Effects of 1800 MHz radiofrequency fields on signal transduction and antioxidant proteins in human A172 glioblastoma cells

Abstract

Purpose

To assess the effects of 1800 MHz radiofrequency electromagnetic field (RF-EMF) exposure on the expression of signal transduction and antioxidant proteins in a human-derived A172 glioblastoma cell line.

Materials and methods

Adherent human-derived A172 glioblastoma cells (1.0 × 10⁵ cells per 35 mm culture dish, containing 2 mL DMEM media) were exposed to 1800 MHz continuous-wave (CW) or GSM-modulated RF fields, in the presence or absence of serum for 5, 30 or 240 min at a specific absorption rate (SAR) of 0 (sham) or 2.0 W/kg. Concurrent negative (vehicle) and positive controls (1 µg/mL anisomycin) were included in each experiment. Cell lysates were collected immediately after exposure, stabilized by protease and phosphatase inhibitors in lysis buffer, then frozen and maintained at −80 °C until analysis. The relative expression levels of phosphorylated- and total-signal transduction proteins (CREB, JNK, NF-κB, ERK1/2, Akt, p70S6K, STAT3 and STAT5) and antioxidant proteins (SOD1, SOD2, CAT, TRX1, PRX2) were assessed using Milliplex magnetic bead array panels and a MagPix Multiplex imaging system.

Results

In cells exposed to 1800 MHz continuous-wave RF-EMF with the presence of serum in the culture medium, CAT expression was statistically significantly decreased after a 30 min exposure, total JNK was decreased at both 30 and 240 min of exposure, STAT3 was decreased after 240 min of exposure and phosphorylated-CREB expression was decreased after 30 min of exposure. In cells exposed to 1800 MHz GSM-modulated RF-EMF in serum-free cultures, the expression level of total STAT5 was decreased after 30 and 240 min of exposure. These observed changes were detected sporadically across time-points, culture conditions and RF-EMF exposure conditions indicating the likelihood of false positive events. When cells were treated with anisomycin for 15 min as a positive control, dramatic increases in the expression of phosphorylated signaling proteins were observed in both serum-starved and serum-fed A172 cells, with larger fold change increases in the serum-free cultures. No statistically significant differences in the expression levels of SOD1, SOD2 or TRX1 were observed under any tested conditions after exposure to RF-EMF.

Conclusions

The current study found no consistent evidence of changes in the expression of antioxidant proteins (SOD1, SOD2, CAT or TRX2) or a variety of signal transductions proteins (CREB, JNK, NF-κB, ERK1/2, Akt, p70S6K, STAT3, STAT5) in a human-derived glioblastoma A172 cell line in response to exposure to 1800 MHz continuous-wave or GSM-modulated RF-EMF for 5, 30 or 240 min in either serum-free or serum-containing cultures.

Keywords:

• Signal transduction
• radiofrequency
• A172
• glioblastoma
• in vitro

Introduction

The continued expansion of wireless networks and personal wireless devices has caused ongoing public concern about the existence of possible health risks associated with radiofrequency electromagnetic field (RF-EMF) exposure. One of the predominant concerns associated with RF-EMF exposure is the possible increased risk of cancer (Hardell and Carlberg 2021; IARC 2013). While RF-EMF radiation does not have sufficient energy to damage DNA directly through ionization, it has been hypothesized that low-level RF-EMF exposure may be capable of inducing oxidative stress within cells and altering the activation of certain signal transduction (ST) pathways (for a review see Yakymenko et al. 2016). Oxidative DNA damage as well as the alteration of ST pathways that regulate DNA damage repair and cell proliferation have both been identified as key events in the development of carcinogenesis (Sosa et al. 2013; Juliano 2020).

Dysregulation of metabolic pathways may result in oxidative stress, whereby increased concentrations of reactive oxygen species (ROS) lead to oxidative damage to DNA, lipids and proteins which in turn can lead to apoptosis, inflammation, and mitochondrial dysfunction (Lin and Beal 2006). Antioxidant enzymes, such as superoxide dismutase (SOD), catalase (CAT), thioredoxin 1 (TRX1) and peroxiredoxin 2 (PRX2) play an important role in balancing ROS levels to prevent these effects. SOD expression is modulated by ST proteins such as nuclear factor kappa B (NF-κB) (Miao and St Clair 2009). CAT also works to balance ROS levels by breaking down hydrogen peroxide. Like SOD, various ST proteins also regulate CAT expression (Glorieux et al. 2015). TRX1 is a cytosolic oxidoreductase protein that has a broad substrate specificity with the ability to reduce thiols (Watson et al. 2004), while PRX2 is a peroxidase protein that is capable of hydrolyzing hydrogen peroxide (Kim and Jang 2019). TRX1 is able to mediate redox-sensitive signal transduction through NF-κB (Nishiyama et al. 2001), while PRX2 can modulate signal transduction through PI3K/Akt and STAT3 pathways (Kim and Jang 2019). An important balance must be maintained between cellular ROS levels, antioxidant enzyme levels and ST protein activation levels to prevent potential harmful effects to the cell.

ST pathways are an innate system of interconnected and controlled cascades of molecular events that allow cells to regulate their internal processes and respond to stressors. Mitogen-activated protein kinases (MAPKs), such as extracellular signal-regulated kinases (ERKs) and c-Jun N-terminal kinases (JNKs), play a key role in ST pathways and they are regulators of many cellular processes such as cell proliferation, metabolism, and the environmental stressor response (Weston et al. 2002). MAPKs are activated through phosphorylation and in turn can phosphorylate and activate transcription factors such as cAMP response element-binding protein (CREB) and signal transducer and activator of transcription 3 (STAT3) either directly or indirectly (Minden and Karin 1997). This process has direct implications in gene transcription, as these molecular cascades result in the regulation of molecules that can directly influence cellular processes. For example, CREB is able to bind to specific areas of DNA in the regulatory region to regulate transcription and is involved in learning and memory in the brain (Carlezon et al. 2005). STAT3 also functions to regulate gene transcription by binding to promoter regions of target genes and is involved in many cellular functions including cell proliferation, cell migration and apoptosis (Guanizo et al. 2018). The PI3K/AKT/mTOR signaling pathway regulates a wide range of cellular functions in response to internal signal and external stressors, including cell survival, growth, differentiation, cellular metabolism, and cytoskeletal reorganization (Noorolyai et al. 2019). Due to involvement in multiple critical cellular functions, Akt dysregulation is often observed in human cancers (Vasudevan et al. 2009).

At levels below the localized exposure limits for the general public recommended by ICNIRP (2020) and IEEE (2019) in the 100 kHz to 6 GHz frequency range, a specific absorption rate (SAR) of 2 W/kg (averaged over 10 g of tissue for the head, neck or trunk), several studies have reported increased levels of ROS (Kim et al. 2017; Marjanovic Cermak et al. 2018) and increased ST protein phosphorylation (Leszczynski et al. 2002; Buttiglione et al. 2007; Friedman et al. 2007; Liu et al. 2014; Lu et al. 2014; Glushkova et al. 2015) in response to RF-EMF exposure in a variety of human and mammalian cell lines. However, other studies have not found similar findings (Lee et al. 2006; Simkó et al. 2006; Valbonesi et al. 2014; Silva et al. 2016; Zielinski et al. 2020), leading to uncertainty as to whether or not exposure to RF-EMF was the causal agent for the observed changes. It is unclear if the conflicting results in the literature are due to differences in the cell lines used, their replication status upon exposure, experimental conditions, serum status or variations in the endpoints assessed.

The purpose of this exploratory study was to assess if exposure of a human glioblastoma-derived A172 cell line to either continuous-wave or GSM-modulated 1800 MHz RF-EMF for 5, 30 or 240 min at a specific absorption rate (SAR) of 2 W/kg resulted in consistent and reproducible changes in the relative protein expression level or activation level of antioxidant and ST proteins in response to RF-EMF exposure. All endpoints in this study were assessed under both serum-containing and serum-free culture conditions, as it has been reported that ST protein and phosphoprotein responses may be more easily detectable under serum-starved culture conditions (Levin et al. 2010).

Methods and materials

Cells

Human glioblastoma-derived A172 cells were obtained from American Type Culture Collection (ATTC) and cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Cedarlane, Canada), supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich, USA), in a standard tissue culture incubator (5% CO₂, 37 °C; NuAire, USA). Cells were maintained in 75 cm² vented culture flasks (Fisher Scientific, Canada) and sub-cultured every 2–3 days. One day prior to experimentation, 1 × 10⁵ cells were sub-cultured into forty-eight 35 mm Nunc culture dishes (Fisher Scientific, Canada) containing 2 mL of DMEM. Cell viability was assessed using a Bio-Rad TC-10 automatic cell counter (Bio-Rad Laboratories, USA) using trypan blue exclusion. At four hours prior to experimentation, the medium was removed from all dishes. For 24 dishes, 2 mL of fresh complete DMEM (containing 10% FBS) was added while for the remaining 24 dishes, 2 mL of FBS-free DMEM (serum starved) was added. The cells were returned to the tissue culture incubator for 4 h prior to RF-EMF exposure. The cells were approximately 50% confluent at the time of RF-EMF exposure.

RF-EMF exposure

The RF-EMF exposure apparatus (sXc-1800) was built and calibrated by the IT’IS Foundation (Zurich, Switzerland). The sXc-1800 exposure system consisted of two identical rectangular waveguide exposure chambers, placed inside a single infrared water-jacketed CO₂ cell culture incubator (NuAire, USA) which ensured similar environmental conditions (37 °C, 80% RH, 5% CO₂) between the sham and RF-EMF exposed samples. The exposure system was fully automated, with exposure settings set prior to experimentation (frequency, modulation, duration, temperature recording, etc.). One of the exposure chambers was set as a sham control whereas the other exposed cell cultures to 1800 MHz RF-EMF at 2 W/kg. The system monitored functionality (incident field strengths, ventilator and exposure set-up functioning), and outlet air temperature (Pt100 temperature sensors, ±0.1 °C) at 10 second intervals for the duration of the exposure period. Temperature in the RF-EMF and sham exposed samples differed by less than 0.1 °C in all experiments. This exposure system and the associated dosimetric characterization has been previously described in detail by Schuderer et al. (2004).

At 4 h after media replacement, three culture dishes containing adherent A172 cells (exponential growth phase) in complete DMEM (with FBS) and three similar culture dishes without FBS (serum-starved) were randomly placed into each of two identical 6-slot customized culture dish holders, which were then inserted into two separate rectangular waveguides. The cell cultures were acclimatized within the exposure system for 15 min prior to initialization of exposure to either 1800 MHz RF-EMF or a time-matched sham condition (0 W/kg) for each of 5, 30 or 240 min, all at a SAR of 2.0 W/kg. Thus, two sham conditions (cells in complete media and cells in media lacking serum) and two RF-EMF exposure conditions (cells in complete media and cells in media lacking serum) were concurrently treated at each time-point for each experiment. One set of 4 independent experiments was conducted using 1800 MHz continuous-wave RF-EMF with a concurrent sham-exposed group. A separate set of 4 independent experiments was conducted using 1800 MHz GSM-modulated RF-EMF with a concurrent sham-exposed group. A separate set of complete or serum-starved cell cultures were concurrently exposed to negative control (15 µL DMSO) or positive control (15 µL anisomycin dissolved in DMSO, final culture concentration of 1 µg/mL) conditions in a separate incubator for 15 min prior to the end of each of the 240 min RF-EMF/sham exposure periods. All samples were coded to ensure blinding of the experimental treatment from the investigators until endpoint analysis was completed.

Protein isolation and coding of samples

Immediately after the RF-EMF/sham exposure period of 5, 30 or 240 min, culture dishes containing the adherent A172 cells were quickly removed from the waveguide apparatus. The culture medium was decanted and the adherent cells were washed with 2 mL of ice-cold Dulbecco’s phosphate buffered saline (PBS) solution (Cedarlane, Canada). The PBS was decanted, then 200 µL of cell lysis buffer, containing a protease and phosphatase inhibitor cocktail (Set III; EMD Millipore, Canada), was added to each culture dish. The dishes were then scraped and the cell lysates from 3 concurrent culture dishes per exposure condition were pooled (total of 600 µL), mixed by pipetting, then distributed into cryovials for storage at −80 °C until analysis. The protein content of each sample was assessed using the BCA Protein Assay Kit (EMD Millipore, Canada).

Analysis of relative total ST protein content and phosphorylation level

Aliquots (35 µL) of cell lysates and standards were thawed, then immediately assayed in triplicate using magnetic multiplexed fluorescence microsphere immunoassays with Milliplex MAP Kits (EMD Millipore, Canada), according to manufacturer’s instructions. The relative ST protein expression levels were assessed using the Milliplex Human Oxidative Stress Magnetic Bead Panel (H0XSTMAG-18K, EMD Millipore, Canada), the Milliplex MAP 9-Plex Multi-Pathway Kit (Total) (Cat#48-681MAG; EMD Millipore, Canada) and the Milliplex MAP 9-plex Multi-Pathway Magnetic Bead Signaling kit (phosphoprotein) (Cat#48-680MAG; EMD Millipore, Canada).

Briefly, the 96-well plate containing the analyte-conjugated beads were washed with assay buffer with a Bio‐Plex Handheld Magnetic Washer (Bio‐Rad, Canada) and incubated with detection antibodies provided by the manufacturer, washed again, then resuspended in sheath fluid and analyzed by a Bio‐Plex MAGPIX™ Multiplex Reader (Bio‐Rad, Canada). The relative expression level of each protein, represented by the median fluorescence intensity (MFI) per bead, was obtained and analyzed with Bio‐Plex Manager MP Software. Relative changes in expression of either phosphorylated or non-phosphorylated samples were quantified by normalizing the mean MFI by the total protein concentration for the sample obtained by the BCA protein assay.

Statistical analysis

All experiments were independently performed four times and the data are presented as the mean ± standard deviation (SD). Data sets were assessed for normality using the Shapiro-Wilk test and data sets that were not normally distributed were log or reciprocal transformed. Data obtained from RF-EMF versus sham at 5, 30 or 240 min were assessed for significance by a repeated-measures one-way ANOVA (effective matching) or an ordinary one-way ANOVA (when effective matching violated), with Bonferroni multiple-comparisons post-hoc tests used to identify pair-wise differences (p ≤ .05). For analysis of positive control (anisomycin) versus negative control (vehicle), the data were assessed for significance by a paired t-test (2-tailed; p ≤ .05).

Results

Antioxidant proteins

The relative expression levels of the antioxidant proteins SOD1, SOD2, CAT, TRX1 and PRX2 in A172 cell culture lysates from cultures exposed for 5, 30 or 240 min to RF-EMF or sham conditions in the presence or absence of serum are depicted in Tables 1 and 2. The expression levels of SOD1, SOD2 and TRX1 in serum-free cultures were elevated in all samples compared to cultures exposed in the presence of serum. PRX2 was not detectable in any culture or under any exposure condition (data not shown).

Table 1. Relative expression levels of antioxidant proteins (MFI/mg protein ± SD) after 1800 MHz continuous-wave radiofrequency field exposure when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	SOD1	SOD2	CAT	TRX1
5	Sham	+	1.72 ± 0.10	1.18 ± 0.23	0.11 ± 0.03	1.24 ± 0.06
5	2 W/kg	+	1.66 ± 0.15	1.12 ± 0.21	0.09 ± 0.03	1.23 ± 0.15
30	Sham	+	1.84 ± 0.13	1.30 ± 0.21	0.13 ± 0.04	1.30 ± 0.11
30	2 W/kg	+	1.49 ± 0.08	1.02 ± 0.22	0.08 ± 0.04^a	1.14 ± 0.05
240	Sham	+	2.00 ± 0.42	1.40 ± 0.33	0.12 ± 0.05	1.42 ± 0.30
240	2 W/kg	+	1.94 ± 0.59	1.36 ± 0.41	0.12 ± 0.03	1.38 ± 0.45
5	Sham	–	2.85 ± 0.28	1.65 ± 0.24	0.11 ± 0.02	2.17 ± 0.27
5	2 W/kg	–	2.39 ± 0.29	1.46 ± 0.42	0.09 ± 0.04	1.87 ± 0.18
30	Sham	–	2.68 ± 0.19	1.63 ± 0.50	0.10 ± 0.04	2.12 ± 0.08
30	2 W/kg	–	2.48 ± 0.25	1.61 ± 0.46	0.10 ± 0.05	2.01 ± 0.08
240	Sham	–	2.57 ± 0.41	1.66 ± 0.48	0.14 ± 0.08	1.87 ± 0.38
240	2 W/kg	–	2.28 ± 0.29	1.47 ± 0.37	0.11 ± 0.06	1.73 ± 0.30

Each data point represents the mean ± SD from 4 independent experiments.

^aDenotes significant difference (p ≤ .05) in relation to the time-matched sham condition.

Table 2. Relative expression levels of antioxidant proteins (MFI/mg protein ± SD) after 1800 MHz GSM-modulated radiofrequency field exposure when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	SOD1	SOD2	CAT	TRX1
5	Sham	+	1.60 ± 0.45	1.14 ± 0.41	0.08 ± 0.04	1.10 ± 0.29
5	2 W/kg	+	1.39 ± 0.48	0.91 ± 0.51	0.06 ± 0.04	0.95 ± 0.35
30	Sham	+	1.36 ± 0.24	0.87 ± 0.27	0.06 ± 0.03	0.91 ± 0.11
30	2 W/kg	+	1.30 ± 0.30	0.81 ± 0.35	0.05 ± 0.04	0.90 ± 0.20
240	Sham	+	1.59 ± 0.30	1.14 ± 0.39	0.09 ± 0.04	1.06 ± 0.20
240	2 W/kg	+	1.61 ± 0.27	1.21 ± 0.31	0.08 ± 0.02	1.10 ± 0.37
5	Sham	–	3.34 ± 0.64	2.06 ± 0.62	0.12 ± 0.05	2.40 ± 0.51
5	2 W/kg	–	3.16 ± 0.61	1.64 ± 0.50	0.08 ± 0.02	2.24 ± 0.48
30	Sham	–	2.97 ± 0.23	1.69 ± 0.46	0.10 ± 0.06	2.17 ± 0.23
30	2 W/kg	–	2.61 ± 0.34	1.24 ± 0.21	0.07 ± 0.02	1.85 ± 0.35
240	Sham	–	3.20 ± 0.33	1.88 ± 0.69	0.12 ± 0.05	2.18 ± 0.28
240	2 W/kg	–	2.97 ± 0.31	2.13 ± 0.56	0.12 ± 0.02	2.11 ± 0.57

Each data point represents the mean ± SD from 4 independent experiments.

When the impact of 1800 MHz continuous-wave RF-EMF exposure after various periods of time were compared to their time-matched sham controls, a tendency for lower relative expression was observed in the RF-EMF samples, however no statistically significant differences were observed at any time points for SOD1, SOD2 or TRX1 (Table 1). A statistically significant decrease in the relative CAT expression was observed after 30 min RF-EMF exposure in serum-containing cultures in relation to the time-matched sham control, but no differences were observed at other time-points or in any of the serum-free cultures.

The relative expression levels in serum and serum-free cultures exposed to sham or 1800 MHz GSM-modulated RF-EMF are depicted in Table 2. Similarly, no statistically significant differences were observed between RF-EMF and sham-exposed cultures at any time-point for any protein.

Signal transduction proteins (total)

The relative expression levels of CREB, JNK, NF-κB, ERK1/2, Akt, p70S6K, STAT3 and STAT5 in A172 cell culture lysates from cultures exposed for 5, 30 or 240 min to RF-EMF or sham conditions in the presence or absence of serum are depicted in Tables 3 and 4. The expression levels of all ST proteins were observed to be elevated in all samples in serum-free culture lysates when compared to lysates from cultures exposed in the presence of serum.

Table 3. Relative expression levels of signal transduction proteins (total; MFI/mg protein ± SD) after 1800 MHz continuous-wave radiofrequency field exposure when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	CREB	JNK	NF-κB	ERK1/2	Akt	p70S6K	STAT3	STAT5
5	Sham	+	0.86 ± 0.12	0.58 ± 0.24	1.09 ± 0.06	1.07 ± 0.23	0.37 ± 0.16	0.21 ± 0.05	0.21 ± 0.05	0.12 ± 0.04
5	2 W/kg	+	0.90 ± 0.17	0.60 ± 0.27	1.15 ± 0.12	1.14 ± 0.27	0.37 ± 0.10	0.21 ± 0.04	0.20 ± 0.05	0.12 ± 0.02
30	Sham	+	0.92 ± 0.16	0.63 ± 0.27	1.17 ± 0.09	1.16 ± 0.29	0.40 ± 0.14	0.24 ± 0.05	0.23 ± 0.05	0.13 ± 0.03
30	2 W/kg	+	0.79 ± 0.11	0.56 ± 0.24^a	1.04 ± 0.09	1.04 ± 0.19	0.32 ± 0.08	0.20 ± 0.03	0.18 ± 0.04	0.11 ± 0.03
240	Sham	+	1.06 ± 0.37	0.70 ± 0.28	1.32 ± 0.32	1.32 ± 0.59	0.50 ± 0.27	0.27 ± 0.09	0.26 ± 0.08	0.14 ± 0.05
240	2 W/kg	+	0.94 ± 0.40	0.61 ± 0.22^a	1.20 ± 0.38	1.25 ± 0.60	0.39 ± 0.21	0.23 ± 0.08	0.20 ± 0.06^a	0.12 ± 0.06
5	Sham	–	1.43 ± 0.34	0.95 ± 0.42	1.96 ± 0.28	1.94 ± 0.54	0.57 ± 0.25	0.34 ± 0.09	0.30 ± 0.09	0.18 ± 0.06
5	2 W/kg	–	1.47 ± 0.33	0.89 ± 0.40	1.90 ± 0.31	1.82 ± 0.59	0.58 ± 0.17	0.35 ± 0.08	0.30 ± 0.03	0.18 ± 0.05
30	Sham	–	1.38 ± 0.27	0.92 ± 0.44	1.89 ± 0.21	1.90 ± 0.37	0.53 ± 0.18	0.33 ± 0.09	0.26 ± 0.06	0.19 ± 0.05
30	2 W/kg	–	1.48 ± 0.29	0.92 ± 0.40	2.03 ± 0.28	1.90 ± 0.56	0.57 ± 0.19	0.35 ± 0.06	0.28 ± 0.06	0.18 ± 0.04
240	Sham	–	1.21 ± 0.14	0.89 ± 0.45	1.62 ± 0.39	1.60 ± 0.26	0.52 ± 0.09	0.30 ± 0.05	0.27 ± 0.04	0.16 ± 0.02
240	2 W/kg	–	1.17 ± 0.14	0.84 ± 0.44	1.53 ± 0.35	1.47 ± 0.20	0.49 ± 0.06	0.28 ± 0.03	0.25 ± 0.04	0.14 ± 0.02

Each data point represents the mean ± SD from 4 independent experiments.

^aDenotes significant difference (p ≤ .05) in relation to the time-matched sham condition.

Table 4. Relative expression levels of signal transduction proteins (total; MFI/mg protein ± SD) after 1800 MHz GSM-modulated radiofrequency field exposure when cells were exposed in the presence or absence of serum.

Time (min)	RF status	Serum	CREB	JNK	NF-κB	ERK1/2	Akt	p70S6K	STAT3	STAT5
5	Sham	+	0.83 ± 0.23	0.32 ± 0.20	1.04 ± 0.27	1.18 ± 0.32	0.43 ± 0.16	0.19 ± 0.04	0.23 ± 0.09	0.12 ± 0.05
5	2 W/kg	+	0.78 ± 0.18	0.31 ± 0.20	0.98 ± 0.24	1.12 ± 0.30	0.39 ± 0.13	0.17 ± 0.03	0.20 ± 0.06	0.10 ± 0.04
30	Sham	+	0.76 ± 0.13	0.30 ± 0.19	0.96 ± 0.16	1.09 ± 0.22	0.40 ± 0.13	0.18 ± 0.02	0.21 ± 0.06	0.12 ± 0.04
30	2 W/kg	+	0.78 ± 0.11	0.31 ± 0.22	0.99 ± 0.16	1.11 ± 0.19	0.37 ± 0.09	0.17 ± 0.02	0.19 ± 0.04	0.11 ± 0.02
240	Sham	+	0.84 ± 0.15	0.36 ± 0.28	1.01 ± 0.19	1.14 ± 0.19	0.47 ± 0.13	0.20 ± 0.04	0.25 ± 0.06	0.12 ± 0.03
240	2 W/kg	+	0.88 ± 0.13	0.38 ± 0.36	1.06 ± 0.21	1.21 ± 0.18	0.45 ± 0.12	0.20 ± 0.03	0.24 ± 0.06	0.12 ± 0.03
5	Sham	–	1.82 ± 0.27	0.81 ± 0.77	2.54 ± 0.54	2.85 ± 0.42	0.88 ± 0.16	0.42 ± 0.11	0.40 ± 0.07	0.25 ± 0.03
5	2 W/kg	–	1.84 ± 0.44	0.79 ± 0.87	2.43 ± 0.82	2.70 ± 0.65	0.86 ± 0.12	0.42 ± 0.13	0.41 ± 0.03	0.24 ± 0.04
30	Sham	–	1.75 ± 0.20	0.78 ± 0.64	2.39 ± 0.45	2.72 ± 0.43	0.84 ± 0.17	0.40 ± 0.08	0.34 ± 0.10	0.25 ± 0.06
30	2 W/kg	–	1.61 ± 0.23	0.68 ± 0.67	2.13 ± 0.50	2.37 ± 0.39	0.76 ± 0.12	0.35 ± 0.05	0.36 ± 0.03	0.21 ± 0.03^a
240	Sham	–	1.75 ± 0.21	0.75 ± 0.71	2.22 ± 0.45	2.52 ± 0.25	0.87 ± 0.19	0.40 ± 0.06	0.41 ± 0.07	0.23 ± 0.28
240	2 W/kg	–	1.52 ± 0.14	0.61 ± 0.61	1.93 ± 0.38	2.17 ± 0.31	0.74 ± 0.18	0.33 ± 0.03	0.37 ± 0.05	0.18 ± 0.04^a

Each data point represents the mean ± SD from 4 independent experiments.

^aDenotes significant difference (p ≤ .05) in relation to the time-matched sham condition.

No statistically significant differences were observed in the relative expression of CREB, NF-κB, ERK1/2, Akt, p70S6K or STAT5 between lysates from serum-containing cultures exposed to 1800 MHz continuous-wave RF-EMF and time-matched sham controls (Table 3). The relative expression of JNK was found to be statistically significantly lower after 30 and 240 min RF-EMF exposure, while STAT3 was statistically significantly lower in the RF-EMF exposed culture lysates after 240 min of exposure. However, similar changes were not observed in serum-free cultures where no statistically significant differences were observed for any protein tested.

When the relative expression of ST proteins (total) was assessed in cultures exposed to 1800 MHz GSM-modulated RF-EMF, no statistically significant differences were observed in serum-containing cultures. However, in serum-free cultures, STAT5 relative expression was found to be statistically significantly lower after 30 and 240 min RF-EMF exposure relative to the time-matched sham controls.

Signal transduction proteins (phosphorylated)

The relative expression levels of phosphorylated CREB, JNK, NF-κB, ERK1/2, Akt, p70S6K, STAT3 and STAT5 in A172 cell culture lysates from cultures exposed for 5, 30 or 240 min to RF-EMF or sham conditions in the presence or absence of serum are depicted in Tables 5 and 6. The relative expression levels of phosphorylated CREB, NF-κB and ERK1/2 were higher, while Akt was lower, in the serum-free culture lysates. Phosphorylated-STAT5 was not detectable in any sample.

Table 5. Relative expression levels of phosphorylated signal transduction proteins (MFI/µg protein ± SD) after 1800 MHz continuous-wave radiofrequency field exposure when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	CREB	JNK	NF-κB	ERK1/2	Akt	p70S6K	STAT3	STAT5
5	Sham	+	20.71 ± 4.69	62.81 ± 18.26	1.38 ± 0.53	852.90 ± 134.55	184.83 ± 48.21	2.32 ± 0.77	14.39 ± 5.00	ND
5	2 W/kg	+	20.17 ± 3.16	66.78 ± 22.33	1.61 ± 0.17	906.09 ± 36.07	193.36 ± 37.11	2.43 ± 0.36	13.93 ± 3.63	ND
30	Sham	+	23.21 ± 4.92	72.33 ± 18.49	1.47 ± 0.31	887.59 ± 117.18	211.88 ± 53.91	3.09 ± 0.43	17.00 ± 4.69	ND
30	2 W/kg	+	18.13 ± 3.62^a	65.70 ± 17.80	1.66 ± 0.22	861.50 ± 132.15	187.71 ± 51.85	2.43 ± 0.55	12.56 ± 3.48	ND
240	Sham	+	20.74 ± 4.52	70.18 ± 18.52	1.64 ± 0.22	843.25 ± 55.11	188.00 ± 71.90	2.32 ± 1.11	13.43 ± 4.72	ND
240	2 W/kg	+	19.05 ± 5.79	68.32 ± 20.55	1.66 ± 0.50	847.72 ± 145.11	180.97 ± 55.00	2.25 ± 0.83	13.84 ± 5.16	ND
Negative control	n/a	+	18.82 ± 3.33	66.90 ± 25.51	1.42 ± 0.22	945.29 ± 159.99	180.05 ± 73.89	2.53 ± 1.01	15.25 ± 6.35	ND
Positive control	n/a	+	109.46 ± 15.30^b	845.71 ± 170.99^b	2.33 ± 0.26^b	1176.41 ± 105.20	236.16 ± 59.93^b	3.86 ± 1.04	23.62 ± 8.23^b	ND
5	Sham	–	32.80 ± 7.33	60.87 ± 15.70	2.21 ± 0.64	2024.70 ± 1018.47	140.33 ± 25.23	1.89 ± 0.33	13.44 ± 5.00	ND
5	2 W/kg	–	29.05 ± 6.24	61.28 ± 16.33	2.70 ± 0.37	1937.31 ± 609.28	129.84 ± 13.16	2.13 ± 0.31	12.89 ± 2.91	ND
30	Sham	–	30.44 ± 6.89	64.26 ± 20.12	2.25 ± 0.21	1670.18 ± 592.63	130.41 ± 25.36	1.77 ± 0.92	12.86 ± 2.79	ND
30	2 W/kg	–	28.82 ± 6.90	63.46 ± 18.79	2.84 ± 0.87	1976.05 ± 689.57	136.93 ± 20.38	2.23 ± 1.11	13.55 ± 4.78	ND
240	Sham	–	23.24 ± 6.75	62.26 ± 25.00	1.89 ± 1.02	1046.76 ± 317.02	129.25 ± 29.09	2.09 ± 0.69	11.74 ± 7.16	ND
240	2 W/kg	–	21.77 ± 2.91	63.70 ± 27.12	2.20 ± 0.56	1173.75 ± 377.03	130.99 ± 27.01	2.52 ± 1.28	12.62 ± 6.93	ND
Negative control	n/a	–	27.06 ± 4.97	62.20 ± 28.74	1.89 ± 0.22	1673.30 ± 380.73	115.79 ± 24.62	1.48 ± 0.73	13.20 ± 5.95	ND
Positive control	n/a	–	152.13 ± 18.47^b	1268.92 ± 132.35^b	4.16 ± 0.73^b	1868.46 ± 696.25	172.23 ± 36.19^b	3.93 ± 0.56^b	20.61 ± 7.84^b	ND

Each data point represents the mean ± SD from 4 independent experiments. ND denotes not detectable.

^aDenotes significant difference (p ≤ .05) in relation to the time-matched sham condition.

^bDenotes significant difference (p ≤ .05) in relation to negative control group.

Table 6. Relative expression levels of phosphorylated signal transduction proteins (MFI/µg protein ± SD) after 1800 MHz GSM-modulated radiofrequency field exposure when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	CREB	JNK	NF-κB	ERK1/2	Akt	p70S6K	STAT3	STAT5
5	Sham	+	19.55 ± 3.22	50.75 ± 5.39	3.15 ± 0.65	883.67 ± 323.09	111.90 ± 78.98	2.91 ± 0.41	14.20 ± 3.83	ND
5	2 W/kg	+	15.66 ± 0.68	45.54 ± 3.39	2.84 ± 1.11	867.14 ± 308.60	95.94 ± 63.37	2.62 ± 1.00	11.49 ± 0.92	ND
30	Sham	+	17.20 ± 1.81	48.71 ± 11.91	2.79 ± 1.16	780.88 ± 191.12	97.04 ± 47.84	2.90 ± 0.93	12.63 ± 2.59	ND
30	2 W/kg	+	17.99 ± 2.89	50.07 ± 8.37	3.20 ± 1.56	805.41 ± 180.17	97.26 ± 52.25	3.46 ± 1.21	12.75 ± 0.95	ND
240	Sham	+	22.60 ± 5.36	58.30 ± 8.51	3.95 ± 1.17	653.72 ± 146.81	71.71 ± 11.13	4.00 ± 0.86	13.57 ± 2.11	ND
240	2 W/kg	+	22.31 ± 4.96	62.37 ± 3.41	4.61 ± 0.80	719.14 ± 156.75	74.46 ± 6.74	4.59 ± 0.70	13.94 ± 0.57	ND
Negative control	n/a	+	20.89 ± 5.74	50.83 ± 4.18	3.71 ± 0.74	761.18 ± 124.70	75.77 ± 14.33	3.58 ± 0.89	12.81 ± 2.62	ND
Positive control	n/a	+	81.89 ± 16.06^a	685.39 ± 153.47^a	5.34 ± 1.16	976.72 ± 239.26	94.97 ± 21.07	5.55 ± 1.17	19.90 ± 3.69^a	ND
5	Sham	–	40.74 ± 11.79	69.88 ± 21.33	7.81 ± 4.90	3034.16 ± 1178.01	102.17 ± 42.68	5.88 ± 4.31	23.41 ± 6.20	ND
5	2 W/kg	–	55.04 ± 27.98	71.39 ± 24.71	9.90 ± 4.39	2607.20 ± 738.61	100.76 ± 41.09	8.31 ± 4.52	23.88 ± 10.71	ND
30	Sham	–	35.32 ± 5.01	61.76 ± 15.00	6.14 ± 3.13	2810.44 ± 953.16	102.70 ± 60.08	4.21 ± 2.04	17.44 ± 7.60	ND
30	2 W/kg	–	42.54 ± 15.58	61.01 ± 17.88	7.49 ± 3.69	2399.41 ± 983.39	111.41 ± 44.49	7.22 ± 4.77	21.83 ± 7.13	ND
240	Sham	–	42.74 ± 10.68	63.61 ± 10.00	7.79 ± 2.40	1868.48 ± 901.09	71.57 ± 6.61	6.06 ± 2.06	16.26 ± 3.90	ND
240	2 W/kg	–	40.63 ± 9.57	58.11 ± 11.56	8.28 ± 1.71	1739.59 ± 341.68	74.10 ± 9.55	6.19 ± 2.25	17.24 ± 3.17	ND
Negative control	n/a	–	38.21 ± 9.39	67.91 ± 18.15	7.84 ± 2.86	2181.83 ± 637.24	77.24 ± 10.80	6.14 ± 2.47	16.43 ± 2.28	ND
Positive control	n/a	–	191.40 ± 78.52^a	1468.44 ± 467.09^a	12.40 ± 6.29	3181.56 ± 1883.95^a	124.19 ± 35.24^a	9.26 ± 2.66^a	32.12 ± 5.86^a	ND

Each data point represents the mean ± SD from 4 independent experiments. ND denotes not detectable.

^aDenotes significant difference (p ≤ .05) in relation to negative control group.

When the relative expression of phosphorylated ST proteins was assessed in serum-containing cultures exposed to 1800 MHz continuous-wave RF-EMF, phosphorylated-CREB was found to be statistically significantly decreased in the RF-EMF group after 30 min exposure relative to the time-matched sham control (Table 5). No statistically significant differences were observed in the relative expression of phosphorylated ST proteins following RF-EMF exposure in serum-free cultures. Treatment with the positive control (anisomycin) caused a statistically significant increase in the relative expression of phosphorylated CREB, JNK, NF-κB, Akt, p70S6K and STAT3 relative to the negative (vehicle) control.

No statistically significant differences were observed in the relative expression level of phosphorylated ST proteins in either serum-containing or serum-free cultures after exposure to 1800 MHz GSM-modulated RF-EMF, relative to time-matched sham controls (Table 6). However, the positive control (anisomycin) was found to elicit a statistically significant increase in the relative expression of the phosphorylation level of most ST proteins in relation to the negative control group.

When the data were analyzed by the ratio of the phosphorylated/total ST protein expression level in relation to the relevant sham exposed group, fold changes of less than 1.5-fold were observed for most exposure times and ST protein targets in both serum-containing and serum-free cultures for both continuous-wave or GSM-modulated 1800 MHz RF-EMF (Tables 7 and 8). In contrast, the positive control (anisomycin) induced large fold changes (greater than 2-fold) for CREB and JNK in both serum-free and serum-containing cultures (Tables 7 and 8), and for NF-κB and p70S6K in the serum-free experiments where cultures were exposed to GSM-modulated RF-EMF (Table 8).

Table 7. Fold change in the expression level of phosphorylated signal transduction proteins ((MFI phosphoprotein/µg total protein)/(MFI protein/mg total protein)) after 1800 MHz continuous-wave radiofrequency field exposure, in relation to the time-matched sham or negative control groups, when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	CREB	JNK	NF-κB	ERK1/2	Akt	p70S6K	STAT3
5	2 W/kg	+	0.93	0.99	1.12	1.01	0.94	1.03	0.99
30	2 W/kg	+	0.91	1.05	1.26	1.07	1.07	0.98	0.94
240	2 W/kg	+	1.01	1.10	1.10	1.07	1.21	1.07	1.26
Positive control	n/a	+	5.13	17.07	1.52	1.14	1.27	1.45	1.49
5	2 W/kg	–	0.86	1.13	1.29	1.03	0.82	1.69	0.94
30	2 W/kg	–	0.91	1.01	1.18	1.21	0.96	1.26	0.95
240	2 W/kg	–	0.98	1.15	1.31	1.22	1.07	1.42	1.17
Positive control	n/a	–	5.67	22.63	2.17	1.18	1.51	2.88	1.53

Table 8. Fold change in the expression level of phosphorylated signal transduction proteins ((MFI phosphoprotein/µg total protein)/(MFI protein/mg total protein)) after 1800 MHz GSM-modulated radiofrequency field exposure, in relation to the time-matched sham or negative control groups, when cells were exposed in the presence and absence of serum.

Time (min)	RF status	Serum	CREB	JNK	NF-κB	ERK1/2	Akt	p70S6K	STAT3
5	2 W/kg	+	0.87	0.97	1.01	1.04	0.96	1.03	0.94
30	2 W/kg	+	1.03	0.94	1.12	1.02	1.07	1.27	1.10
240	2 W/kg	+	0.95	1.16	1.15	1.03	1.10	1.20	1.07
Positive control	n/a	+	4.18	11.63	1.56	1.41	1.61	1.80	1.72
5	2 W/kg	–	1.29	1.17	1.33	0.92	1.00	1.47	0.95
30	2 W/kg	–	1.28	1.16	1.32	0.98	1.24	1.85	1.18
240	2 W/kg	–	1.09	1.19	1.25	1.10	1.23	1.24	1.19
Positive control	n/a	–	4.63	25.63	1.53	1.46	1.49	1.46	1.61

Discussion

International human exposure limits for RF-EMF in the 100 kHz to 300 GHz frequency range have been established since the 1980s and have been recently updated (IEEE 2019; ICNIRP 2020). However, concerns persist that adverse health effects may be experienced from exposure to RF-EMF at levels below these recommended limits. Several researchers have hypothesized that RF-EMF exposure at levels below these limits may alter the metabolic balance within cells, leading to oxidative stress and the irregular activation of certain ST pathways (for review see Yakymenko et al. 2016). In the current study, a human-derived glioblastoma A172 cell line was exposed to 1800 MHz continuous-wave and GSM-modulated RF-EMF at 2 W/kg for 5, 30 or 240 min and then the relative expression of a broad spectrum of ST proteins and their phosphorylated forms were assessed. While statistically significant changes in some targets were observed at some time-points, such changes were not observed at other time-points, were not consistent between continuous-wave and GSM-modulated conditions or in serum-free versus serum-containing conditions. We postulate that the changes observed in the current study likely represented false-positive events due to their lack of consistency.

Several in vitro studies have found increased production of ROS and oxidative stress levels in cell cultures in response to RF-EMF exposure. Marjanovic Cermak et al. (2018) reported statistically significantly higher ROS levels in human neuroblastoma-derived SH-SY5Y cells following exposure to 1800 MHz RF-EMF at a SAR of 1.6 W/kg for exposure times ranging from 10 to 60 min. The authors also reported a statistically significant increase in malondialdehyde content in RF-EMF exposed samples after a 60 min exposure period, however temperature within the cell cultures was not monitored during RF-EMF exposure and thermal confounding could not be excluded as a possible artifact in this study. Kim et al. (2017) examined the ability of RF-EMF to modify the level of oxidative stress induced by glutamate in a mouse hippocampus-derived HT22 cell line. The authors reported that co-exposure of these cells to glutamate and 1.95 GHz WCDMA-modulated RF-EMF for 1–12 h at a SAR of 4 or 6 W/kg resulted in an increased level of ROS in relation to cells exposed to glutamate alone. No effect was observed at a SAR of 2 W/kg. Other studies have not found similar effects. Silva et al. (2016) exposed primary human thyroid cells to 900 MHz RF-EMF for 3 to 65 h at a SAR of 0.08–0.17 W/kg and reported no change in the ROS levels in the RF-EMF exposed cultures in relation to the sham control. Simkó et al. (2006) exposed a human monocyte-derived Mono-Mac 6 cell line to 1800 MHz continuous-wave or GSM-modulated RF-EMF for 1 h at a SAR of 2 W/kg and found no change in superoxide anion production. In the current study, the level of antioxidant proteins remained relatively unchanged, indicating that a statistically significant level of oxidative stress was not induced by RF-EMF exposure.

Several studies have also reported the ability of RF-EMF to modulate the expression and activation levels of a variety of ST proteins. Buttiglione et al. (2007) exposed a human neuroblastoma cell line (SH-SY5Y) to 900 MHz RF-EMF for 5, 15, 30, 360 min or 24 h at a SAR of 1 W/kg then assessed Egr-1 mRNA expression and the expression of total and phosphorylated ERK1/2 and SAPK/JNK protein levels. The authors reported that RF-EMF exposure caused an increase in Egr-1 after a 5–30 min exposure time, but no changes were observed after 6 or 24 h exposure. A concomitant increase in phosphorylated-ERK1/2 was observed after a 5–15 min exposure and an increase in phosphorylated-SAPK/JNK was observed after a 5–30 min exposure. No changes were observed in phosphorylated-p38MAPK or the total expression level of ERK1/2, SAPK/JNK or p38MAPK. Friedman et al. (2007) exposed serum-starved Rat1 and HeLa cells to 800, 875 or 950 MHz RF-EMF for 5–30 min at a power density of 0.05–3.1 W/m², then assessed the activation of ERK1/2, JNK1/2 and p38MAPK. The authors reported that a short (5–30 min) RF-EMF exposure at a power density of 0.7 W/m² caused a statistically significant increase in the level of phosphorylated-ERK1/2, but no change was observed in the level of phosphorylated-JNK or p38MAPK. The authors also found that ROS scavengers could inhibit this response leading them to conclude that RF-EMF may be leading to ST activation via NADH-oxidase mediated oxidative stress. Glushkova et al. (2015) exposed murine RAW 264.7 macrophages to broad spectrum RF-EMF in the 8.15–18 GHz frequency range for 1 h at a power density of 14 mW/m². The authors reported that RF-EMF exposure caused increased expression of phosphorylated-NF-κB and SAPK/JNK. Leszczynski et al. (2002) exposed a human endothelial cell line (EA.hy926) to 900 MHz GSM-modulated RF-EMF for 1 h at a SAR of 2 W/kg, then assessed the protein expression of activated p38MAPK. The authors reported that RF-EMF exposure resulted in an increased expression of phosphorylated-HSP27 and p38MAPK immediately after exposure. Liu et al. (2014) exposed a mouse spermatocyte-derived GC-2 cell line to 1800 MHz GSM-modulated RF-EMF for 24 h at either 1, 2 or 4 W/kg. The authors reported increased intracellular ROS levels and increased levels of phosphorylated-ERK1/2. Lu et al. (2014) assessed STAT3 activation in N9 mouse microglial cells and astroglial C8-D1A cells after exposure to 1800 MHz RF-EMF for 1–24 h at a SAR of 2 W/kg. The authors reported increased levels of phosphorylated-STAT3 in microglial cells, but not astroglial cells. No changes were observed in either cell line for the total expression level of STAT3.

In contrast to these studies, other researchers have found no effect on the expression or activation of ST proteins after RF-EMF exposure. Lee et al. (2006) exposed human T-lymphocyte derived Jurkat cells to 1.763 GHz CDMA RF-EMF for 30 min or 1 h at a SAR of 2 or 20 W/kg, then assessed the expression level of HSP27, HSP70, HSP90, ERK1/2, JNK1/2 and p38MAPK at 0, 1 or 2 h post-exposure. The authors reported no statistically significant differences in the expression level of total ERK1/2, JNK or p38MAPK or HSP27, HSP70 or HSP90 after RF-EMF exposure. When co-treated with RF-EMF and the tumor promoter 12-O-tetradecanoylphorbol 13-acetate (TPA), the authors observed no changes in the expression of either total or phosphorylated-ERK1/2, JNK or p38MAPK. Valbonesi et al. (2014) assessed HSP70 gene and protein expression and expression of activated MAPK after exposure of rat PC12 cells to 1800 MHz GSM-modulated RF-EMF for 4, 16 or 24 h at a SAR of 2 W/kg. The authors reported increased HSP70 transcript levels after a 16 or 24 h RF-EMF exposure but no statistically significant differences were observed for HSP70 protein expression or MAPK phosphorylation. Zielinski et al. (2020) exposed murine N9 microglial cells and human neuroblastoma-derived SH-SY5Y cells to 935 MHz GSM-modulated RF-EMF for 2 or 24 h at a SAR of 4 W/kg, then assessed for activation of ERK1/2 and hydrogen peroxide production. The authors reported no changes in ERK1/2 phosphorylation after RF-EMF exposure in either cell line. The findings of the current study do not provide convincing evidence to support the hypothesis that low-level RF-EMF exposure alters the levels or activation status of a range of ST proteins.

While the current study assessed a large number of antioxidant and signal transduction proteins simultaneously at multiple time-points after both continuous-wave and GSM-modulated 1800 MHz RF-EMF exposure, it is possible that the exposure times chosen were not ideal for the detection of effects. Future studies should consider assessing a wide range of exposure times to further investigate this aspect. Furthermore, while consistent changes in either antioxidant or ST proteins were not observed in the current study, subtle but biologically meaningful changes in response cannot be ruled out. Indeed, there was a general tendency for decreased expression of antioxidant proteins and increased expression of the phosphorylated signal transduction proteins. The consequences of such small fold changes on cellular and human health is unknown. It is not clear how serum starvation affected the general outcome of this study, as no consistent effects were observed across the endpoints studied. While changes were observed between the serum-containing and serum-free cultures, the degree of effect varied and was dependant on the pathway studied, similar to the observations of Pirkmajer and Chibalin (2011).

In summary, the current study did not demonstrate consistent changes in the expression of antioxidant proteins (SOD1, SOD2, CAT or TRX2) or a variety of signal transduction proteins (CREB, JNK, NF-κB, ERK1/2, Akt, p70S6K, STAT3, STAT5) in a human glioblastoma-derived A172 cell line in response to 1800 MHz continuous-wave or GSM-modulated RF-EMF exposure for 5, 30, or 240 min in either serum-free or serum-containing cell cultures. Future studies examining additional cell lines, time-points and exposure levels under tightly controlled RF-EMF exposure conditions would be beneficial to understanding and interpreting the possible interaction between wireless communication signals and biological systems.

Disclosure statement

The authors report no conflicts of interest. This study was funded entirely by the Government of Canada.

Additional information

Notes on contributors

James P. McNamee

James P. McNamee, PhD is a Research Scientist at the Consumer and Clinical Radiation Protection Bureau, Health Canada, Ottawa, Ontario, Canada and an Adjunct Assistant Professor at the Department of Biomedical and Molecular Medicine, Queen’s University, Kingston, Ontario, Canada.

Veronica S. Grybas

Veronica S. Grybas is an undergraduate (co-op) student at Carleton University, Ottawa, Ontario, Canada.

Sami S. Qutob

Sami S. Qutob, PhD is a Research Scientist at the Consumer and Clinical Radiation Protection Bureau, Health Canada, Ottawa, Ontario, Canada.

Pascale V. Bellier

Pascale V. Bellier, MSc is a Biologist at the Consumer and Clinical Radiation Protection Bureau, Health Canada, Ottawa, Ontario, Canada.