Cytogenetic analysis of the effects of 2 . 5 and 10 . 5 GHz microwaves on human lymphocytes

The biological effects of microwaves on living organisms remain highly controversial. Although some reports have suggested that microwaves may be directly or indirectly genotoxic, a direct action is unlikely because the low energy of microwave photons makes them unable to cause single-strand breaks in DNA. In this work, we examined the possible clastogenic properties of microwaves (2.5 and 10.5 GHz) on blood lymphocytes in vitro by monitoring the frequency of chromosomal aberrations. We also investigated whether blood cells showed increased radiosensitivity or radioresistance when pretreated with the microwaves and then irradiated with gamma radiation. There was no significant difference in the frequency of chromosomal aberrations between cells which had or had not been treated with microwaves. Control cells had a mean frequency of 0.013 aberrations per cell compared to 0.010 and 0.011 aberrations per cell in the microwave-exposed samples. Nor was there any alteration in the radiosensitivity of cells pretreated with microwaves. Gamma irradiated cells showed a mean frequency of 0.279 aberrations per cell compared to 0.343 and 0.310 aberrations per cell in samples pretreated with microwaves. However, cell mortality increased markedly after exposure to microwaves. The results suggest that microwaves do not interact directly or indirectly with chromosomes, although they may target other cell structures, such as cell membranes.


Introduction
In recent decades, there have been considerable advances in the development of sources of non-ionizing radiation, such as microwaves.The widespread use of such energy sources and the increase in the number of devices emitting microwaves and radiofrequencies (RF), including mobile phones, has become a matter of concern for regulatory authorities and non-regulatory bodies (IEGMP, 2000).
Structural chromosomal aberrations may involve the chromosomes or chromatids.Most chemical mutagens and non-ionizing mutagenic radiations are unable to cause double-strand breaks in DNA and act mainly in the S phase of the cell cycle.Such agents are only indirectly clastogenic and produce mainly chromatid-type aberrations (OECD, 1997).In contrast, chromosome-type aberrations are induced directly by agents such as ionizing radiation that can produce double strand breaks in DNA (IAEA, 2001).
The biological effects of microwaves on living organisms are highly controversial (Maes et al., 1993).A direct genotoxic action is unlikely because of the low energy of microwave photons which are unable to cause strand breaks in DNA.However, despite this general conviction that microwaves are not sufficiently energetic to be able to directly damage DNA, there is considerable evidence indicating that microwaves can be directly and indirectly clastogenic, with a significant increase in chromosome damage (Sagripanti and Swicord, 1986;Garaj-Vrhovac et al., 1991, 1992;Maes et al. 1993;Haidler et al., 1994;Sarkar et al., 1994;Lai andSingh 1995, 1996;Timchenko and Ianchevskaia, 1995;Balode, 1996;Verschaeve et al., 1994;Vijayalaxmi et al., 1997;Phillips et al., 1998;Tice et al., 1999).In addition, cell phone radiation can alter protooncogene activity (Ivaschuk et al., 1997;Goswami et al., 1999).However, a similar number of studies have failed to detect obvious clastogenic effects following microwave irradiation of isolated animal cells in vitro (Alam et al., 1978;Lloyd et al., 1984Lloyd et al., , 1986;;Wolff et al., 1985;Meltz et al., 1987Meltz et al., , 1989Meltz et al., , 1990;;Kerbacher et al., 1990;Maes et al., 1997Maes et al., , 2001)).Thus, there is still no conclusive answer as to whether exposure to microwaves is clastogenic, i.e., whether they can direct or indirectly increase the frequency of chromosomal aberrations.
A further question is whether microwaves can act as epigenetic factors to influence the genotoxicity of other environmental "pollutants" (Maes et al., 2001).Cancer is generally considered to be initiated by alterations in DNA.However, some non-genotoxic chemicals and processes (known as epigenetic carcinogens) are unable to damage DNA and are usually not clastogenic in vitro, but can enhance the progress of cells towards malignancy in vivo.Several studies have suggested that radiofrequency radiation (RF) has an epigenetic effect in vivo, and can enhance the genotoxic effects of ionizing radiation or cancerinducing substances, or potentiate other epigenetic factors (ICNIRP, 1998).
The aim of this study was to investigate the clastogenic effects of 2.5 and 10.5 GHz microwave fields, alone and in combination with ionizing radiation, on peripheral blood lymphocytes.The combination of microwaves with ionizing radiation ("synergy" test) was designed to screen for joint effects of those two types of radiation.For this, we assessed whether blood samples pretreated with microwaves would be more sensitive (or resistant) to damage by gamma radiation.In all cases, the chromosomal damage was assessed using conventional cytogenetic techniques.The chromosome aberration test is often used to identify physical or chemical agents that cause structural chromosomal aberrations in cultured mammalian cells (OECD, 1997).

Materials and Methods
The microwave sources used were: (a) a 2450 MHz microwave thermal oven (model MARS 5, CEM Corporation) with a power output of up to 1200 W and controls for regulating power and temperature by ventilation, and (b) a 10.5 GHz, 15 mW, linearly polarized, non-thermal microwave source (model WA-9314B, PASCO).The gamma radiation source was cobalt-60 (0.034 Gy.min -1 ), with the absorbed dose being 1.5 Gy (4.5 mJ) per blood aliquot.
Initially, whole blood samples were exposed to the microwave sources for varying periods of time in order to determine the best exposure time for each source.Long periods of exposure resulted in a high cell mortality, seen as a high rate of cell lysis and a low number of metaphases after culturing.Based on these preliminary experiments, exposure times of 40 s at 3 W for the 2.5 GHz oven, and 5 min for the 10.5 GHz device were used.To prevent overheating in the 2.5 GHz oven, the temperature of the blood samples was kept below 36 °C, (starting temperature was 28 °C and reached 33 °C after 40 s of exposure).In the case of the 10.5 GHz device, initial tests showed that there was no increase in the temperature of water samples, even after hours of exposure to this device.
The energies transmitted to each blood sample were calculated to be 75,310 and 230 mJ for the 2.5 and 10.5 GHz sources, respectively.When expressed as the specific energy absorption rate (SAR), these energies cor-responded to 626.67 W.kg -1 and 0.25 W.kg -1 , respectively.The SAR expresses the energy absorbed and is a function of the power absorbed in the sample (in Watts) per kg of sample mass.
A 10 mL blood sample was collected into a heparinized vacutainer and immediately divided into six blood aliquots.An equal volume of culture medium (1.5 mL) without phytohemaglutinin (PHA) was added to each blood aliquot before the treatment (irradiation with microwaves and/or gamma radiation).One 3 mL aliquot served as the untreated control, another served as the 1.5 Gy gamma-irradiated control, and the remaining aliquots were treated with microwaves, with or without subsequent 1.5 Gy gamma irradiation.All aliquots were held at 37 °C during the irradiations and incubations.A 2 h interval was allowed between the treatment with microwaves and exposure to gamma radiation.All control samples were handled in the same way as the exposed ones, but without exposure to microwaves or radiation.
Lymphocytes from all blood samples were cultured under identical conditions using standard methods (IAEA, 2001), with modifications.Briefly, 10 mL of Ham's F-10 medium (Cultilab, Campinas, SP, Brazil) supplemented with 25% fetal calf serum (Cultilab) and 0.5 mL of phytohemagglutinin M (Gibco -Invitrogen Corporation, Carlsbad, CA, USA) was used.The cells were incubated for 48 h and 0.04 mg of colchicine (Sigma Chemical Co., St. Louis, MO, USA) was added 3 h before harvesting.After treatment with hypotonic saline solution (0.075 M KCl) for 15 min, the lymphocytes were fixed in methanol:acetic acid (3:1, v/v) and transferred to clean microscope slides followed by staining with 3% Giemsa.
The chromosome aberration test was done using blood samples from four healthy volunteers, both sexes, different ages and not under the use of medications (age and sex in parentheses): donor 1 (44 y, F), donor 2 (28 y, M), donor 3 (23 y, F) and donor 4 (36 y, F).After exposure of the blood aliquots to the different treatments, phytohemagglutinin-stimulated (48 h) lymphocyte cultures were started to obtain chromosomal preparations.The samples were scored blind, except during the initial experiments to estimate the appropriate exposure times, during which the viability of the cultures was evaluated.
In all of the experiments, the maximum possible number of cells per sample was scored using a Nikon Labophot light microscope.Based on the guidelines for the in vitro mammalian chromosome aberration test issued by the OECD (1997), at least 200 well-spread metaphases per sample were scored for structural chromosome-and chromatid-type aberrations.The frequency of polyploid cells was also examined since an increase in polyploidy may indicate that a chemical or physical agent has the potential to induce numerical aberrations.
The results were expressed as the aberration yield (Y ± S.E.), with the standard errors calculated using a Poisson distribution.The distribution of aberrations among the scored cells was tested for conformity to the Poisson distribution.Values of U higher than 1.96 indicated that the distribution was overdispersed (IAEA, 2001).All statistical comparisons were done using the Kruskal-Wallis H-test, within 95% confidence limits.

Results
Table 1 shows the chromosome-and chromatid-type aberrations seen among lymphocytes from blood samples exposed to the microwave fields.There was no significant difference between control cells and those exposed to microwave fields.Control cells had a mean frequency of 0.013 aberrations per cell compared to 0.010 and 0.011 aberrations per cell in the microwave-exposed samples.Statistical comparison of these results using the Kruskal-Wallis H-test revealed no significant differences within 95% confidence limits.The distribution of the aberrations among cells is shown in Table 2. Subject 1 had a somewhat higher than normal and overdispersed frequency of aberrations, probably because of previous partial-body irradiation.This donor was therefore used in subsequent experiments (Table 3).
Table 3 shows the results of the "synergy" test.Cells were exposed to microwaves and subsequently to 1.5 Gy of 60 Co gamma radiation.There was no significant difference between microwave-treated or non-treated cells.Gamma irradiated cells showed a mean frequency of 0.279 aberrations per cell compared to 0.343 and 0.310 aberrations per cell in samples pretreated with microwaves.Statistical comparison of these results using the Kruskal-Wallis H-test revealed no significant differences within 95% confidence limits.The distribution of the aberrations among the cells scored is shown in Table 4. Acentric fragments tended to be slightly overdispersed, as normal (IAEA, 2001).
Table 5 summarizes the results of the numerical aberrations observed according to the different treatments.Again, there were no significant differences among the various groups, according to the Kruskal-Wallis H-test, within 95% confidence limits.

Discussion
Following exposure to microwaves from both sources, there was a high rate of cell mortality that increased with the amount of energy transferred to the cells.This mortality was reflected in the high degree of cell lysis and the low number of metaphases after culturing.For blood samples treated in the 2450 MHz oven, the cell lysis was attributed to thermal effects ("cooking").This phenomenon was also observed by Lloyd et al. (1984) in similar experiments.To prevent hyperthemia in the present experiments, the temperature of the blood samples was kept below 36 °C (the starting temperature was 28 °C and reached 33 °C after 40 s of exposure).However, in the case of the 10.5 GHz device, no thermal effects were observed since there was no increase in the temperature of the water samples, even after hours of exposure to this device.Thus, the high level of cell lysis and mortality seen following exposure to the 10.5 GHz device was attributable to other non-thermal processes.

462
Cytogenetic studies of microwaves in lymphocytes    There were no significant differences in the frequencies of chromosomal aberrations between microwave-treated or untreated samples, despite the fact that samples treated with microwaves received huge amounts of transferred energy that were 50-17,000 times greater than the energy transferred by 1.5 Gy of ionizing radiation.The intracellular targets for ionizing radiation are the chromosomes in the nucleus (IAEA, 2001).The results shown here suggest that microwaves do not interact directly or indirectly with chromosomes, although they may target other cell structures, such as cell membranes.This would explain the high degree of lysis seen in the microwave experiments.
Although direct genetic effects from microwave exposure were not expected to occur, indirect effects of microwaves would be more likely, because of the influence of eletromagnetic fields on the free radical system (Maes et al., 1997), but this was not seen in the present chromosome aberration test.These findings agree with other reports showing that the microwave irradiation of human lymphocytes in vitro has no direct or indirect clastogenic effects (Lloyd et al., 1984;Maes et al., 1997).In addition to this lack of a direct or indirect effect of microwaves on chromosomes, pretreating cells with microwaves also failed to affect their sensitivity to ionizing radiation.
A long-standing dogma in radiation science has been that energy from radiation must be deposited in the nucleus to elicit its biological effects.In recent years, a number of epigenetic effects have been described that challenge this dogma.Epigenetic factors, although not themselves genotoxic, act synergistically to enhance the carcinogenic effects of other agents.Several studies (Szmigielski et al., 1982;Scarfi et al., 1996;Maes et al., 1997;Pakhomova et al., 1997) have suggested that microwaves can have an epigenetic effect in vivo, and that they can exacerbate the genotoxicity of ionizing radiation or cancer-inducing sub-stances, or potentiate other epigenetic factors (IEGMP, 2000).However, the evidence for an epigenetic effect of microwaves is equivocal since some studies have failed to reproduce the positive results reported by others (Ciaravino et al., 1987(Ciaravino et al., , 1991;;Meltz et al., 1989Meltz et al., , 1990;;Cain et al., 1997).
In conclusion, the results described here do not support the hypothesis that microwaves enhance the direct effect of gamma radiation or cause cells to respond differently to ionizing radiation in vitro.It is possible that some of the epigenetic responses to microwaves in vivo could be the result of thermal effects (IEGMP, 2000), as concluded by Pakhomova et al. (1997), who found that high frequency microwaves (61 GHz) enhanced DNA recombination, but not mutagenesis, in yeast cells exposed to ultraviolet radiation.Our findings indicate that further investigations are needed to examine the influence of microwaves in vivo and in vitro.Cytogenetic studies of microwaves in lymphocytes

Table 1 -
Number and frequencies of structural chromosomal aberrations in human lymphocytes exposed to microwave fields.

Table 2 -
Distribution of the chromosome-type structural aberrations among the scored cells indicated in Table1and according to the treatment given.

Table 3 -
Number and frequencies of structural chromosomal aberrations in human lymphocytes exposed to microwave fields and subsequently to 1.5 Gy of gamma radiation.

Table 4 -
Distribution of the chromosome-type structural aberrations among the scored cells in Table3.

Table 5 -
Number and frequencies of chromosomal aberrations in human lymphocytes according to the different treatments.Samples from the different subjects were pooled for each treatment.