Mechanism of Plasma-Induced Immunogenic Cancer Cell Death

Experimental Design: The overall objective of the study was to uncover the mechanism by which non‐thermal plasma elicits immunogenic cancer cell death by evaluating the role of RONS generated. This was accomplished by 1) determining an ICD‐inducing regime of plasma, 2) identifying the RONS generated in that regime, and 3) delineating their effect by comparing direct DBD plasma treatment and treatment with exogenously prepared RONS solutions. Our initial screenings of RONS effect on ICD were performed in vitro on two melanoma cell lines and validated in an in vivo vaccination assay using syngeneic mice.
Cell Lines: The B16F10 murine melanoma cell line and the A375 human melanoma cell line were purchased from the American Type Culture Collection. Both cell lines were cultured in complete Dulbecco's modified Eagle medium containing 10% fetal bovine serum, 100 U mL⁻¹ penicillin, 100 µL streptomycin, and 4 × 10⁻³m l‐glutamine. Cells were cultured in a humidified environment at 37 °C with 5% CO₂.
Microsecond‐Pulsed DBD Plasma Parameters for Direct Treatment: A microsecond‐pulsed power supply was purchased from Advanced Plasma Solutions, and a 1.25 cm diameter copper DBD electrode was used for treatment in 24‐well plates. The copper electrode was covered with a 0.5 mm fused‐silica dielectric (Technical Glass) to prevent current arching. The microsecond‐pulsed power supply generated 17 kV pulses with ≈5 µs rise times and ≈1.5 µs pulse widths. The duty cycle was fixed at 100%.
Both cell lines were seeded into 24‐well plates at 3 × 10⁵ cells mL⁻¹ (0.5 mL per well) 1 d prior to plasma treatment. On the day of plasma treatment, medium was removed and cells were washed twice with PBS to remove serum and other organics from cell culture medium. PBS from the second wash was left in the well until right before plasma treatment. For DBD plasma treatment, PBS was removed and the DBD electrode was lowered into the well and positioned 1 mm above the cells with a z‐positioner (Figure 2A,B). Plasma was then discharged directly on the cells for 10 s at various pulse frequencies (50, 100, 250, and 500 Hz). Following plasma treatment, 500 µL of fresh cell culture medium was immediately added back into the well. Cells were incubated at 37 °C with 5% CO₂ for 24 h until further analysis. Mitoxantrone dihydrochloride (Sigma‐Aldrich, ≥97%, M6545), a chemotherapeutic used as a positive control, was diluted to a 5 mg mL⁻¹ stock solution and prepared into a working solution of 2 µg mL⁻¹ in complete medium. Cells were incubated with mitoxantrone for 24 h before collection and analysis.
Treatment with Pulsed Electric Fields: On the day of treatment, cells were washed with PBS and 1 mL of PBS or RONS solution (700 × 10⁻⁶m of H₂O₂, 1770 × 10⁻⁶m of NO₂⁻, and 35 × 10⁻⁶m of ONOO⁻) was added to each well immediately before treatment. The DBD electrode was then submerged into the liquid and operated at 17 kV and 500 Hz for 10 s, 1 mm above the cells. This method was to subject the cells to PEF generated from the microsecond‐pulsed power supply and DBD electrode without the production of plasma (Figure 2C). Following 10 s treatment, all the liquid was removed from the well and fresh cell culture medium was added. Cells were then incubated at 37 °C with 5% CO₂ for 24 h until further analysis.
DBD Plasma Treatment of Liquid: For liquid analysis, 50 µL of liquid (PBS or deionized water) was added into a 24‐well plate and distributed evenly across the bottom. The DBD electrode was positioned 1 mm above the surface of the liquid using the z‐positioner (Figure 2D). DBD plasma was generated at fixed voltage (17 kV), while the pulse frequency and treatment time were varied.
For preparation of plasma‐treated PBS for treatment of cells, PBS was treated for 100 s at 500 Hz. Prior to finishing the 100 s treatment, PBS was removed from the wells containing cells. All 50 µL of plasma‐treated PBS was then collected and added onto cells and 45 µL was removed. After 10 s treatment with 5 µL of remaining plasma‐treated PBS, 500 µL of complete medium was added back into the well and the cells were incubated for 24 h until further analysis.
Preparation of RONS Solutions and Treatment: The solutions of H₂O₂, NO₂⁻, and NO₃⁻ were prepared from commercially available H₂O₂ (Sigma‐Aldrich, ≥30%, 95321), sodium nitrite (NaNO₂) (Sigma‐Aldrich, ≥97%, 237213), and potassium nitrate (KNO₃) (Sigma‐Aldrich, ≥99%, P8394) dissolved in PBS (without iron, calcium, and magnesium). ONOO⁻ was prepared from NaOONO (Cayman Chemicals, ≥90% solution in 0.3 m sodium hydroxide, 14042‐01‐4) dissolved in PBS. Four different RONS solutions were prepared1)

H₂O₂/NO₂⁻/NO₃⁻—H₂O₂: 700 × 10⁻⁶m; NO₃⁻: 410 × 10⁻⁶m; NO₂⁻: 1360 × 10⁻⁶m

2)

H₂O₂/NO₂⁻—H₂O₂: 700 × 10⁻⁶m; NO₂⁻: 1770 × 10⁻⁶m

3)

ONOO⁻—35 × 10⁻⁶m

4)

H₂O₂/NO₂⁻/ONOO⁻—H₂O₂: 700 × 10⁻⁶m; NO₂⁻: 1770 × 10⁻⁶m; ONOO⁻: 35 × 10⁻⁶m

On the day of treatment, the cells were washed twice with PBS, to follow identical handling procedures with plasma treatment. PBS from the second wash was removed immediately before treatment and 50 µL of RONS solution was added into the well and rocked to ensure even distribution on cells. 45 µL was removed and the remaining 5 µL was left on the cells for 10 s. Following treatment, 500 µL of fresh, complete medium was added to the well and cells were incubated at 37 °C with 5% CO₂ for 24 h until further analysis. This procedure most closely mimics the process of direct DBD plasma treatment and is most realistic to the concentration of RONS generated by plasma and experienced by the cells.
Cell Survival Assay: Cell survival was quantified with a trypan blue exclusion test. Following 24 h incubation, cell supernatant was collected. Cells were then washed with 0.5 mL of PBS and detached with 200 µL of accutase. PBS from the wash was also collected with the cell supernatant. The cell suspension was collected, pooled with their supernatant and PBS wash, and homogenized by pipetting. A 50 µL sample was acquired and equal parts 0.4% trypan blue (Gibco, 15250‐061) was added to the sample. Cell counts were performed using a TC20 Automated Cell Counter (Bio‐Rad). The live cell concentration was recorded, and data were represented as a normalization to control.
CRT Expression from Cell Lines: CRT was measured using dual staining of PI and a monoclonal CRT antibody. Following 24 h after incubation, cells were washed with PBS, detached with 200 µL of accutase, and washed twice with 2 mL of FACS buffer (500 mL sheath fluid (BD Biosciences, 342003) + 2 g bovine serum albumin (Sigma, A9418) + 1 g NaN₃ (Merck, 1.06688.0100) in 100 mL H₂O). Each sample was split into two vials and one was stained with monoclonal primary rabbit anti‐CRT antibody (Abcam, ab196158) while the other was stained with rabbit IgG, monoclonal isotype control (Abcam, ab199091) for 40 min at 4 °C. Cells were then washed once with FACS buffer. 0.5 µL of PI was added to each sample immediately before being quantified with a flow cytometer. Fifteen thousand events were collected and only the PI− cells were analyzed for surface CRT expression. Data were analyzed and gated using the FlowJo software (FlowJo LLC, version 10). Data were expressed as percent CRT positive after accounting for nonspecific binding with their corresponding isotype. The gating strategy is described in detail in Figure S9 in the Supporting Information.
Mice and Antitumor Vaccination Assay: Thirty‐two 8‐week‐old female C57BL/6J mice were purchased from Charles River and housed in a pathogen‐free room at the Animal Center of the University of Antwerp. The sample size of this study (eight mice per group) was chosen using information in the literature7 and running an a priori power analysis using G*Power software (version 3.0.10). Input parameters included effect size (large, 0.8), α error probability (0.05), power (0.8), and number of groups (4). A total sample size of 24 was calculated to give an actual power of 0.859. Eight mice were randomly assigned to one of four groups, and housed during the whole of the experiment in four separate cages. Two mice from each group were housed in each cage. Investigators were not blind to the groups.
The vaccines for this assay were prepared from B16F10 melanoma cells exposed to 1) DBD plasma (500 Hz), 2) PEF + RONS (RONS: 700 × 10⁻⁶m of H₂O₂, 1770 × 10⁻⁶m of NO₂⁻, and 35 × 10⁻⁶m of ONOO⁻), or 3) mitoxantrone (2 µg mL⁻¹) in vitro, while untreated cells were used as a negative control. After treatment, cells were collected, washed twice with PBS, and resuspended in PBS at 10⁶ cells mL⁻¹. Cell suspension was incubated for 24 h at 37 °C with 5% CO₂ to reduce the viability of the cells and prevent subsequent tumor growth at the vaccination site.
On the day of vaccination, mice were shaved with electric clippers (to help visualize tumors) and injected with vaccine (10⁵ cells in ≈100 µL) on the right dorsal side. After 7 d, each mouse was injected with 10⁴ live B16F10 cells (in ≈100 µL) on the left dorsal side. Tumor size and growth were followed up to day 50 as defined prior to the start of the experiment. Three orthogonal diameters were measured using a digital caliper, and volumes were calculated using (4/3π)r₁ × r₂ × r₃. The humane study endpoint was set to when the total tumor volume exceeded 1500 mm³ or if tumors began to ulcerate. All animal experiments were approved by the University of Antwerp Animal Research Ethical Committee (ECD‐dossier 2017‐53).
Detection of H₂O₂: The H₂O₂ concentration was detected using potassium oxotitanate dehydrate (Alfa Aesar, 89620) solution in H₂O and H₂SO₄ (Sigma‐Aldrich, 95–98%, 258105M). Concentration of H₂O₂ in plasma‐treated samples was determined by UV–vis measurements performed on a Genesys 6 (Thermo Fischer) spectrophotometer with quartz cuvettes (10 mm light path, 2 mm internal width). Titanium(IV) reagent (0.1 m Ti, 5 m H₂SO₄) was prepared by dissolving 0.354 g of potassium bis(oxalato)oxotitanate(IV) dihydrate in a mixture of 2.72 mL of sulfuric acid and diluted to 10 mL with Milli‐Q water. 50 µL of plasma‐treated sample was added to the cuvette and diluted with 150 µL of PBS. 50 µL of sodium azide (NaN₃) (Sigma‐Aldrich, ≥99.5%, S2002) was added to the cuvette and thoroughly mixed. Afterward, 50 µL of Ti sulfate solution was added and homogenized. Air bubbles in the cuvette were eliminated with a sonicator (Branson 3200 ultrasonic bath) and water droplets were wiped from the cuvette before reading at 400 nm.
Detection of NO₂⁻ and NO₃⁻: A nitrate/nitrite colorimetric assay kit (Cayman Chemical, 780001) was used according to the provided protocol. To detect NO₂⁻ only, 50 µL of Griess reagent 1 (sulfanilamide) was added to each sample in a 96‐well plate, and 50 µL of Griess reagent 2 (N‐(1‐naphthyl)ethylenediamine) was immediately added afterward. The absorbance wavelength was read with a microplate reader Infinite 200 Pro (Tecan) at 540 nm. To detect NO₃⁻ and NO₂⁻, a nitrate reductase mixture (Cayman Chemical, 780010) and an enzyme cofactor mixture (Cayman Chemical, 780012) were added to each sample prior to the addition of Griess reagents. This allowed for the conversion of nitrate into nitrite. The absorbance was measured in duplicates and the samples were prepared in triplicates. The concentrations were calculated based on the obtained calibration curve.
EPR Spectroscopy Analysis: 50 µL capillaries (Ringcaps) were used to collect plasma‐treated samples, and a MiniScope MS200 spectrometer (Magnettech) was used to perform the analysis. After each plasma exposure experiment, the samples were immediately placed into a capillary tube. The overall time between exposure and analysis was 1 min. The general EPR parameters were as follows: frequency 9.4 GHz, power 3.16 mW (31.6 mW in case of (MGD)₂Fe₂⁺–NO), modulation frequency 100 kHz, modulation amplitude 0.1 mT, sweep time 30 s, time constant 0.1, and sweep width 15 mT. The simulated spectrum was double integrated to determine the concentrations reported here. Simulations were performed using hyperfine values obtained from the literature in the Spin Trab Database (National Institute of Environmental Health Sciences, 2018). EPR calibration was performed using solutions of 4‐hydroxy‐TEMPO (Sigma‐Aldrich, 97%, 176141) as reported elsewhere.33 Specific spin traps and other molecules were used to detect RONS in the liquid (Table2). All recorded experimental EPR spectra and simulations are shown in Figure S1 in the Supporting Information, along with the corresponding hyperfine values used. All fixed‐pulsed experiments presented in the results were performed with three to five replicates unless otherwise specified.
Detection of O/¹O₂/O₃: TEMP spin trap was dissolved in PBS (50 × 10⁻³m) to detect O/¹O₂/O₃ following DBD plasma treatment (Sigma‐Aldrich, ≥99%, 115754). TEMP reacts with these oxygen species to form the spin adduct TEMPO, which can be detected by means of EPR spectrometry. To determine the contribution of O/O₃, 100 × 10⁻³m sodium azide (NaN₃) (Sigma‐Aldrich, ≥99.5%, S2002) was added to the TEMP solution before plasma treatment to quench ¹O₂. Therefore, the collected spectrum of TEMPO was a result of the remaining oxygen species.
Detection of ^•OH and O₂^•− with EPR Spectroscopy: DEPMPO spin trap (Enzo Life Sciences, ≥99%, ALX‐430‐093) was dissolved in PBS (100 × 10⁻³m) to detect ^•OH and O₂^•−. DEPMPO reacts with O₂^•− to produce DEPMPO–OOH while it reacts with ^•OH to produce DEPMPO–OH. Experiments using 50 Hz pulse frequency were performed once with treatment times of 10 s or more.
Detection of ^•NO with EPR Spectroscopy: The spin probe PTIO (Enzo Life Sciences, ≥98%, ALX‐430‐007) was dissolved in PBS (200 × 10⁻⁶m) to detect ^•NO from DBD plasma treatment. ^•NO reacts with PTIO to form PTI that can be detected through EPR spectroscopy. When pulse frequency was varied from 50 to 500 Hz, treatment time was fixed at 50 s in order to generate detectable levels of ^•NO.
The MGD spin trap (Enzo Life Sciences, ≥98%, ALX‐400‐014) was also used to detect ^•NO. MGD was dissolved in deionized water (20 × 10⁻³m) and combined with Fe(II)SO₄·7H₂O (4 × 10⁻³m) (Sigma‐Aldrich, ≥99%, 215422). This solution was treated with DBD plasma and Na₂S₂O₃ (100 × 10⁻³m in deionized water degassed with argon) (Sigma‐Aldrich, ≥98%, 72049) was immediately added to the sample and collected for EPR analysis. When pulse frequency was varied from 50 to 500 Hz, treatment time was fixed at 120 s in order to generate detectable levels of ^•NO.
Detection of ONOO⁻with LC–MS: Solutions of 100 × 10⁻⁶m l‐tyrosine (Sigma‐Aldrich, ≥98%, T‐3754) and 100 × 10⁻⁶m diethylenetriaminepentaacetic acid (Sigma‐Aldrich, ≥98%, D1133) in 2 × PBS were exposed to plasma for a given period of time, as described by Wende et al.31 The solutions were collected and flash frozen immediately after exposure.
The separation and detection of 3‐nitrotyrosine was done by a Waters ACQUITY ultraperformance liquid chromatograph (UPLC) coupled to a Waters triple quadrupole mass spectrometer (Xevo TQ MS). The used column was a Waters ACQUITY UPLC HSS T3 2.1 mm × 100 mm column (1.8 µm particles), heated to 40 °C. The 9 min gradient was used for separation with A) water containing 0.1% formic acid and B) acetonitrile containing 01% formic acid, at a flow rate of 0.6 mL min⁻¹: 0–1.0 min 2% B, 1.0–4.0 min 2% to 18% B, 4.0–5.0 min 18% to 97% B, 5.0–6.0 min 97% B, 6.0–7.0 min 97% to 2% B, and 7.0–9.0 min 2% B. The parameters used in electrospray ionization tandem mass spectrometry in positive mode were as follows: capillary, 0.5 kV; cone, 22 V; extractor, 3 V; source temperature, 150 °C; desolvation temperature, 600 °C; desolvation gas flow, 1000 L h⁻¹; cone gas flow, 0 L h⁻¹; collision gas flow, 0.15 mL min⁻¹; collision energy, 2 V.
A multiple reaction monitoring (MRM) method application was optimized for the detection of tyrosine (transition m/z 182–136) and 3‐nitrotyrosine (transition m/z 227–181). For calibration of 3‐nitrotyrosine, eight standard solutions were made ranging from 0 to 10 × 10⁻⁶m and analyzed using the MRM method. The samples were diluted to a starting concentration of 10 × 10⁻⁶m tyrosine in 95% water and 5% acetonitrile containing 0.1% formic acid.
Statistical Analysis: Statistical differences for cell survival and CRT expression were analyzed using the linear mixed model with JMP Pro 13 (SAS software). The fixed effect was the treatment, and the random effects included were the different dates the experiment was performed and the flasks the cells used were split from. The interactions between the treatment and the date as well as interactions between the treatment and the flasks were tested. The random slope model was used when the interactions were significant (P < 0.05) and the random intercept model was used in all other cases. The fixed effect tests determine whether there was a significant difference between treatments (P < 0.05). When the difference is significant, the Dunnett's test for statistical significance was used to calculate adjusted P value compared to the control. A P value of <0.05 was considered statistically significant. For all in vitro experiments, treatment conditions were performed in duplicates on the same day and repeated on three separate days as a minimum. The total number of observations for each treatment group is defined in the figure or figure legend. The survival curve of the vaccination study was prepared in Graphpad Prism and compared using the log‐rank (Mantel–Cox) test. A P value of <0.05 was considered statistically significant. All figures were prepared in Graphpad Prism (Graphpad Software). For all chemical species, a nonlinear regression was used to determine the best‐fit line and R² value with a y‐intercept constraint at zero. Analysis was performed and figures were prepared in Graphpad Prism (Graphpad Software). No data were excluded.