Bioelectromagnetics 29:380^386 (2008)

NewTheoretical Treatment of Ion Resonance Phenomena

G. Vincze,1 A. Szasz,1 and A.R. Liboff2*
2

Biotechnics Department, Szent Istvan University, Budapest, Hungary Center for Molecular Biology and Biotechnology, Florida Atlantic University, Boca Raton, Florida 33431

Despite experimental evidence supporting ICR-like interactions in biological systems, to date there is no reasonable theoretical explanation for this phenomenon. The parametric resonance approach introduced by Lednev has enjoyed limited success in predicting the response as a function of the ratio of AC magnetic intensity to that of the DC eld, explaining the results in terms of magnetically induced changes in the transition probability of calcium binding states. In the present work, we derive an expression for the velocity of a damped ion with arbitrary q/m under the inuence of the Lorentz force. Series solutions to the differential equations reveal transient responses as well as resonance-like terms. One fascinating result is that the expressions for ionic drift velocity include a somewhat similar Bessel function dependence as was previously obtained for the transition probability in parametric resonance. However, in the present work, not only is there an explicit effect due to damping, but the previous Bessel dependence now occurs as a subset of a more general solution, including not only the magnetic eld AC/DC ratio as an independent variable, but also the ratio of the cyclotronic frequency O to the applied AC frequency o. In effect, this removes the necessity to explain the ICR interaction as stemming from ionprotein binding sites. We hypothesize that the selectively enhanced drift velocity predicted in this model can explain ICR-like phenomena as resulting from increased interaction probabilities in the vicinity of ion channel gates. Bioelectromagnetics 29:380386, 2008. 2008 Wiley-Liss, Inc. Key words: ion cyclotron resonance; parametric resonance; ion mobility; theoretical ICR models

INTRODUCTION There is good evidence [Liboff, 2005] that lowfrequency magnetic elds can interact with biological systems when the elds are tuned to the cyclotron resonance frequencies of critical ions. For the most part the effects of sinusoidal elds within the 3 300 ELF range have been studied in this regard. Most remarkably, the elds that appear to be interactive are very weak, with intensities of the order 1100 mT, placing them well below the intensity thresholds that have been estimated as potentially interactive via Faraday induction [Weaver and Astumian, 1990; Adair, 1991]. The original explanation for these effects [Liboff, 1985; McLeod and Liboff, 1986] maintained that ion cyclotron resonance (ICR) was possible in ion membrane bound ion channels for weak magnetic elds, rst because of the convenient helical structure within channels, and second, the vanishingly small intra-luminal damping. However, this explanation was shown to be inconsistent [Liboff, 2003] with predicted transit times for low-frequency resonant ions moving through ion channels. These are on the order of a factor
2008 Wiley-Liss, Inc.

of 106 slower than well-established times obtained from measurements on ion channel permeabilities [Kuyucak et al., 2001]. Still another criticism of the original ICR explanation was that the orbits of cyclotronic motion, as derived from the kT-energy value, were substantially greater than the ion channel diameter. A number of alternative theoretical explanations have been proposed [Zhadin and Fessenko, 1990; Lednev, 1991; Male, 1992; Edmonds, 1993; Blanchard and Blackman, 1994; Zhadin, 1998; Binhi, 2000; Del
This work was rst presented at the 27th annual meeting of the Bioelectromagnetics Society, Dublin, June 2005. *Correspondence to: A.R. Liboff, Center for Molecular Biology and Biotechnology, Florida Atlantic University, Boca Raton, FL 33431. E-mail: arliboff@aol.com Received for review 26 April 2006; Final revision received 29 June 2007 DOI 10.1002/bem.20406 Published online 20 February 2008 in Wiley InterScience (www.interscience.wiley.com).

Model for Ion Resonance Phenomena

381

Giudice et al., 2002; Zhadin and Barnes, 2005]. The concept of parametric resonance, rst introduced by Lednev [1991] and later elaborated on by Blanchard and Blackman [1994] utilizes an approach originally developed in connection with the theory of atomic spectroscopy. One key advantage, intrinsic to parametric resonance but lacking in the earlier simple Lorentz models [McLeod and Liboff, 1986; Durney et al., 1988; Bianco and Chiabrera, 1992] is the predicted dependence on AC magnetic intensity. This appears in terms of the probability p(BAC) of transitions between adjacent binding states for calcium (or other metallic) ions to calcium binding proteins such as calmodulin. This probability can be written as pBAC 1n KJn ne 1

between collisions in a region carrying an electric eld E and a magnetic eld B. The magnetic eld includes both a static component B0 and one that is time varying, such that B B0 BAC eiot 2

If v(t) is the particles instantaneous velocity vector its equation of motion is m dv v m qE v B dt t 3

where e is the ratio of AC to DC magnetic intensities, K is a constant and Jn is the nth order Bessel function. The fact that these solutions are framed in terms of the dependence on AC intensity is of some signicance, since this provides a possible way to characterize the interaction strength of 50/60 Hz magnetic elds down to the 110 mT levels implicated in epidemiological studies [Wertheimer and Leeper, 1979]. Furthermore, the explicit dependence on intensity allows one to design experiments to test Equation (1). A number of observers [Shuvalova et al., 1991; Blackman et al., 1994; Trillo et al., 1996; Prato et al., 2000; BerdenZrimec and Jerman, 2001; Koch et al., 2003; Sarimov et al., 2005] have reported results that are in part supportive of Equation (1) or the similar formulation by Blanchard and Blackman [1994]. Nevertheless, objections have arisen based on fundamental energy constraints [Adair, 1992] inasmuch as Equation (1) describes the probability for sub-thermal transitions between a pair of deeply energetic adjacent binding states. Thus, even though there has been some limited success in using this parametric resonance expression to predict the outcomes of ICR experiments, the underlying molecular explanation, framed in terms of calcium binding, is still uncertain. In the present work, we reexamine the original approach [Liboff, 1985; McLeod and Liboff, 1986] that focused on the kinetics of ions under the inuence of the Lorentz force. We show below that a more complete analysis of this problem than was originally attempted yields a solution that is somewhat consistent with Equation (1), but is not encumbered by explanations related to calcium binding states. THEORETICAL TREATMENT We assume the motion of a particle characterized by charge q and mass m moving with a mean time t

where the right side of this equation is the Lorentz force. Equation (3), describing the motion of a single particle, can also be interpreted as the equation of motion for an average particle within an ensemble of identical particles that are all moving within the region permeated by E and B. In this case v plays the role of drift velocity for this ensemble. We choose E and B arbitrarily to be at right angles (not without the loss of generality) with E (Ex, 0, 0) and B (0, 0, Bz). Then Equation (3) can be rewritten m ddvtx m vtx qEx qBz vy dv v m dty m ty qBz vx m ddvtz m vtz 0 4

from which it is clear that the effect of E and B on the drift velocity is limited to the xy-plane. Accordingly, the drift velocity is a plane vector that is readily represented as the complex expression v vx i vy. Substituting this complex term for v into Equation (4) and making use of Equation (2), we obtain the complex differential equation dv 2bv G iO1 e cos otv dt where 1 b 2t O q m B0 BAC e B0 Ex GO B0   6 The methods by which detailed solutions to Equation (5) are obtained are shown in Appendix I. It is shown there that the velocity v contains a number of components, including transient terms that decay in time as well as periodic terms. The periodic terms include both resonance-like terms and nonresonant periodicities. Dening d as the ratio of cyclotronic frequency O to applied frequency o, and considering only resonant-like terms, the velocity is given in terms
Bioelectromagnetics

382

Vincze et al.

of the nth-order Bessel series. vG

1 X Jn edJn ed 2b iod n n1

For Bessel order n 1 at resonance such that d 1, the real part of the velocity reduces to v G 2 J e 2b 1 8

Equation (8) shows that at resonance the drift velocity is always positive and varies with a dependence on a Bessel function whose argument is the ratio of AC magnetic eld intensity to DC magnetic eld intensity. Further, there is an explicit dependence on damping as well as on the electric eld and the charge to mass ratio of the ion. When the frequency is not at resonance, such that d 6 1, the drift velocity is vG
2 J1 ed

2b o2 =2bd 12

In this case we see that the argument of the Bessel function is augmented by d, the ratio of the cyclotronic frequency O to the applied frequency o. Thus, we have obtained an explicit dependence for the behavior of the drift velocity not only in terms of the AC magnetic intensity but also in terms of the applied frequency. DISCUSSION Despite the fact that Equations (1) and (8) are derived using strikingly different physical assumptions and describe different physical variables, transition probability on the one hand, and drift velocity on the other, it is nevertheless instructive to compare the two results. In the nal analysis what matters is whether the physiological response to magnetic elds follows the functional dependence of variables such as q/m, BAC, and/or o as these are expressed in one theoretical model or another. The various experimental reports [Blackman et al., 1995; Koch et al., 2003; Sarimov et al., 2005] that have explored this question unfortunately were not designed to test the present model. Nevertheless, it is clear that our model lends itself to straightforward experimental trials. For example, examining Figure 2 one could design experiments around two different sets of points in the ed plane, one set corresponding to maxima and the other to minima. In effect this was attempted by Prato et al. [2000] along the line d 1 when testing the parametric resonance model. They examined the experimental biological response as a function of AC magnetic eld intensity with special
Bioelectromagnetics

consideration given to determining the argument of the Bessel function corresponding to the rst maximum, that is, when e, the ratio of AC magnetic intensity to DC magnetic intensity is equal to 1.84. We note that the same value of e is predicted whether one explores 2 the response in terms of J1(e) or in terms of J1 e (see Fig. 1). One unresolved difculty with parametric resonance is the arbitrary use of absolute values of the function J1(e), necessitated by the fact that J1(x) includes negative as well as positive terms. In the present model, the dependence is 2 in terms of J1 e which removes this problem (see Fig. 1). One essential aspect to Equation (8) is the explicit dependence on damping, indicated by the factor (1/2b). There is no dependence on damping in the parametric resonance model. The present work uses ion drift velocity (or similarly, ion mobility) as the variable of interest. This allows one to more directly involve the damping factor b as obtained, say, from experimentally derived resonance curves [Liboff, 2003]. On the other hand, an explanation based on ion kinematics immediately faces the same daunting problems as encountered in earlier attempts to use the Lorentz force [Liboff, 1985; McLeod and Liboff, 1986; Durney et al., 1988]. We can rephrase this problem in the present context. The mobility m is normally dened as the factor connecting the velocity of a charged particle to its surrounding electric eld, such that v mE. In this work the mobility as obtained from Equation (8) can be 2 written as m q=mJ1 et. This can be compared to the mobility of a charged particle in the absence of a magnetic eld, m (q/m)t [Bleaney and Beaney, 1976]. At rst glance it appears that the presence of the 2 magnetic eld adds the factor J1 e. However, this says nothing about the time between collisions t, which is the determining factor for damping. Therefore, whatever the effect of the magnetic eld on the mobility,

Fig. 1. Comparison of the two Bessel functions J1(e) (dashed line) and J2 1 e, both plotted against e.

Model for Ion Resonance Phenomena

383

it cannot be detected by the biological system. The original argument raised by Weaver and Astumian [1990] suggested that biological systems were unable to respond to electrical signals that were lost in thermal background noise. However, a good deal of experimental evidence [Liboff, 2005] has since accumulated showing that biosystems do indeed respond to magnetic signals, specically ICR signals, that in principle should not be recoverable from the thermal background. In addition, the nature of this response is often functionally dependent on jJ1(e)j (or 2 J1 e). The original discussions concerning thermal noise constraints [Weaver and Astumian, 1990; Adair, 1991] assumed that the biological system is passive, that is, contains no special qualities that might enable it to recognize or utilize E- and B-elds whose effects would ordinarily be lost in the thermal noise. Emphasis was placed on the noise developed in the lipid membrane, whose electrical properties are much better characterized than those of proteins, notwithstanding the fact that proteins are more likely to act as eld-sensitive interactive units. More recent attempts to explore this question [Kaune, 2002; Bier, 2005; Vincze et al., 2005] viewed the membrane as merely one component of a more inclusive biological complex. Vincze et al. [2005] suggested that heating of the extracellular matrix by external elds might conceivably lead to diffusion gradients having biological signicance, and Bier [2005] pointed out that the nonequilibrium nature of membrane proteins such as ion pumps may enable them to capture electromagnetic energy to perform work that is metabolically signicant. In the present work, Equation (8) indicates that the velocity of a charged particle is a direct function of not only the collision time but also the electric eld, a variable dependent on the local physical characteristics within the biological medium. Thus, we suggest that magnetic elds with intensities lower than the thresholds previously predicted as detectable using thermal energy constraints [Weaver and Astumian, 1990; Adair, 1991] may be detected in those selected regions of biological systems that enjoy sufciently large local electric eld intensities. We can estimate the order of magnitude of electric eld required to allow the biological system to detect a weak magnetic eld. If Equation (8) is rearranged it shows that the velocity is directly dependent on the product of electric eld and collision time: q 2 v e 10 Ex tJ1 m

Apart from the dimensionless Bessel factor, a particle moving at this speed has the kinetic energy mv2 q q  11 Ex t2 2 m 2 This can be compared to the thermal energy 3kT/2, which in SI units after inserting Boltzmanns constant k and the temperature T (310 K) is equal to 6.4 1021 J. Substituting for the charge and chargeto-mass ratio of the calcium ion, namely q 2 1.6 1019 C and q/m 4.8 106 C/kg, respectively, the kinetic energy equals the thermal energy when the product (Ex t) equals 9.8 105 volts/m/s. If one assumes a collision time t 1010 s, we nd that this requires that the electric eld intensity be approximately 106 V/m, within an order of magnitude of the electric eld usually associated with the cell plasma membrane, 107 V/m. This lends credence to the hypothesis that structures within or at the surface of cell membranes are the sites for ELF magnetic eld interactions. One additional difference in the present model is found in Equation (9) where the Bessel argument is not merely e, but instead the product ed. In other words, the dependence on frequency now assumes equal importance to that of the AC intensity. In Figure 2 we 2 plot J1 ed as a function of both e and d, over the ranges 04 and 16, respectively. These ranges are chosen arbitrarily to illustrate the profound changes in predicted response from that expected in parametric resonance when there is an added dependence on frequency. The rst Bessel maximum for the frequency

Fig. 2. The function J2 1 ed plotted over the ranges e 0 ^ 4 and d 1^ 6. [The color figure for this article is available online at www.interscience.wiley.com.] Bioelectromagnetics

384

Vincze et al.

ratio d 1 is clearly evident at e 1.84. Note that this maximum response to AC magnetic elds shifts drastically to lower AC intensities as the ratio d increases. In addition, the width of the resonance narrows considerably as one proceeds to lower AC magnetic eld intensities. Because the detection of increased values of velocity at given values of e and/or d are dependent on the level of damping, there will be a oor or threshold below which the response pattern shown in Figure 2 will not be detectable. This is illustrated in Figure 3 where we have arbitrarily set this threshold at a level that allows only the largest two maxima in velocity to result in a response. In this situation, where the ion drift velocity lies outside of the noise threshold, it is reasonable to think that the enhanced ionic velocity can result in greater interaction cross-sections with, say, the gating apparatus attached to ion channel structures. Our model also predicts a Bessel function dependence on frequency when the magnetic eld ratio e is kept the same. This is shown in Figure 4. CONCLUSIONS It is possible to formulate a model to explain the ICR response that has a functional dependence that is similar to the form seen in parametric resonance models. However, the assumptions underlying this model are not dependent on molecular transition probabilities but instead are connected to Lorentz force considerations. The model response that we have formulated differs from that of parametric resonance in that:

Fig. 4. Velocity as a function of the frequency ratio O/o when BAC B0. [The color figure for this article is available online at www.interscience.wiley.com.]

(a) damping is explicitly involved; (b) effectively, damping is inversely related to the strength of endogenous electric elds; (c) the predicted variations in response are always positive; (d) it is a function of both the AC frequency as well as the AC magnetic intensity; (e) the predicted response is in terms of ionic drift velocity rather than transition probability.

APPENDIX We seek the general solution to the homogeneous equation dv 2b iO1 e cos otv 0 dt 12

Let y(t) be the general solution satisfying the condition y(0) C, such that Y (1/C)y, and v y Y t Zt
0

Y 1 t0 Gt0 dt0

13

Simple substitution shows that Equation (12) is satised by the periodic function Y e2biOt eied sin ot 14

If this is substituted into Equation (13), we obtain vt yt Ge

Fig. 3. Possible effect of noise threshold on response. [The color figure for this article is available online at www.interscience. wiley.com.] Bioelectromagnetics
2biot ied sin ot

Zt
0

e2biOt eied sin ot dt0 15

where it is assumed that G is not a function of time.

Model for Ion Resonance Phenomena

385

We transform Equation (15), making use of the JacobiAnger transformation eied sin ot
1 X n1

vG
1 X

1 X Jn edJn ed G 2b iO no n1

Jn edein ot

16

1mn Jm edJn ed inmot km;n e 2b iO no m;n1

22

where Jn(ed) is the nth order Bessel function of the rst kind, and also making use of the Bessel function identity Jn x 1n Jn x Consequently Zt
0

17

where km,n is the Kronecker delta, equal to 1 when m 6 n and equal to zero when m n. Disregarding the second term because it is periodic, we nd that the drift velocity, as derived in Equation (22), is the same as shown in Equation (7).

e2biOt eied sin ot dt0

1 X

REFERENCES 18
Adair RK. 1991. Constraints on biological effects of weak extremely-low-frequency electromagnetic elds. Phys Rev A 43:10391048. Adair RK. 1992. Criticism of Lednevs mechanism for the inuence of weak magnetic elds on biological systems. Bioelectromagnetics 13:231235. Berden-Zrimec M, Jerman I. 2001. New biological detection system for weak ELF magnetic elds and testing of the parametric resonance model (Lednev 1991). Electro Magnetobiol 20: 2741. Bianco B, Chiabrera A. 1992. From the Langevin-Lorentz to the Zeeman model of electromagnetic effects on ligand-receptor binding. Bioelectrochem Bioenerg 28:355365. Bier M. 2005. Gauging the strengths of power frequency elds against membrane electrical noise. Bioelectromagnetics 26: 595609. Binhi VN. 2000. Amplitude and frequency dissociation spectra of ion-protein complexes rotating in magnetic elds. Bioelectromagnetics 21:3435. Blackman CF, Blanchard JP, Benane SG, House DE. 1994. Empirical test of an ion parametric resonance model for magnetic eld interactions with PC-12 cells. Bioelectromagnetics 15:239250. Blackman CF, Blanchard JP, Benane SG, House DE. 1995. The ion parametric resonance model predicts magnetic eld parameters that affect nerve cells. FASEB J 9:547 551. Blanchard JP, Blackman CF. 1994. Clarication and application of an ion parametric resonance model for magnetic eld interactions with biological systems. Bioelectromagnetics 15:217238. Bleaney BI, Beaney B. 1976. Electricity and magnetism, 3rd edition. London: Oxford. p 85. Del Giudice E, Fleischmann M, Preparata G, Talpo G. 2002. On the unreasonable effects of ELF magnetic elds upon a system of ions. Bioelectromagnetics 23:522530. Durney CH, Rushforth CK, Anderson AA. 1988. Resonant a.c.-d.c. magnetic elds: Calculated response. Bioelectromagnetics 9:315336. Edmonds DT. 1993. Larmor precession as a mechanism for the detection of static and alternating magnetic elds. Bioelectrochem Bioenerg 30:312. Kaune WT. 2002. Thermal noise limit on the sensitivity of cellular membranes to power frequency electric and magnetic elds. Bioelectromagnetics 23:622628. Bioelectromagnetics

Jn ed fe2biOnot 1g 2 b i O n o n1 This allows us to write the general solution vt yt Geied sin ot 1 X Jn ed ein ot Ge2biot eied sin ot 2 b i O n o n1 1 X Jn ed
n1

2b iO no

19

The rst and third terms in Equation (19) decay in time, leaving us with the remaining periodic part: v Geied sin ot Jn ed ein ot 2 b i O n o n1
1 X

20

This solution can be further decomposed into resonant and nonresonant terms by again applying the transformation given in Equation (16). v Geied sin ot
1 X

Jn ed ein ot Geied sin ot 2 b i O n o n0 21

1 X 1mn Jm edJn ed in ot e 2b iO no m;n1

m6n

The second term in this expression is the resonant term. After discarding the rst term the resonant remainder can be expanded allowing the velocity to be written as

386

Vincze et al.
Sarimov R, Markova E, Johansson F, Jenssen D, Belyaev I. 2005. Exposure to ELF magnetic eld tuned to Zn inhibits growth of cancer cells. Bioelectromagnetics 26:631638. Shuvalova LA, Ovstrovskaja MV, Sosunov EA, Lednev VV. 1991. Effect of weak magnetic eld in the parametric resonance mode on the rate of calmodulin-dependent phosphorylation of myosin in the solution. Doklady Akademii Nauk SSSR (Reports of the Academy of Sciences of the SSSR, in Russian) 317:227230. Trillo MA, Ubeda A, Blanchard JP, House DE, Blackman CF. 1996. Magnetic elds at resonant conditions for the hydrogen ion affect neurite outgrowth in PC-12 cells: A test of the ion parametric resonance model. Bioelectromagnetics 17:1020. Vincze G, Szasz N, Szasz A. 2005. On the thermal noise limit of cellular membranes. Bioelectromagnetics 26:2835. Weaver JC, Astumian RD. 1990. The response of living cells to very weak magnetic elds: The thermal noise limit. Science 247: 459462. Wertheimer N, Leeper E. 1979. Electrical wiring congurations and childhood cancer. Am J Epidemiol 109:273284. Zhadin M. 1998. Combined action of static and alternating magnetic elds on ion motion in a macromolecule: Theoretical aspects. Bioelectromagnetics 19:279292. Zhadin M, Barnes F. 2005. Frequency and amplitude windows in the combined action of DC and Low frequency AC magnetic elds on ion thermal motion in a macromolecule: Theoretical analysis. Bioelectromagnetics 26:323330. Zhadin M, Fessenko EE. 1990. Ionic cyclotron resonance in biomolecules. Biomed Sci 1:245250.

Koch CLMB, Sommarin M, Persson BRR, Salford LG, Eberhardt JL. 2003. Interaction between weak low frequency magnetic elds and cell membranes. Bioelectromagnetics 24:395 402. Kuyucak S, Anderson OS, Chung S-H. 2001. Models of permeation in ion channels. Rep Prog Phys 64:14271472. Lednev VV. 1991. Possible mechanism for the inuence of weak magnetic elds on biological systems. Bioelectromagnetics 12:7175. Liboff AR. 1985. Geomagnetic cyclotron resonance in living cells. J Biol Phys 13:99102. Liboff AR. 2003. Ion cyclotron resonance in biological systems: Experimental evidence. In: Stavroulakis P, editor. Biological effects of electromagnetic elds. Berlin: Springer. Chapter 2.4, pp 76113. Liboff AR. 2005. The charge-to-mass ICR signature in weak ELF bioelectromagnetic effects. In: Lin JC, editor. Advances in electromagnetic elds in living systems, Vol. 4. New York: Springer. Chapter 6, pp 189218. Male J. 1992. Biological effects of magnetic elds: A possible mechanism? Biologist 39:8789. McLeod BR, Liboff AR. 1986. Dynamic characteristics of membrane ions in multield congurations of low-frequency electromagnetic radiation. Bioelectromagnetics 7:177 189. Prato FS, Kavaliers M, Thomas AW. 2000. Extremely low frequency magnetic elds can either increase or decrease analgesia in the land snail depending on eld and light conditions. Bioelectromagnetics 21:287301.

Bioelectromagnetics