' Balanced Salt Solution (HBSS) generate considerable background signal with DHR123, and increasing DMSO concentration greatly increases the fluorescence signal in these buffers. However, after UV irradiation the amount of DHR123 that remains unoxidized generates sufficient fluorescence signal to measure the ROS produced by H2O2 and peroxidase in vitro or Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase‐mediated ROS production within HL‐60 neutrophils or primary human neutrophils. We conclude that UV irradiation oxidizes DHR123 to generate Rhodamine 123 (R123) green fluorescence signal, and that the R123 present in the culture supernatant could give erroneous results in plate reader assays. However, flow cytometry and fluorescence microscopy reliably detect ROS in cells such as neutrophils. Overall, avoiding false‐positive results when detecting ROS using DHR123 requires selection of, agonists, the correct buffers, media, cell types, and measurement techniques.


Introduction
Dihydrorhodamine 123 (DHR123) is a fluorescence probe commonly used for measuring reactive oxygen species (ROS) (Crow 1997). Several types of cells produce ROS or reactive nitrogen species during oxidative stress, cell death, exposure to UV irradiation or agonists such as PMA in the form of superoxide, peroxynitrite, H 2 O 2 , and singlet oxygen molecules (Crow 1997;Kalyanaraman et al. 2012). In the presence of ROS, nonfluorescent DHR123 is oxidized to rhodamine (R123). When R123 is excited at 488 nm, it emits a green fluorescence signal with peak intensity at 525 nm (Qian and Buettner 1999). Therefore, ROS production is routinely measured with DHR123 using fluorescence plate reader assays, flow cytometry or fluorescence microscopy. DHR123 has been used for measuring ROS both in cell-free conditions (Shen et al. 2014) and in cell-based systems with human neutrophils, neutrophils differentiated from HL-60 cell line (Chen and Junger 2012;Douda et al. 2014Douda et al. , 2015, eosinophils (Elsner et al. 1996), lymphocytes (Poniedzialek et al. 2015), epithelial cells (Wilhelm et al. 2009;Kelley et al. 2013) and mixed cell cultures (Hanson and Clegg 2002). However, some of these cell types do not have the peroxidases needed to convert H 2 O 2 to HOCl, a substrate necessary to effectively convert DHR123 to fluorescence R123 (Furtmuller et al. 2000). Therefore, using DHR123 in all of these cells to measure UV-or PMA-mediated ROS production requires careful re-evaluation.
In this study, we aimed to analyze the direct effect of UV irradiation on DHR123 dissolved in commonly used buffers and media, and in four different cell types: T cells, fibroblasts, epithelial cells, and neutrophils. We found that UV irradiation alone increases the green fluorescence signal of DHR123 and thus could lead to erroneous conclusions in plate reader based assays. However, the oxidative effect of UV light on DHR123 does not interfere with flow cytometry-and fluorescent microscopy-based analysis (e.g., neutrophils). DHR123 is appropriate for measuring PMA-mediated ROS production in neutrophils, but not in epithelial, fibroblast and T cells. Overall, great care should be taken in the selection of buffers, media, and cell types when using DHR123 to measure ROS production. It is particularly important because background buffers and cells are not UV irradiated with DHR123. Therefore, correcting buffer or cell background would not correct the background effect of UV on DHR123mediated increase in fluorescence in these assays.

Reagents and buffers
Dulbecco's Modified Eagle's Medium (DMEM) and Roswell Park Memorial Institute medium (RPMI) were purchased from Wisent Inc. (Quebec, Canada). Hanks' Balanced Salt Solution (HSBB) and Phosphate Buffered Solution (PBS) were purchased from Life Technologies (California, United States). DHR123 and Dimethyl Sulfoxide (DMSO) were purchased from Sigma-Aldrich (Missouri, United States). Hydrogen peroxide (H 2 O 2 ) was purchased from Thermo Fisher Scientific (Massachusetts, United States).

Cell culture
Jurkat T cells (an immortalized line of human T lymphocytes) were cultured in RPMI. A549 cells (a human lung epithelial cells line, ATCC), mouse NIH/3T3 fibroblast cells (a fibroblast from mouse embryo, ATCC) and HL-60 cells (a human promyelocytic leukemia cell line, ATCC) were cultured in DMEM. All media were supplemented with 10% (v/v) fetal bovine serum (Invitrogen), and the cells were grown at 37°C in the presence of 5% CO 2 (v/v). A concentration of 1 9 10 6 cells per mL was maintained throughout the experiments.

Human primary T cell and neutrophils isolation
Peripheral blood from healthy donors was obtained with informed consent, and the study was approved by the Research Ethics Board of the Hospital for Sick Children. Peripheral blood mononuclear cells (PBMCs) were isolated by gradient centrifugation in Lymphoprep (Axis-Shiel, Oslo, Norway) following the manufacturer's instructions. The isolation of human T cells was performed using the Human T-cell Enrichment Kit (STEM-CELL Technologies Inc, BC, Canada). Isolated T cells were re-suspended in RPMI supplemented with 10 mmol/ L HEPES for UV irradiation experiments.
Human peripheral neutrophils were isolated from blood collected in K2 EDTA blood collection tubes (Becton, Dickinson and Co.) using PolymorphPrep (Axis-Shield) according to the manufacturer's recommendation. Isolated neutrophils were then re-suspended in RPMI medium (Invitrogen) supplemented with Hepes buffer (10 mmol/L, pH 7.2) for UV irradiation and PMAmediated ROS production experiments.

UV irradiation
BD 96-well plates were used for irradiation experiments and irradiation with UV light was performed using the Stratagene UV Stratalinker 2400 which emits UV light at 254 nm wavelengths and delivers 75 J/sec.

Fluorescence plate reader assays
The Spectra Max Gemini EM plate reader (Molecular Devices) was used for fluorescence measurement. To detect and quantify ROS production from cells, 25 lmol/L DHR123 was used. DHR123 was either incubated with the buffer alone or with the cells at 37°C for 15 min, followed by three washes at 400g for 5 min. The Spectra Max Gemini EM plate reader settings, 488 nm for excitation and 525 nm for emission, were used for measuring the R123 green fluorescence signal, unless otherwise stated. BD Falcon 96-well plates and 100 lL of solution per well were used in all experiments.
Evaluation of the effect of increased irradiation with UV light on DHR signal DHR123 (different concentrations, 0-20 lmol/L) in 100 lL HSBB was placed in the wells of a 96-well plate. The wells were irradiated with UV light for various periods of time ranging from 0 to 5 min. Fluorescence intensity was measured using the Spectra Max Gemini EM plate reader.

Evaluation of the effect of media, buffer and DMSO concentration on DHR signal
Solutions with different concentrations of DMSO (0-1.25%, v/v) with and without 25 lmol/L DHR123 were prepared in two media (RPMI and DMEM) and two buffers (PBS and HSBB). The volume of 100 lL of each buffer solution was loaded onto 96-well plates and irradiated for 30 sec with UV light. Fluorescence intensity was measured using the Spectra Max Gemini EM plate reader.
Measuring ROS production in cells incubated with DHR123 before UV irradiation A549 epithelial cells, 3T3 fibroblast cells, Jurkat T cells, primary T cells, HL-60 differentiated to neutrophils and primary neutrophils (10 6 cells/mL) were incubated with 25 lmol/L DHR123 for 15 min. Cells were then centrifuged twice at 400g for 5 min and re-suspended in fresh RPMI media. A volume of 100 lL of cells suspension was plated in a well within a 96-well plate (10 5 cells per well). The cells were treated with and without PMA and irradiated for 30 sec with UV light. Oxidized R123 green fluorescence signal was measured using the Spectra Max Gemini EM plate reader.
Measuring ROS production in cells incubated in DHR123 after UV irradiation A549 epithelial cells, 3T3 fibroblast cells, Jurkat T cells, primary T cells, HL-60 differentiated to neutrophils and primary neutrophils (10 6 cells/mL) were plated onto BD Flacon 24-well plates (10 6 cells per well) and irradiated for 30 sec with UV light. The cells were incubated for 15 min with 25 lmol/L DHR123. The cells were then centrifuged at 400g for 5 min and resuspended in fresh RPMI media. The procedure was repeated twice. The cells were plated onto 96-well plates (10 5 cells per well). ROS signal was measured using the Spectra Max Gemini EM plate reader.

Amount of DHR remaining after irradiation with UV light
Solutions with various concentrations of H 2 O 2 , 1:10,000 donkey-anti rabbit HRP-conjugated antibody as a source of peroxidases and 25 lmol/L DHR123 were prepared in HSBB. A volume of 100 lL solutions was loaded onto 96-well plates and irradiated for 30 sec with UV light. Fluorescence intensity was measured using the Spectra Max Gemini EM plate reader.

Flow Cytometry
To evaluate ROS production in Jurkat T cells and HL-60 cells, cells were suspended in RPMI media supplemented with 10 mmol/L HEPES. Cells were further treated with 25 lmol/L DHR123 by incubation at 37°C for 15 min followed by three washes. One batch of cells was treated with 25 nmol/L PMA and the other batch of cells was irradiated for 30 sec. Irradiated, PMA-treated and control cells without irradiation or PMA treatment were cultured over 1 h, and ROS production was analyzed by counting 10 4 events using the 488 filter and the FITC laser of the Gallios (Beckman Coultler Inc) flow cytometer.

Fluorescence Microscopy
One-half of the cells used for flow cytometry were used for fluorescence microscopy, and the experiment (UV or

Irradiating DHR123 with UV light generates green fluorescence
In order to determine the direct effect of UV irradiation on DHR123, we measured the intensity of the peak fluorescence in the green range of the spectrum (525 nm) that was generated after irradiating various concentrations of DHR123 with UV light for different time periods. At each DHR123 concentration (5-20 lmol/L), exposure of the probe to UV light for increasing durations (0-5 min) increased the green fluorescence signal in direct proportion to the duration of exposure to UV light (Fig. 1A). The slopes of these graphs increased with increasing DHR123 concentration, showing that more DHR123 is oxidized at higher concentrations ( Fig. 1B).
To determine the effect of a variety of buffers and media on UV-induced oxidative effect on DHR123, a concentration of 25 lmol/L DHR123 was diluted in RPMI, DMEM, HBSS and PBS, and the solutions were irradiated with UV for 30 sec. Green fluorescence (525 nm) measurements showed higher fluorescence intensity for RPMI medium, followed by HBSS and PBS buffers (P < 0.05). DMEM media showed the lowest fluorescence response after irradiating DHR123 with UV light ( Fig. 2A).
DMSO is often used as a solvent for DHR123 and other compounds. To determine the effect of DMSO on the UV-induced increase in green fluorescence signal of DHR123, experiments were repeated with various concentrations of DMSO in these four buffers and media. The presence of DMSO in DMEM increased the amount of green fluorescence; however, further increases in DMSO concentration did not affect the green fluorescence signal. Furthermore, the presence of DMSO at any of the tested concentrations (0-1.25%, v/v) had little effect on DHR fluorescence signal in RPMI media. By contrast, increasing concentrations of DMSO steadily increased the green fluorescence signal of DHR123 in PBS and HBSS buffers (Fig. 2B). Therefore, concentration of dye, duration of UV irradiation, types of media, and types of buffers all differentially affect DHR123 green fluorescence signal. This point is important because during experiments conducted with cells, the background is typically subtracted using buffers and cells with no UV irradiation. Since these control buffers and cells do not emit green fluorescence without UV irradiation, any UV-induced DHR123 background signal will not be corrected.
To determine whether UV irradiation depletes the DHR123 available to react with ROS, we conducted a solution-based assay. The DHR123 was irradiated with UV as shown above (Fig. 1) and further oxidized with H 2 O 2 in the presence of peroxidase enzyme. UV irradiated DHR123 increased green signal compared to nonirradiated DHR123 at the 0 time point. In both of these samples (UV irradiated or non-irradiated DHR123), green fluorescence signal increased over time. Compared to the 0 time point, the signal increased by a factor of >3.75 at the 3-h time point (Fig. 3A, B). Thus, there are large amounts of unoxidized DHR123 available to react with ROS even after irradiating the dye with UV. Furthermore, at the later time points, the differences between the UVirradiated and non-irradiated samples became small, suggesting that when large amounts of ROS are produced, the presence of pre-oxidized DHR123 has little effect on the ability to detect ROS by the remaining DHR123.
Rapid increase in green fluorescence in DHR123 is attributable to UV irradiation DHR123 is often used for measuring ROS production in various cells. Therefore, to determine whether UV irradiation affects DHR123 signal in the context of cell-based experiments, we used A549 epithelial cells (Fig. 4A), 3T3 fibroblasts (Fig. 4B), Jurkat T cells (Fig. 5A), and primary T cells (Fig. 5B). These cells were cultured in DMEM or RPMI, incubated with or without DHR123, washed and re-suspended in RPMI, irradiated with UV light, and the green fluorescence signal was measured. DHR123-loaded    cells, 3T3 fibroblasts, Jurkat and primary T cells did not generate ROS detectable by DHR123, instead UV irradiating the cells that contained DHR123 showed a spike in increased green signal. This initial signal increase is attributable to the direct effect of UV light on DHR123. Therefore, this assay gives false positive results when UV irradiating DHR123-loaded epithelial, fibroblast and T cells.

DHR123 is good at detecting large amounts of ROS in neutrophils
Neutrophils are known to generate large amounts of ROS after activating their NADPH oxidase with PMA. Therefore, to determine the effect of UV irradiation and NADPH oxidase-dependent activation of neutrophils, we differentiated promyelocytic HL-60 cells into neutrophils (Douda et al. 2015). Differentiated HL-60 (Fig. 6A) and primary neutrophils freshly isolated from blood (Fig. 6B) were loaded with DHR123 and either irradiated with UV or incubated with PMA or both. Monitoring green fluorescence signal showed that neutrophils, but not Jurkat T cells (Fig. 5), generated large amounts of ROS that are detectable by DHR123 (Fig. 6). Similar to Jurkat and primary T cells, UV irradiation of DHR123-loaded neutrophils also showed the initial spike in green fluorescence (note the scale difference between Figs. 5, 6). However, generation of a large amount of ROS still allows the detection of PMA-induced NADPH oxidase-mediated production of ROS by this method.

Flow cytometry and fluorescence microscopy are more suitable for detecting ROS in cells compared to plate reader assays
The presence of extra DHR123 in the culture solutions or inside the cells during UV irradiation or PMA treatment could give erroneous results in plate-reader assays. To evaluate whether flow cytometry and fluorescence microscopy are better methods, we conducted experiments with Jurkat T cells and neutrophils differentiated from HL-60 cells. These cells were first incubated with 25 lmol/L DHR123, washed, and then either treated with PMA or exposed to UV light. Cells were subsequently analyzed by flow cytometry after incubating them at 37°C for 1-h. UV-irradiated Jurkat T cells showed no significant difference in fluorescence intensity compared to non-irradiated or PMA treated cells (Fig. 7A). By contrast, HL-60 neutrophils showed a high level of ROS production that is readily measurable by DHR123 when treated with PMA, compared to non-treated cells or cells irradiated with UV (Fig. 7B). Median fluorescence intensity values show that DHR123-loaded HL-60 neutrophils incubated with PMA have the highest fluorescence intensity after 1-h incubation period (Fig. 7C).
Imaging these cells by fluorescence microscopy showed that incubation of Jurkat T cells and HL-60 cells with DHR after UV irradiation, or without any stimulation for 1-h resulted in the detection of very little signal (Fig. 8).
By contrast, large numbers of HL-60 neutrophils, but not Jurkat cells, show intense green signal after stimulating the DHR123-loaded cells with PMA. These data indicate that fluorescent microscopy and flow cytometry are suitable tools to study ROS production inside the cells that produce large amounts of ROS using DHR123.

Discussion
DHR123 is a common fluorescence probe used for analyzing the production of ROS in cells undergoing oxidative stress (Gomes et al. 2005). However, the effects of using DHR123 in experiments involving UV irradiation of a variety of cell types in various buffer conditions have not been clearly established. This study shows that the DHR123 concentration and the duration of UV irradiation each directly increase the intensity of DHR123 green fluorescence, and that DMSO further increases the green fluorescence in certain buffers and media. These are key points because typical subtraction of values for buffers and cells (without UV irradiation) would not account for the observed effect of UV on DRH123. A spike in green fluorescence signal was observed immediately after UV irradiation of DHR123-loaded A549 epithelial cells, 3T3 fibroblasts, and T cells. This initial increase is attributable to the direct effect of UV irradiation on DHR123, but not de novo ROS production. However, DHR123 can be used in HL-60 neutrophils and primary neutrophils to detect ROS production. In terms of techniques, flow cytometry and fluorescence microscopy are superior to plate reader assays for measuring ROS production. Therefore, using DHR123 for measuring ROS in UV-based experiments could give false-positive results, and require great care in selecting experimental approaches. Data obtained from a cell-free system clearly show that UV irradiation directly increases the green fluorescence of DHR123 in concentration-and duration of UV irradiation-dependent manners (Fig. 1). This increase is attributable to UV light oxidizing DHR123 to R123, consistent with the data obtained from a previous study, which suggests that UV irradiation can increase DHR123 oxidation without the presence of any ROS (Boulton et al. 2011). The type of UV used for irradiation (e.g., UV-B, UV-C) could also differentially affect DHR123 fluorescence. UV-B in particular has been shown to generate a higher green signal from DHR123 compared to UV-C (Horikawa-Miura et al. 2007). Higher fluorescence intensity of DHR123 detected in various buffers is attributable to the different types of pigments and buffer salts present in the media (Fig. 2) because some of these components could increase the oxidation of DHR123 (Gomes et al. 2005). DMSO is a reagent commonly used as a carrier of various drugs and compounds in both in vitro and in vivo experiments (Champelovier et al. 2013;Douda et al. 2014Douda et al. , 2015. The fluorescence intensity remains unchanged for RPMI and DMEM compared to PBS and HBSS in the presence of various DMSO concentrations (0-1.25%, v/v). A previous study showed that DMEM could increase green fluorescence of DHR123 (Cunningham et al. 1985) and that DMSO increases the fluorescence of DHR123 in   Redox-active metals such as Fe 2+ , and Cytochrome C in mitochondria during UV-induced apoptosis also oxidize DHR123 (Qian and Buettner 1999). Moreover, compounds such as riboflavin and some ions present in commonly used culture media can oxidize DHR123 and interfere with DHR123-mediated detection of ROS. Furthermore, HOCl and other compounds, but not H 2 O 2 itself can oxidize DHR123 (Crow 1997). Therefore, buffers have to be carefully chosen when studying ROS with DHR123 and UV irradiation.
It is a common practice to pre-load DHR123 into cells and tissues, and UV irradiate them or treat them with PMA to monitor ROS production during different types of cell deaths (Douda et al. 2014(Douda et al. , 2015Poniedzialek et al. 2015). This study shows that UV irradiation of DHR123preloaded A549 epithelial cells, 3T3 fibroblasts, Jurkat T cells, or primary T cells results in a rapid increase in fluorescence intensity compared to control cells (Figs. 4, 5). Nevertheless, adding DHR123 after UV irradiation shows no such increase in fluorescence signal. Thus, DHR123 is not useful for measuring ROS in A549, 3T3, Jurkat T cells, and primary T cells in the context of cell death. DHR123 has been used in lymphocytes to measure ROS (Poniedzialek et al. 2015); however, the data presented here  indicate that caution is needed in the use of DHR123 in such studies. Interpreting DHR123 results would be expected to be more complicated in cultures with mix cell populations. This is because different cells induce different degrees of ROS, and that soluble or pre-loaded DHR123 present in the cultures will contribute to erroneous interpretations of the degree of ROS production, particularly when plate reader assays are used.
Although the use of DHR123 is problematic in many cells that produce relatively small amount of ROS via mitochondrial pathways during UV irradiation, neutrophils generate large amount of ROS when treated with an NADPH oxidase activator such as PMA (Douda et al. 2014(Douda et al. , 2015. Similar to non-neutrophil cells (A549 epithelial cells, 3T3 fibroblasts and T cells), neutrophils (HL-60 and primary cells) also generate nondetectable ROS upon UV irradiation (Figs. 4-6). However, a large amount of ROS is detectable by the DHR123-based method in HL-60 neutrophils and primary neutrophils activated with PMA (Fig. 6). In fact, H 2 O 2 does not directly react with DHR123 and it requires peroxidase to convert H 2 O 2 into other intermediates that react with DHR123 and generate R123 (Gomes et al. 2005). Cell-free experiments show that even though some of the DHR123 molecules were oxidized upon UV irradiation, the remaining dye molecules are sufficient for detecting ROS in cells such as neutrophils (Figs. 3, 6). The amount of DHR123 molecules remaining in human skin HaCaT keratinocytes after UV irradiation is reported to be not sufficient to react with ROS (Boulton  In addition to plate reader assays, flow cytometry and fluorescence microscopy have been used for measuring ROS production using DHR123. This study shows that flow cytometry and fluorescence microscopy provide reliable results (Figs. 7, 8). The use of multiple techniques provides additional assurance of the correct interpretation of ROS production data measured with DHR123 (Douda et al. 2014(Douda et al. , 2015. Collectively, to correctly interpret the DHR123-based ROS data, several factors such as buffers, media, DMSO concentration, UV irradiation, and the types of cells should be taken into consideration.