1. Introduction
Magnetic fields are constantly present in the environment. The natural geomagnetic field and its fluctuations accompanied the evolution of life on Earth. The current strength of the geomagnetic field varies from the equator to the poles from approximately 25 to 65 μT, and its direction varies in space in a rather complex way. The alternating constituent of natural magnetic fields is represented by low-frequency fluctuations, mainly due to regular and irregular processes in the Earth's magnetosphere-ionosphere current system. The range of these magnetic fluctuations rarely exceeds hundreds of nT, and most natural variable magnetic fields are recorded in the range of up to 10 Hz [
1]. Anthropogenic magnetic fields are superposed on the natural magnetic background due to the development of technology over the past century. These fields vary significantly in their characteristics [
2]. The overall intensity of anthropogenic magnetic fields is steadily increasing year by year. In this regard, it is essential to understand the influence of the magnetic environment on organisms.
Weak (conditionally not higher than 100 μT) low-frequency magnetic fields (LFMF) are of great interest among the fluctuating magnetic fields. Their energy does not exceed the thermal noise, i.e., LFMF cannot directly affect an elementary act of chemical reactions of kT order [
3]. At the same time, the biological effects of weak LFMF are often described in the literature [
4]. The researchers working in this field apply their efforts to clarify the mechanisms by which magnetic influence with energy not exceeding the thermal noise of molecules causes biological effects. Over the last decades, several hypotheses have been proposed in this regard. However, there is no definitive understanding of the formation of a specific biological response to LFMF.
This review presents some of the effects of LFMF and the most developed hypotheses trying to explain these effects.
3. Radical Pair Magnetoreception and Its Application to the Effects of Low-Frequency Magnetic Fields
Another promising hypothesis considers singlet-triplet transitions in radical pairs as a primary target for the influence of magnetic fields of geomagnetic order. Research in this direction originates from the discovery of the light-dependent orientation of birds in the geomagnetic field in the 1980s [
42]. More precisely, it was found that birds can use the geomagnetic field for choosing migrational direction only under blue-green light with a wavelength of 424-565 nm. The birds were disoriented under the illumination of the experimental area with yellow or red light with a wavelength of 590-650 nm [
43,
44]. Accurate experiments with robins show that the threshold of light-dependent perception of the magnetic field in this species lies in the interval between the wavelengths of 561-568 nm [
45].
The necessity of photons with energy above a certain threshold for bird orientation in the geomagnetic field prompted Ritz and co-authors [
46] to the idea that the primary magnetodetection associated with the effect of magnetic fields on the electron spin state in radical pairs [
47,
48]. According to Ritz and co-authors, such radical pairs occur in cryptochrome molecules in the bird retina [
46]. Cryptochromes are a class of blue light-sensitive flavoproteins found in various tissues in many plants and animals [
49]. Cryptochrome molecules are structurally similar to the bacterial enzyme photolyase involved in DNA repair processes [
50]. Cryptochromes and photolyase utilize flavin adenine dinucleotide (FAD) as a light-sensitive chromophore [
51]. After the absorption of a photon with sufficient energy, the FAD into an excited state. An electron from the excited FAD is sequentially transferred between tryptophan residues [
52]. This electron transport reaction produces a deprotonated last tryptophan residue in this chain and a reduced FADH. It was suggested that cryptochrome with reduced FADH is in its active (or signal) state, and long-lived radical pairs are formed during the process described above [
53]. The inclination and/or intensity of an external magnetic field should affect the electron spin state of such radical pairs [
46]. I.e., the ratio of cryptochrome molecules in the signal and non-signal states should change depending on the direction of a static magnetic field [
54]. This mechanism is currently viewed as the most likely basis for the magnetic orientation in birds and other animals since the inclination of the geomagnetic field is place-specific depending on the latitude and some features of the Earth's crust.
A lot of evidence for the crucial role of cryptochromes in biological magnetoreception has been accumulated to date [
52,
55,
56,
57,
58,
59,
60,
61,
62]. Experiments with transgenic
Drosophila melanogaster showed that the knockout of cryptochrome genes leads to the disorder of behavioral responses to magnetic fields [
63]. Moreover, the expression of human cryptochrome (hCRY2) instead of the knocked-out led to the restoration of magnetosensitivity [
64]. The amount of available data leaves no doubt that cryptochromes are the key proteins in the evolutionarily formed mechanism of animal orientation in the geomagnetic field, the so-called “chemical magnetic compass.” At the same time, significant magnetic field effects in the isolated molecules in vitro have been only detected at fields more than twenty times exceeding the geomagnetic one [
52,
56,
58]. Even the most likely candidate for the role of an avian magnetosensor, cryptochrome 4, responded in vitro to a magnetic field on the order of thousands of microtesla [
58]. Responses to the fields of the geomagnetic order are shown so far on a model molecular triad consisting of covalently linked carotenoid, porphyrin, and fullerene moieties [
65]. Probably, there must be an evolutionarily formed mechanism for enhancing the magnetic sensitivity of cryptochromes in the cells responsible for magnetic perception. The molecular environment and conformational rearrangements during magnetic perception may be of importance. Several scientific groups are currently working on this problem. In addition, attempts are ongoing to find alternative radical pairs in cryptochromes sensitive to weak magnetic fields [
66]. For example, the radical pair formation between flavin and a superoxide radical [
67,
68,
69] or ascorbyl radical [
70] has been suggested.
It should be said that cryptochromes are a special case, and radical pairs are constantly present in biomolecules and affect almost all vital biochemical processes [
71,
72,
73]. It allows considering radical pairs as primary targets for the influence of alternating magnetic fields regardless of their application to the magnetic orientation of birds and other animals. The lifetime of a radical pair in the individual free radical events takes place in the nanosecond to microsecond time scale [
74,
75]. The period of LFMF oscillations that influence organisms is often more than 16.67 ms (for LFMF fields with a frequency of 60 Hz or less). Therefore, LFMF can be viewed as static during the lifetime of a radical pair [
75]. In other words, each radical pair experiences a different "quasistatic" magnetic field whose intensity (
BQS) depends on the phase of the LFMF (α), which barely changes during the lifetime of the pair, with α randomly distributed between 0 and π [
74]:
Aspects of the possible impact of such a quasistatic LFMF on the radical pairs have been viewed by Peter Hore [
74]. It is important that the dependence between singlet and triplet yields and the intensity of the applied magnetic field in the range of
BDC ± amplitude of
BAC must be nonlinear for the LFMF to affect the course of a radical-pair reaction. In this case, LFMF can shift the equilibrium between singlet and triplet yields relative to the equilibrium state in a static geomagnetic field (
Figure 2). Another essential thing is the so-called "weak field effect" that appears under magnetic fields of the order of tens and hundreds of microtesla. This effect has been shown in experiments with different radical pair reactions [
52,
76,
77,
78,
79]. It provides the nonlinearity of the dependence of the yield of a radical-pair reaction on the strength of a constant magnetic field [
80]. That is, the probability of compliance with the "nonlinear dependence" condition is maximal in magnetic fields of the geomagnetic order.
4. Some Inconsistencies
Among the results of experiments, some data do not agree with the current biophysical models of the LFMF influence on biomolecules. The radical-pair mechanism described in the previous section can explain a significant part of the effects of low-frequency magnetic fields with an amplitude of tens to hundreds of microtesla. However, it cannot explain the effects of weaker fields. The simulation performed by Peter Hore [
74] for a radical pair, [FAD•− Z•] with a lifetime of 1 µs, under
BDC = 50 µT and LFMF of 1 µT, 50 Hz showed the largest change in the reaction yield caused by the LFMF is −14 ppm parts per million (ppm). The radical pair [FAD•− Z•] is identical to [FAD•− TrpH•+] except that the second hypothetical radical, Z•, has no hyperfine interactions and is more sensitive to weak magnetic fields. For the [FAD•− TrpH•+] radical pair, such a simulation resulted in −1.2 ppm [
74]. These values seem small, and it is difficult to imagine that changes in the accumulation of singlet and triplet products of radical-pair products are directly related to the effects of LFMF with induction of less than 1 μT. Although the radical-pair mechanism can clarify the dependence of biological effects on the amplitude of LFMF [
5,
8,
12,
13,
14], it is hard to directly explain the frequency effectiveness windows of LFMF [
8,
12,
24,
33,
81].
The same applies to "resonant" biophysical models. For example, significant biological effects caused by alternating magnetic fields directed perpendicular to a static (geomagnetic) field have been described [
82,
83,
84,
85]. Garcia-Sancho et al. [
84] have estimated the uptake of radioactive
42K in several cell types (red blood cells, thymocytes, Ehrlich ascites tumor cells, HL60 human leukemia cells, and U937 human leukemia cells) exposed to parallel and perpendicular configurations of the static magnetic field and LFMF with the cyclotron frequency for sodium, calcium, and potassium ions. Only U937 human leukemia cells responded to both parallel and perpendicular configurations with an increase in uptake of
42K [
84]. In another work, Blackman et al. tested various mutual orientations of the static magnetic field and LFMF when studying their resonance model [
83]. Both perpendicular and parallel configurations of static and alternating magnetic fields cause neurite outgrowth responses. Besides, these responses differed: the enhancement and inhibition of neurite outgrowth registered in perpendicular and parallel configurations correspondingly [
83]. Noteworthy that the significant effects of the perpendicular static magnetic field and LFMF were described in tumor cell lines. It emphasizes the dependence of LFMF effects on the cell type described earlier.
Moreover, the dependence of the effects on the presence of light was shown when studying the influence of LFMF with resonance parameters for calcium and potassium ions on the nociceptive response in the terrestrial snail
Cepaea nemoralis [
23]. Thermal response latencies of snails injected with the enkephalinase inhibitor SCH34826 and exposed to calcium or potassium resonance-tuned magnetic fields were significantly higher in the light compared to the control (sham exposure). At the same time, there was no significant difference in responses between exposed and unexposed individuals in the dark [
23]. Such dependence on light is characteristic of the radical-pair mechanism of the magnetic influence on living systems.
Thereby, the results of some experiments are not consistent with current hypotheses.
5. Future Prospects
The contradictions indicated in the previous section are the consequence of our incomprehension of the perceived magnetic impact transforming into registered biological effects. All resonance-like models were developed on experimental results at the cellular, organismal, and even population levels. A network of signaling pathways and molecular interactions between the putative primary targets and biological effects remain obscured. These pathways and the intracellular environment of the magnetosensitive molecules can play a crucial role in the occurrence of magnetobiological effects. The same challenges face researchers investigating the influence of the geomagnetic field on radical-pair reactions in cryptochrome. Protein-protein interactions and conformational changes of molecules in the process of radical-pair magnetoreception can be the amplifier allowing the detection of minor changes in the geomagnetic field. For example, Qin et al. [
86] reported that the protein encoded by the CG8198 gene can bind to cryptochrome and promote its magnetoreceptor function. Further search for long-life radical pairs in biomolecules and pathways for signal transmission is required.
Probably, several different primary targets perceive LFMF in cells. This assumption complicates the interpretation of emerging biological responses. Therefore, magnetobiological experiments with test systems that evaluate target magneto-sensitive biochemical processes are required. Unlike biological parameters at the organismal level, such indicators registered "closer" to the primary targets will give a more unambiguous interpretation. At the same time, the dependence of the manifestation of the effects of LFMF with resonance parameters for calcium and potassium ions on the presence of light [
23] makes one think about the radical-pair nature of resonance-tuned LFMF effects. However, the question arises: how can the frequency-amplitude windows of LFMF biological efficiency be explained? A general hypothesis, which combines the resonance-like responses of biological systems to LFMF and the effects of magnetic fields on radical pairs, is given below.
1. Radical pairs are the main target for LFMF's influence on organisms. If radical pairs emerge in biochemical processes oscillating in cells with the frequency f OSC, and this emerging occurs in a specific phase of the oscillations, then the manifestation of biological effects can depend on the frequency of LFMF.
2. The altered or "signal" state of a given oscillating process depends on the ratio of singlet and triplet yields of the radical-pair reaction included in it.
3. If an external LFMF with frequency f = f OSC and amplitude BAC parallel to the static (geomagnetic) field (BDC) is applied to such an oscillating chemical process, then due to a negligible lifetime of radical pairs and depending on phase coincidence, some radical pairs of the biochemical oscillators will be at resulting magnetic field BDC + BAC throughout the whole LFMF exposure. Another approximately the same part of the oscillators will generate radical pairs exposed to BDC - BAC throughout the whole LFMF exposure. Most of the radical pairs will be under a "quasistatic" magnetic field with the intensity from BDC - BAC to BDC + BAC throughout the exposure.
4. According to Hore [
74], changes in the ratio of singlet and triplet yields of a biradical reaction in response to LFMF occur if there are non-linear dependence between the singlet-triplet reaction yields and the magnetic field strength within limits from
BDC -
BAC to
BDC +
BAC (Figure 2B). Synchronization of LFMF frequency with the frequency of chemical oscillations provides a quasistatic "effective" magnetic field for radical pairs in a portion of chemical oscillators. The ratio of triplet and singlet yields for this portion of oscillators will differ from the state for the rest of the oscillators throughout the whole LFMF exposure due to the non-linear dependence between the triplet and singlet yields and magnetic field intensity as the "low field effect". A notable change in LFMF frequency (
f ≠
f OSC) leads to a condition where radical pairs regularly generated by a chemical oscillator will experience quasistatic magnetic fields of different intensities at different moments. The disappearance of the biological effect at changed non-resonant LFMF frequency can be a consequence of the inability to maintain a "signal" state of the portion of the biochemical oscillators throughout the LFMF exposure. It ensures the appearance of frequency windows of magnetobiological effects.
5. The “low field effect” [
80] provides “non-linear dependence” conditions; therefore, the biologically effective amplitude of the LFMF exists for a specific radical-pair reaction. A change in this amplitude can shift the magnetic field intensity values to the area of linear dependence, which leads to the absence of a biological effect (Figure 2A). It explains the amplitude windows of the LFMF efficiency.
Consider a specific example to show the viability of the above model describing amplitude-frequency effective windows in responses of biological systems to LFMF within the framework of a radical-pair mechanism. Most experiments testing ion-related resonance models are performed with fields tuned to the cyclotron frequency for the calcium ion [
7,
8,
12,
17,
20,
24]. This frequency for calcium ions in the geomagnetic field at temperate latitudes of 40-55 µT will be about 30-42 Hz. These values have often been described as Bdc and fac in the experiments testing resonance biophysical models for calcium ions [
17,
20,
21,
24,
81,
87]. The 30-42 Hz frequency should correspond to a chemical oscillation period of approximately 24-33 ms. The biologically effective amplitude of the LFMF tuned to calcium ions provides a quasi-static magnetic field relative to the lifetime of radical pairs from approximately zero to 0.1 mT (for
BAC/
BDC ~ 0.9 in the case of the Blackman’s model) or from roughly -0.05 mT through zero up to +0.15 mT (for
BAC/
BDC ~ 1.8 in the case of the Lednev’s model).
The described effects of calcium-tuned LFMF can be explained by the above model applied to radical pairs in chemical oscillators on the mitochondrial membrane [
88]. Under unstressed “physiological” conditions mitochondrial membrane potential (ΔΨm) fluctuates with a period close to 25 ms (
Figure 3A). This oscillation is associated with fluctuating levels of cytoplasmic superoxide anion in the nM range. Its period can vary from 34 ms to 16 ms depending on the balance between reactive oxygen species production and scavenging (
Figure 3B). These oscillations become low-frequency and high-amplitude with the transition of mitochondria to the “pathophysiological” state [
89]. Scilicet, oscillatory processes with the 30-42 Hz frequency that formally corresponds to the cyclotron resonance frequency for calcium ions in the geomagnetic field at temperate latitudes (40-55 µT) exist in mitochondria under normal conditions. The chemical composition of this oscillatory process includes radical-pair reactions as a shunt of electrons of the respiratory chain towards the generation of superoxide anion, following the transport and scavenging of superoxide radical by superoxide dismutase [
88]. These radical-pair processes may be responsible for the biological effects of “calcium-tuned” fields with characteristic frequency-amplitude windows with two assumptions: 1) the ratio of triplet and singlet yields depend non-linearly on the magnetic field within 0-0.1 (or 0-0.15) mT; 2) the ratio of triplet and singlet yields at some stage of the oscillation affects the functional parameters of mitochondria. Moreover, changes in mitochondria functioning can explain most of the biological effects obtained in earlier experiments with “calcium-tuned” LFMF, such as changes in diatom mobility [
8], neurite outgrowth [
81], planarian regeneration [
12,
20,
24], gravitropic reaction [
87], etc.
This example explains the frequency-amplitude efficiency windows of LFMF with a frequency “tuned to calcium ions” from the standpoint of the magnetic field effect on radical pairs in the 30-42 Hz biochemical oscillator in mitochondria. One can test the proposed assumptions since it is possible to register the period of mitochondrial oscillations, and the frequency of these oscillations can be changed with inhibitors [
88,
89]. The effective frequency of LFMF should change when the frequency of fast low-amplitude mitochondrial oscillations alters. If the hypothesis is supported, a further research direction could be the search for additional biochemical oscillators [
91] that generate radical pairs. It probably could help to explain the biological effects of LFMF with other frequencies and amplitudes.