J. theor. Biol. (2000) 206, 291}298
doi:10.1006/jtbi.2000.2123, available online at http://www.idealibrary.com on
Biological E4ects of Electromagnetic Fields=Mechanisms for the E4ects of
Pulsed Microwave Radiation on Protein Conformation
JOCELYN A. LAURENCE*, PETER W. FRENCH-,
ROBYN A. LINDNER? AND DAVID R. MCKENZIE*A
*School of Physics, ;niversity of Sydney, Sydney NS= 2006, Australia, -Centre for Immunology,
St. <incent1s Hospital, Sydney, NS= 2010, Australia, ?School of Chemistry,
;niversity of =ollongong, =ollongong, NS=, Australia and AAustralian Key Centre of Microscopy
and Microanalysis, ;niversity of Sydney, Sydney NS= 2006, Australia
(Received on 23 December 1999, Accepted in revised form on 2 June 2000)
Microwave exposure under &&athermal'' conditions occurs when no temperature rise can
be measured by conventional thermometry. The existence of biological e!ects arising from
the athermal exposure is still controversial, partly because of a lack of the linear dose response
relation. We propose a model in which pulsed microwave radiation causes a triggering of the
heat shock or stress response by altering the conformation of proteins through a transient
heating of the protein and its close environment. We support this by modelling using the heat
di!usion equation and show that pulsed exposure even when athermal can lead to transient
temperature excursions outside the normal range. We propose that the power window
phenomenon in which biological e!ects are observed at low power levels may be caused by an
incomplete triggering of the heat shock response.
2000 Academic Press
Introduction
A considerable body of experimental work has
The safety of exposure to low-level radio-
revealed that e!ects can occur under conditions
frequency
in which the macroscopic temperature as mea-
"elds has been a subject of debate and
concern in both the scienti
sured by conventional thermometry does not
"c and lay communi-
ties. This issue has been brought sharply into
exceed normal levels. These studies include the
focus by the enormous recent expansion of mo-
well-accepted physiological e!ect known as the
bile telecommunications which employ a trans-
microwave hearing phenomenon which occurs in
mitting antenna held close to the head. It is well
individuals of normal hearing when short micro-
accepted that exposure to the electromagnetic
wave pulses are administered to the head (Frey
& Messenger, 1972) as well as in vivo reports of
"elds in the microwave region imposes stresses
on the living cells when the speci
increased cancer incidence following microwave
"c absorption
rate (SAR) is beyond a level su
exposure (Repacholi et al., 1997) and in vitro
$cient to cause
the temperature of the cells and tissues to be sig-
e!ects on cell growth, morphology and secretion
ni
(French et al., 1997; Donnellan et al., 1997).
"cantly elevated. In this case there is a clear
mechanism, namely a heat stress which causes
However, at apparent &&athermal'' exposures
e!ects.
Daniells et al. (1998) showed that using a pulsed
0022}5193/00/180291#08 $35.00/0
2000 Academic Press

---

292
J. A. LAURENCE E¹ A¸.
microwave exposure at 750 MHz, heat shock
genes were activated. Importantly, they showed
that lower power levels tended, in general, to
induce larger responses, which is contrary to the
results expected if the e!ect were simply due to
heating. They concluded that microwave radi-
ation causes the activation of cellular stress re-
FIG. 1. The protein folding/unfolding (
) and o!-
folding pathways (
) (adapted from Lindner et al., 1997).
sponses, presumably re#ecting increased levels of
The former is a reversible pathway and involves interaction
protein structural alteration by a mechanism
with chaperone proteins such as HSP70 which leads to
other than heating.
correctly folded proteins. The latter is an irreversible path-
way leading to precipitation which involves the interaction
In this paper we examine mechanisms based
with sHSP chaperone proteins.
on the hypothesis that exposure to electromag-
netic "elds can cause changes in the conforma-
tion of biologically active macromolecules. The
possibility of direct e!ects of electromagnetic
"elds
on the conformation of macromolecules
also increased in Alzheimers disease tissues (Perez
has been suggested previously (Porcelli et al.,
et al., 1991).
1997; Taylor 1981; Byus et al., 1984). This sugges-
tion arises from the fact that small changes in
protein conformation can result in signi
A Practical De5nition of Thermal and
"cant ef-
Athermal Exposures
fects on the biological function (Koshland, 1998).
Neshev & Kirilova (1996) have discussed the po-
It is important to establish a basis for distin-
tential ability of pulsed microwave radiation at
guishing between thermal and athermal expo-
very low power levels to alter the conformation of
sures. This can be done by considering the
enzymes. A cell under some conditions may re-
principles of temperature and its measurement.
spond to this stress by activating a stress re-
Strictly speaking, the temperature of a system is
sponse. This stress response acts through the
meaningful only when all the parts of the system
so-called heat shock proteins (HSPs) and is essen-
are in thermal equilibrium, that is, there is no
tial to normal cellular function. HSPs act as
heat #owing between one part of the system and
&&molecular chaperones'', binding to and stabilis-
another. In order to determine the temperature,
ing partially unfolded proteins, thus providing
there is an implication of a measurement (or
the cell with protection from imposed environ-
thermometer response) time being the time re-
mental stresses. Without the chaperone e!ect
quired to bring the system into thermal equilib-
of HSPs, cells would be vulnerable to small
rium with a thermometer. No meaningful
changes in their environment. The most abun-
measurement of temperature can be made on
dant are the small HSPs (sHSPs) and HSP70
a time scale shorter than the response time. The
(and its homologues). Lindner et al. (1997),
International
Commission
on
Non-Ionising
using NMR spectroscopy, have shown that
Radiation Protection (ICNIRP) guidelines (1998)
sHSPs (e.g. HSP27) only interact with the unsta-
for limiting the exposure to time-varying electro-
ble, disordered molten globule states of de-
magnetic radiation specify a measurement time
stabilised proteins that are on an irreversible
of 6 min during which time-varying doses may
pathway towards an aggregated and precipitated
be averaged. This enables a distinction to be
state. On the other hand, large HSPs (e.g. HSP70)
made between a thermal exposure, that is one
interact
with
proteins
along
their
folding
which causes a signi"cant temperature rise as
pathway (see Fig. 1).
determined by a convenient thermometer with
Increased expression of heat shock proteins
a 6 min averaging time, and an athermal expo-
has been associated with dementia and gliosis.
sure, one which does not cause a signi"cant tem-
HSP27 alpha B-crystallin, for example, has been
perature rise in this sense. The 6 min value is
found in reactive gliosis, in Alzheimer's disease
useful as this enables readily available thermo-
(Renkavek et al., 1993) HSP72 and HSP73 were
meters such as liquid-in-glass type to be used.

---

BIOLOGICAL EFFECTS OF ELECTROMAGNETIC FIELDS
293
In order to measure the temperature any ther-
Temperature variations occurring only over such
mometer will only be in thermal equilibrium with
small sizes and time periods are of course ather-
a certain volume of the subject system and this
mal under the de"nition of the previous section.
volume needs to be speci"ed. The ICNIRP stan-
Let us now suppose that the energy is injected
dard provides a practical de"nition. It speci"es
directly into a protein molecule, assumed spheri-
a 10 g mass of tissue which gives a characteristic
cal, at time t"0. The temperature distribution
system size of approximately 10 cm. Therefore,
¹ of the molecule and its aqueous environment
only when a 10 g mass of the tissue shows a sig-
can be calculated from the heat di!usion equa-
ni"cant temperature rise in the above sense is the
tion
exposure considered in this work to be thermal,

c *¹
¹"
(1)
otherwise we shall consider it to be athermal.
*t
where c is the heat capacity,
is the density and
Characteristic Sizes and Times for Important
is the speci"c thermal conductivity. By solving
Cell Processes
this equation, Marks (1997) showed that the
In this section we will carry out some calcu-
characteristic time for cooling of spherical re-
lations to determine the approximate size and
gions scales as
time scales relevant to initial processes related to
the interaction of proteins with electromagnetic
c
tJ
r
(2)
"elds in order to categorise them as occurring in
the thermal or athermal size and time regions, as
de"ned in the previous section, in the context of
This scaling rule is consistent with the result
common industrial applications of pulsed elec-
tromagnetic "elds. Before doing so, it is useful to
c
consider the characteristic times of pulses from
t+0.5r
(3)
the common sources.
derived by NoKlting (1995) for the pulse length of
SOME CHARACTERISTIC TIMES FOR EXPOSURES
a light pulse absorbed by a spherical body above
which heat exchange with the environment be-
The characteristic time for a burst of informa-
comes important. Beginning with Marks
tion carried on the GSM digital mobile telecom-
' result
for a cooling time of 0.052 ps for 0.6 nm spherical
munications system is 577 s (Gibson, 1996). This
region in diamond, and using values of c"
contains 156 individual pulses each of 3.69 s.
4.18;10 JK\ kg\,
"0.6 Wm\ K\ and
The 3.69 s is taken as typical of a pulsed expo-
"1 g cm\ for an aqueous medium, we
sure to microwaves and is shown in Fig. 2. Pulses
"nd
a cooling time constant of 1 ns for a protein with
of microwaves are also used in radar. A typical
a diameter of 10 nm immersed in an aqueous
pulse width in modern high repetition rate
medium. This diameter corresponds to a samp-
systems is 1 s (Blackman & Popoli, 1999).
ling volume of approximately 10\ cm.
CHARACTERISTIC COOLING TIME OF A PROTEIN IN AN
CHARACTERISTIC TIME FOR ENERGY TO DIFFUSE
AQUEOUS ENVIRONMENT
FROM A VIBRATIONAL MODE IN A MACROMOLECULE
In the previous section we have chosen rela-
It is conceivable that direct energy absorption
tively large characteristic sizes and times in order
could occur from an electromagnetic "eld into
to provide a practical distinction between ther-
a vibrational mode of a protein molecule. Such
mal and athermal exposures. In this section we
a mode could be, for example, a mode with strong
will assume that it is in principle still possible to
infrared activity. The time taken for the energy to
de"ne temperature on a length scale comparable
spread into other vibrational modes and reach
to the dimensions of a protein molecule and
thermal equilibrium can be estimated. The en-
on time scale of the order of nanoseconds.
ergy in one vibrational mode spreads to another

---

294
J. A. LAURENCE E¹ A¸.
FIG. 2. Shows characteristic times and dimensions of some critical biological processes and common industrial applica-
tions of pulsed electromagnetic "elds. The lines delineate four regions, a thermal region (1) in which temperature excursions
occur on a scale both in space and time so that conventional thermometry can detect temperature changes. Regions (2) and (3)
are athermal since only one criterion is set. Region (4) is athermal on both criteria.
as a result of the anharmonic terms in the in-
time for the di!usion of heat away from a protein
teratomic potential. The time taken is necessarily
molecule but somewhat smaller than the time for
longer than one vibrational period of the reson-
protein unfolding.
ant mode because otherwise the mode would be
The characteristic volume for this interaction
overdamped and not resonant with the "eld. One
is the approximate volume of a protein molecule,
period is therefore a lower bound to the charac-
10\ cm. The di!usion of heat from a vibra-
teristic time for energy di!usion but it is likely
tional mode of a protein is therefore in the ather-
that at least 10 periods would be necessary. It is
mal region on both the time and the length scales.
di$cult to provide an upper bound without a
detailed investigation of the dynamics of the
molecule. For small molecules with vibrational
CHARACTERISTIC TIME FOR UNFOLDING OF
modes in the THz range it is exceptional for the
PROTEINS
lifetime of a vibrational excited state to exceed
While the time required for the initial unfold-
1000 oscillations (Chronister & Crowell, 1991).
ing of a protein from its native state to its lowest
The rate of migration of energy is dependent
energy intermediate state is unknown, we can
upon the anharmonic terms and these are likely
place an upper limit on this time by considering
to be very large for macromolecule vibrations.
the most rapid events in the folding}unfolding
Damping by collision with solvent will also en-
process. A time constant of 300 s was observed
hance the rate of migration. Consequently,
experimentally for a rapid folding step in mutant
a good upper band on the lifetime of a vibration-
barstar (Schreiber & Firsht, 1993). Experimental
ally excited state in a macromolecule would
measurements
of
di!usion
limited
contact
be 1000 oscillations. For a resonant absorption
formation in unfolded cytochrome c show that
in the GHz range an upper bound for the time
35}40 s is required to form an encounter
scale for energy di!usion would therefore be
complex (Hagan et al., 1996). The latter authors
0.4 ms. This time is larger than the characteristic
also "nd, from theoretical considerations of

---

BIOLOGICAL EFFECTS OF ELECTROMAGNETIC FIELDS
295
di!usion-controlled contact formation in poly-
conclude that such deviations from thermal equi-
mers, that the time to form a loop in a short
librium between the protein and its environment
protein containing 6}10 residues is 1}3 s. The
are unlikely to be of su$cient duration to cause
process of loop formation, when reversed, should
a protein conformation change and that confor-
be a good model for the simplest and therefore
mational change occurs with approximate equi-
most rapid step in a protein unfolding event. If
librium between the protein and its environment.
di!usion e!ects are excluded molecular dynamics
Surrounding water molecules exposed to hydro-
simulations show that loop formation can pro-
phobic regions on protein molecules adopt hy-
ceed rapidly. For example, the molecular dynam-
drogen-bonded pentagonal structures. Cameron
ics simulation of Wu & Sung (1998) showed loop
et al. (1997) have modelled the formation of mul-
formation in an isolated 16 residue polypeptide
tiple water molecule pentagons into a closed
occurring in as little as 0.5 ns. However, it is
&&clathrate'' structure that can conform to regular
important that di!usion processes are included
surfaces and even enclose a purely hydrophobic
since these will be the rate-limiting processes for
molecule or polymer. An increase in the temper-
a protein in an aqueous environment. Therefore,
ature of the water can have signi"cant conse-
a microsecond timescale is the most likely. This is
quences for protein conformation. The breaking
a long-time relative to the time taken for thermal
of hydrogen bonds within the water pentagonal
equilibrium to be established. It is, therefore, un-
structure may cause it to give way to the bulk
likely that direct heating of a protein by the
water state (Chattopadhyay et al., 1997). As a re-
electromagnetic "eld relative to its surroundings
sult, hydrophobic groups on the surface of the
could be responsible for protein conformational
protein may be able to interact with each other to
changes. Any conformational change must pro-
a greater extent and the protein may adopt a dif-
ceed in an approximate equilibrium between the
ferent conformation. In several protein}ligand
protein and its immediate environment.
complexes, single water molecules lie between the
protein and the ligand "lling gaps and mediating
hydrogen bonds between the two molecules
A Mechanism for Microwave Radiation
(Rarey et al., 1999).
to Act as a Cell Stressor
We therefore conclude that an analysis of
On a time scale of nanoseconds and longer,
a speci"c exposure situation using Maxwell's
a protein molecule is in approximate thermal
equations to determine the energy absorption
equilibrium with its aqueous environment. Times
and the heat di!usion equation to determine the
longer than this are required for protein unfold-
consequent temperature distributions will be suf-
ing events because of the rate-limiting e!ects of
"cient to determine the consequences of the expo-
di!usion. We therefore propose that the micro-
sure for protein conformation. There is risk of
wave pulses can cause unfolding to occur only
protein conformational changes during temper-
through transient heating of both the molecule
ature excursions exceeding the minimum time for
itself and its aqueous environment. We propose
protein unfolding for which we estimate 1 s is
that this can occur either through the resonant
a lower bound.
absorption of energy initially in the peptide chain
(with subsequent di!usion of energy to the
aqueous environment) or through resonant
Modelling Pulsed Exposures with the Heat
absorption of the aqueous environment (with
Di4usion Equation
subsequent di!usion of energy into the protein).
Pulsed exposure appears to be more e!ective
Chattopadhyay et al. (1997) have discussed the
in eliciting a biological response than a continu-
resonance of microwave radiation at very low
ous exposure of the same SAR value. An example
power levels with small groups of amino acids
of a known biological e!ect is the &&microwave
within proteins, e.g. in the active sites of key
hearing'' e!ect, attributed to a thermoelastic ef-
enzymes. At higher power levels, they propose
fect in the auditory cortex of the brain. It arises
that the entire macromolecule is forced into a
for pulses shorter than 30 ms at 2450 MHz,
vibratory mode. From our analysis above, we
giving a speci"c absorption (SA) greater than

---

296
J. A. LAURENCE E¹ A¸.
4 mJ kg\ or an SAR greater than 133 W kg\
considered to be in contact with a thermal reser-
(Frey & Messenger, 1972).
voir, the base of the cell culture and its environ-
The di$culty in understanding the e!ects of
ment, and the surface at x"0 was considered to
pulsed energy inputs, as opposed to continuous
be in contact with a thermal insulator. The upper
ones, is that rapid temperature changes result,
surface of the medium is in contact with air but
and spatially non-uniform temperature distribu-
excepting the rapid pulsed exposures of interest,
tions arise. This gives rise to the experimental
this will be essentially a non-convective bound-
di$culties in measuring temperature changes
ary since the convective currents need time to
since most methods of temperature measurement
establish. Because of the small value of air con-
have signi"cantly longer response times than the
ductivity we assume this to be an adiabatic
duration of the microwave pulses of interest. Spa-
boundary.
tial non-uniformity is also di$cult to quantify
The rate of heat delivery per unit volume at
because of the "nite size of the most exposed
a distance x into an absorbing medium decreases
volume and the perturbing e!ects of thermo-
exponentially with x. We assume an expression of
meters, even low thermal mass ones like #uorop-
the form
tic probes. Modelling work is therefore essential
to determine the temperature as a function of
qR"Ae\RRe\V*
(6)
time and position. To quantify the di!usion of
heat from the delivery point to the surround-
ings, the second-order di!erential equation
where the constant A was initially chosen to an
known as the heat equation is the governing
exposure of 0.17 W kg\ and had the value
equation. For a one-dimensional system this is
A"5.96;10 W m\ which gives a pulse energy
expressed as
delivery (measured at x"0) of 59.6 J kg\ and
a maximum SAR (measured at t"0 and x"0)
*
of 7.2;10 W kg\ assuming a tissue culture
¹ (x, t)
qR
1 *¹(x, t)
#
"
(4)
medium density of 1000 kg. The constant ¸
*x
*x
determines the rate at which the heat delivery
decreases with distance into the medium as a re-
where x is the single spatial variable, t is the time,
sult of the attenuation of the microwave signal.
¹ (x, t) is the temperature, qR is the rate of heat
A value of ¸"1 cm is used in these calculations.
delivery per unit volume,
is the thermal con-
The constant t determines the duration of the
ductivity, and the di!usivity is given by
pulse. A value of 1 ms was chosen. Note the pulse
delivers an energy of 7.2;10 J m\, su$cient to
place it well within the microwave hearing range.
"
(5)
The pulse has a maximum intensity at the onset
cN
and decays exponentially with time. The temper-
ature as a function of time is shown in Fig. 3 and
where cN is the heat capacity at constant pressure shows that for a very short time, the temperature
and
is the density.
has an excursion well outside the normal operat-
It is useful to apply the analysis to a common
ing range of living cells. The temperature returns
mode of exposure used in studying the biological
quickly to normal after the pulse, and so the
e!ects of electromagnetic "elds and typical cul-
excursion would be di$cult to measure using
tures of cell medium contained in a #ask so that
normal thermometry.
the absorbing medium lies in the form of a slab
This large, transient increase in temperature
which is thin compared to its lateral dimensions.
may result in the alteration of protein conforma-
The one-dimensional heat di!usion equation is
tion. Indeed, this time scale is of the same order
applicable to this case and to a "rst approxima-
of magnitude as that of the dynamics of protein
tion the edge e!ects can be neglected. The bound-
conformational change demonstrated for the
ary conditions for the medium were chosen as
interaction of nuclease with DNA (Ha et al.,
follows. In the plane at x"1 the medium was
1999).

---

BIOLOGICAL EFFECTS OF ELECTROMAGNETIC FIELDS
297
in enzymes can lead to large functional e!ects
(Koshland, 1998). This model explains the power
window phenomenon, widely reported in the
electromagnetic bioe!ects literature, where biolo-
gical e!ects are seen at lower, but not at higher
power levels (e.g. French et al., 1997).
Conclusion
An athermal exposure to microwaves is one
which does not cause the temperature of a speci-
FIG. 3. The rate of heat delivery per unit volume (W m\)
as a function of time (dotted line) for a pulse of microwave
"ed volume of the subject averaged over a speci-
energy corresponding to an average SAR of 0.17 W kg\
"ed time period to exceed the normal value. Since
delivered to a sample of water with its top surface insulated
the unfolding time of proteins is much larger than
and its bottom surface in contact with a thermal reservoir at
the time to reach equilibrium between a protein
373C. The solid line shows the temperature at the surface of
molecule and its environment, cell exposures,
the sample.
whether thermal or athermal can be analysed by
the heat-di!usion equation. We "nd that signi"-
cant temperature transients which could cause
The Dependence of E4ects on Absorbed
protein conformational changes are induced by
Power Level
pulsed microwave exposures in realistic situ-
The hypothesis that the microwave exposure
ations. We proposed an explanation for apparent
can a!ect the protein conformation can be re-
nonlinear dose response relationships. At some
"ned to di!erentiate the e!ects at di!erent power
point, the power is su$cient to induce conforma-
levels.
tional change in some target proteins, but will be
At low power levels, a partial unfolding of spe-
insu$cient to induce the stress response, so a bio-
ci"c target protein(s) occurs, which will be insu$-
logical e!ect could occur unprotected by the
cient to induce the stress response, but su$cient
stress response. At higher power levels, the con-
to alter protein function. A biological e!ect (e.g.
formational change will be great enough to acti-
on cell proliferation) will be observed.
vate the stress response, reducing or nullifying
At higher power levels a more unfolded (molten
the e!ect by protecting against further protein
globule) conformation is induced. The stress re-
unfolding. At still higher power levels irreversible
sponse will be activated, protecting the protein,
damage will be done to a range of biological
and preventing an observable biological e!ect.
systems which the stress response is incapable of
At very high power levels, protein aggregation
preventing.
and precipitation occurs, and despite the acti-
vation of the entire stress response, a catastro-
The authors wish to acknowledge the role of Associ-
phic biological e!ect (e.g. cell death) will be
ate Prof. John Carver of the University of Wollongong
observed.
in the formation of the ideas underlying this work.
The above hypotheses explain many of the
One of us (R.A.L.) thanks the National Health and
apparently inconsistent observations in the sci-
Medical Research Council for "nancial support.
enti"c literature regarding microwave exposure.
As there would be multiple thresholds at a num-
REFERENCES
ber of power levels, there would be a nonlinear
dose-response e
BLACKMAN, S. & POPOLI, R. (1999). Design and Analysis of
!ect. The threshold power levels
Modern ¹racking Systems. P56 Boston: Artech House.
for biological e!ects depend on the degree of
BYUS, C. V., LUNDAK, R. L., FLETCHER, R. M. & ADEY, W.
protein unfolding needed to elicit a conforma-
R. (1984). Alterations in protein kinase activity following
tional change in a key protein molecule or a
exposure of cultured human lymphocytes to modulated
partial or complete heat shock response. As
microwave "elds. Bioelectromagnetics 5, 341}351.
CAMERON, I. L., KANAL, K. M., KREENER, C. R. & FULLER-
discussed above, small conformational changes
TON, G. D. (1997). A mechanistic view of the non-ideal

---

298
J. A. LAURENCE E¹ A¸.
osmotic and motional behavior of intracellular water. Cell
KOSHLAND, D. E. (1998). Conformational changes: how
Biol. Intl. 21, 99}113.
small is big enough? Natl. Med. 4, 1110}1114.
CHATTOPADHYAY, S. K., TOEWS, K. A., BUTT, S., BARLETT,
LINDNER, R. A., KAPUR, A. & CARVER, J. A. (1997). The
R. & BROWN, H. D. (1997). Reverse-micelle model: pH,
interaction of the molecular chaperone, -crystallin, with
electromagnetic "eld and inhibitor enzyme interaction.
molten globule states of bovine
-lactalbumin. J. Biol.
Cancer Biochem. Biophys. 15, 245}255.
Chem. 272, 27 722}27 729.
CHRONISTER, E. L. & CROWELL, R. A. (1991). Vibrational
MARKS, N. A. (1997). Evidence for subpicenoid thermal
coherence decay and population relaxation in weakly dis-
spikes in the formation of tetrahedral amorphous carbon.
ordered ionic molecular crystals. J. Phys. Chem. 95,
Phys. Rev. B 56, 2441.
2420}2426.
NESHEV, N. N. & KIRILOVA, E. I. (1996). Environmental-
DANIELLS, C., DUCE, I., THOMAS, D., SEWELL, P., TATTER-
health aspects of pulse-modulated microwaves. Rev. En-
SALL, J. & DE POMERAI, D. (1998). Transgenic nematodes
viron. Health 11, 85}88.
as biomonitors of microwave-induced stress. Mut. Res.
NOGLTING, B. (1995). Distribution of temperature in globular
399, 55}64.
molecules, cells, or droplets in temperature jump, sound
DONNELLAN, M., MCKENZIE, D. R. & FRENCH, P. W.
velocity, and pulsed LASER experiments. J. Phys. Chem. 102.
(1997). E!ects of exposure to electromagnetic radiation at
PEREZ, N., SUGAR, J., CHARYA, S., JOHNSON, G., MERRIL,
835 MHz on growth, morphology and secretory character-
C., BIERER, L., PERL, D., HAROUTUNIAN, V. & WALLACE,
istics of a mast cell analogue, RBL-2H3. Cell Biol. Intl. 21,
W. (1991). Increased synthesis and accumulation of heat
427}439.
shock 70 proteins in Alzheimer's disease. Brain Res. Mol.
FRENCH, P. W., DONNELLAN, M. & MCKENZIE, D. R.
Brain Res. 11, 249}254.
(1997). Electromagnetic radiation at 835 MHz changes
PORCELLI, M., CACCIAPUOTI, G., FUSCO, S., MASSA, R.,
the morphology and inhibits proliferation of a human
D'AMBROSIO, G., BERTOLDO, C., DE ROSA, M. & ZAPPIA,
astrocytoma cell line. Bioelectrochem. Bioenerg. 43,
V. (1997). Non-thermal e!ects of microwaves on proteins:
13}18.
thermophilic enzymes as model system. FEBS ¸ett. 402,
FREY, A. H. & MESSENGER, R. (1972). Psychophysical data
102}106.
on the RF hearing e!ect. 7th Annual Microwave Power
RAREY, M., KRAMER, B. & LENGAUER, T. (1999). The par-
Symposium, International Microwave Power Institute, 25,
ticle concept: Placing discrete water molecules during pro-
New >ork, ;SA.
tein-ligand docking predictions. PRO¹EINS: Struct.
GIBSON, J. D. (ed.) (1996). ¹he Mobile Communications
Funct. Genetics 34, 17}28.
Handbook. Boca Raton: CRC Press, New York: IEEE
RENKAWEK, K., BOSMAN, G. J. & GAESTEL, M. (1993). In-
Press.
creased expression of heat-shock protein 27 kDa in
HA, T., TING, Y., LIANG, J., CALDWELL, W. B., DENIZ, A. A.,
Alzheimer disease: a preliminary study. Neuroreport 5, 14}26.
CHEMLA, D. S., SCHULTZ, P. G. & WEISS, S. (1999). Single-
REPACHOLI, M., BASTEN, A., GEBSKI, V., NOONAN,
molecule #uorescence spectroscopy of enzyme conforma-
D., FINNIE, J., HARRIS, A. W. (1997). Lymphomas in Em-
tional dynamics and cleavage mechanism. Proc. Natl.
Pim1 transgenic mice exposed to pulsed 900 MHz electro-
Acad. Sci. ;.S.A. 96, 893}898.
magnetic "elds. Rad. Res. 147, 631}640.
HAGAN, S. J., HOFRICHTER, J., SZABO, A. & EATON, W. A.
SCHREIBER, G. & FERSHT, A. R. (1993). The refolding of cis-
(1996). Di!usion-limited contact formation in unfolded
and trans-peptidylprolyl isomers of barstar. Biochemistry
cytochrome c: Estimating the maximum rate of protein
32, 11 195}11 203.
folding. Proc. Natl. Acad. Sci. ;.S.A. 93, 11 615}11 617.
TAYLOR, L. S. (1981). The mechanisms of athermal micro-
INTERNATIONAL COMMISSION ON NON-IONISING RADIA-
wave biological e!ects. Bioelectromagnetics 2, 259}267.
TION (ICNIRP) (1998). Guidelines for limiting exposure to
WU, X. & SUNG, S. (1998). Constraint dynamics algorithm
time-varying electric, magnetic, and electromagnetic "elds
for simulation of semi#exible macromolecules. J. Comp.
(up to 300 GHz). Health Phys. 74, 494}522.
Chem. 19, 1555}1566.

---

Document Outline

• Introduction
  • FIGURE 1
• A Practical Definition of Thermal and Athermal Exposures
• Characteristic Sizes and Times for Important Cell Processes
  • FIGURE 2
• A Mechanism for Microwave Radiation to Act as a Cell Stressor
• Modelling Pulsed Exposures with the Heat Diffusion Equation
  • FIGURE 3
• The Dependence of Effects on Absorbed Power Level
• REFERENCES

---