Since a very large number of cellular components, cellular processes, and cellular systems can conceivably be affected by EMF, mechanistic studies are essential to interpret and help guide the experimental work. Because evidence from previous theoretical and experimental studies suggested that EMF are unlikely to induce DNA damage directly, most studies have been conducted to examine its effect on the cellular membrane, general and specific gene expression, and signal transduction pathways. More recently, studies specifically addressing the genotoxic effect of exposure to magnetic fields have been pursued.
While some studies were conducted with samples from animals exposed to EMF in vivo, most results have come from studies of cultured cells exposed in vitro. The great advantage of in-vitro exposure is its precision, since the geometry and physical properties of the system can be well controlled.
While a major concern of the possible effects of EMF on public health is cancer, only a small number of in-vitro studies were designed to address particular hypotheses generated from previous studies of carcinogenesis. Thus, while EMF have been tested in a number of established assays for genetic toxicity, the effects primarily studied have been non-genotoxic mechanisms of carcinogenesis, including induction of cell proliferation, inhibition of intercellular communication, modulation of cell differentiation, and aberrant control of proto-oncogene expression. These cellular effects and subcellular effects such as active oxygen production and modulation of protein kinase C signal transduction are well-documented activities of non-genotoxic carcinogens such as the phorbol esters.
In this section, the in vitro effects of exposure to EMF are documented, including theoretical and physical considerations of mechanisms by which EMF can induce biological effects. The in vitro responses considered are mainly genotoxicity, regulation of gene expression, cell signaling and proliferation, and cell differentiation.
The interest in assessing the effects of EMF at the cellular level in order to understand the mechanism(s) by which exposure may be harmful and/or beneficial to human health has temporally coincided with a rapid increase in knowledge about regulation of gene expression and the role that specific gene products play in cell replication, proliferation, differentiation, and pathological processes.
The genotoxic effects of EMF have been extensively reviewed by McCann et al. (McCann et al., 1998; McCann et al., 1993).
It is generally accepted that ELF EMF do not transfer energy to cells in sufficient amounts to directly damage DNA; however, it is possible that certain cellular processes altered by exposure to ELF EMF, such as free radicals, indirectly affect the structure of DNA. Most investigators have looked for strand breaks and other chromosomal aberrations, including sister chromatid exchange, formation of micronuclei, and/or effects on DNA repair.
Cultured human peripheral blood lymphocytes from volunteers were used to assess the effects of power-frequency sine wave or pulsed magnetic fields on cytogenetic events. Magnetic fields of 1-7.5 mT and exposure durations of 48-72 h produced no significant chromosomal effects (Paile et al., 1995; Rosenthal & Obe, 1989; Scarfi et al., 1994). Simultaneous exposure to an electric field (60 Hz, 30 µA/cm2) and a magnetic field (0.1 or 0.2 mT) did not cause sister chromatid exchange or chromosomal breaks (Cohen et al., 1986a; Cohen et al., 1986b) In one study (Khalil & Qassem, 1991), however, EMF were found to induce chromosomal aberrations in human lymphocytes. In this study, exposure of human lymphocytes to a pulsed field with a peak magnetic flux density of 1.05 mT for 24-72 h induced a statistically significant, twofold increase in the frequency of chromosomal aberrations and a simultaneous lowering of the mitotic index. In human lymphocytes exposed to a 4.4 kHz pulsed field with a 14 Hz repetition rate, no effect on sister chromatid exchange rate was seen (Garcia-Sagredo et al., 1990). [With only one exception Rosenthal & Obe (1989), positive controls were not used in these studies.]
In another model, human amniotic fluid cells were exposed to a 50 Hz sine-wave field (30 µT rms) for 72 h. They showed a modest but consistent increase in the frequency of total chromosomal aberrations when compared with controls (Nordenson et al., 1992). While higher flux densities (300 µT) did not affect the cells in this manner, intermittent exposure (15 s on-off for 72 h) to the 30 µT field was as effective as continuous exposure in increasing aberration frequencies (Nordenson et al., 1994). In a study to replicate these findings, Galt et al. (Galt et al., 1995) reported a nonsignificant decrease in the number of cells with chromosomal aberrations after exposure to 30 µT. [A striking difference between these two investigations was the level of aberrations in the control cells, which was fourfold higher in the study of Galt et al; the authors did not comment on this point. Furthermore, the ambient DC field differed substantially between the two laboratories.] In established cell lines, however, exposure to EMF does not appear to induce chromosomal aberrations (Fairbairn & O'Neill, 1994; Fiorani et al., 1992; Livingston et al., 1991; Reese et al., 1988; Takahashi et al., 1987).
The 'Comet' assay (DNA microelectrophoresis) has been used to detect DNA fragmentation in single cells (Lai & Singh, 1997a; Lai & Singh, 1997c; Singh & Lai, 1998). Increases in both single- and double-strand breaks were reported in brain cells of rats exposed in vivo to a 60 Hz field at 0.1-0.5 mT. Radical scavengers like melatonin and N-tertbutylphenynitrone counteracted the effects of exposure, suggesting that magnetic fields can affect the levels or lifetimes of certain free radical species (Lai & Singh, 1997c). The authors further suggested that exposure to ELF magnetic fields (60 Hz, 0.5 mT, 2 h) can cross-link DNA and also DNA and proteins, in a manner similar to mitomycin C (Singh & Lai, 1998). [The authors used mitomycin C as a positive control only in human lymphocytes; in addition, the report does not mention ambient DC fields.]
Scarfi et al. (Scarfi et al., 1991; Scarfi et al., 1993; Scarfi et al., 1994) were consistently unable to detect an increase in micronuclei in normal cultured human lymphocytes exposed to a 50 Hz pulsed magnetic field or a sinusoidal E field; however, cells from patients with Turner's syndrome showed an 80-100% increase in micronuclei after similar exposure (Scarfi et al., 1997b). After exposure to a 100 Hz magnetic field, the number of micronuclei increased in cultured human lymphocytes (Scarfi et al., 1997a).
Tofani et al. (Tofani et al., 1995) found that the DC field is important, since exposure of peripheral lymphocytes to 32 or 50 Hz (75 or 100 µT) with no DC component was ineffective in inducing micronuclei, whereas cells exposed to the 32 Hz field had a higher incidence of micronuclei when a DC component (42 µT) was introduced parallel to the AC field. In Chinese hamster V79 cells, however, exposure to pulsed magnetic fields (10 or 100 Hz, 20 or 80 µT) for 24 h increased 3H-thymidine incorporation, while exposure to 0.4 mT decreased incorporation (Takahashi et al., 1986). Cadossi et al. (Cadossi et al., 1992) using the same assay conditions, reported that lymphocytes from aged donors and from patients with B-cell chronic lymphocytic leukemia responded better to the pulsed magnetic field (2.5 mT, 2 mV PEMF with a 50 Hz repetition rate, 24 or 48 h) than lymphocytes from young healthy donors. Exposure of human T-cell leukemia Jurkat cells to a magnetic field (1.8 mT) bone healing signal reduced 3H-thymidine incorporation by 50%, whereas a 60 Hz sine wave (0.1 or 0.4 mT) applied for 20 min reduced incorporation by 20-25 % (Nindl et al., 1997). Simko et al. (Simko et al., 1998) found increased micronucleus formation in a human squamous-cell carcinoma cell line (SCL II), but not in human amniotic fluid cells, exposed to a 50 Hz magnetic field (0.1-1.0 mT) for 24, 48, or 72 h. The results of most other studies on micronucleus formation have been negative (Lagroye & Poncy, 1997; Livingston et al., 1991), although the latter authors showed micronuclei formation in rat tracheal epithelial cells treated with ionizing radiation (6 Gy) prior to a 50 Hz, 100 µT field for 24 h, the magnetic field alone had no effect. Very strong static fields (4.7 T) were shown to decrease micronuclei frequency in mitomycin-treated Chinese hamster lung/IU cells (Okonogi et al., 1996).
Several authors have studied the ability of ELF EMF to alter the repair of strand breaks induced by hydrogen peroxide or by radiation. No effects were seen with exposure to either B or E fields (Bersani et al., 1989; Cantoni et al., 1995; Frazier et al., 1990; Whitson et al., 1986).
Relatively few studies have been reported of the effects of EMF on gene mutation (McCann et al., 1998). As is often the case, the exposure protocols differ widely, and no consistent picture has evolved on possible effects. Exposure to electric and/or magnetic fields at various frequencies did not induce mutation in Salmonella typhimurium (Morandi et al., 1996; Nafziger et al., 1993). Similarly, Ager and Radul (Ager & Radul, 1992) found that exposure of yeast cells to a magnetic field (1 mT) alone or in combination with ultraviolet irradiation (2-50 J/m2) had no effect on mutations; and Pakhomova et al. (Pakhomova et al., 1998) saw no mutations in Saccharomyces cerevisiae after exposure to an ultra-wide bandwidth pulsed field (101-104 kV/m, impulses repeated at 16 or 600 Hz). In contrast, Koana et al.(Koana et al., 1997) reported that a strong static field (5 T, 24-h exposure) increased the frequency of somatic recombinations in third-instar larvae of Drosophila melanogaster, which was blocked by vitamin E treatment, suggesting involvement of oxygen radicals.
Prior and concurrent treatment of a rat embryo fibroblast cell line carrying the Escherichia coli lacI gene with N-methyl-N-nitrosurea (MNU) or menedione and exposure to a 60 Hz magnetic field (3 mT) for 120 h did not increase the mutation frequency (Suri et al., 1996). The hprt gene mutation was induced in human MeWo cells exposed to 50 Hz, 400 mT for 2 h (Miyakoshi et al., 1996). Subsequently, these authors also reported (Miyakoshi et al., 1998) enhanced mutagenicity in a human osteosarcoma cell line, Saos-LP-12, exposed to a 50 Hz, 400 mT magnetic field. When this cell line, which does not carry the p53 gene, was transfected with an inducible wild-type p53 gene, the enhanced mutagenicity associated with exposure to EMF was suppressed, suggesting a role for the wild-type p53 gene in guarding the genome from DNA damage. No marked differences in the mutation spectrum in the hprt gene was observed in those cells with or without the p53 gene after exposure to magnetic fields (Miyakoshi et al., 1998). Nafziger et al. (Nafziger et al., 1993), however, did not observe induction of hprt mutations in Chinese hamster V79 cells after exposure to a 50 Hz magnetic field at lower field strengths (1 or 10 µT). An effect of EMF on hprt gene mutations was reported in Chinese hamster ovary cells after exposure to ionizing radiation in two laboratories. Exposure to EMF alone did not induce mutations in either study. A dose-related increase in mutation frequency induced by EMF (60 Hz, 0.47-0.7 mT) was reported after pre-exposure to g-irradiation (2 Gy) (Walleczek et al., 1998). A small increase in mutation rate was also observed when Chinese hamster ovary cells were exposed to EMF (60 Hz, 5 mT) after pre-exposure to X-rays (3 Gy) (Miyakashi et al., 1998). Similarly, EMF (60 Hz, 400 mT) increased the mutation rate in MeWo cells after prior exposure to X-ray (3 Gy) (Miyakoshi et al., 1996). [A clear effect of magnetic fields on mutation has so far been seen only at field strengths well above those that occur in the environment.]
Several studies have addressed the general question of whether ELF EMF can affect RNA synthesis, with both positive and negative outcomes. This subject has been reviewed (Lacy-Hulbert et al., 1998).
Effects on gene expression, particularly at the transcriptional level, after exposure to ELF EMF was first investigated by Goodman and coworkers. The main focus of their research was the effect of EMF on c-myc mRNA levels in human HL60 cells, although in earlier papers (Goodman et al., 1992; Goodman et al., 1989) they reported increased expression of several specific transcripts (including b-actin, b-tubulin, histone H2B, c-myc, and c-src) as deduced by dot-blot analysis. Goodman et al. (Goodman et al., 1989) found that several types of fields (sinusoidal 60 or 72 Hz; pulsed fields at 1.5, 15, or 72 Hz) at field strengths of 0.38-3.5 mT could increase mRNA levels by two- to threefold after 20 min exposure, as measured by dot-blot analysis. The strongest response was to 60 Hz sinusoidal fields with a 1.5 mT peak B value. In contrast, this group reported later (Goodman et al., 1992) that the most prominent effects appeared after 20 min exposure to a 60 Hz, 5.7 µT field. Stronger fields or longer exposure times tended to diminish the observed effects. The paper included a report of northern blot analysis to confirm the identity of the investigated transcripts [The group used dot-blot analysis without internal controls for quantification of mRNA levels]. Rao and Henderson (Rao & Henderson, 1996)] transfected HL60 cells with a c-fos promoter construct upstream of the bacterial CAT gene to obtain a reporter gene assay system. The positive control, TPA, increased CAT activity by 10-40% above control levels, while a 20-min exposure to a 60 Hz, 6 µT (rms) field increased the activity by 5-20 % above the control level. [The induction of promotor activity by TPA was rather small.] Increased expression of the large T antigen mRNA and protein in SV40-transformed fibroblasts exposed to a 60 Hz, 8 µT field for 20 min (the only exposure conditions employed) was reported (Gold et al., 1994).
EMF (pulsed 72 Hz field, 3.5 mT peak value) stimulated 3H-uridine incorporation, mRNA synthesis, and protein synthesis but not DNA synthesis in the human leukemia cell line CCRF-CEM (Phillips & McChesney, 1991). While exposure for 30 min to 4 h was effective, longer exposure diminished the response. The same cells were used in subsequent studies of transcription of c-myc, c-fos, jun, and PKC-b mRNA; nuclear run-off assays showed two- to threefold increases in mRNA transcription after exposure to a 60 Hz, 100 µT magnetic field for 15, 30, 60, or 120 min. The effects were dependent on time cell density (Phillips et al., 1992). Exposure of CCRF-CEM cells to a pulsed magnetic field (72 Hz, 3.5 mT) induced a modest decrease in p21-ras mRNA and protein levels (Phillips et al., 1993).
Possible effects of EMF on the expression of specific genes have been studied by other groups. Lagroye and Poncy (Lagroye & Poncy, 1998) reported that exposure to a 50 Hz, 100 µT magnetic field for 5 h up-regulated c-jun expression in primary and immortalized rat tracheal cells, while expression of c-fos increased in the immortalized cells and decreased in the primary culture. Down-regulation of the multi-drug resistance gene MDR1 mRNA was found after E-field stimulation (10-3000 mV/cm, 60 Hz, 16 h) (Walter et al., 1997). Strong static fields (0.18-0.2 T) were shown to induce c-fos transcription in HeLa cells (Hiraoka et al., 1992). In several studies on the effects of 60 Hz fields, however, no effects on transcripts have been seen, including those of c-myc, c-fos, and c-jun, in HL60 cells (Balcer-Kubiczek et al., 1996; Greene et al., 1993), HeLa cells, MCF7 breast cancer cells (Dees et al., 1996; Harrison et al., 1997), and several other human and rodent cell lines (Parker & Winters, 1992). [These studies were performed under more stringent conditions than earlier ones; they included better control conditions and stricter exposure metrics.] In a study on HeLa cells stably transfected with an HIV-LTR-CAT construct (Libertin et al., 1994), combined AC and DC fields (10 Hz to 1.6 kHz; 35-70 µT; plus a 0.17 mT DC field) did not induce reporter gene expression. A 50 Hz (200 or 400 mT) sine-wave field induced expression of a reporter gene containing the human VIP promoter in front of a b-galactosidase gene only when cells were simultaneously treated with forskolin, which activates adenylyl cyclase. These effects were blocked by the PKC inhibitor calphostin C and also by nifedipine and dentrolen, suggesting that the effect is mediated by calcium (Ohtsu et al., 1995).
The effect of EMF on the general transcriptional level has been studied in the yeast S. cerevisiae and in cultured human cells (Binninger & Ungvichian, 1997; Woloschak et al., 1998). After exposure of the yeast cells for 15 generations (24 h) to a 60 Hz sinusoidal field at 20 µT, Binninger and Ungvichian (Binninger & Ungvichian, 1997) found that most of the mRNA (44-67% in four separate experiments) was not affected by exposure, although 26-38% of the investigated mRNA species were increased twofold or more by exposure; 7-18% seemed to be down-regulated by at least 50%. The relatively large variations between experiments are possibly due to the unusually long exposure (several generations), thus reflecting changes in properties that biological materials undergo with time. Another genome-wide approach was adopted by Woloschak et al. (Woloschak et al., 1998), who used the so-called differential display reverse transcription-polymerase chain reaction protocol. This procedure allows comparisons of material from two experimental conditions and identification of virtually any change in mRNA pattern. The authors exposed human HeLa cells for 24 h to a 60 Hz, 0.10 mT magnetic field. After analysis of roughly 10% of all transcripts, they found two genes that were specifically induced by exposure. Their identity is unknown. The authors deduced that no more than 20 genes would be affected by their specific exposure regimen.
Although a fourfold increase in c-myc mRNA was initially found in HL60 cells exposed to various EMF, as measured by dot-blot analysis (Goodman et al., 1989), the magnitude of the increase has since been reported to be only 15% above control levels when analyzed by northern blots and including proper controls after a 20 min, 8 µT, 60 Hz exposure (Karabakhtsian et al., 1994). In transfected HeLa cells, both endogenous c-myc mRNA and the activity of the reporter gene were increased after exposure to a 60 Hz, 80 µT magnetic field for 20 min (Lin et al., 1994). An additional group (Liburdy et al., 1993a) has reported a three-fold increase in c-myc mRNA levels after exposure of rat thymocytes to 60 Hz (22 mT).
Recently, several groups have tried to reproduce or replicate the findings of Goodman and co-workers on c-myc induction by power-frequency magnetic field in HL60 cells. Exposure to 50 Hz sinusoidal fields (10 µT or 1 mT; 20 min to 72 h) was used by Desjobert et al. (Desjobert et al., 1995) and a 60 Hz sine wave (0.57 µT to 10 mT, 20-60-min exposure by Lacy-Hulbert et al. (Lacy-Hulbert et al., 1995), Saffer and Thurston (Saffer & Thurston, 1995), Owen et al. (Owen, 1998), and Miyakoshi et al. (Miyakoshi et al., 1996). The latter group used only a 5 mT 60 Hz field for 30, 60, or 180 min. None of these studies showed an effect on c-myc levels, b-actin level (Lacy-Hulbert et al., 1995), or on any other transcripts as evaluated by the mRNA differential display method (Saffer & Thurston, 1995). In a follow-up study, Goodman and co-workers (Jin et al., 1997), demonstrated that different strains of the HL60 cell line have different responses to the 60 Hz field. Only cells obtained from Columbia University showed increased c-myc after exposure, whereas cells from the ATCC were unresponsive. Neither Lacy-Hulbert et al. nor Saffer and Thurston used the Columbia University cells; however, Owen et al. (Owen, 1998) performed a similar comparison of these cell line variants. They found differences in several characteristics, as did Goodman and colleagues, but could not replicate the finding of changes in c-myc after exposure in any of the strains. Owen et al. also analyzed the cells in Goodman's laboratory and found no effect of magnetic fields, irrespective of the exposure location.
It has been suggested that cells might respond to EMF with changes in transcription and translation of heat-shock proteins, as they do to some other environmental stresses (Goodman et al., 1994a; Lin et al., 1997; Weisbrot et al., 1993). In support of this suggestion, exposure to a magnetic field (60 Hz sine wave, 8 µT peak B field, 11 µV /m induced field, 20 min) increased hsp70 mRNA in cultured human cells (Goodman et al., 1994b). Exposure to EMF (60 Hz sine wave; 0.8, 8, and 80 µT with an induced E electric field of 1.1 µV/m, 11 µV /m, and 110 µV/m, respectively) had a similar effect on SSA1 mRNA, corresponding to hsp70 in S. cerevisiae (Weisbrot et al., 1993). Additionally, HL60 cells exposed to 60 Hz, 8 µT (peak-to-peak) for 20 min showed activation of heat-shock transcription factor (hsf1) and subsequent DNA binding (Lin et al., 1997); however, exposure of cultured mouse and human cells to rotating 60 Hz fields (0.1 mT for 24, 48, or 72 h) did not induce hsp70 mRNA synthesis (Parker & Winters, 1992).
The effect of EMF on protein synthesis both in general and more specifically has been studied in several laboratories. Increases in protein synthesis were reported in prokaryotes and eukaryotes after exposure to EMF (Lacy-Hulbert et al., 1998). Only a few studies of the effects of EMF on the synthesis of specific proteins are available. Increased synthesis of some proteins such as IGF II was reported (Fitzsimmons et al., 1992).
Despite the large number of studies of molecular effects in cells exposed to EMF, no consistent picture emerges. The studies differ substantially in their biological approach and the techniques used; even within the same laboratory, some studies could not subsequently be reproduced.
The results of studies of primary cultures of peripheral blood cells from volunteers indicate that even quite strong fields, in the millitesla range, do not generally cause chromosomal aberrations; positive results were reported in only one study.
As no consistency was seen in the exposure parameters required for such effects, the observed findings have little predictive value.
Mutation is an area of molecular research in which consistency among results appears to be developing. In numerous studies, 48-h exposure to flux densities below approximately 0.1-1 mT have consistently shown no effect on mutagenesis in Salmonella typhymurium. As discussed elsewhere, there is little evidence that EMF below 0.1 mT damage DNA or induce cytogenetic damage; these effects are usually associated with mutation. Fields of 400 mT have, however, been reported to enhance mutagenicity at the HPRT gene locus in a human cell line after pre-exposure to ionizing radiation. In addition, a 2-h exposure to a 400 mT field enhanced mutagenesis in the absence of ionizing radiation in this same cell line and in a human osteosarcoma cell line. Expression of a regulated p53 transgene in the osteosarcoma cell line suppressed the effect of EMF on mutation. At lower field intensities (0.23-5 mT), exposure was reported by two laboratories to enhance mutagenicity in Chinese hamster ovary cells after exposure to ionizing radiation, and in one laboratory the mutation response was found to depend on the intensity of the field. Finally, in a single study, a static field of 5 T was shown to increase the frequency of somatic recombination in Drosophila in a protocol that implicates the involvement of oxygen radicals.
A few research groups have contributed most of the scientific literature on EMF-related effects on transcription and gene expression. Overall, the results are conflicting, especially with regard to the effects of power-frequency fields. Attempts to replicate the findings of other laboratories have met with little success, although reproducibility has been reported within a single laboratory.
The steady-state mRNA levels of several genes, including proto-oncogenes such as c-myc, c-fos, and c-jun, were reported by a single laboratory to be increased after exposure to EMF. The reports indicated that very low field strengths are effective and that specific time-windows are needed to induce effects. The effect has been seen for several years. Specific studies of replication regarding c-myc expression in other laboratories have not been able to confirm this finding. As the reported response was low-a 10-20% increase-replication will be difficult, if this effect is real. It is furthermore unclear whether such small effects on c-myc are biologically important for downstream events. The group that reported induction of proto-oncogene expression also tried to follow the specific sequence of events that leads to induction of hsp70 transcription and translation. Their findings indicate that the primary effect is a stress response to the field, but this is difficult to reconcile with an increase in RNA synthesis and mRNA transcription in general.
Membrane-mediated signal transduction by hormones and other signaling agents involves transmission across the plasma membrane that does not require penetration of the membrane. Signals associated with a conformational change in a membrane protein are propagated across the cell membrane by three well-understood mechanisms: opening and closing of ion channels and resultant current flow; changes in an intrinsic enzymatic activity of the receptor; and changes in the affinities of the receptor for intracellular proteins, which might have enzymatic activity.
In nearly all cases, the mechanism of signal transduction distal to the receptor involves effects on intracellular pathways due to changes in the ionic composition of the cytosol (e.g. changes in intracellular calcium) or changes in phosphorylation of intracellular proteins (e.g. enzyme activity, enzyme regulators, and factors transcriptional regulatory). Cellular responses to signals are either short-term, with little or no persistence of the effect after removal of the signal, or long-term, involving persistent changes in the function of cells, such as increased or decreased proliferation, changes in gene expression or differentiation, and, in some cases, apoptosis (programmed cell death). The short-term changes are generally mediated by modification of enzyme activity in the cytosol or membrane of the cell. The long-term changes invariably involve alteration of nuclear function, such as transcription, cell division, and cell-cycle regulation.
Walleczek and Liburdy (Walleczek & Liburdy, 1990) observed that 60 Hz magnetic fields increased 45Ca influx during concanavalin A-induced signal transduction in lymphocytes. Rat thymocytes were exposed to a 22 mT magnetic field (induced electric field, 1.0 mV/cm) for 60 min at 37 °C in the presence or absence of concanavalin A. In the absence of the mitogen, the cells were unresponsive to the magnetic field: the 45Ca influx was not altered. In its presence, the magnetic field increased the 45Ca influx by 50-200%.
Even though the ion cyclotron resonance mechanism (Liboff et al., 1990) is no longer accepted as a plausible biophysical explanation, a number of experiments were performed under exposure conditions corresponding to the calculated ion cyclotron resonance frequency for calcium, with variable results. For example, some investigators (Liburdy, 1992; Walleczek & Budinger, 1992; Yost & Liburdy, 1992) found alterations when these exposure conditions were combined with stimuli such as mitogens. Parkinson and Hanks (Parkinson & Hanks, 1989), however, saw no changes in cytosolic calcium concentrations with resonant and nonresonant EMF in Balb/c#3T3, L929, V79, or ROS cells.
Lindstrom et al. (Lindstrom et al., 1993) showed that application of a 50 Hz 100 µT magnetic field increased intracellular Ca++ signaling in the Jurkat T-cell line, and the effect was similar to that obtained with an anti-CD3 monoclonal antibody, used as a positive control. Additional work by this laboratory showed that exposure to this EMF resulted in a significant increase in inositol 1,4,5-trisphosphate concentration (Korzh-Sleptsova et al., 1995). As chelation of intracellular Ca++ did not block the increase, magnetic fields may affect signal transduction events upstream of inositol 1,4,5-triphosphate. Lyle et al. (Lyle et al., 1997) failed to replicate these studies on calcium signaling. [Differences in the strain of Jurkat cells may have affected the response.]
It is well established that intracellular calcium concentrations can oscillate in response to an external stimulus (Fewtrell, 1993; Meyer & Stryer, 1991), the period of the oscillations typically being between 1 s and several min. A model based on nonlinear dynamics and the theory of self-sustained (limit cycle) oscillators was developed which shows how a small change in the signal pathway at an early stage can lead to large changes in calcium metabolism in the cell (Eichwald & Kaiser, 1993); however, experimental verification of this model is lacking.
More recent experimental work has taken advantage of technical advances in the study of cell calcium metabolism, including the development of intracellular calcium probes, sensitive imaging procedures, and calcium-selective microelectrodes (Borle, 1990; McLeod, 1992). These studies also involve improved techniques for estimating exposure to EMF.
Walleczek et al. (Walleczek, 1995) developed a dual-chamber real-time fluorescence spectroscopy system that allows sham exposure to study the effects of low-level ELF fields. Exposure to a 60 Hz 2 mT magnetic field inducing an electric field of 1.8 mV/m had no effect on calcium influx in cells with a high initial calcium influx rate (n = 65), but in cells with a low pre-exposure flux rate, field exposure resulted in a significant increase (3.5%, n = 88; p < 0.001) in comparison with sham-exposed samples after a single 2-min exposure. The authors concluded that the cellular response to field exposure is strictly dependent on the cell state.
McLeod (McLeod, 1992) studied changes in calcium transient activity in a transformed bone cell line (ROS17/2.8) using a pure electric field exposure system. Aqueorin-loaded cells growing in a monolayer on glass or polystyrene substrates were exposed to 60 Hz electric fields (0.1-10 mV/m). Exposure resulted in a significant, dose-dependent decrease frequency of Ca2+ pulses or amplitude. activity that was different for the two substrates. The lowest threshold for a detectable field effect was estimated by the investigators by extrapolation to be approximately 0.4 mV/m for the cells growing on polystyrene. Sisken et al. (Sisken, 1998), using a similar approach for measuring calcium but a significantly different exposure system based on magnetic induction, were unable to replicate these effects.
Sontag (Sontag, 1998) used HL-60 cells differentiated into granulocytes to study the effect on cytosolic free calcium of sinusoidal electric fields at selected frequencies between 0 and 100 Hz with the calcium indicator fluo-3. Field strengths of 1-2000 Vpp/m (external field) or 0.1-1000 Vpp/m (in medium) had no significant effect. [As the highest of these intensities would depolarize the cells by more than 10 mV, the sensitivity of this assay to detect intracellular calcium changes is questionable.]
Although many early studies on bone cell cultures were performed with complex waveforms, it has since been demonstrated that sine-wave fields can have similar effects. The cellular mechanism of bone stimulation by sine-wave fields appears to be similar to that of pulsed fields. Exposure to a 0.1 mT, 60 Hz sinusoidal magnetic field (Luben, 1993; Luben, 1994) significantly inhibited cAMP accumulation in osteoblasts in response to 1 nM parathyroid hormone. This inhibitory activity was decreased at higher doses of parathyroid hormone, with no significant inhibition at 100 nM or 0.1 µM -7 mol/L. Frequency spectrum analysis of the Electro-Biology TM pulsed fields indicated that the most widely used clinical device has a significant component of magnetic field strength in the vicinity of 60 Hz (Polk, 1995). Subsequent experiments with the same 0.1 mT 60 Hz exposure field (Luben, 1994) also showed that exposure for 30 min to 24 h caused translocation of protein kinase C (PKC) to the membrane fraction of bone cells, followed by a progressive down-regulation of PKC activity. Down-regulation of PKC is often seen after treatment of cells with agents such as hormones and phorbol esters which activate it (Kikkawa et al., 1989). It has been known for some time that agents that regulate parathyroid hormone receptor responses in osteoblasts also induce PKC translocation and down-regulation (Abou-Samra et al. 1989; Fujimori et al., 1992).
PKC is believed to be the receptor for tumor-promoting phorbol esters (Kikkawa et al., 1989), and several investigators have examined the effects of magnetic fields on PKC. Monti et al. (Monti et al., 1991) showed that HL-60 lymphocytes exposed to a 50 Hz, 8 mT magnetic field have increased binding of the PKC-specific phorbol ester PDBu, suggesting that these relatively strong magnetic fields may modify the cellular response to tumor promoters. Similar results were obtained by Holian et al. (Holian et al., 1996), who showed that PKC activity in the cytosol of HL-60 human leukemia cells is regulated by exposure to 60 Hz electric fields at 100-1000 mV/cm. The effects of EMF treatment were additive with those of 2µM TPA.
Miller et al. (Miller & Moulder, 1998) used the human promonocytic leukemia cell line U937 to evaluate the hypothesis that exposure to a 60 Hz EMF amplifies the PKC-dependent signal transduction pathway that mediates the activation of NF-kB or AP-1-dependent reporter gene expression. In comparison with well-understood chemical and biological agents, EMF of 80, 100, or 1300 µT for 0.5-24 h had no effect on the NF-kB or AP-1 signaling pathway. This finding raises the possibility that the membrane receptor-linked protein tyrosine kinase signaling pathway operating in B-lineage leukemia cells may be important for the effects of EMF.
Uckun et al. (Uckun et al., 1995) also reported that PKC activity is increased in human pre-B leukemia cells exposed to a 60 Hz 100µT magnetic field. Moreover, activation of PKC was dependent on the activation of lyn kinase, a tyrosine protein kinase of the src family which is known to be involved in proliferation of leukemia cell clones. In recent studies, the same group used the DT-40 chicken lymphoma B-cell model to demonstrate that lyn kinase is essential for phospholipase c-2 activation and inositol 1,4,5-triphosphate turnover stimulated by exposure to 60 Hz, 100µT EMF (Dibirdik et al., 1998). In a third paper, this group reported that phospholipase c-2 activation in EMF-stimulated cells in the DT-40 model system is mediated by stimulation of Bruton's tyrosine kinase (Kristupaitis et al., 1998). These three studies provide evidence that a delicate balance of growth regulation in B-lineage lymphoid cells might be altered by EMF.
[The findings of Uckun's group are of interest as they show that EMF affect a well-understood, important signal transduction pathway involved in signaling through the B-cell receptor. Simultaneous, blinded, real and sham exposure was not carried out in the recent studies of Dibirdik et al. and Kristupaitis et al. since both reports indicate that the unexposed cells were placed in a 'control incubator' with a measured field of 0.8 µT. This is a critical procedure for quality control. Their observation of apparently robust effects occurring within 15 s of exposure to a 100µT 60 Hz field, coupled with the uncertain role of the biological and exposure conditions required for a robust effect, indicate that replication in other laboratories is essential.]
Miller and Furniss (Miller & Furniss, 1998) attempted to replicate the effect of a 100µT, 60 Hz EMF on inositol 1,4,5-trisphosphate shown by Dibirdik et al. and Kristupaitis et al., using the DT40 genetic model obtained directly from the Uckun group and with an experimental design shared between the two laboratories. A rigorous experiment, with blinded sham and field exposure and B-cell receptor-induced signaling as a positive control, showed no effect. [The failure to replicate the findings may be due to a number of reasons. Miller's group presented a detailed protocol for exposure in a well-characterized system and showed no causal effect, whereas Uckun's group showed an apparently robust effect and presented multiple types of evidence for an EMF-induced signaling pathway but without sham exposure.]
Cell proliferation is a complex process which is under the control of cellular signal transduction pathways. Altered proliferation of cells in vitro has been observed in a number of studies, but in none were sham controls used and none have been independently replicated. For example, Rosenthal and Obe (Rosenthal & Obe, 1989) showed a 10-15% increase in cell-cycle progression of human lymphocytes exposed to a 5 mT, 50 Hz field. West et al. (West et al., 1994) demonstrated increased colony growth in anchorage-independent JB6 cells after 10-14 days' exposure to a 1.1 mT 60 Hz magnetic field. No effect of the induced electric field was reported. Antonopoulos et al. (Antonopoulos et al., 1995) independently confirmed the effects of a 5 mT 50 Hz field reported by Rosenthal and Obe (Rosenthal & Obe, 1989). Using two exposure systems with temperature control, they showed significant acceleration of the cycle of human peripheral lymphocytes. Schimmelpfeng and Dertinger (Schimmelpfeng & Dertinger, 1993)) reported a reduction in cell number after exposure of SV40-3T3 cells to a 2 mT, 50 Hz field. In a subsequent study of proliferation, Schimmelpfeng and Dertinger (Schimmelpfeng & Dertinger, 1997) used organ culture dishes with an inner and outer compartment and flow cytometry to demonstrate that the reduction in cell number after a 1h exposure to a 2 mT, 50 Hz field was due to the induced electric field. The peak induced electric field was calculated to be 8-12 mV/m.
In the most recently reported study on proliferation, Katsir et al. (Katsir et al., 1998) demonstrated an increase in cell proliferation with exposure over the frequency range of 50-100 Hz and intensity range of 0.1-0.7 mT, as assayed by cell counts, 3H-thymidine incorporation, and MTT. Both frequency- and intensity-dependent responses were observed, with a maximum enhancement of proliferation of 70% seen with exposure to 100 Hz at 0.7 mT (p < 0.05). At 50 and 60 Hz, cell proliferation was enhanced by 13% and 26%, respectively. A blinded experimental design was used which included sham-sham studies, although active sham exposure (double-wound coils) was not included.
A potential correlation between cell proliferation and exposure to magnetic fields was described by Liburdy et al. (Liburdy et al., 1993b) in human estrogen-responsive breast cancer cells (MCF-7 cell line). These cells grow rapidly in the presence of normal concentrations of estrogens, but their growth rate decreased in the presence of melatonin, a hormone produced by the pineal gland. It has been proposed that disruptions of the normal daily cycle of melatonin synthesis are risk factors for human breast cancer (Stevens, 1987). Melatonin synthesis in whole animals has been shown to be altered by exposure to ELF EMF (Wilson et al., 1990). Liburdy et al. confirmed previous studies by Blask (Blask, 1993) that melatonin at a normal physiological concentration (10-9 mol/L) can decrease the growth rate of a specific strain of MCF-7 cells; however, application of a 1.2 µT sinusoidal magnetic field at 60 Hz prevented this action of melatonin. A field of 0.2 µT had no significant effect, suggesting that a threshold might exist between 0.2 and 1.2 µT. In more recent work, Liburdy's group extended these findings to demonstrate in experiments similar to those carried out using melatonin that the antiproliferative action of tamoxifen (an anti-cancer drug) is blocked by a 1.2 µT field (Harland & Liburdy, 1997). They also showed that the effect was due to the magnetic field and not due to the induced electric field.
Two other laboratories replicated the effects of 1.2 µT fields on inhibition of MCF-7 cell growth by both melatonin and tamoxifen (Blackman et al., 1998; Liburdy & Levine, 1998). Liburdy's group reported similar effects of 1.2 µT on inhibition by tamoxifen of a second human breast cancer cell line, T47B (Harland et al., 1998) and a human glioma cell line, SF-757 (Afzal & Liburdy, 1998). [While these results have been replicated, the extremely small effects observed (10-20% growth over 7 d) and the nature of the experimental design raise serious concerns about the robustness of this effect.]
Ornithine decarboxylase (ODC) activity is modulated by membrane-mediated signaling events, and its activation during carcinogenesis is associated with the activity of mitogens and tumor-promoting agents of various types. Byus et al. (Byus et al., 1987) reported that ODC activity in three cell lines-human lymphoma cells (CEM), mouse myeloma cells (P3), and rat hepatoma (Reuber H35) cells-was increased by up to 500% when they were exposed to a sinusoidal 60 Hz electric field at 10 mV/cm. Increased ODC activity in Reuber H35 cells was detected at fields as low as 0.1 mV/cm. In comparison, phorbol ester at doses associated with tumor promotion activated ODC by more than 1000%. The investigators concluded that electric fields act on the cell membrane, resulting in an effect of signal transduction on ODC activation by mechanisms that were not directly investigated in these or subsequent studies. These findings have been used as a basis for the hypothesis that low electric fields act as a co-promoter with tumor-promoting agents, resulting in more activation of ODC and more growth promotion of carcinogen-induced cells than in the absence of electric fields. Litovitz et al. (Litovitz et al., 1991) also reported enhancement of ODC activity in mouse L929 cells by exposure to a 60 Hz magnetic field for 8 h at a strength of 1, 10, or 100 µT. Maximal enhancement of approximately 100% above the control level was produced by 4 h of exposure to a magnetic field at 10 µT.
Effects of EMF on ODC activity have also been reported by other laboratories, although the conditions and signaling agents varied. Mevissen et al. (Mevissen et al., 1995) showed that exposure of rats in vivo to a 50Hz, 50 µT field for six weeks doubled the ODC activity in mammary tissue. Valtersson et al. (Valtersson et al., 1997) found that both ODC activity and polyamine levels were increased in the Jurkat human leukemia cell line, but not in the non-leukemic CEM lymphocyte line, after exposure to a 50 Hz, 0.1 mT magnetic field. Little effort was made in these studies to isolate the specific change in membrane receptor mechanisms that resulted in the observed change in ODC activity.
Litovitz et al. (Litovitz et al., 1994) and Farrell et al. (Farrell et al., 1998) extended their earlier observation of increased ODC activity in mouse L929 cells at 60 Hz, 10 µT, by showing that an incoherent noise field could block the effect of the coherent 60 Hz field at equal flux density. In contrast, Azadniv et al. (Azadniv et al., 1995) showed that a 4h exposure of L929 cells to a 60 Hz 10 µT magnetic field had no statistically significant effect on ODC activity. Moreover, Cress et al. (Cress et al., 1995) failed to show an effect when using the cells, methods, and exposure system of Litovitz' group. [The function generator and power amplifier used in this study were not, however, those used by Litovitz' group.]
Scarfi et al. (Scarfi et al., 1991) showed that micronucleus formation in human lymphocytes was not affected by exposure to a pulsed magnetic field of 2.5 mT (peak pulse) at 50 Hz. This field also had no effect on mitomycin C-induced micronucleus formation. This finding was consistent with that of a previous study from the same laboratory (Cossarizza et al., 1989), in which the same pulsed magnetic field produced no change in cell survival or unscheduled DNA synthesis in human lymphocytes with or without treatment with ionizing radiation.
Tofani (Tofani et al., 1995) reported a synergistic effect on micronucleus formation in human peripheral lymphocytes after concurrent exposure to mitomycin C and EMF at a Ca++ resonance condition consisting of a 32 Hz 75 µT AC field with a 42 µT static field. In these studies, neither a 75 µT nor a 150 µT AC field affected micronucleus formation in the absence of mitomycin C or an applied static magnetic field.
Lagroye and Poncy (Lagroye & Poncy, 1997) observed no change in micronucleus formation in a spontaneously transformed tracheal epithelial cell line exposed to EMF alone (50 Hz, 100 µT, sinusoidal), but when the cells were exposed to EMF plus 6 Gy of ionizing (gamma) radiation, the number of binucleated cells with micronuclei increased by approximately 10% (p < 0.05).
Simko et al. (Simko et al., 1998) examined micronucleus formation and other apoptotic morphological changes in two human cell lines exposed to 0.1-1.0 mT 50 Hz AC magnetic fields. In the SCLII transformed squamous-cell line, a dose-dependent increase in micronucleus and apoptosis was seen after 48 and 72 h continuous exposure to 50 Hz (0.8 and 1.0 mT). In contrast, in a non-transformed amniotic fluid cell line, no significant changes were noted, suggesting that different cell lines react differently to the same field. In a brief communication, Ismael et al. (Ismael et al., 1998) reported use of a specific assay system for apoptotic cells (TUNEL assay) to observe in increase in apoptosis of mouse thymocytes treated with dexamethasone. The increase was observed only in thymocytes and not in splenic T-cells from animals exposed to a 0.4-1.0 µT 60 Hz magnetic field. The levels of apoptosis in thymocytes and spleen cells from mice exposed to an 8-20 µT DC magnetic field were similar to those in controls.
[None of these recent studies of apoptosis has been replicated. The wide variety of effects, cell lines, and exposure conditions and the relatively qualitative techniques used (e.g. micronucleus counting) cast doubt on any robust effect. Use of more accurate techniques such as the TUNEL and other quantitative assays may result in more definitive answers in the future.]
Cells undergoing embryogenesis, somatic differentiation, and some pathways in carcinogenesis show changes in cytological markers that indicate the expression of lineage-specific genes. These markers include changes in the kinetics of embryonic staging, changes in matrix protein synthesis, changes in cell surface characteristics, changes in cell morphology, and gap-junctional communication.
Changes in the kinetics of staging in embryogenesis are a well-accepted measure of coordinated development and can be readily measured in a number species in vitro. A drawback to this assay is that there is little evidence that changes in the onset of these stages result in aberrant development.
Changes in matrix protein synthesis, including both extracellular matrix molecules and cell adhesion molecules, are also well-accepted measures of changes in a cell population's differentiated state. Not only does the matrix modulate cell differentiation, but many cell types, particularly those of mesenchymal lineage, can significantly modify the characteristics of the matrix by synthesizing extracellular matrix molecules or matrix metalloproteases. In addition, by changing integrin expression, cells incapable of altering the substrate characteristics can modulate the degree and nature of their attachment to that substrate (Shumaker et al., 1994) Differential adhesion can affect cell associative preferences in heterogeneous aggregates and is therefore considered to play an important role in malignant invasion (Steinberg & Foty, 1997).
Closely related to the process of cell adhesion and alterations in the extracellular matrix are alterations in cell surface characteristics. The glycocalyx serves as the interface between the cell and its environment, and changes in the charge density, enzyme content, or activity of this layer and the distribution of the various glycosaminoglycans are accepted markers of differentiation.
There is accumulating evidence that cell shape is a major determinant of differentiation (Folkman & Moscona, 1978) and is a distinct marker of differentiation in many cell types. Alterations in cell morphology, size, and orientation can reflect the cellular response to the extracellular matrix environment but also influence the way in which the cell affects its environment; examples are the orientation of forces in wound contraction and the polarity of a tissue. Mobility is also closely related to orientation and is a commonly used end-point of differentiation.
Intercellular communication has been postulated as a necessary condition for cells to progress through normal differentiation. Gap-junctional communication is considered to play a key role in tissue homeostasis, and its disturbance has been associated with various health problems.
Many studies have been undertaken to assess the developmental effects associated with embryonic exposure to PEMF. The most rigorously designed of these was an international collaboration involving six laboratories in four countries (Berman et al., 1990). Each laboratory used two identical egg incubators equipped to produce a 500-ms pulse of 1 µT with a 2-ms rise-and-fall time. The pulse was repeated at a frequency of 100 Hz. [The induced electric field within the egg can be estimated to have been about 10 µV/m.] Eggs were sham exposed by incubation without energizing the magnetic field coil. Incubated eggs were exposed for 48 h at 37.6-38 °C prior to evaluation for fertility, stage of development, and presence of abnormalities. The evaluations were performed blindly. The principal observation from the combined results of these studies was a 6% increase in the number of abnormal embryos (control, 19%; exposed, 25%; p < 0.001); however, significant inter-laboratory differences were observed. No significant effect on embryonic staging was observed. [This study suffers from a small dynamic range of response and inadequate control, given the small flux densities used.]
Zimmerman et al. (Zimmerman et al., 1990) investigated the effect of a 60 Hz magnetic flux on the development of sea-urchins. The exposure system consisted of two identical chambers equipped with two pairs of Helmholtz coils 1.5 m in diameter in an orthogonal orientation. The two coil pairs were driven 90o out of phase to obtain rms flux densities up to 500 µT. The geomagnetic field was measured at 43.8 µT; for sham exposure, the two pairs of coils were not energized in one system. Fertilized eggs were maintained in 20 ml of seawater at 18 °C in 250-ml beakers [suggesting that the maximum induced rms electric field intensities were approximately 5 mV/m]. The developmental progress of the embryos was assessed by counting the number of cells in the embryo at 10, 16, and 22 h after fertilization. A significant delay (p < 0.001) in development was observed, which showed a dose-dependent correlation with flux density. The investigators concluded that the field had caused a temporary delay at the morula stage of development.
Using a similar exposure system, Cameron et al. (Cameron, 1993) investigated the effect of a 60 Hz field on mouse embryogenesis. Two-cell-stage mouse embryos were cultured for 36 h after fertilization and then exposed to either a 10 or a 50 µT flux for 48-68 h at 37 °C. [As the size of the exposure chambers was not reported, the induced electric field intensity cannot be estimated.] The developmental stage was assayed by counting the number of cells in 9-10 embryos under a phase-contrast microscope. Exposure at 10 µT or 50 µT significantly delayed development (p < 0.025 and 0.01, respectively by binomial test).
Levin and Ernst (Levin & Ernst, 1995) also investigated the effect of exposure to ELF magnetic fields on early development of sea-urchins in a study which spanned the frequency range of DC to 600 kHz. The exposure apparatus consisted of a 10 cm vertically orientated solenoid, 7 cm in diameter, sufficient to permit insertion of a 250 ml beaker, and exposure was to 60 Hz over the rms flux density range of 1.7-8.8 mT [corresponding to a maximum induced rms electric field intensity of 50 mV/m]. The control and exposed samples were held in the same incubator at 12 °C, and the embryos were stirred continuously with thermally isolated stirring motors. By analyzing 200 control and exposed embryos every 15 min, a significant advance in the time of both the first and second cell divisions was observed. Moreover, the degree of advance appeared to be linear with flux density and duration of exposure but to have a complex relationship with the frequency of exposure. A monotonic decrease in efficacy was observed over frequencies ranging from DC to 600 kHz, with a maximum response at DC, suggesting a thermal effect.
In an investigation of chick embryo development, sinusoidal field exposure was used (Veicsteinas et al., 1996). Groups of 210 exposed and sham-exposed (no coil excitation) eggs were incubated for up to 18 d with 2 h on, 22 h off intermittent exposure to a 200 µT, 50 Hz field [consistent with an induced electric field intensity of about 1 mV/m]. The embryos were examined for developmental stage after 48 h. Immunocytochemical analysis of extracellular components was performed on day 7, and histological examinations on days 7, 12, and 18 of incubation. No difference in developmental progression, extracellular components, or malformations in tissues was observed. In addition, follow-up 90 d after hatching showed no difference between the exposed and unexposed chick populations.
In a recent replication of the chick embryo studies, both pulsed and sinusoidal fields were used (Farrell et al., 1997). A total of 2500 chick embryos were examined in five experimental series spanning five years. The pulsed field exposure comprised a 500-ms pulse, 100 pulses per second, reaching a maximum flux of 1 µT in 2 ms. The sinusoidal exposure was at 60 Hz, with a peak flux density of 4 µT. [This gives an estimated peak induced electric field intensity of 10 mV/m. No sham exposure appears to have been used.] In four of the five series, a significant increase (p < 0.01) in the number of abnormally formed embryos was seen in a blinded fashion after 48 h of continuous exposure. The investigators note, however, that there was also a nearly 10-fold variation in the rate of apparent abnormality in the control samples in the five experimental series, suggesting that there would be no effect by analysis of variance.
In association with a longstanding interest in the promotion of wound and fracture healing, numerous investigators have addressed the ability of low-frequency magnetic fields to alter extracellular matrix protein synthesis. In two related studies, Murray and Farndale (Farndale & Murray, 1985; Murray & Farndale, 1985) investigated the effect of pulsed-field exposure on collagen production in primary and cultured fibroblasts. Primary chick tendon fibroblasts were exposed in 30-mm culture dishes to a Helmholtz coil pair arrangement producing a 2.2 mT flux at 4 kHz lasting 4.8 ms, with the pulse burst repeated at a 15 Hz rate. [The peak induced fields in the dishes were calculated to be 2.3 V/m. Sham exposure was not incorporated in the experimental design.] Total protein, collagen synthesis, and collagenolytic activity were assayed by radioisotope incorporation after 6 d of exposure (6 h on, 6 h off). Exposure was found to increase collagen production relative to total protein and to induce a significant, two-fold reduction in collagen turnover. The effect of exposure was observed only after the cultures had reached confluency. Similar results were obtained when cultured fibroblasts from rabbit bone-marrow stroma were used. Thermal effects were specifically addressed and dismissed as a confounding factor.
Fitzsimmons et al. (Fitzsimmons et al., 1986) reported the results of studies with a capacitively coupled exposure system to stimulate bone matrix formation in tissue culture. Samples maintained in tissue culture were placed within a pair of capacitor plates separated by 2 cm. Tritiated hydroxyproline incorporation in mouse calvaria and tibia was assayed over the last 24 h of a 3 d exposure. Exposure to a 16 Hz electric field at a calculated electric field intensity of 10 µV/m resulted in a twofold increase in incorporation in the calvaria but no significant change in the tibial preparation.
McLeod et al. (McLeod et al., 1987c) also investigated the effect of exposure to electric fields on matrix protein synthesis. Primary bovine fibroblasts were incorporated into three-dimensional collagen gels which, when contracted, permitted transfer to an apparatus in which a controlled current density could be applied through agar bridges. Field intensities of 1 mV/m to 3 V/m were used over a frequency range of 0.1 Hz to 1 Khz. Sham exposure consisted of exposing the cell system in an identical system with no current excitation. Tritiated proline incorporation into extracellular protein was assayed in cultures sustained in serum-free medium and exposed for 12 h. Peak sensitivity of the cell system was found near 10 Hz, with significant inhibition of proline incorporation at 6.5 mV/m rms (p < 0.02). Enhanced sensitivity (two-fold increase) of the cell system was seen when the cells were orientated in the direction of the applied electric field.
MacGinitie et al. (MacGinitie et al., 1994) investigated field-stimulated matrix synthesis in a cartilage explant model. In this system, 9.5-mm disks of cartilage from the femuropatellar groove of one- to two-week-old calves were maintained in culture for 5-8 d and then exposed to 10-30 mA/cm2 at frequencies of 1 Hz to 10 kHz. The explants were maintained in 0.1% defined serum-supplemented medium during exposure. After 12 h of exposure, radiolabelled methionine (a measure of glycosaminoglycan synthesis) was assayed. Peak sensitivity was observed at 100 Hz, with enhanced incorporation at 24 and 30 mA/cm2 (p < 0.05). No significant increase was observed at 10 mA/cm2, a current density associated with an induced electric field intensity on the order of 100 V/m.
The sensitivity of cartilage explants to fields was also evaluated by Liu et al. (Liu et al., 1996), who exposed 16-d-old chick embryo sternal cartilage explants to 30-ms pulse bursts (pulse rate, 4 kHz) repeated at 1.5 Hz for 3 h/d for 2 d. Exposure was given from a Helmholtz coil pair (14 cm in diameter), permitting a peak flux density of 100 µT with a rise time of 230 µs and fall time of 30 µs. For sham exposure, the coils in a similar system were not energized. Glycosaminoglycan production was assayed by 35S incorporation. The exposed explants showed a 21% increase in glycosaminoglycan content (p < 0.05) and, correspondingly, a 30% decrease in glycosaminoglycan released into the medium (p < 0.02). Exposure was also found to significantly decrease the amount of newly synthesized glycosaminoglycan (76%, p < 0.001), leading the investigators to conclude that the exposure also significantly suppressed pre-existing glycosaminoglycan degradation in the explants.
Rodemann et al. (Rodemann et al., 1989) reported dramatic increases in protean synthesis in fibroblasts in response to exposure to a 6 mT, 20 Hz magnetic field. Normal human fibroblast cell lines (HH-8 and WI38) and virally transformed fibroblasts (WI385V40) were exposed twice a day for 6 h in a water-cooled solenoid exposure system. Control cultures were maintained in a separate incubator. S35-Methionine incorporation, 3H-proline incorporation, total protein, and the mitotic index were assayed. After 21 d of exposure, a 5- to 13-fold increase in total protein synthesis was observed, accompanied by a significant (p < 0.05) increase in collagen synthesis and a shift in the mitotic index. The authors concluded that field exposure induced differentiation in these cell lines. [The lack of a sham-exposed group in a system that required water cooling to maintain the incubation temperature raises concerns about the reproducibility of these studies.]
Interest in the effects of pulsed magnetic fields has led to numerous studies of phenotypic changes in exposed cell populations. Most cell types have a distinct coat (glycocalyx) which presents a highly negatively charged surface to the environment. Smith et al. (Smith et al., 1991a) investigated the effect of exposure to PEMF on this cell coat by exposing a monocyte-like, non-adherent, mammalian cell line (U937) to a 25 Hz pulse-burst field from a Helmholtz coil pair (7.5 cm radius). A peak flux density of 0.63 mT were used, with a rise time of 200 µs and a fall time of 20 µs. [The induced electric field intensity at the periphery of the culture vessel was estimated at 160 mV/m.] For sham exposure, the Helmholtz coils were not energized. Surface charge density was assayed by partition chromatography. Cells exposed for 48 h had a significantly higher partition coefficient than control cells (p < 0.03), consistent with an increase in the negative surface charge density on the cells. In the context of previous experiments undertaken by this group, the investigators concluded that the observed effect was due to the induced electric field.
Differentiation of bone cells is commonly characterized by an abrupt rise in the activity of alkaline phosphatase, an ectoenzyme believed to be associated with the mineralization process. McLeod et al. (McLeod & Guilak, 1993) investigated the effect of field exposure, from three identical solenoid systems installed in a single incubator, on alkaline phosphatase activity in a rat osteosarcoma cell line (ROS-17/2.8). Bifilar winding of the solenoids permitted excitation of the sham-exposed samples, and a third unexcited coil system was used as the control. The tissue culture dishes were exposed for 72 h to a rms flux density of 1.8 mT at 30 Hz [sufficient to induce a maximum electric field of 600 µV/m]. Cells were maintained in a normal growth medium with 10% serum, and enzyme activity was assayed from the conversion of para-nitrophenylphosphate to para-nitrophenol, which was quantified spectrophotometrically. While no effect on enzyme activity was found in exposed cell populations plated sparsely, a doubling of activity (p < 0.001) was observed for cells plated densely (3000, or use 3x104 cells/cm2) as compared with sham-exposed samples. As these results were associated with a significant depression in cell numbers in the exposed population, the investigators concluded that the normal differentiation response of these cells accelerates as they attain confluency.
Neural cell adhesion molecules are associated with neuronal differentiation. Horton et al. (Horton, 1993) investigated the ability of a sinusoidal flux to affect the onset of expression of these molecules in rat pheochromocytoma cells (PC12, ATCC), a well-investigated neuronal differentiation model. Field exposure was accomplished from a Helmholtz coil pair (30 cm diameter) to produce a vertical field with a peak-peak flux density of 40 µT at 16 Hz. A DC magnetic flux was imposed at 20 µT, colinear with the applied AC flux. [The estimated induced electric field intensity would be 50 µV/m. Sham exposure conditions were not reported.] Cells were exposed to the field for up to 72 h while maintained in a low serum (1%) medium in T-75 flasks treated with polylysine or polyornithine. Neural cell adhesion molecule expression was assayed by monoclonal antibody labeling and immunoprecipitation. A transient effect of field exposure was observed, as neural cell adhesion molecule expression was found to be significantly elevated at 24 h but not at 72 h. Moreover, the effect was observed only in cultures exposed for 30 min/d and not in those exposed continuously.
Kula and Drozdz (Kula, 1996) investigated alterations in the character of the cell coat of fibroblasts caused by field exposure. Balb/c mouse fibroblasts maintained in 5% heat-inactivated serum in 43-cm3 flasks were exposed to a 50 Hz, 20 mT flux from a solenoid coil; for sham exposure, a similar apparatus was used which was not energized. [The estimated induced electric field intensity would be 100 mV/m.] The length of exposure ranged from 2 to 64 min/d from day 5 to day 8 of culture. Glycosaminoglycan distribution was assayed in the cells, the medium, and cell coats by 35S incorporation. Rapid changes in the distribution of glycosaminoglycans in the cell coat were observed. 35S incorporation into heparin sulfate was found to decrease by two-thirds after exposure for only 16 min/d, although longer exposure had little additional effect. Conversely, 35S incorporation into chondroitin sulfate nearly doubled with similar lengths of exposure. Control exposure with static magnetic fields of up to 0.5 T showed no effect.
McLeod et al. (McLeod, 1998) used the upregulation of alkaline phosphatase activity in differentiating bone cells to determine the dose-response characteristics of cells to an ELF (30 Hz) sinusoidal field. A non-transformed bone-cell line which undergoes terminal differentiation (MC-3T3-E1) was plated into confluent culture and maintained for 4 d in a normal growth medium with 10% serum. These cultures were then exposed to fields from a solenoidal apparatus, which provides an active sham exposure for comparison. Exposure at 2.5 mT resulted in a non-uniform electric field distribution at the cell monolayer surface, varying from approximately 100 to 6000 µV/m peak. Alkaline phosphatase activity was then assayed in situ at periods of 4-64 h with an optical scanning technique which allows spatial mapping of the overall cell response. Alkaline phosphatase distribution was changed in comparison with that in the sham-exposed samples within 16 h, reflecting suppression of activity correlated to the intensity of the induced electric field. By 64 h of exposure, a uniform 25% suppression of alkaline phosphatase activity was observed across the culture dish. These investigators concluded that the cellular response depends on the magnitude of the induced electric field with a threshold intensity in the range of 100 µV/m.
In a series of investigations, Blackman et al. (Blackman et al., 1993a; Blackman et al., 1993b; Blackman et al., 1994; Blackman et al., 1998) studied the effect of ELF EMF on neurite outgrowth in a well-developed model of neuronal differentiation. PC-12 cells were primed for growth by plating them on collagen-coated dishes and were maintained on growth medium with 15% serum for one week; they were then replated, and nerve growth factor (5 ng/ml) was added to initiate differentiation. The cells were exposed in a Helmholtz coil pair to produce either a uniform flux of 0-30 µT or a non-uniform flux of 0-50 µT, in order to obtain data for determining a dose-response relationship in a single experiment. Sham exposure was not undertaken, but concurrent control samples were placed in an adjacent magnetically shielded enclosure. The percentage of cells undergoing differentiation was assayed by counting all cells in randomly selected areas in which neurite growth was greater than the cell body length. Blackman et al. (Blackman et al., 1993b) observed suppression of neurite outgrowth (normally about 60% of the cells), with maximal inhibition of approximately 20% after 22 h of exposure to a 50 Hz magnetic field. A threshold-like response in neurite outgrowth was observed, with a transition in the range of 5-10 µT; higher flux densities induced no additional inhibition. In additional experiments, uniform suppression was observed in cells in dishes of different sizes resulting in induced electric field intensities of 3.7-45 µV/m, for an applied flux density of 7.9 µT, suggesting that the observed effects were strictly dependent on the magnetic field.
In a subsequent experiment with identical exposure conditions but using the PC-12D cell line, Blackman et al. (Blackman et al., 1993a) showed that exposure to a 50 Hz magnetic field enhanced neurite outgrowth by more than two-fold (maximum, 55%) in the absence of nerve growth factor. This effect was also found to be dependent on the magnetic flux density and not on the induced electric field intensity, 1-10 µV/m. Blackman (Blackman, 1994) then investigated the response of the PC-12 cell line to flux densities as high as 50 µT with a nerve growth factor concentration of 50 ng/ml to achieve 100% neurite outgrowth in the absence of fields. Exposure of the cells at 45 Hz had no effect up to flux densities of 5 µT, but outgrowth was inhibited at flux densities of 5-30 µT, with a maximum inhibition of up to 60%. With flux densities greater than 30 µT, the efficacy of the field to inhibit neurite outgrowth appeared to be diminished. Blackman et al. (Blackman et al., 1998) recently replicated these effects in a blinded fashion [although sham exposure was not incorporated in the experimental design].
A trial to replicate the findings of Blackman et al. was undertaken at the Oak Ridge National Laboratory (Griffin et al., 1998). Populations of PC-12 cells previously primed by treatment with nerve growth factor were treated with graded doses of this factor alone (2.5 mg/ml) or combined with EMF (2.38 mT at 45 Hz plus 3.66 mT DC). Appropriate sham controls were included. No significant difference in neurite outgrowth was observed. Different nerve growth factor-primed populations showed differential response to challenge with the factor.
Greenebaum et al. (Greenebaum et al., 1996) reported the effects of exposure to pulsed fields on neurite outgrowth. Chick dorsal root ganglia were isolated from 6 d-l d chick embryos and placed in 60-mm Primeria culture dishes. Neurite outgrowth was initiated by treatment with 50-100 ng/ml nerve growth factor, and the cells were exposed to fields from a Helmholtz coil pair, which produced a 4.0 mT flux pulse with a 200µs rise time and a 20µs fall time. [The peak induced electric field intensity at the edge of the culture dish was estimated to be 3 V/m.] A series of 22 pulses in 4.8 ms were repeated at a rate of 15 or 25 Hz. Exposure lasted 18 h, and the cells were then left undisturbed for an additional 6 h before fixation and assessment. Controls samples were held in a similar incubator with unenergized coils. Neurite outgrowth was determined by the length of the growth process from the dorsal root ganglion body. In addition, the positions of the ganglia within the dish and the angular orientation of the growth processes were recorded. In their study of 808 ganglia, the authors reported a high level of inter-experiment variation but were able to identify a significant increase in neurite growth in the field-exposed population that was independent of nerve growth factor concentration. Moreover, field exposure led to greater directional growth. No dependence on induced electric field intensity or frequency was identified.
Spadinger et al. (Spadinger et al., 1995) investigated motility and morphological changes in cells during field exposure, using an automated tracking system to measure and record the field-induced changes. NIH-3T3 fibroblasts were maintained in normal growth medium with 10% serum and exposed in 6-cm petri dishes. They were exposed to vertical magnetic fields in the frequency range of 10-1000 Hz, at rms flux densities of 10-800 µT, from a short solenoid coil encased in plastic. [No sham exposure was performed.] No DC field was applied, but the geomagnetic field was 29 µT. [The induced electric field intensity at 60 Hz for a 100 µT flux was calculated to be 0.4 mV/m.] Motility was assayed as cell speed, direction, and persistence, and the morphological measurements included cell area and circularity. These assays were performed during 3-4-h field exposures interspersed with periods of no exposure. The investigators found no evidence of significant changes in motility or morphology that could be attributed to the applied magnetic field.
Osteoclast formation is a multistep process involving extensive cell-matrix and cell-cell interactions, in that cell migration, aggregation, and fusion are required. Rubin et al. (Rubin et al., 1996) investigated the effect of exposure to ELF fields on osteoclast formation in a mixed murine marrow culture system. Marrow cells were obtained from the femurs and tibias of 4-8-week-old male C57Bl/6 mice and plated on chamber slides with medium and vitamin D to promote osteoclast recruitment. Exposure was begun immediately with the solenoid exposure system described by McLeod et al. (McLeod & Guilak, 1993), which provides matched field and sham exposure. The cultures were exposed for 8 d to 30- and 60 Hz fluxes of 1.8 mT [corresponding to an induced rms electric field intensity of 600 µV/m at 30 Hz]. Osteoclast recruitment was assayed by counting all multinucleated, tartrate-resistant, acid phosphatase-positive cells in each 0.9-cm2 culture well. The counts were performed in a blinded fashion. A 22% reduction (p < 0.0001) in osteoclast number was found with exposure to 60 Hz and a 28% reduction with 30 Hz. The investigators concluded that the fields may affect the pool of proliferating precursor cells.
Lee and McLeod (Lee & McLeod, 1998) reported the results of a study of the morphological adaptation of osteoblast-like cells (MC-3T3-E1) to a 60 Hz field. Cells maintained in culture medium with 10% serum were exposed for 24 h to fields from a solenoid exposure system (McLeod & Guilak, 1993), with a 0.7 mT rms flux density, sufficient to induce a 0.5-mV/m electric field intensity in the culture wells. Cell length, width, area, perimeter, circularity, and angular orientation were calculated from digitized phase-contrast images of the cells with commercial image-analysis software. As reported by Spadinger et al. (Spadinger et al., 1995), no significant effect of exposure could be identified when the cell measurements were pooled. When the morphological measurements were analyzed with respect to the orientation of the cells (angular orientation with respect to the induced electric field vector), however, the cells orientated parallel to the induced electric field were found to be significantly longer, with larger areas and perimeters, and more spindle shaped (p < 0.001). The investigators concluded that these effects were due to the induced electric field rather than the magnetic field.
Alterations in gap-junctional coupling provide two distinct views of the interactions of EMF with cells and tissues. First, gap junctionally coupled cells theoretically interact differently from isolated cells with induced electric currents. Second, gap-junctional competence can serve as an assay of cell-cell communication, a factor considered to be essential in the normal development and function of tissue.
Ubeda et al. (Ubeda et al., 1995) investigated changes in gap-junctional competence in dye-coupling experiments. C3H 10T1/2 cells were plated sparsely onto tissue culture dishes [size not reported] and allowed to proliferate for 18 d until reaching confluence. Melatonin was then added to the medium at a concentration of 0.1µM, and the cells were allowed to incubate for another 27 h. At the end of this incubation period, they were exposed for 30 min to a magnetic field of 50 Hz at 1.6 mT from a Helmholtz coil pair. Controls were sham exposed by not energizing the coils. A subset of the cells in each dish was then loaded with Lucifer yellow dye by a single scrape across the culture dish, and transfer of the dye to adjacent cells was determined by counting fluorescently labeled cells. Exposure to the field blocked the up-regulation of coupling caused by the melatonin treatment.
Conversely, Schimmelpfeng et al. (Schimmelpfeng et al., 1995) showed in dye microinjection studies with NIH 3T3 cells that a 50 Hz field could up-regulate intercellular coupling. NIH 3T3 fibroblasts which had obtained a monolayer density of 1-3 x 105 cells/cm2 and aggregates of cells 200-400 µm in diameter were exposed in 60-mm culture dishes to a sinusoidal 50 Hz field at 2 mT by placing the dishes within the air gap of an iron core electromagnet. [The maximum induced electric field intensity can be calculated to be approximately 10 mV/m.] Control cultures were placed in the same incubator at a sufficient distance from the electromagnet that the flux densities were less than 2-3 µT. After exposure, three to nine cells were assayed for gap-junctional coupling by microinjection with Lucifer yellow. The dye that had spread into neighboring cells was evaluated by counting fluorescent cells 2 min after the injection. Exposure to the field for 5 min was reported to cause a 1.6-fold increase in the average number of coupled cells in the monolayer configuration. Coupling of cells in aggregates was not readily quantified. [Four independent experiments with three to nine cells per experiment provides a questionable statistical basis for comparison.]
Vander Molen (Vander Molen, 1997) reported a series of experiments designed to test the altered sensitivity of a cell population to EMF in the absence of gap-junctional coupling. An osteosarcoma cell line (ROS 17/2.8) was transfected with anti-sense connexin-43 to yield a cell line that was intercellular communication-deficient. Communication-competent and -incompetent cells at similar confluent densities were then exposed to a 1.8 mT rms flux at 30 Hz [sufficient to induce a 600µV/m electric field intensity in the culture wells] for up to 72 h from the exposure system of McLeod et al. (McLeod & Guilak, 1993). Cell proliferation and alkaline phosphatase activity were assayed. The communication-competent and -incompetent cell lines had similar sensitivity to EMF exposure. These results were confirmed by exposing sparsely plated cell populations (precluding significant numbers of gap-junctional couplings). The investigators conclude that gap-junctional coupling does not enhance cell sensitivity to EMF.
Griffin et al. (Griffin et al., 1998) examined the effects of EMF on the inhibition of gap-junctional communication by chloral hydrate. Clone 9 rat liver cells were treated with 1.5 mmol/L of chloral hydrate for 24 h; they were then left for 30 min for CO2 and temperature to reach equilibrium and then treated with EMF for 30 min. After exposure to a combined field of 2.38 mT at 45 Hz and 3.66 mT DC, cell monolayers were assayed for dye transfer by monolayer assay techniques. Mezerein was used as a positive control, and sham-exposed controls were available. No significant effects of EMF on dye transfer were observed.
The results of the studies described in this section do not show a clear pattern of effects of EMF on cytological markers. Two of the studies on development showed no effect of exposure to fields, and two had questionable protocols; the two remaining studies show a significant effect of exposure, with a delay in the kinetics of progression of embryogenesis. The delays were achieved, however, only after extended exposure (10-70 h) to 50-500 µT, inducing electric field intensities of 0.5-5 mV/m. In none of these studies was an active sham-exposure system used, as is now standard in research on EMF sponsored by the NIEHS.
The results of studies on alterations in extracellular matrix synthesis are interesting, for three reasons. First, they include two investigations of only electric fields, which established remarkably low-intensity thresholds (10 µV/m and 6 mV/m). Second, a reasonable level of correlation is seen between dose and the differentiated state of the cells. The lowest reported threshold (10 µV/m) was observed for differentiating calvaria, while the highest (100 V/m) was seen for fully differentiated articular cartilage. Cultures at intermediate stages of differentiation (primary cells, cell lines, and embryonic sternal cartilage) had intermediate sensitivity to exposure to fields. Each of these experiments involved long durations of exposure, ranging from a minimum of 12 h to 21 d. Third, it should be noted that none of the exposure systems incorporated an active sham design.
Several important observations can be made on the basis of the results of the studies of cell surface markers. In this series, an active sham-exposure system, calibrated according to NIEHS guidelines, was used, and significant cellular responses were observed in comparison with cells exposed to the active sham. In addition, while high flux densities that give rise to high-intensity induced electric fields were reported to alter cellular activity, two investigations with significantly different flux densities (40 µT and 2.5 mT) but inducing similar electric field intensities (50 µV/m and 100 µV/m) showed cellular effects. Perhaps of equal interest, the field frequencies used (16 and 30 Hz) and exposure durations required (30 and 240 min) were similar in these two studies. Moreover, both cell systems (PC-12 and MC-3T3-E1) were exposed during well-established stages of differentiation. A critical observation in two of the studies was that the induced electric field was the active agent.
Similar observations were made in studies of cell morphology. In a terminally differentiating osteoclast model, with active sham exposure, an 8-d exposure to a 1.8 mT, 30 Hz magnetic flux, inducing a 0.6-mV/m electric field, resulted in a significant delay. Consistent with this observation were the morphological changes observed after a 24-h exposure to a 60 Hz, 0.7 mT flux (inducing a 0.5-mV/m electric field), whereas no significant morphological changes were observed with a 60 Hz, 0.1 mT flux applied for 3 h. Flux magnitudes and induced electric field intensities smaller than these induced differential responses, while values well above these consistently produced robust responses.
Few studies have been reported of the effect of EMF on gap-junctional communication, and, of those studies that have been completed, fewer have been reported fully. The two studies reported here provide little or no support for a role of gap junctions in cellular responses to EMF.
Exposure to EMF can be expressed either as induced electric field intensity or imposed magnetic flux density. Studies of pure electric fields showed that only ELF fields can affect cytological responses, with a required threshold intensity in the range of 0.1 mV/m plus or minus one order of magnitude. This threshold is consistent with those found in several experiments on magnetic induction. Moreover, no response has been found with exposure shorter than 30 min, and most reported cellular responses require exposure of tens of hours. Finally, the threshold field intensity with exposure to 10-100 Hz appears to be lower than that for broader-spectrum or higher-frequency exposure.
The patterns associated with magnetic fluxes are more difficult to identify. Flux densities ranging from 1 µT to 2 mT have been reported as threshold values, although the lowest are associated with stimuli with a higher frequency content. Assuming an electric field threshold near 0.1 mV/m, it is difficult to separate the effects of magnetic flux from those of electric fields when the induced electric fields exceed this level. Only a few studies reported cytological responses to induced electric fields below 0.01 mV/m, and these observations have not been replicated.
In-vitro experiments permit the testing of potentially toxic exposure under controlled conditions, typically at doses well above those encountered in the environment. Studies of the genotoxic effect of such exposure and effects on cell proliferation, alteration of signal transduction pathways, and modification of differentiation processes can serve to identify agents with potential carcinogenic effects or effects on other health end-points. This toxicological approach may be applied to EMF, although careful consideration must be given to the range over which the 'dose' or intensity of EMF is varied; unlike many chemical agents, EMF may have different mechanisms of field-cell coupling over different ranges of field intensity.
The generally accepted theoretical limits for effects of EMF are one to two orders of magnitude lower than they were a decade ago. The reasons for this change are that new mechanisms have been considered that were not studied in detail previously and that slightly more realistic biological models have been constructed. There is no controversy about the theoretical basis and experimental evidence for biological effects at magnetic flux densities greater than 0.1 mT or internal electric field strengths greater than approximately 1 mV/m. Similarly, there is general agreement about the lack of theoretical models and experimental evidence for effects at magnetic flux densities of less than 0.1 µT, and theoretical models for effects at densities less than 0.1 mT, and particularly less than 5 µT, are controversial. It is important to note that most of the theoretical results reported to date are based on single-cell models. Realistic modeling of temporal and spatial averaging across functional groups of cells (e.g. synchronized neurons) is a newly developing area of research, which may serve to expand the range of physical mechanisms of interaction. Existing models and theoretical thresholds are only as good as the biological data used to construct them; advances in biology and biochemistry can therefore be expected to serve as a basis for advances in our understanding of the mechanisms of interactions with EMF.
Three critical factors were considered in evaluating the contribution that in-vitro research can make to our understanding of the potential effect of EMF on human health: whether an observed (positive or negative) response has been independently validated, whether there is a demonstrated physical mechanism for the response at the field intensities used, and whether the end-points evaluated are widely considered to be predictive of potential health effects.
A series of recent (1996-98) studies demonstrated the effects of fields on gene mutations. Studies of ELF exposure at flux densities below 0.1 mT have consistently shown no effect on mutation rates; however, exposure to 0.2-400 mT reproducibly and significantly enhanced the mutation rate after X-ray or gamma-ray initiation. Moreover, exposure to 400 mT increased the number of mutations in the absence of ionizing radiation in two human cell lines. Thus, multiple, self-consistent reports demonstrate a dose-dependent effect on a process or end-point commonly considered to be associated with carcinogenesis. Importantly, the flux densities used in all of these studies (> 0.1 mT) are within the range of a single physical transduction mechanism, specifically magneto-chemical transduction. The potential for magneto-chemical effects at flux densities greater than 0.1 mT has been firmly established in both theoretical analyses and biochemical investigations.
In addition to these reports on genotoxicity, numerous well-programmed studies have shown strong effects on other end-points commonly associated with carcinogenic agents. These include significantly increased cell proliferation, disruption of signal transduction pathways, and inhibition of differentiation, all of these effects being seen at field levels >0.1 mT. These investigations were also performed at sufficiently high intensities that magneto-chemical transduction is a plausible mechanism of field-cell transduction, although this does not preclude other mechanisms of interaction. Indeed, several well-controlled studies of physical transduction have clearly demonstrated that the induced electric field can alter cell behavior. In addition to these studies, a large number of well-performed studies have also shown biological effects of EMF that are not associated with cancer end-points, suggesting that exposure may affect other disease end-points.
A limited number of well-performed studies provide moderate evidence for mechanistically plausible effects of EMF greater than 0.1mT in vitro at end-points generally regarded as reflecting the action of toxic agents.
[This conclusion was supported by 27 Working Group members; there were 2 abstentions.]
There is weak evidence for an effect of fields lower than approximately 0.1mT.
[This conclusion was supported by 26 Working Group members; there were 3 abstentions.]
The potential role of magneto-chemical processes in the coupling of ELF fields to biological systems suggests that extension of both theoretical and experimental studies of intensity regimes below 0.1 mT should be conducted.