A Puzzle of the Effect of Magnetic Field on Biological Cells

}I1~~

Life Science Journal, 2 (1 ) ,2005, Teng, A Puzzle of the Effect of Magnetic Field on Biological Cells

MINlREVIEW .I.

A Puzzle of the Effect of Magnetic Field on Biological Cells~

jj

tic human in tenns of average and maximum elec-tric field intensities. The calculation revealed the

induced dosimetry amount for various major organsupon exposing to extremely low-frequency electro-magnetic fields. The whole body induced maximumcurrent density is about 21 micro amp per metersquare from a 50 Hz to a 60 Hz source frequencyand from a 1 amp per meter to a' O. 1 Gauss sourcestrength. Comparatively, the cell culture dosimetryfor low frequency magnetic fields study by Hart(Hart, 1996) has shown the induced current densi-

ty averagely about O. 9 micro amp per meter squarefrom a 60 Hz source of frequency and a 1 Gausssource of field strength. Accordingly, the inducedeffect between the human body and the cells inculture clarified the interaction mechanisms in-volved could be different. The mechanisms where-

by ELF electromagnetic field stimulates changes inbiological functions of cells in culture may notguarantee that would cause the same biological ef-fect in vivo. Many researchers therefore switched

their interest to the biological effect of very highfrequency band of wireless communication protocoljust because of the support funding being hard tocontinuous for ELF biological effect study.

.,r)J

JI

J

jJJJ

I

,

1IJ

jJJJ

j

JJI

Hsien Chiao Teng

Department of Electrical Engineering, Chinese Military Academy,

Fengshan,Kaohsiung, Taiwan 830, ROC;scteng@cc. cma. edu. tw

Abstract: The mechanism for interactions of magnetic field, particularly in extremely low frequency with biologicalcells is still puzzle to the scientific researchers. Our investigations have guided to the speculation of roles of the mag-netic field for the cell-cell communication in physiological responses. To explain the complexity, the experimentalevidences and observations and their associated theories have been included in this review. The result of a disruptionof the homeostatic regulation of the cells responding to a specific strength and frequency of magnetic field can be asthe signal that triggers signal transduction to modulate the cell-cell communication. This review leads into thestochastic systems in cell, for instance, ion channels on cell membrane, providing a basis for signal amplification todisrupt the cell homeostatic regulation. [Life ScienceJournal. 2005; 2(1) :16 - 21] (ISSN: 1097- 8135).

Keywords: biological effect; cell; magnetic field

1 Introduction

.I

Biological cells have been shown to respondlow frequencies electromagnetic fields as well as inchemical and biochemical reactions ( Dobson,1996). Puzzle of the energy of the fields is too lowto the noise in cellular level to produce observableeffects (Adair, 1991; 1992; 1994). Therefore, thephysical mechanism of primary interaction betweenthe magnetic fields and the biological target sites,such as electrical charge in motion, molecules struc-ture with magnetic moments and the application ofFaraday law, generating local electrical current by avarying magnetic field has mainly concluded onlyamplification mechanisms can solve the puzzle. Aswe know, the interaction should be very weak atfield strength less than 10 Gauss and the frequencybelow 100 Hz. If considering the stochastic reso-nance model, the amplification of weak electromag-netic interaction signals can be modulated by exter-nal fields (J ung, 1993). By considering ferromag-netic transduction model, the minimum appliedmagnetic fields would produce a torque on a bio-genic magnetite particle coupling via the cytoskele-ton to the ion gates. The deformation of the cellmembrane and the closing of the ion gates wouldoccur quickly enough to compensate for the forcedopening of the gate as long as the frequency of theforcing field was below 100 Hz, regardless of thestrength of the applied field. Dawson et al. (Daw-son, 1996) used scalar potential finite differencecode for low frequency electromagnetic computa-tions and to model induction in anatomically realis-

http://www.sciencepub. org

JI~

-J

2 The ELF Biological Effect

~I

-~I

-,.I

,.~

~~

Jj

iThe concern of the exposure to extremely low

frequency (ELF) electromagnetic fields may pre-sent a health hazard to workers and the public. Thecontroversial and contradictory finding in the scien-tific research, especially from epidemiological stud-ies is puzzle (EPRI, 1994). Research of ELF elec-

-...

JI

...

J

JJ

JI

. 16 . editor@sciencepub. net

IrI.

L.

llL

Life Science Journal, 2 (1),2005, Teng, A Puzzle of the Effect of Magnetic Field on Biological Cells

tromagnetic fields interaction caused biological ef-feet in biological system includes experimental in-vestigations on both in vivo and in vitro. In pre-sent, the researchers have been interested in bothto the electric fields and magnetic fields as well as i-dentifying a possible mechanism for ELF acting as acancer initiator, promoter or co-promoter. On theother hand, epidemiological studies have shownvery weak connection between ELF exposure andleukemia, brain cancers, breast cancer and lungcancer. Almost all related research studies were in

flaws, for instance, numbers of cases were too lowto look at cancer subtypes, lack of specific expo-sure, lack of reliability of data, lack of statisticalpower and lack of control for repeaters. There is noreliable supporting data for an association betweenELF exposure and cancer risk in the public. Thereis no conclusive evidence so far from the epidemio-logical evidence that electric or magnetic fieldscause a risk of cancer. The researcher only holds apossibility of the risk of cancer of ELF electromag-netic fields occupational exposures. Oppositely, ex-perimental investigations with cellular systems haveshown that electromagnetic fields can interact withbiological systems. Several researchers have con-firmed cellular effects involving the movement ofcalcium ions through cellular membranes underELF interaction (Fewtrell, 1994). The signifi-cance of this effect as it relates to possible adversehealth outcomes is still not understood. Direct ef-

fects on significant cellular molecules, such asDNA, have not been observed. No direct muta-genic or carcinogenic effects on animals have beenobserved. Current research has shifted to focusingon the role of ELF (particularly magnetic fields) asa tumor promoter or co-promoter. However, it isstill no effects were observed on mice exposed with-out the chemical promoter yet. The overall viewobtained from the research literature indicates that

while some biological effects of exposure to ELFelectric and magnetic fields occur, there are no re-sulting adverse health effects from these exposures.In the United States, a large number of researchpapers and overview reports have been produced a-long with numerous conferences over the past 15years. Unfortunately, the findings remain contro-versial and contradictory. There is insufficient datato determine if a cause and effect relationship ex-ists. The National Council on Radiation Protection

and Measurements (NCRP) in Bethesda, Mary-land, set up a committee chaired by Dr. W. RossAdey to review the possible health effects of ELF.The National Academy of Sciences committee,chaired by Dr. Charles Stevens, Salk Institute,California released a report in 1996 concluding no

~

\.L

[..

L.

tI

l

..I

~

l....'-

llL..

~...~

L..

'L

l

...

l

llI..

ltl-

~...

~\..lI10I

http://www.sciencepub.org

to

clear, convincing evidence exists to show that resi-dential exposures to electric and magnetic fields(EMFs) are a threat to human health. The UnitedKingdom's National Radiological Protection Board(NRPB) established an Advisory Group on Non-Ionizing Radiation in 1990 to review the scientificevidence and determine the extent to which this ev-idence suggests possible health risks. The Interna-tional Non-Ionizing Radiation Committee (INIRC)of the International Radiation Protection Associa-tion (IRPA) in cooperation with the EnvironmentalHealth Division of the World Health Organization(WHO) developed recommendations for 50/60 Hzelectric and magnetic field exposure limits. At the8th Congress of the IRPA in May 1992, the IRPAestablished a new independent scientific organiza-tion, the International Commission on Non-Ioniz-ing Radiation Protection (ICNIRP) as a continua-tion of the former IRPA/INIRC. In April 1998,ICNIRP published guidelines for limiting electro-magnetic field exposures for frequencies up to 300GHz, including 50/60 Hz.

3 Energy Transduction in Cell

The mitochondria in cell contain the series of

catalysts known as the respiratory chain collecttransport reducing equivalents to react with oxygenfor forming water. The reducing equivalents,-H orelectrons, are made from oxidation of carbohy-drate, fatty acids and amino acids. The cell systemcouples respiration to generate the high energy in-termediate A TP, termed oxidative phosphorylationinvolved NADH, Succinate, Ubiquinol, Ferrocy-tochrome and ATP synthase five protein lipid en-zyme complexes. Three different mechanisms in-cluding chemical coupling hypothesis, the confor-mational coupling hypothesis and the chemiosmotichypothesis have been proposed to explain the ener-gy transfer between electron transport and ATPsynthesis (Menendez, 1996). Accordingly, elec-tron transport along the respiratory chain can be thesource of electromagnetic field to exert forces on theproton. Through these forces, electromagnetic cou-pling hypothesis is proposed to describing the pro-tons are moved from the mitochondrial matrix to

the exterior. In the mean time, the protons movedby the field toward the inner membrane passthrough its protein components plays the role ofproton channel. Just as if in a motor, ELF triggersthe electrical energy, transfer of electrons along therespiratory chain, to produce proton translocation,which is mechanical energy, be used in cell system.


channel and be transported to plasma membrane ofthe cell. Before pairing process, hemi channels areclosed to avoid leakage of cellular contents and en-try of extra-cellular materials. During the pairing ofconnexons and aggregation into plaques at the plas-ma membrane, connexin 43 is phosphorylated atleast twice and connexons are attracted to those lo-cated on the adjacent cells. Two connexons join inan end-to-end manner to form a complete channel.The channels aggregate into large gap junctionplaques open to connect two cells for cell-to-cellcommunication and is called gap junctional intracel-lular communication (GJIC), which can be modu-lated by environmental factors, such as drugs, X-ray, electromagnetic fields etc. Since the functionof the GJIC, cultured cells coupled in vitro exceptthe stem cells and cancer cells (Trosko, 1991;2001) . Furthermore, in basic signal transduction,we want to know what the gap junction growthregulatory signal within cells is. The cAMP is avery important gap junction signal molecule. Thissignal molecule can pass between cells through gapjunction channels and affects cell growth. The lev-els of the cAMP can oscillate within the cell andgenerates oscillation in growth control. The ampli-tude of the oscillations would be dampened byGJIc. However, cAMP is not the only signalmolecule that may affect GJIc. Increasing drugpenetration and dispersal in tumors would increaseGJIc. Gap junction channels in cell membrane cancreate a stochastic two-state system, opening orclosing of the channels, to amplify weak signalcaused by ELF fields.

}

JJJI

~II

J

1

)

J

1

)

1

JJJ

J

J)~

jJ

J

.~I

JI

.J

Life Science Journal ,2(1 ) ,2005, Teng ,A Puzzle of the Effect of Magnetic Field on Biological Cells

4 Stochastic Resonance

~

Several resonance theories, for instance, ioncryotron resonance (Liboff, 1991), parametric res-onance were proposed to describe the biological ef-fect caused by ELF fields. For the problem of syn-chronization of the relaxation time and the noise, itis hard to find the evidence of resonance in cell sys-tem. Nevertheless, because of ionic channels of cellmembranes are elemental molecular switches forchannel states, open or close, in spite of the smallsize of the channel, high resolution ion current de-tection allows us to watch electrical activity of asingle ionic channel isolated within a micrometerpatch of a cell membrane. The state-transition be-havior between the open and close states of an ionchannel characterizes many biological processes.The experimental results suggested that the poten-tial oscillation could be driven by the oscillation inthe intracellular concentrations of cyclic AMP andcalcium in two-state model system (channel closingand opening). The signal-to-noise ratio increaseswhen the state-transition rate of membrane channelinfluenced by the frequency response of the intra-cellular sensing system. An applied ELF (electricor magnetic field) field is an important factor toperturb the rate of state-transition of membranechannel in opening or closing. Jung (1993) has de-veloped stochastic resonance driven process for mul-tichannel systems and shown the optimal choice ofparameters can lead to signal amplification. Fromthe stochastic resonance theory, time average of thetotal current through a set of N channels can becalculated. To access the magnitude of amplifica-tion of the responding signal caused by externalELF field in cell, the membrane with N voltagegated channels and biochemical oscillator (calciumoscillation) must be considered. Because of thismodulation, the external ELF field signal is trans-formed into a periodic component in the ion currentacross the membrane. As we have focused in thestudy on the primary mechanisms that a biologicalcell can use to amplify weak external influences, asystem of identical ion channels embedded in amembrane and synchronously modulated can signif-icantly amplify the original signal. The amplifica-tion of the signal to noise ratio is proportional to thesquare root of the number of channels modulatedand inversely proportional to the square root of thesum of the average channel relaxation times.

5 Role of Gap Junctions

In cell, six connexin 43 subunits oligomerze inthe Golgi apparatus into a connexon, called hemi


- - -

,I

6JJj

1-I

Physics Correlation between GJIC and ELFMagnetic Field

The diffusive current equation for connexin 43channels can be written as J

Ij

J..

_J

j

i

< I> = f,kP(k)k=!

where probability P(k) indicates total m channelsis taken into account for opening k channels fromall cell-to-cell communications on the surface of thecell mono layer. Therefore,,

P( k ) = m : (PO)k ( pe ) m - k (2 )k!(m-k)!

dpo = repe - rOpO (3)dt

dpe = rOpo- repe (4)dt

where re is the rate of changing from c-state to 0-

state and rO is the rate of changing from o-state toc-state of the connexin 43 channels activating total-

lyon the cells mono layer surface. Generally, rO

(1)

-\;

....

J

jJI


"J

[ll

t

!

lttlL

llllll~~L.

lL

l

llllllll~

lllll

.J.

Life ScienceJournal, 2 (1 ) ,2005, Teng, A Puzzle of the Effect of Magnetic Field on BiologicalCeUs

does not have to be same with rCsince the life timesof the o-state and c-state may vary. To clarify thephysical meaning, we further assume the currentthrough an open channel as i. The diffusive currentcaused by GnC channels can be rewritten as

< I> = miP~ (5)where P~ is the modulated probability for o-state byexternal ELF field signal. According to theory ofJung (1993), the power spectral component origi-nated from the signal is given by

(mi)2 00 I ISk = ~ {;1 Cq 8(w - qWk)

Cq is the Fourier expansion coefficients of P~.In comparison with equations (3) and (6),

the signal-to-noise ratio (SNR) of the characteristicfrequency of the cell system can be depicted (Gal-vanovskis, 1997).

SNR =I

signal amplitude1

2

background amplitude

=A2I Jm ~ rOrc 1

2 (7)t:..w(rO + rC)

where m is the number of channels, A is the ampli-tude and t:..wis the bandwidth of the external ELFfield signal.

7 Specific Inhibit and Promote of GJIC

Many researchers have developed new tech-niques to inhibit or promote the expression of a tar-get gene in culture cells and animals. ELF magneticfield treatment is one of the main factors that scien-tists are interested. We proposed a new way to sep-arate the background and signal and revealed theevidence of existing stochastic resonance systemburied in biological cells by GnC essay under ELFmagnetic fields. Specific inhibit of GnC withinmouse osteoblast cells in culture under the exposureof ELF magnetic field (Teng, 2002) depicted sev-eral possibilities, blockage of connexin gene expres-sion, connexin gene knockout and transfection ofdefective connexin genes. Additionally, specific en-hancement of GnC within mouse osteoblast cellsunder the exposure of different doses of ELF mag-netic field (Hart, 1996) can be transfected of func-tional connexin genes.

8 SNR Spectrum

~l

ltll

By using of the probe of Gauss-meter, cells-in-duced magnetic fluctuation can be shown as

IB;I= IBl ,Bl"" ,Bloool (8)Equation (8) contains the cellular response signal ofthe reaction to the external ELF magnetic field(Teng, 2003). The sampling time was 0.0005

http://www.sciencepub .org

(6)

second. The probe was located at the distance 10- 4

m perpendicularly to the center of single layer ofthe cell surface in culture dish. The Gauss-meterwas manufactured by F. W. Bell Company (seriesof 9550) in Florida of USA. Oscilloscope was man-ufactured by Agilent Company (54621A) and canbe used to convert IB;I to voltage sequenceIv;1as

I V;I = I VI, V1,"', v1oool (9)Matlab and Fortran programming were used forpower density spectrum analysis of these voltage se-quences. IBi I for the first control was taken 2000

times per second at the distance 10- 4 m perpendic-

ularly to the culture dish with only medium in it.Ceo-field control IBi I was taken with the same

sample rate at the distance 10- 4 m perpendicular tothe empty culture dish recording local geomagneticfield fluctuation. The corresponding IVi I andIVi I can be obtained by the same way as IVi Ipreviously. Furthermore, trial signals,

ni ( n) = Ai x sin( Win) ,lHz:(;wi:(;60Hz,where signal amplitude Ai = F X V rnax' F is theadjustablefraction factor and Vmax is such, V rnax=max( IV;I), as to the maximum value of the se-quence IV; I.By taking into consideration of signalamplitudes at F values, for instance,F=O. 7,0.4,and O. 03 respectively, the corresponding ampli-tudes would be

AO.7=0. 7X V max'AO.4=0.4x Vrnax'AO.O3=0.03x Vmax

for a given trial signal at ELF Wi(1 Hz:(; Wi:(;60Hz). We computed autocorrelation function ofIV;:t ni ( n ) I and its Fourier transforms to obtaintheir corresponding signal-to-ratio ratio. The SNRspectrum, then, for IV;:t ni ( n ) Iat frequencycould be simply a second order equation as

aX(Swi(F»2+bxSwi(F)+c=0 (10)Accordingly, substituting the SNR at different

F (different amplitudes at frequency) into the e-quation, we can solve unknowns a, band c. If c-value is much bigger than zero, then, the SNR ofthe intrinsic signal peaked at Wi is detected (Teng,2005). The value of c is about O.07.

In the paper by (Galvanovskis, 1997), theSNR could be written as

SWi(F) =(FxVmax)2XmXQX (2: )Wi r ° r C

when the life time of c-state and o-state equal to

10 - 6 second. Under optimal condition, the quality

factor Q = ::approximately equals to 100 at 60

~~


jLife Science Journal, 2 (1 ) ,2005, Teng, A Puzzle of the Effect of Magnetic Field on Biological C€lls

vealed the possibility of signal transduction path-ways perturbed by ELF magnetic fields at differentcharacteristic frequencies and the stochastic reso-nance frequencies within' the cells. The result of amodulation of the GnC within the cells respondingto a specific strength and frequency of magneticfield can be as the signal that triggers signal trans-duction in cell.

JjJJ

i}Jj1

jJJ

JJJ)~JJ

J

jJ

Hz with bandwidth ~w = 0.6 Hz and F = O. 6.The numbers of GnC channels are taken 1000 percell (Galvanovskisz, 1997). The SNR value is fitto the value calculated from SNR spectrum suggest-ed by Teng (Teng, 2005).

9 Discussion

We examined the role of the ELF magneticfield in the biological cells system. It is hard toimagine how the forces of evolution could have ledto biological mechanisms that respond to ELF mag-netic field in cells, because low frequency fieldshave not been present during virtually all of evolu-tion. So far, there is still no direct evidence thatELF magnetic field can affect the physiological end-points within cells in culture. However, enoughevidences have shown the biological effect that thecell responded to the interaction of externally ap-plied ELF magnetic field (Kaiser, 1996). SinceGnC affiliates with many physiological endpoints,changing of the characteristics of GnC causes thebiological effect within cells in culture. A confluentcell culture in a vertical magnetic field could receiveseveral different exposures. If the cells are nottightly jojnted to each other and to the walls, thenthe system is homogeneous. Different cell typescould e~perience very different induced current andelectric field distributions under the same magneticfield exposure in similar dishes. Gap junctions jointthe cells and are open to allow current paths con-nect the cell interiors. They produce the currentpaths throughout the interiors of the cells in the cellculture. Most of the time, two dimensional model

is oversimplify the current density distribution thatis actually experienced by cells in a culture dish.Value for the induced current density, in effect, isthe top surface perpendicular to the applied mag-netic field. When the magnetic field is applied a-long the surface of the cells, perpendicular to thenonnal direction of the surface, the homogeneoussystem would be inapplicable around the centralplanes, but it could be used away from the centralregion.

10 Conclusion

~

If a cell culture is exposed to a vertical mag-netic field for a relatively long period, the exposurereceived by the cells may change during the courseof the experiment. Trusko et al. (Trosko, 2001;Upham, 1998) originated the perfusion of the dyeproduced by the function of the GnC in cell cultureto study the cell physiology. T eng et al. first pro-posed the signal-to-noise SNR spectrum calculationof the near magnetic field in 2002. They have re-


Correspondenceto:Hsien Chiao TengDepartment of Electrical EngineeringChinese Military AcademyFengshan, Taiwan 830, ROCTelephone: 011886-7747-9510 ext 134Email: [email protected]

References1. Adair RK. Constraints on biologicaleffects of weak ex-

tremely low frequencyelectromagneticfield. Physical RevA 1991;43:1039-48.

2. Adair RK. Criticism of Lednev' s mechanism for the in-fluence of weak magnetic fields on biological systems.Bioelctromagneics 1992; 13 :231 - 5 .

3. Adair RK. Biological responses to weak 60 Hz electricaland magnetic fields must vary as the square of the fieldstrength. Proc Natl Acad Sci 1994;91 :9422 - 5.

4. Dobson JP, Grassi P. Magnetic properties of human hip-pocampal tissue: evidence for biogenic magnetite in thehumanbrain. BrainResBull1996;39:255- 9.

5. Dawson T, Stuchly MA. Analytic ~alidation of a threedimensional scalar potential finite difference code for lowfrequency magnetic induction. Appl Comput ElrctonmagnSoc] 1996; 11 (3) : 63 -71.

6. EPRI. Biology and electric and magnetic fields: biologicalmechanisms of interaction. Printed by Gradint Corpora-tion, Cambridg, Massachusetts. 1994.

7. Fewtrell C. Calcium oscillation in non-excitable cells, inAnn Rev of Physiology, Vol. 55, Editor Hoffman JF,Annual Review, Inc. Palo Alto. 1994; 55: 427 - 54.

8 .Bruce N. Biomedical signal processing and signal model-ing. Wiley-Interscience Publication, John Wiley & Sons,Inc. New York, ISBN 0-471-34540-7. 2001.

9. Glaser R, Michalsky M, Schamek R. Is the ci + trans-

port of human erythrocytes influenced by ELF- and MF-electromagnetic fields? Bioelectrochemistry and Bioener-getics 1998 ;47: 311 - 8.

10. Hart F. Cell culture dosimetry for low frequency magnet-ic fields. Bioelectromagnetics1996; 17:48 - 57.

11. J ung P. Periodically driven stochastic systems. Phys Rep(Phys Lett) 1993 ;234: 175 - 95.

12. Kaiser F. External signals and internal oscillation dynam-ics: biophysical aspects and modeling approaches for in-teractions of weak electromagnetic fields at the cellularlevel. Bioelectrochemistry and Bioenergetics 1996; 41 : 3-18.

13. Liboff AR, Parkinson WC. Search for ion cyclotron reso-nance in an Na+ transport system. Bioelectromagnetics1991;12:77-83.

14. Menendez RG. An electromagnetic coupling hypothesis

JI

."

I,...4J

-}

J

1~

JI

-}

J

L"!,

JJ

1

jJ


l

~

llL

~

tr.

ltlLlllll~LL..ltlL

L

llll,lllLL

lll

..J.

Life Science Journal, 2 ( 1 ) ,2005, Teng, A Puzzle of the Effect of Magnetic Field on Biological Cells

to explain the proton translocation mechanism in mito-chondria, bacteria, and chloroplasts. Medical Hypotheses1996 ;47 : 179 - 82.

15. Takb H, Shiga T, Kato M, Masada E. Biological andHealth Effects from E~pcsure to Power Line FrequencyElectromagnetic Fields-confinuation of Absence of AnyEffects at Environmental Field Strengths. IOS Press,Ohmsha, ISBN 4-274-90402-4C3047. 1999.

16. Teng HC, Cherng S, Trosko JE, Chang CC, Upham BL.Sensitivity of Osteoblast Cells to Inhibition of Gap Junc-tional Intercellular C.ommunication By ELF-EMF at 14Hz (The 24th Annual Meeting of the BioelectromagneticsSociety). 2002.

17. Teng HC, Cherng S. Mouse osteoblast cell sensitivity tothe AC magnetic field at 14 Hz. Nature and Science

~. lllll


2003;l (1) :27 - 31.18. Te!1g HC. The molecular biological application of the

theory of stochaStic resonance: the cellular response to theELF AC magnetic field. Nature and Science 2005; 3 ( 1) :37 ~ 43.

19. Trosko JE,Chang CC,Madhukar BV. Modulation of In-tercellular Communication during Radiation and ChemicalCarcinogenesis. Radiation Research 1990; 123 :241 - 51.

20. Trosko JE, Chang CC. Role of stem cells and gap junc-tional intercellular communication in human carcinogen-sis. Radiation Research 20tH; 155: 175 - 80.

21. Upham BL,Deocampo ND, Wurl B, Trosko JE. Inhibitionof Gap Junctional Intracellular Communication by perfluo-rinated fatty acids is dependent on the chain length of thefluorinated taiL Int J Cancer 1998;78: 491 - 5.

'"


Documents

A Puzzle of the Effect of Magnetic Field on Biological Cells