- Email: [email protected]
Genotoxic effects of exposure to radiofrequency electromagnetic fields (RF-EMF) in cultured mammalian cells are not independently reproducible
Mutation Research 626 (2007) 42–47
Genotoxic effects of exposure to radiofrequency electromagnetic fields (RF-EMF) in cultured mammalian cells are not independently reproducible G¨unter Speit ∗ , Petra Sch¨utz, Heike Hoffmann Universit¨at Ulm, Abteilung Humangenetik, D-89070 Ulm, Germany Received 22 June 2006; received in revised form 3 August 2006; accepted 4 August 2006 Available online 25 September 2006
Abstract Conflicting results have been published regarding the induction of genotoxic effects by exposure to radiofrequency electromagnetic fields (RF-EMF). Using the comet assay, the micronucleus test and the chromosome aberration test with human fibroblasts (ES1 cells), the EU-funded “REFLEX” project (Risk Evaluation of Potential Environmental Hazards From Low Energy Electromagnetic Field Exposure Using Sensitive in vitro Methods) reported clearly positive effects for various exposure conditions. Because of the ongoing discussion on the biological significance of the effects observed, it was the aim of the present study to independently repeat the results using the same cells, the same equipment and the same exposure conditions. We therefore exposed ES1 cells to RF-EMF (1800 MHz; SAR 2 W/kg, continuous wave with intermittent exposure) for different time periods and then performed the alkaline (pH > 13) comet assay and the micronucleus test (MNT). For both tests, clearly negative results were obtained in independently repeated experiments. We also performed these experiments with V79 cells, a sensitive Chinese hamster cell line that is frequently used in genotoxicity testing, and also did not measure any genotoxic effect in the comet assay and the MNT. Appropriate measures of quality control were considered to exclude variations in the test performance, failure of the RF-EMF exposure or an evaluation bias. The reasons for the difference between the results reported by the REFLEX project and our experiments remain unclear. © 2006 Elsevier B.V. All rights reserved. Keywords: Comet assay; Micronucleus test; Low energy electromagnetic field; DNA damage; Mutation
1. Introduction During the last years, a large number of studies were performed using cultured human and other mammalian cells to determine a potential genotoxic effect of exposure to radiofrequency electromagnetic fields (RF-EMF). Conflicting results were reported but the reasons for the contradictory findings remained unclear.
∗ Corresponding author. Tel.: +49 731 500 23429; fax: +49 731 500 23438. E-mail address: [email protected] (G. Speit).
1383-5718/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.mrgentox.2006.08.003
The investigations published in scientific journals between 1990 and 2003 were critically evaluated in a recent review [1]. Various results indicating a genotoxic effect of RF-EMF were recently produced by the collaborative EU-funded REFLEX (Risk Evaluation of Potential Environmental Hazards From Low Energy Electromagnetic Field Exposure Using Sensitive in vitro Methods) project [2]. One laboratory taking part in the REFLEX project published genotoxic effects in cultured human fibroblasts exposed to RF-EMF (1800 MHz, SAR 2 W/kg, continuous or intermittent exposure, with or without modulations of the signal) for 16 or 24 h [3]. DNA damage was assessed by use of the comet assay, in
G. Speit et al. / Mutation Research 626 (2007) 42–47
both alkaline and neutral versions. Intermittent exposure caused a stronger effect than continuous exposure. The positive comet assay results were supported by positive findings in the micronucleus test (MNT) and the chromosome aberration test [2]. Due to the importance of these results in general, and the critical comments particularly with regard to the assessment of comet assay effects and the biological significance of the comet assay results [4,5], an independent replication of the experiments and confirmation of the results was desirable. Due to our experience with the comet assay and other genotoxicity tests, we have been asked by the coordinator of the REFLEX project to repeat parts of the study in our laboratory using the same cell cultures and the same exposure conditions as in the previous experiments [2,3]. We were provided with ES1 human fibroblasts and a system for RF-EMF exposure under strictly controlled conditions [6]. We decided to perform the alkaline comet assay using exposure conditions (1800 MHz, SAR 2 W/kg, continuous wave, intermittent exposure, 5 min field on/10 min field off) that led to the strongest effects in the previous investigations [3]. The MNT should be performed in parallel to confirm effects with a second genetic endpoint indicating true mutagenic effects on the chromosomal level. We also decided to conduct experiments with V79 cells because this rapidly proliferating Chinese hamster cell line is very sensitive towards a broad spectrum of genotoxic effects and routinely used in genotoxicity testing. It was the goal of our study to independently confirm the results of the REFLEX project. 2. Materials and methods 2.1. Cell culture and RF-EMF exposure conditions Human diploid fibroblasts (ES1 cells) and a Chinese hamster cell line (V79 cells) were cultured in DMEM and MEM, respectively, supplemented with 10% fetal calf serum (FCS) and antibiotics. Cells were maintained in a humidified incubator at 37 ◦ C with 5% CO2 and harvested with 0.15% trypsin and 0.08% EDTA. For the experiments, cells were seeded into 35-mm Petri dishes 24 h prior to RF-EMF exposure. The exposure system was built, provided and checked for correct function by the Foundation for Research on Information Technologies in Society, Zurich, Switzerland (IT’IS Foundation). For this setup R18 hollow rectangular waveguides operating at a frequency of 1800 MHz are used as previously described in detail [3,6]. Two waveguide-based exposure chambers containing the cell-culture dishes (exposed, sham-exposed) were placed inside a commercial incubator to ensure constant environmental conditions (37 ◦ C, 5% CO2 , 95% humidity). In addition, the incident field was monitored inside the waveguides. The air temperature difference between the chambers measured at the air-outlet did not exceed 0.1 ◦ C.
43
To enable blind exposure and evaluation, a computer randomly determined which of the two waveguides was exposed. This information was provided to us by the IT’IS Foundation via e-mail after transmission of our test results. The fields as well as all sensors were continuously monitored. The experiments were generally performed at a carrier frequency of 1800 MHz at intermittent exposure (5 min field on/10 min field off). The field was applied without modulation of the signal (=continuous wave). In some experiments, we performed a continuous exposure with modulation (GSM basic) of the signal (see Diem et al. [3] for technical details). The average specific absorption rate (SAR) was 2 W/kg because it is the safety limit for the mobile-phone microwave-radiation emission and was used in the previous experiments [2,3]. The exposure time varied between 1 and 24 h. After the end of the exposure period, cells were detached with trypsin and directly used for the comet assay. For the MNT, cells were either used directly after exposure or suspended in fresh culture medium and cultured further. 2.2. Comet assay The comet assay was performed according to the protocol that is usually used previously in our laboratory [7]. Aliquots of 10-l cell suspension (about 15,000 cells) were mixed with 120 l low melting-point agarose (0.5% in PBS) and added onto microscope slides (with frosted ends), which had been covered with a bottom layer of 1.5% agarose. Slides were lysed (pH 10; 4 ◦ C) for at least 1 h and processed as previously described in detail [7] using a time of alkali denaturation of 25 min and electrophoresis (0.86 V/cm) of 25 min at a pH > 13. Slides were coded and stained with ethidium bromide (10 g/ml). Measurements were made for 50 randomly selected cells per slide by image analysis (Comet Assay II, Perceptive Instruments). For all experiments, we evaluated three image analysis parameters: tail migration, tail intensity and tail moment. In none of the experiments there was a significant difference between these parameters. Therefore, we chose one parameter (tail moment) for the presentation of the results. Tail moment (TM) is calculated according to the formula: TM = (tail intensity/total comet intensity) × (tail centre of gravity − peak position). In one experiment the “comet tailfactor” was comparatively determined according to Diem et al. [3]. For each data point 500 cells were scored and assigned to five categories corresponding to the amount of DNA in the tail. For a positive control, a parallel culture was exposed to 2 Gy 137 Cs gamma-rays at 4 Gy/min at the end of the RF-EMF exposure period of the experimental cultures and immediately processed in the comet assay together with the RF-EMF-exposed cultures. 2.3. Micronucleus test (MNT) The MNT was basically performed as described by the REFLEX project [2] in accordance with standard procedures. After RF-EMF exposure, ES1 cells were cultured for another 44 h in the presence of cytochalasin B (final concentration
44
G. Speit et al. / Mutation Research 626 (2007) 42–47
3 g/ml). As a positive control, a parallel culture was exposed to 2 Gy 137 Cs gamma-rays at 4 Gy/min at the end of the 24-h RF-EMF exposure period of the experimental cultures, and further processed in parallel. V79 cells were either cultured after exposure (without cytochalasin B) for another 18 h, or they were directly processed after an exposure period of 18–22 h. As a positive control, a parallel culture was exposed to 2 Gy 137 Cs gamma-rays at 4 Gy/min and further processed in parallel to the RF-EMF-exposed cultures (i.e. cultivation for 18 or 22 h, respectively). Cells were detached by trypsin, treated after centrifugation with a hypotonic solution (0.56% KCl) and fixed three times with methanol/glacial acetic acid (5:1 for ES1 cells; 3:1 for V79 cells). Air-dried slides were stained with acridine orange (125 g/ml in phosphate buffer). The frequency of micronuclei (MN) was determined by analysing 500 binucleated cells (ES1) or 1000 cells (V79) from coded slides. 2.4. Statistical analysis The experiments with ES1 cells were independently performed three times and differences between mean values were tested for statistical significance (P < 0.05) using Student’s ttest. Experiments with V79 cells were confirmed in a second independently performed test and results are presented as the mean of the two tests. All results were evaluated with regard to their biological relevance and in comparison with the previously reported effects under the same experimental conditions.
3. Results and discussion In a previously published study [3], RF-EMF induced genotoxic effects in the alkaline comet assay with ES1 cells. Significant differences between exposed and sham-exposed cultures were measured after exposure periods of 16 and 24 h, but not after 4 h. Genotoxic effects were found independently of whether the field was applied with or without modulation and whether exposure was continuous or intermittent. We decided to repeat the experiments using a field without modulation (continuous wave) and intermittent exposure (5 min field on/10 min field off) because this protocol had produced the strongest effects (i.e. doubling of the comet tail-factor). We used the same exposure conditions (1800 MHz, SAR 2 W/kg) but exposure times of 1, 4 and 24 h. The published data did not show any difference between the time points 16 and 24 h and we preferred to substitute the time point 16 h by a very early time point, because the comet assay usually detects primary DNA damage that may occur immediately and may be repaired rapidly. Our results are summarized in Fig. 1A and clearly show that under the same experimental conditions as used by Diem et al. [3] we did not detect a genotoxic effect of RF-EMF on ES1 cells at the three time points tested.
Fig. 1. The effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on DNA migration (tail moment) in the comet assay (A) and on the frequency of micronuclei (B) in ES1 cells after exposure for different time periods (1, 4 and 24 h) in comparison to sham-exposed cultures and a negative (Co) and a positive (2 Gy gamma irradiation after cultivation for 24 h) control. Mean ± S.E.M. of three independent experiments.
Increased DNA effects after exposure for 16 or 24 h but not after exposure for 4 h as reported by Diem et al. [3] is an uncommon finding with the comet assay because most genotoxins induce effects rapidly and such a delayed effect may be indicative of an indirect effect related to DNA repair or disturbed physiological conditions. Furthermore, there is no apparent explanation as to why intermittent exposure should induce more DNA damage than continuous exposure. Diem et al. [3] also used a neutral version of the comet assay and show results that are nearly identical to those of the alkaline comet assay. The fact that there was nearly no difference between the two versions with regard to the absolute values of the controls and the exposed cultures casts doubt on the specificity of the neutral version of the comet assay used and the biological significance of these findings. Altogether, the study of Diem et al. [3] suggests the induction of a genotoxic effect by RF-EMF in the comet
G. Speit et al. / Mutation Research 626 (2007) 42–47
assay with human fibroblasts. However, as we were not able to independently reproduce the results under the same experimental conditions, the relevance of the observed genotoxic effects remains unclear. To exclude a difference between the exposure conditions, the exposure systems used in Ulm and Vienna were checked by the experts from the IT’IS Foundation (Zurich) and comparative measurements did not reveal any significant difference between the two exposure units. As part of the REFLEX project the induction of micronuclei in ES1 cells was also studied and a clear effect was measured [2]. The results of the MNT are of particular importance because the genetic endpoint is a chromosomal mutation transmitted to a daughter cell. While comet assay effects can be questioned with regard to their genetic relevance (i.e. the DNA effects seen in the comet assay might be DNA damage of transient nature and/or intermediates in the repair of DNA damage), the biological relevance of micronuclei is not a matter of discussion. Furthermore, while the evaluation of comet assay effects may lead to conflicting results due to the high assay variability and the different ways to determine an effect, micronuclei are detected with high reliability and there is no serious concern that a positive effect might be missed under appropriate test conditions. We performed the micronucleus test using the same exposure conditions as used in our comet assay experiments (SAR 2 W/kg, intermittent exposure, 5 min on/10 min off for 24 h) and did not measure any effect on the frequency of micronuclei (Fig. 1B). The comet assay results reported for the REFLEX study indicated that the different modes of exposure do not have a fundamental influence on the outcome of the test. However, to exclude an influence of the mode of exposure and to exclude differences in the test performance between the laboratories in Ulm and Vienna, colleagues from Vienna came to Ulm to conduct experiments together. In these experiments we used our standard exposure conditions but also GSM basic modulation, 1800 MHz, 1 or 2 W/kg for 24 h. We performed the comet assay and the MNT (only single experiments without independent repetition), prepared slides in parallel and evaluated the coded slides independently. All these experiments did not reveal any genotoxic effect of RF-EMF and the negative results were confirmed by the evaluation of these slides by our colleagues from Vienna (data not shown). It is surprising that the clear effect in the MNT (20-fold increase, in the range of the positive control, 10 g/ml bleomycin) could not be reproduced under nearly identical experimental conditions in our laboratory, but there is no obvious reason for this discrepancy. A clastogenic effect of RF-EMF on ES1 cells has further been demonstrated in the REFLEX project
45
by a positive result from a chromosome aberration test. Interestingly, a strong induction of chromosometype aberrations (i.e. dicentric chromosomes and acentric fragments) was measured. Because of the clearly negative MNT in our laboratory, we refrained from conducting a chromosome aberration test. In previously published studies with cultured human lymphocytes no induction of chromosome aberrations, micronuclei or sister chromatid exchange was measured after exposure to a 900-MHz GSM signal (SAR between 0.2 and 10 W/kg) for various periods of time [8–10]. Regarding the comet assay effects, it is interesting to note that the different ways of evaluating the comet assay (i.e. determination of the tail moment of 50 cells by image analysis and determination of the comet tailfactor of 500 cells by defining five categories of cells corresponding to the amount of DNA in the tail [3]) led to very similar results (Fig. 2). It has been discussed, whether the tail-factor used by Diem et al. [3] is a reliable measure of DNA effects in the comet assay [5,11]. This comparative evaluation demonstrates that the calculation of the tail-factor does not reveal a difference that is missed by our evaluation using image analysis. It confirms that the results obtained in our laboratory with the comet assay are in fact negative. Our negative findings with ES1 cells were confirmed by comprehensive experiments with V79 cells. We performed additional experiments with V79 cells because this cell line is well known for its sensitivity for the detection of genotoxic effects and is frequently used in routine genotoxicity testing. We performed the comet assay immediately after an exposure period of 1, 4 or 24 h (Fig. 3A) and did not measure any significant increase
Fig. 2. Parallel evaluation of the effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off; 24 h) on DNA migration (tail moment) in the comet assay with ES1 cells using either image analysis (tail moment) or arbitrary damage classes (tailfactor).
46
G. Speit et al. / Mutation Research 626 (2007) 42–47
Fig. 4. The effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on the frequency of micronuclei in V79 cells after exposure for 18 h and analysis at the end of the exposure period in comparison with sham-exposed cultures and a negative (Co) and a positive (2 Gy gamma irradiation at the beginning of the exposure period) control. Mean of two independent experiments.
Fig. 3. The effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on DNA migration (tail moment) in the comet assay (A) and on the frequency of micronuclei (B) in V79 cells after exposure for different time periods (1, 4 and 24 h) and further cultivation for 18 h in comparison with sham-exposed cultures and a negative (Co) and a positive control (2 Gy gamma irradiation after cultivation for 24 h, immediate analysis in the comet assay, further cultivation for 18 h in the MNT). Mean ± S.E.M. of three independent experiments.
in DNA migration (tail moment). Parallel cultures were further cultured for 18 h after the end of the exposure and then prepared for the analysis of micronuclei. In accordance with the results with ES1 cells there was no induction of micronuclei in any of the experiments (Fig. 3B). We then used a modification of the time schedule and exposed V79 cells for 18 h and prepared the cells for analysis of micronuclei immediately after the end of the exposure. This protocol, which is also suggested for routine genotoxicity testing with the MNT, should detect micronuclei that are formed during the exposure period. Again, the result of the modification of the MNT was also clearly negative (Fig. 4). We also analysed V79 cells after exposure for 4 h (with 18 h post-incubation) or for 22 h to RF-EMF with a SAR of 1 and 2 W/kg and obtained negative results (Fig. 5). In our opinion, the negative
results with V79 cells confirm the negative findings with ES1 cells. Effects for negative and positive controls were similar to those measured in ES1 cells. However, fundamental differences between different cell types with regard to the induction of genotoxic effects by electromagnetic fields cannot be excluded. In a study applying extremely low-frequency electromagnetic fields (ELFEMF) it was shown that a genotoxic effect can be found in some types of cultured mammalian cells but not in others [12]. The cause for this difference is unknown but might just be a matter of different sensitivity as suggested by positive effects of RF signals (837 MHz) in lympho-
Fig. 5. The effect of RF-EMF (1800 MHz, SAR 1 and 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on the frequency of micronuclei in V79 cells after exposure for 4 h and further cultivation for 18 h and after exposure for 22 h and analysis at the end of the exposure period in comparison with sham-exposed cultures and a negative (Co) and a positive (2 Gy gamma irradiation and cultivation for 22 h) control. Mean of two independent experiments.
G. Speit et al. / Mutation Research 626 (2007) 42–47
cytes after extended exposure conditions (5 and 10 W/kg for 24 h) [13]. However, a recent inter-laboratory study using lymphocyte cultures from 10 healthy donors did not find a genotoxic effect of exposure to 900 MHz (GSM signal) at SARs of 1, 5 and 10 W/kg [10]. These inconsistencies may also indicate that the observed genotoxic effect of electromagnetic fields in some studies is not a direct genotoxic effect (i.e. due to a direct interaction with the genetic material, which should occur in every directly exposed cell type), but may be due to unknown indirect effects. In conclusion, we were not able to independently confirm some of the genotoxic effects of RF-EMF reported by the REFLEX project. However, our results do not exclude a genotoxic effect of RF-EMF in general. We did not test other exposure conditions and the role of the exposure conditions for the expression of a genotoxic effect is not yet understood. Further mechanistic studies are necessary to elucidate the biological basis of the genotoxic effects reported for RF-EMF and to clarify the contradictory findings in different studies. Acknowledgements We gratefully acknowledge the excellent cooperation with Prof. F. Adlkofer (REFLEX project; VERUM foundation) and Prof. H.W. R¨udiger and his co-workers (Vienna). We thank Prof. N. Kuster, Denis Sp¨at and Albert Roman (IT’IS Foundation, Zurich) for providing us with the exposure system, for checking the function of the system and for supporting the blind evaluation of the experiments. The ES1 cells were kindly provided by Prof. R¨udiger and the study was financially supported by the VERUM foundation. References [1] Vijayalaxmi, G. Obe, Controversial cytogenetic observations in mammalian somatic cells exposed to radiofrequency radiation, Radiat. Res. 162 (2004) 481–496.
47
[2] REFLEX, Risk Evaluation Of Potential Environmental Hazards From Low Energy Electromagnetic Field Exposure Using Sensitive in vitro Methods, Final Report, 2004 (http://www.verumfoundation.de). [3] E. Diem, C. Schwarz, F. Adlkofer, O. Jahn, H. R¨udiger, Nonthermal DNA breakage by mobile-phone radiation (1800 MHz) in human fibroblasts and in transformed GFSH-R17 rat granulosa cells in vitro, Mutat. Res. 583 (2005) 178–183. [4] M.J. Crumpton, A.R. Collins, Are environmental electromagnetic fields genotoxic? DNA Repair (Amst.) 3 (2004) 1385–1387. [5] Vijayalaxmi, J.P. McNamee, M.R. Scarfi, Comments on: “DNA strand breaks” by Diem et al. [Mutat. Res. 583 (2005) 178–183] and Ivancsits et al. [Mutat. Res. 583 (2005) 184–188], Mutat. Res. 603 (2006) 104–106. [6] J. Schuderer, T. Samaras, High peak SAR exposure unit with tight exposure and environmental control for in vitro experiments at 1800 MHz, IEEE Trans. Microw. Theory Tech. 52 (2006) 2057–2066. [7] G. Speit, A. Hartmann, The comet assay: a sensitive genotoxicity test for the detection of DNA damage and repair, Methods Mol. Biol. 314 (2006) 275–286. [8] O. Zeni, A.S. Chiavoni, A. Sannino, A. Antolini, D. Forigo, F. Bersani, M.R. Scarfi, Lack of genotoxic effects (micronucleus induction) in human lymphocytes exposed in vitro to 900 MHz electromagnetic fields, Radiat. Res. 160 (2003) 152–158. [9] O. Zeni, M. Romano, A. Perrotta, M.B. Lioi, R. Barbieri, G. d’Ambrosio, R. Massa, M.R. Scarfi, Evaluation of genotoxic effects in human peripheral blood leukocytes following an acute in vitro exposure to 900 MHz radiofrequency fields, Bioelectromagnetics 26 (2005) 258–265. [10] M.R. Scarfi, A.M. Fresegna, P. Villani, R. Pinto, C. Marino, M. Sarti, P. Altavista, A. Sannino, G.A. Lovisolo, Exposure to radiofrequency radiation (900 MHz, GSM signal) does not affect micronucleus frequency and cell proliferation in human peripheral blood lymphocytes: an interlaboratory study, Radiat. Res. 165 (2006) 655–663. [11] H.W. R¨udiger, E. Kratochvil, A. Pilger, Reply to the letter to the editor: reply to the letter by Vijayalaxmi et al. [Mutat. Res. 583 (2005) 178–183], Mutat. Res. 603 (2006) 107–109. [12] S. Ivancsits, A. Pilger, E. Diem, O. Jahn, H.W. R¨udiger, Cell type-specific genotoxic effects of intermittent extremely low-frequency electromagnetic fields, Mutat. Res. 583 (2005) 184–188. [13] R.R. Tice, G.G. Hook, M. Donner, D.I. McRee, A.W. Guy, Genotoxicity of radiofrequency signals. I. Investigation of DNA damage and micronuclei induction in cultured human blood cells, Bioelectromagnetics 23 (2002) 113–126.
Genotoxic effects of exposure to radiofrequency electromagnetic fields (RF-EMF) in cultured mammalian cells are not independently reproducible G¨unter Speit ∗ , Petra Sch¨utz, Heike Hoffmann Universit¨at Ulm, Abteilung Humangenetik, D-89070 Ulm, Germany Received 22 June 2006; received in revised form 3 August 2006; accepted 4 August 2006 Available online 25 September 2006
Abstract Conflicting results have been published regarding the induction of genotoxic effects by exposure to radiofrequency electromagnetic fields (RF-EMF). Using the comet assay, the micronucleus test and the chromosome aberration test with human fibroblasts (ES1 cells), the EU-funded “REFLEX” project (Risk Evaluation of Potential Environmental Hazards From Low Energy Electromagnetic Field Exposure Using Sensitive in vitro Methods) reported clearly positive effects for various exposure conditions. Because of the ongoing discussion on the biological significance of the effects observed, it was the aim of the present study to independently repeat the results using the same cells, the same equipment and the same exposure conditions. We therefore exposed ES1 cells to RF-EMF (1800 MHz; SAR 2 W/kg, continuous wave with intermittent exposure) for different time periods and then performed the alkaline (pH > 13) comet assay and the micronucleus test (MNT). For both tests, clearly negative results were obtained in independently repeated experiments. We also performed these experiments with V79 cells, a sensitive Chinese hamster cell line that is frequently used in genotoxicity testing, and also did not measure any genotoxic effect in the comet assay and the MNT. Appropriate measures of quality control were considered to exclude variations in the test performance, failure of the RF-EMF exposure or an evaluation bias. The reasons for the difference between the results reported by the REFLEX project and our experiments remain unclear. © 2006 Elsevier B.V. All rights reserved. Keywords: Comet assay; Micronucleus test; Low energy electromagnetic field; DNA damage; Mutation
1. Introduction During the last years, a large number of studies were performed using cultured human and other mammalian cells to determine a potential genotoxic effect of exposure to radiofrequency electromagnetic fields (RF-EMF). Conflicting results were reported but the reasons for the contradictory findings remained unclear.
∗ Corresponding author. Tel.: +49 731 500 23429; fax: +49 731 500 23438. E-mail address: [email protected] (G. Speit).
1383-5718/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.mrgentox.2006.08.003
The investigations published in scientific journals between 1990 and 2003 were critically evaluated in a recent review [1]. Various results indicating a genotoxic effect of RF-EMF were recently produced by the collaborative EU-funded REFLEX (Risk Evaluation of Potential Environmental Hazards From Low Energy Electromagnetic Field Exposure Using Sensitive in vitro Methods) project [2]. One laboratory taking part in the REFLEX project published genotoxic effects in cultured human fibroblasts exposed to RF-EMF (1800 MHz, SAR 2 W/kg, continuous or intermittent exposure, with or without modulations of the signal) for 16 or 24 h [3]. DNA damage was assessed by use of the comet assay, in
G. Speit et al. / Mutation Research 626 (2007) 42–47
both alkaline and neutral versions. Intermittent exposure caused a stronger effect than continuous exposure. The positive comet assay results were supported by positive findings in the micronucleus test (MNT) and the chromosome aberration test [2]. Due to the importance of these results in general, and the critical comments particularly with regard to the assessment of comet assay effects and the biological significance of the comet assay results [4,5], an independent replication of the experiments and confirmation of the results was desirable. Due to our experience with the comet assay and other genotoxicity tests, we have been asked by the coordinator of the REFLEX project to repeat parts of the study in our laboratory using the same cell cultures and the same exposure conditions as in the previous experiments [2,3]. We were provided with ES1 human fibroblasts and a system for RF-EMF exposure under strictly controlled conditions [6]. We decided to perform the alkaline comet assay using exposure conditions (1800 MHz, SAR 2 W/kg, continuous wave, intermittent exposure, 5 min field on/10 min field off) that led to the strongest effects in the previous investigations [3]. The MNT should be performed in parallel to confirm effects with a second genetic endpoint indicating true mutagenic effects on the chromosomal level. We also decided to conduct experiments with V79 cells because this rapidly proliferating Chinese hamster cell line is very sensitive towards a broad spectrum of genotoxic effects and routinely used in genotoxicity testing. It was the goal of our study to independently confirm the results of the REFLEX project. 2. Materials and methods 2.1. Cell culture and RF-EMF exposure conditions Human diploid fibroblasts (ES1 cells) and a Chinese hamster cell line (V79 cells) were cultured in DMEM and MEM, respectively, supplemented with 10% fetal calf serum (FCS) and antibiotics. Cells were maintained in a humidified incubator at 37 ◦ C with 5% CO2 and harvested with 0.15% trypsin and 0.08% EDTA. For the experiments, cells were seeded into 35-mm Petri dishes 24 h prior to RF-EMF exposure. The exposure system was built, provided and checked for correct function by the Foundation for Research on Information Technologies in Society, Zurich, Switzerland (IT’IS Foundation). For this setup R18 hollow rectangular waveguides operating at a frequency of 1800 MHz are used as previously described in detail [3,6]. Two waveguide-based exposure chambers containing the cell-culture dishes (exposed, sham-exposed) were placed inside a commercial incubator to ensure constant environmental conditions (37 ◦ C, 5% CO2 , 95% humidity). In addition, the incident field was monitored inside the waveguides. The air temperature difference between the chambers measured at the air-outlet did not exceed 0.1 ◦ C.
43
To enable blind exposure and evaluation, a computer randomly determined which of the two waveguides was exposed. This information was provided to us by the IT’IS Foundation via e-mail after transmission of our test results. The fields as well as all sensors were continuously monitored. The experiments were generally performed at a carrier frequency of 1800 MHz at intermittent exposure (5 min field on/10 min field off). The field was applied without modulation of the signal (=continuous wave). In some experiments, we performed a continuous exposure with modulation (GSM basic) of the signal (see Diem et al. [3] for technical details). The average specific absorption rate (SAR) was 2 W/kg because it is the safety limit for the mobile-phone microwave-radiation emission and was used in the previous experiments [2,3]. The exposure time varied between 1 and 24 h. After the end of the exposure period, cells were detached with trypsin and directly used for the comet assay. For the MNT, cells were either used directly after exposure or suspended in fresh culture medium and cultured further. 2.2. Comet assay The comet assay was performed according to the protocol that is usually used previously in our laboratory [7]. Aliquots of 10-l cell suspension (about 15,000 cells) were mixed with 120 l low melting-point agarose (0.5% in PBS) and added onto microscope slides (with frosted ends), which had been covered with a bottom layer of 1.5% agarose. Slides were lysed (pH 10; 4 ◦ C) for at least 1 h and processed as previously described in detail [7] using a time of alkali denaturation of 25 min and electrophoresis (0.86 V/cm) of 25 min at a pH > 13. Slides were coded and stained with ethidium bromide (10 g/ml). Measurements were made for 50 randomly selected cells per slide by image analysis (Comet Assay II, Perceptive Instruments). For all experiments, we evaluated three image analysis parameters: tail migration, tail intensity and tail moment. In none of the experiments there was a significant difference between these parameters. Therefore, we chose one parameter (tail moment) for the presentation of the results. Tail moment (TM) is calculated according to the formula: TM = (tail intensity/total comet intensity) × (tail centre of gravity − peak position). In one experiment the “comet tailfactor” was comparatively determined according to Diem et al. [3]. For each data point 500 cells were scored and assigned to five categories corresponding to the amount of DNA in the tail. For a positive control, a parallel culture was exposed to 2 Gy 137 Cs gamma-rays at 4 Gy/min at the end of the RF-EMF exposure period of the experimental cultures and immediately processed in the comet assay together with the RF-EMF-exposed cultures. 2.3. Micronucleus test (MNT) The MNT was basically performed as described by the REFLEX project [2] in accordance with standard procedures. After RF-EMF exposure, ES1 cells were cultured for another 44 h in the presence of cytochalasin B (final concentration
44
G. Speit et al. / Mutation Research 626 (2007) 42–47
3 g/ml). As a positive control, a parallel culture was exposed to 2 Gy 137 Cs gamma-rays at 4 Gy/min at the end of the 24-h RF-EMF exposure period of the experimental cultures, and further processed in parallel. V79 cells were either cultured after exposure (without cytochalasin B) for another 18 h, or they were directly processed after an exposure period of 18–22 h. As a positive control, a parallel culture was exposed to 2 Gy 137 Cs gamma-rays at 4 Gy/min and further processed in parallel to the RF-EMF-exposed cultures (i.e. cultivation for 18 or 22 h, respectively). Cells were detached by trypsin, treated after centrifugation with a hypotonic solution (0.56% KCl) and fixed three times with methanol/glacial acetic acid (5:1 for ES1 cells; 3:1 for V79 cells). Air-dried slides were stained with acridine orange (125 g/ml in phosphate buffer). The frequency of micronuclei (MN) was determined by analysing 500 binucleated cells (ES1) or 1000 cells (V79) from coded slides. 2.4. Statistical analysis The experiments with ES1 cells were independently performed three times and differences between mean values were tested for statistical significance (P < 0.05) using Student’s ttest. Experiments with V79 cells were confirmed in a second independently performed test and results are presented as the mean of the two tests. All results were evaluated with regard to their biological relevance and in comparison with the previously reported effects under the same experimental conditions.
3. Results and discussion In a previously published study [3], RF-EMF induced genotoxic effects in the alkaline comet assay with ES1 cells. Significant differences between exposed and sham-exposed cultures were measured after exposure periods of 16 and 24 h, but not after 4 h. Genotoxic effects were found independently of whether the field was applied with or without modulation and whether exposure was continuous or intermittent. We decided to repeat the experiments using a field without modulation (continuous wave) and intermittent exposure (5 min field on/10 min field off) because this protocol had produced the strongest effects (i.e. doubling of the comet tail-factor). We used the same exposure conditions (1800 MHz, SAR 2 W/kg) but exposure times of 1, 4 and 24 h. The published data did not show any difference between the time points 16 and 24 h and we preferred to substitute the time point 16 h by a very early time point, because the comet assay usually detects primary DNA damage that may occur immediately and may be repaired rapidly. Our results are summarized in Fig. 1A and clearly show that under the same experimental conditions as used by Diem et al. [3] we did not detect a genotoxic effect of RF-EMF on ES1 cells at the three time points tested.
Fig. 1. The effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on DNA migration (tail moment) in the comet assay (A) and on the frequency of micronuclei (B) in ES1 cells after exposure for different time periods (1, 4 and 24 h) in comparison to sham-exposed cultures and a negative (Co) and a positive (2 Gy gamma irradiation after cultivation for 24 h) control. Mean ± S.E.M. of three independent experiments.
Increased DNA effects after exposure for 16 or 24 h but not after exposure for 4 h as reported by Diem et al. [3] is an uncommon finding with the comet assay because most genotoxins induce effects rapidly and such a delayed effect may be indicative of an indirect effect related to DNA repair or disturbed physiological conditions. Furthermore, there is no apparent explanation as to why intermittent exposure should induce more DNA damage than continuous exposure. Diem et al. [3] also used a neutral version of the comet assay and show results that are nearly identical to those of the alkaline comet assay. The fact that there was nearly no difference between the two versions with regard to the absolute values of the controls and the exposed cultures casts doubt on the specificity of the neutral version of the comet assay used and the biological significance of these findings. Altogether, the study of Diem et al. [3] suggests the induction of a genotoxic effect by RF-EMF in the comet
G. Speit et al. / Mutation Research 626 (2007) 42–47
assay with human fibroblasts. However, as we were not able to independently reproduce the results under the same experimental conditions, the relevance of the observed genotoxic effects remains unclear. To exclude a difference between the exposure conditions, the exposure systems used in Ulm and Vienna were checked by the experts from the IT’IS Foundation (Zurich) and comparative measurements did not reveal any significant difference between the two exposure units. As part of the REFLEX project the induction of micronuclei in ES1 cells was also studied and a clear effect was measured [2]. The results of the MNT are of particular importance because the genetic endpoint is a chromosomal mutation transmitted to a daughter cell. While comet assay effects can be questioned with regard to their genetic relevance (i.e. the DNA effects seen in the comet assay might be DNA damage of transient nature and/or intermediates in the repair of DNA damage), the biological relevance of micronuclei is not a matter of discussion. Furthermore, while the evaluation of comet assay effects may lead to conflicting results due to the high assay variability and the different ways to determine an effect, micronuclei are detected with high reliability and there is no serious concern that a positive effect might be missed under appropriate test conditions. We performed the micronucleus test using the same exposure conditions as used in our comet assay experiments (SAR 2 W/kg, intermittent exposure, 5 min on/10 min off for 24 h) and did not measure any effect on the frequency of micronuclei (Fig. 1B). The comet assay results reported for the REFLEX study indicated that the different modes of exposure do not have a fundamental influence on the outcome of the test. However, to exclude an influence of the mode of exposure and to exclude differences in the test performance between the laboratories in Ulm and Vienna, colleagues from Vienna came to Ulm to conduct experiments together. In these experiments we used our standard exposure conditions but also GSM basic modulation, 1800 MHz, 1 or 2 W/kg for 24 h. We performed the comet assay and the MNT (only single experiments without independent repetition), prepared slides in parallel and evaluated the coded slides independently. All these experiments did not reveal any genotoxic effect of RF-EMF and the negative results were confirmed by the evaluation of these slides by our colleagues from Vienna (data not shown). It is surprising that the clear effect in the MNT (20-fold increase, in the range of the positive control, 10 g/ml bleomycin) could not be reproduced under nearly identical experimental conditions in our laboratory, but there is no obvious reason for this discrepancy. A clastogenic effect of RF-EMF on ES1 cells has further been demonstrated in the REFLEX project
45
by a positive result from a chromosome aberration test. Interestingly, a strong induction of chromosometype aberrations (i.e. dicentric chromosomes and acentric fragments) was measured. Because of the clearly negative MNT in our laboratory, we refrained from conducting a chromosome aberration test. In previously published studies with cultured human lymphocytes no induction of chromosome aberrations, micronuclei or sister chromatid exchange was measured after exposure to a 900-MHz GSM signal (SAR between 0.2 and 10 W/kg) for various periods of time [8–10]. Regarding the comet assay effects, it is interesting to note that the different ways of evaluating the comet assay (i.e. determination of the tail moment of 50 cells by image analysis and determination of the comet tailfactor of 500 cells by defining five categories of cells corresponding to the amount of DNA in the tail [3]) led to very similar results (Fig. 2). It has been discussed, whether the tail-factor used by Diem et al. [3] is a reliable measure of DNA effects in the comet assay [5,11]. This comparative evaluation demonstrates that the calculation of the tail-factor does not reveal a difference that is missed by our evaluation using image analysis. It confirms that the results obtained in our laboratory with the comet assay are in fact negative. Our negative findings with ES1 cells were confirmed by comprehensive experiments with V79 cells. We performed additional experiments with V79 cells because this cell line is well known for its sensitivity for the detection of genotoxic effects and is frequently used in routine genotoxicity testing. We performed the comet assay immediately after an exposure period of 1, 4 or 24 h (Fig. 3A) and did not measure any significant increase
Fig. 2. Parallel evaluation of the effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off; 24 h) on DNA migration (tail moment) in the comet assay with ES1 cells using either image analysis (tail moment) or arbitrary damage classes (tailfactor).
46
G. Speit et al. / Mutation Research 626 (2007) 42–47
Fig. 4. The effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on the frequency of micronuclei in V79 cells after exposure for 18 h and analysis at the end of the exposure period in comparison with sham-exposed cultures and a negative (Co) and a positive (2 Gy gamma irradiation at the beginning of the exposure period) control. Mean of two independent experiments.
Fig. 3. The effect of RF-EMF (1800 MHz, SAR 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on DNA migration (tail moment) in the comet assay (A) and on the frequency of micronuclei (B) in V79 cells after exposure for different time periods (1, 4 and 24 h) and further cultivation for 18 h in comparison with sham-exposed cultures and a negative (Co) and a positive control (2 Gy gamma irradiation after cultivation for 24 h, immediate analysis in the comet assay, further cultivation for 18 h in the MNT). Mean ± S.E.M. of three independent experiments.
in DNA migration (tail moment). Parallel cultures were further cultured for 18 h after the end of the exposure and then prepared for the analysis of micronuclei. In accordance with the results with ES1 cells there was no induction of micronuclei in any of the experiments (Fig. 3B). We then used a modification of the time schedule and exposed V79 cells for 18 h and prepared the cells for analysis of micronuclei immediately after the end of the exposure. This protocol, which is also suggested for routine genotoxicity testing with the MNT, should detect micronuclei that are formed during the exposure period. Again, the result of the modification of the MNT was also clearly negative (Fig. 4). We also analysed V79 cells after exposure for 4 h (with 18 h post-incubation) or for 22 h to RF-EMF with a SAR of 1 and 2 W/kg and obtained negative results (Fig. 5). In our opinion, the negative
results with V79 cells confirm the negative findings with ES1 cells. Effects for negative and positive controls were similar to those measured in ES1 cells. However, fundamental differences between different cell types with regard to the induction of genotoxic effects by electromagnetic fields cannot be excluded. In a study applying extremely low-frequency electromagnetic fields (ELFEMF) it was shown that a genotoxic effect can be found in some types of cultured mammalian cells but not in others [12]. The cause for this difference is unknown but might just be a matter of different sensitivity as suggested by positive effects of RF signals (837 MHz) in lympho-
Fig. 5. The effect of RF-EMF (1800 MHz, SAR 1 and 2 W/kg; continuous wave, intermittent 5 min on/10 min off) on the frequency of micronuclei in V79 cells after exposure for 4 h and further cultivation for 18 h and after exposure for 22 h and analysis at the end of the exposure period in comparison with sham-exposed cultures and a negative (Co) and a positive (2 Gy gamma irradiation and cultivation for 22 h) control. Mean of two independent experiments.
G. Speit et al. / Mutation Research 626 (2007) 42–47
cytes after extended exposure conditions (5 and 10 W/kg for 24 h) [13]. However, a recent inter-laboratory study using lymphocyte cultures from 10 healthy donors did not find a genotoxic effect of exposure to 900 MHz (GSM signal) at SARs of 1, 5 and 10 W/kg [10]. These inconsistencies may also indicate that the observed genotoxic effect of electromagnetic fields in some studies is not a direct genotoxic effect (i.e. due to a direct interaction with the genetic material, which should occur in every directly exposed cell type), but may be due to unknown indirect effects. In conclusion, we were not able to independently confirm some of the genotoxic effects of RF-EMF reported by the REFLEX project. However, our results do not exclude a genotoxic effect of RF-EMF in general. We did not test other exposure conditions and the role of the exposure conditions for the expression of a genotoxic effect is not yet understood. Further mechanistic studies are necessary to elucidate the biological basis of the genotoxic effects reported for RF-EMF and to clarify the contradictory findings in different studies. Acknowledgements We gratefully acknowledge the excellent cooperation with Prof. F. Adlkofer (REFLEX project; VERUM foundation) and Prof. H.W. R¨udiger and his co-workers (Vienna). We thank Prof. N. Kuster, Denis Sp¨at and Albert Roman (IT’IS Foundation, Zurich) for providing us with the exposure system, for checking the function of the system and for supporting the blind evaluation of the experiments. The ES1 cells were kindly provided by Prof. R¨udiger and the study was financially supported by the VERUM foundation. References [1] Vijayalaxmi, G. Obe, Controversial cytogenetic observations in mammalian somatic cells exposed to radiofrequency radiation, Radiat. Res. 162 (2004) 481–496.
47
[2] REFLEX, Risk Evaluation Of Potential Environmental Hazards From Low Energy Electromagnetic Field Exposure Using Sensitive in vitro Methods, Final Report, 2004 (http://www.verumfoundation.de). [3] E. Diem, C. Schwarz, F. Adlkofer, O. Jahn, H. R¨udiger, Nonthermal DNA breakage by mobile-phone radiation (1800 MHz) in human fibroblasts and in transformed GFSH-R17 rat granulosa cells in vitro, Mutat. Res. 583 (2005) 178–183. [4] M.J. Crumpton, A.R. Collins, Are environmental electromagnetic fields genotoxic? DNA Repair (Amst.) 3 (2004) 1385–1387. [5] Vijayalaxmi, J.P. McNamee, M.R. Scarfi, Comments on: “DNA strand breaks” by Diem et al. [Mutat. Res. 583 (2005) 178–183] and Ivancsits et al. [Mutat. Res. 583 (2005) 184–188], Mutat. Res. 603 (2006) 104–106. [6] J. Schuderer, T. Samaras, High peak SAR exposure unit with tight exposure and environmental control for in vitro experiments at 1800 MHz, IEEE Trans. Microw. Theory Tech. 52 (2006) 2057–2066. [7] G. Speit, A. Hartmann, The comet assay: a sensitive genotoxicity test for the detection of DNA damage and repair, Methods Mol. Biol. 314 (2006) 275–286. [8] O. Zeni, A.S. Chiavoni, A. Sannino, A. Antolini, D. Forigo, F. Bersani, M.R. Scarfi, Lack of genotoxic effects (micronucleus induction) in human lymphocytes exposed in vitro to 900 MHz electromagnetic fields, Radiat. Res. 160 (2003) 152–158. [9] O. Zeni, M. Romano, A. Perrotta, M.B. Lioi, R. Barbieri, G. d’Ambrosio, R. Massa, M.R. Scarfi, Evaluation of genotoxic effects in human peripheral blood leukocytes following an acute in vitro exposure to 900 MHz radiofrequency fields, Bioelectromagnetics 26 (2005) 258–265. [10] M.R. Scarfi, A.M. Fresegna, P. Villani, R. Pinto, C. Marino, M. Sarti, P. Altavista, A. Sannino, G.A. Lovisolo, Exposure to radiofrequency radiation (900 MHz, GSM signal) does not affect micronucleus frequency and cell proliferation in human peripheral blood lymphocytes: an interlaboratory study, Radiat. Res. 165 (2006) 655–663. [11] H.W. R¨udiger, E. Kratochvil, A. Pilger, Reply to the letter to the editor: reply to the letter by Vijayalaxmi et al. [Mutat. Res. 583 (2005) 178–183], Mutat. Res. 603 (2006) 107–109. [12] S. Ivancsits, A. Pilger, E. Diem, O. Jahn, H.W. R¨udiger, Cell type-specific genotoxic effects of intermittent extremely low-frequency electromagnetic fields, Mutat. Res. 583 (2005) 184–188. [13] R.R. Tice, G.G. Hook, M. Donner, D.I. McRee, A.W. Guy, Genotoxicity of radiofrequency signals. I. Investigation of DNA damage and micronuclei induction in cultured human blood cells, Bioelectromagnetics 23 (2002) 113–126.