University of California

CELL AND MOLECULAR BIOLOGY ASSOCIATED WITH RADIATION FIELDS OF MOBILE TELEPHONES
by Dr W. Ross Adey
Department of Biochemistry, University of California, Riverside Calif. 92521 USA

[PART ONE]

1. Introduction
In the perspective of human health, exposure to power-frequency electromagnetic fields (EMFs) has been universal in civilized societies for more than a century.

Developments in radio engineering technology over the past decade have created striking new options in personal communication devices and systems. In addition to extremely- low-frequency (ELF) electric power fields, many millions of mobile phone users worldwide are now also exposed daily to radiofrequency fields under near-field conditions. We may expect that these newly evolved behavioral patterns will be lifelong, with intermittent exposures at the phone user’s head making yet one more contribution to an already complex daily EMF exposure arising in an aggregate of multiple and disparate sources.

There are important biomedical considerations associated with long term exposure to any environmental factor capable of tissue interactions. In the case of EMFs, they include effects specifically attributable to intermittency of exposure, frequency of recurrent exposure, interactions involving simultaneous exposure to multiple fields, age at onset of exposure, and even considerations of ethnicity that may determine individual susceptibility.

Proximity of the phone to the user’s head has raised possibilities of direct interactions with brain functions. Modifications in sleep patterns have been reported, and changes in electroencephalographic records have been observed in the left hemisphere in the presence of a cell phone field during tone recognition [Eulitz et al., 1998]. Life-long exposure to digital and analog cell phone fields has been tested for possible brain tumor promotion in a rat model [Adey et al., 1998a and b]. These reports of essentially global effects would no doubt achieve wider acceptance if they could be cast against defined mechanistic substrates at the far finer levels of cell and molecular biology. Much current research in EM bioeffects is directed towards this goal.

This review will consider recent evidence pointing to a key role for a sequence of events in bioeffects of EM fields that begins at cell membranes and initiates a series of intracellular enzyme cascades. In the short term, their most basic functions, as expressed in sensitivities to microwave fields, can be defined in such actions as essential metabolic and messenger activity, and in cell growth regulation. But in the longer term, a broader perspective has defined probable continuing roles for these same enzyme cascades in programmed-cell-death (apoptosis) [Polyak et al.,, 1997], in the experimental modeling of ELF magnetic field influences in acute lymphoblastic leukemia of childhood [Uckun et al., 1995], in the promotional phase of tumor formation [Byus et al., 1987; O’Brien et al., 1994, 1998], and in the pathophysiology of certain neurodegenerative diseases, including Parkinson’s [Holshouser et al., 1995] and Alzheimer’s [Sobel et al., 1995].

Recurring criticisms of EMF research have addressed reported inconsistencies in experiments within a single laboratory, or between laboratories that have sought to replicate a single reported bioeffect. We shall review these concerns, considering the essentially universal distribution of EM fields along most intercellular gutters as preferred pathways. Not sufficiently recognized are the ensuing options for simultaneous EMF interactions with large numbers of surface receptors on individual cells. In turn, they are functionally connected to intracellular pathways serving a variety of physiological mechanisms. Absence of predictable one-on-one stimulus-response patterns in these parallel and composite pathways becomes possible, with evidence for varying amounts of „cross-talk” between cascading enzyme pathways at successive signaling levels in their passage from cell membranes to cell nuclei and other intracellular structures [Hill, 1998].

2. Athermal tissue sensitivities; evidence for role of cell membranes in first detection of microwave fields.
Collective evidence points to cell membrane receptors as the probable site of first tissue interactions with both ELF and microwave fields for many neurotransmitters [Kolomytkin et al., 1994], hormones [Liburdy et al., 1996], growth-regulating enzyme expression [Byus et al., 1987; Litovitz et al., 1993; Penafiel et al., 1997] and cancer- promoting chemicals [Cain et al., 1993]. In none of these exposures does tissue heating appear involved causally in the responses, and indeed, additional evidence for modulation frequency-dependence in these and related studies appears to support the case for mediation through athermal mechanisms. Physicists and engineers have continued to offer microthermal, rather than athermal, models for these phenomena [Barnes, 1997; Astumian et al., 1995], with views that exclude consideration of cooperative organization and coherent charge states; but it is difficult to reconcile experimental evidence for factors such as modulation frequency-dependence and required duration of an amplitude- modulated signal to elicit a response (coherence time) [Litovitz et al., 1993] with models based on the equilibrium thermodynamics of tissue heating.

2.1 May tissue electrochemical thresholds reside in ensemble properties of tissue elements?
An emergent concept.

Research in sensory physiology has suggested that some threshold properties may reside in highly cooperative properties of a population of elements, rather than in a single detector (see Adey, 1998 for review) as for example, in auditory vibration thresholds around 10-11m, or approximately the diameter of a hydrogen atom; or human olfactory thresholds for musk at 10-13M, with odorant molecules distributed over 240 mm2 [Adey, 1959]; or detection of single photons of blue-green light at energies of 2.5eV [Hagins, 1979). Bialek [Bialek,1983; Bialek and Wit, 1984) has addressed this problem of the auditory receptor in quantum mechanical terms. He evaluated two distinct classes of quantum effects: 1) a macroquantum effect, typified by the ability of the sensory system to detect signals near the quantum limits to measurement; and 2) a microscopic quantum effect, in which „the dynamics of individual biological macromolecules depart from predictions of a semi-classical theory.”

Bialek concluded that:

1) quantum-limited sensitivity occurs in several biological systems, including displacements of sensors of the inner ear. Quantum limits to detection are reached in the ear in spite of a seemingly insurmountable level of thermal noise.
2) To reach the quantum limit, the receptor cells of the inner ear must possess amplifiers with noise performance approaching limits set by the uncertainty principle.
It is impressive that suppression of intrinsic thermal noise allows the ear to function as though at close to 0K, again suggestive of system properties inherent in the detection sequence. These „perfect” amplifiers could not be described by any chemical kinetic model, nor by any quantum mechanical theory in which the random phase approximation is valid. The molecular dynamics of amplifiers in Bialek’s models would require that quantum mechanical coherence be preserved for times comparable to integration times of the detector. In the context of this review, it remains a matter for speculation whether comparable mechanisms may determine electromagnetic sensitivities as a more general biological property in tissues at cellular and subcellular levels.

2.2 Significance of ELF modulation-dependent effects;
the SAR concept revisited
A considerable number of studies have reported RF and microwave field effects in animals and cell cultures. In cell preparations with SARs less than 5W/kg, and therefore in the presumed absence of heating as a mediating stimulus, cellular responses occurred primarily from exposures to microwaves that were amplitude- or pulse-modulated at ELF frequencies [Adey, 1990, 1992; 1997; Penafiel et al., 1997].

Reported effects included changes in calcium ion efflux, [Bawin et al., 1975; Blackman et al., 1979, 1985; Dutta et al., 1984, 1989; Lin-Liu and Adey, 1982];
altered enzyme activity [Byus et al., 1984; 1988; Litovitz et al., 1993; Penafiel et al., 1997];
altered cerebral neurotransmitter binding [Kolomytkin et al., 1994]; and
induction of cellular transformations [Balcer- Kubiczek and Harrison, 1985, 1989, 1991; Czerska et al., 1992].
In contrast, effects on cell cultures have been reported with CW microwave fields [Cleary et al., 1990; Saffer and Profenno, 1992; Garaj-Vhrovac et al., 1992], all using SARs equal to or greater than 10 W/kg. Thus, the aggregate evidence suggests an important biological role for ELF amplitude- or pulse-modulated microwave fields with absorbed energies below 5 W/kg, where heating is not the mediator of these bioeffects.

Observed bioeffects have been typically to coherent signals associated with ELF fields, or to RF fields with coherent ELF amplitude- or pulse-modulation [Adey, 1990].

Temporally incoherent magnetic fields mitigate responses of biological systems to temporally coherent ELF magnetic fields [Litovitz et al., 1994]. Tissue threshold gradients have yet to be reliably defined for ELF electric fields, but neurobehavioral effects have been reported in submammalian marine vertebrates with environmental fields as weak as 10-6 V/m [Kalmijn, 1971]. Litovitz et al. [1993] have proposed that these oscillating fields must be present for a certain minimum duration to be an effective stimulus in activation of the enzyme ornithine decarboxylase (ODC) in cultured cells. For an RF field amplitude-modulated at 60 Hz, this coherence time was reported as between 1 and 10 sec.

Sensitivities may be windowed with respect to ELF field frequency [Walleczek, 1994], or to ELF modulation frequency. First reports of modulation frequency-dependent bioeffects related to calcium efflux from brain tissue [Bawin et al., 1975; Adey, 1980], in tissue preparations covering a wide size range. Increased calcium efflux from the chick cerebral hemisphere described a „tuning curve” , peaking around 16 Hz and diminishing at higher and lower frequencies [Bawin et al., 1975; Blackman et al., 1979]. As a function of size, a similar behavior was reported in preparations of brain synaptosomes, micron-sized subcellular particles derived from nerve fiber terminals in cerebral cortex [Lin-Liu and Adey, 1982]. Calcium efflux increased with exposure to a 450 MHz field modulated at 16 Hz, but not to an identical field modulated at 50 Hz, nor to an unmodulated (CW) field.

Most enzymes, the great class of proteins that act as biological catalysts, are calcium- dependent. Enzymes differing widely in structure and function have also shown modulation frequency-dependence in RF/microwave exposures. Messenger enzymes, the protein kinases, are acutely but transiently reduced in activity in human leukocytes when exposed to a 450 MHz field sinusoidally modulated at 16 Hz field (Byus et al., 1984).

Ornithine decarboxylase (ODC) regulates cell growth through synthesis of polyamines necessary for protein and DNA synthesis. Its activity in fibroblasts and liver cells is stimulated by microwave fields with ELF modulation, but not by unmodulated fields [Byus et al., 1988; Penafiel et al., 1997].

Little is known about factors that might determine upper and lower bounds for this modulation frequency-dependence, since much of this research has used spot frequencies, rather than examining a spectral continuum. Cerebral calcium efflux studies cited above have suggested a peak around 16-20 Hz. For ODC activation with sinusoidal amplitude- modulated 835 MHz fields, Penafiel et al. [1997] reported a much broader responsive spectrum from 6 to 600 Hz, with sharp boundaries; but no response to a similar field with 60 Hz frequency modulation at + 60 kHz deviation.

Observation of these modulation frequency-dependent bioeffects would appear to raise significant questions concerning validity of continued use of the thermally based Specific Absorption Rate (SAR) as a universally valid predictor of bioeffects attributable to RF field exposure. It is clearly a blanket measure blind to specifics of field characteristics that may be critical determinants of the biological response in many athermal exposures.

2.3 Models of tissue detection of ELF modulation of RF/microwave fields
If the concept of modulation frequency-dependence continues to gain support in further research, answers must be sought as to the manner of its detection. For ELF fields, models based on joint static/oscillating magnetic fields have been hypothesized. They include ion cyclotron resonance models [Liboff, 1992], where mono- and divalent cations, such as calcium and potassium, abundant in the cellular environment, may exhibit cyclotron resonance at ELF frequencies in the presence of an ambient static field of less than 100 ?T, such as the geomagnetic field. Other models describing ELF frequency dependence have considered phase transitions [Lednev, 1991] and ion parametric resonance [Blackman et al., 1994], but interpretation of this frequency dependence based on ion parametric resonance remains unclear [Adair, 1998].

For amplitude- or pulse-modulated RF fields, there is the implication that some form of envelope demodulation occurs in tissue recognition of ELF modulation components, but the tissue remains essentially transparent to the same signal as an unmodulated carrier [Adey, 1981]. In that case, is the biological low frequency dependence established at the transductive step in the first tissue detection of the field, or does it reside at some higher level in an hierarchical sequence of signal coupling to the biological detection system [Engstrom, 1997]? For ELF magnetic fields, experimental evidence points to a slow time scale of interaction in inhibition of tamoxifen’s antiproliferative action in human breast cancer cells [Harland and Liburdy, 1998]. In the following sections, we consider evidence for free radical involvement in magnetic field bioeffects, but their characteristically short lifetimes (typically in nanoseconds or less, though longer if a geminate radical pair is contained ) would appear to preclude their role in direct recognition of low frequency oscillations in any imposed electromagnetic field, and thus, their direct involvement in a modulation frequency-dependent transductive step.

A suggested basis for envelope demodulation at cell surfaces may reside in the intensely anionic charge distribution on strands of glycoprotein protruding from the cell interior. They attract a surrounding cationic atmosphere of calcium and hydrogen ions, the charge separation creating a Debye layer. In models and in experimental data from charged resin particles, Einolf and Carstensen [1971] concluded that this physical relationship creates a large virtual surface capacitance, with dielectric constants as high as 106 at frequencies below 1 kHz. Displacement currents induced in this region by ELF modulation of an RF field may then result in demodulation.

3. Evidence for free radical participation in the transductive step.
An understanding of biological mechanisms that mediate detection of oscillating electromagnetic fields at levels below tissue thermal collision energies must await future research. However, one line of current research has suggested free radical mechanisms to account for these sensitivities (see Adey, 1993, 1997 for summaries).

3.1 Magnetic fields interact with free radicals formed in ongoing chemical reactions
Chemical bonds constantly break and reform in the course of all chemical reactions.

These bonds are magnetic, formed between atoms with paired electrons having opposite spins and thus magnetically attracted. With breaking of a bond in a chemical reaction, each atomic partner reclaims its electron, moving away as a free radical to seek another partner with an opposite electron spin. The brief lifetime of a free radical is typically in the nanosecond range or less, before once again forming a singlet pair with a partner having an opposite spin; or for electrons with similar spins, having options to unite in three ways, forming triplet pairs. During this brief lifetime, imposed magnetic fields may delay may delay return to the singlet pair condition, thus influencing the rate and amount of product of an on-going chemical reaction [McLauchlan, 1992].

Although evidence already cited is against a possible role for free radicals as direct mediators of the first transductive step in ELF modulation-frequency-dependent bioeffects, they may mediate responses to static magnetic fields, even at the lowest levels.

Moreover, when alternating fields (ELF to microwave) are imposed on static magnetic fields, they can further affect reactions by providing quanta of energy equal to the gap between singlet and triplet states, allowing transition of radicals and thus increasing reaction probability [Lacy-Hulbert et al.,1998]. Static magnetic fields in the low mT range can affect enzymatic reactions based on radical intermediates [Grissom, 1995].

3.2. Static magnetic field effects at energies below biosystem thermal energies.
McLauchlan [1992] has noted that static fields in the low mT range can exert „an enormous effect” on a chemical reaction, and that this effect begins at the lowest applied field strength, at energy levels very much smaller than the thermal energy of the system.

Spin-mixing of orbital electrons and nuclear spins is a possible mechanism for biosensitivities at extremely low magnetic field levels [McLauchlan and Steiner, 1991].

But the highest level of free radical sensitivity may reside in hyperfine-dependent singlet- triplet state mixing in radical pairs with a small number of hyperfine states that describe their coupling to nearby nuclei. There are further requirements in conditions of maximum electron-nuclear spin alignments and zero applied field. Since electron and nuclear spins have different moments, application of any field causes loss of spin correlation, due to their differing Larmor frequencies. Also, singlet-triplet interconversion must be sufficiently fast to occur before diffusion reduces probability of radical re-encounter to negligible levels. Studies in magnetochemistry have suggested that a form of cooperative behavior may exist in populations of free radicals that remain spin-correlated after initial separation of a singlet pair [Grundler et al., 1992]. Considered in a biological context, these are some of the constraints that currently limit acceptance of free radical models as general descriptors of threshold events.

3.3 Role of reactive-oxygen (ROS) and reactive-nitrogen (RON) species in biomolecular systems.
A broadening perspective on actions of free radicals in all living systems emphasizes a dual role: first, as messengers and mediators in many key processes that regulate cell functions throughout life; and second, in the pathophysiology of oxidative stress diseases.

Lander [1997] has emphasized that we are at an early stage of understanding free radical signal transduction. „Future work may place free radical signaling beside classical intra- and intercellular messengers and uncover a woven fabric of communication that has evolved to yield exquisite specificity.” In normal regulatory processes from brain tissue to blood vessels, they are essential in certain hormone actions, ion transport (including calcium-regulated cell membrane potassium channels), neuromodulation of EEG rhythms, programmed-cell-death (apoptosis), and in vascular control of penile erection [Lander, 1997]. On the other hand, in excess they may irreversibly damage many biomolecules, including proteins and phospholipids. Reactive oxygen intermediates can eventually kill cells. Parkinson’s and Alzheimer’s diseases, some forms of epilepsy, coronary artery disease, aging and cancer may be attributable to oxidative stress from accumulated free radical damage [Das, 1995].

3.4 Free radicals in regulation of specificity of cell membrane receptors.
At cell membranes, free radicals may play an essential role in regulation of receptor specificity, but not necessarily through „lock and key” mechanisms [Lander, 1997]. As an example, cysteine molecules may be strategically exposed on the surface of P21-ras proteins at cell membranes. There, they become selective targets for nitrogen- and oxygen-free radicals, thereby inducing covalent modifications. At this point, the redox potential of the target molecule would become the critical determinant of highly specific interactions with free radicals. As a free radical scavenger, glutathione is present at the cell surface in concentrations as high as 10 mM, and may modulate ensuing cascades of enzyme activation within cells.

4.0 Free radicals as mediators in enzyme cascades initiated at cell membranes; parallel inward paths from cell surface receptors
Although still at a pioneering stage, recent studies have identified enzyme systems associated with cell membranes that are sensitive to imposed magnetic fields in the course of normal enzymatic activation. Their signaling pathways are centrally directed from cell membranes to intracellular organelles, including the nucleus. Products of some of their actions may also move centrifugally, transported from the cell of origin to neighboring cells. As examples, we may review

1) glutamate receptors and the synthesis of nitric oxide (NO);
2) synthesis and export of polyamines by ornithine decarboxylase (ODC); and
3) protein kinase activity as a messenger function in cell growth regulation.
In the central progression of signals along these pathways, options for crosstalk between them also occur, as reflected in degrees of unpredictability in stimulus outcome. The studies have tested both ELF and ELF-modulated RF fields.

4.1. Glutamate receptor activation in brain tissue with synthesis of NO;
modulation of an enzyme cascade by ELF magnetic fields
Glutamate has widespread functions in all vertebrate nervous systems as an excitatory neurotransmitter. In the mammalian brain, it regulates the rhythmic patterns of the electroencephalogram [Bawin et al., 1994]. Disorders of glutamate metabolism in the human brain occur in Parkinson’s disease [Holshouser et al., 1995]. An enzymatic cascade is initiated within brain cells when glutamate receptors are activated, beginning with a small inward calcium current. This leads to synthesis of NO from the amino acid arginine by the enzyme nitric oxide synthase. As a gaseous molecule, NO diffuses readily into cells surrounding its cell of origin. By further enzymatic steps, NO stimulates conversion of the chemical fuel substance guanylate triphosphate (GTP, the analog of ATP) into cyclic-guanylate monophosphate (cGMP, the analog of cAMP). cGMP is a powerful messenger molecule, initiating a further spectrum of biochemical paths.

Experimental evidence supports a role for NO in controlling regularity of patterns of EEG waves in rat brain hippocampal tissue. Suppression of NO synthesis is associated with shorter and more stable intervals between successive burst of rhythmic waves [Bawin et al., 1994]. Their rate of occurrence is also disrupted by exposure to weak 1 Hz magnetic fields (peak amplitudes, 0.08 and 0.8 mT). Normal operation of these regulatory processes requires NO synthesis in tissue [Bawin et al., 1996].

4.2 Modulated RF fields influence the cell growth regulating enzyme ornithine decarboxylase (ODC)
Recent research in the cancer field has led to increasing acceptance of new concepts about essential steps in the unregulated cell growth of tumor formation. Damage to DNA in cell nuclei is seen as an essential but no longer as a sufficient step in tumorigenesis.

Instead, initiation (transformation, mutation) of a cell involves damage to nuclear DNA, possibly as a single event. But for tumor formation to occur, initiation must be followed by a long period of promotion, in man possibly involving a large part of the individual’s lifetime. Typically, this promotion phase involves repeated, intermittent exposure to the promoting agent. Recent research now implicates overexpression of a single enzyme in an initiated cell as sufficient for tumorigenesis. Overexpression of ornithine decarboxylase (ODC) appears to meet this condition in initiated cells in the absence of any other promoter [O’Brien, 1976; O’Brien et al., 1994]. Activation of ODC as a response to ELF fields [Byus et al., 1987; Litovitz et al., 1993] and to ELF-modulated RF fields [Byus et al., 1988; Penafiel et al., 1997] has already been noted.

ODC synthesizes polyamine molecules from ornithine [Pegg et al., 1994]. These long chain molecules (spermine, spermatidine, putrescine and cadaverine) are polycationic.

They have amongst the highest charge/mass ratios found in biomolecules. They are essential for protein and DNA synthesis, and are exported differentially from their cells of origin [Tjandawinata et al., 1994a and b; Tjandrawinata and Byus, 1995].

Polycationic putrescine exported into intercellular gutters is reported to modulate excitation of glutamate receptors located in the polyanionic atmosphere of the cell membrane surface [McBain and Meyer, 1995; Bowie et al., 1998].

Taken together, these reports suggest options for imposed EM fields to influence concurrent interactions at cell membranes between two quite distinct enzyme cascades, the glutamate cascade directly involved in NO free radical expression, and an ODC cascade modulating activity at glutamate receptor sites. Moreover, the efflux of putrescine and cadaverine from rapidly growing cells is regulated by putrescine already exported from the cell interior [Tjandrawinata and Byus, 1995].