Magnetoreception in microorganisms and fungi

Magnetoreception in microorganisms and fungi

(Parte 1 de 7)

Review article CEJB 2(4) 2007 597–659

Magnetoreception in microorganisms and fungi

1 Department of Biology I

Ludwig-Maximilian University Munchen, D-80638 Munchen, Germany

2 Department of General Microbiology and Microbial Genetics

Friedrich-Schiller-University Jena D-07743 Jena, Germany

3 Faculty of Biology

Philipps-University Marburg D-35032 Marburg, Germany

Received 14 May 2007; accepted 09 July 2007

Abstract: The ability to respond to magnetic fields is ubiquitous among the five kingdoms of organisms. Apart from the mechanisms that are at work in bacterial magnetotaxis, none of the innumerable magnetobiologicaleffects are as yet completely understood in terms of their underlying physical principles. Physical theories on magnetoreception, which draw on classical electrodynamics as well as on quantum electrodynamics, have greatly advanced during the past twenty years, and provide a basis for biological experimentation. This review places major emphasis on theories, and magnetobiological effects that occur in response to weak and moderate magnetic fields, and that are not related to magnetotaxis and magnetosomes. While knowledge relating to bacterial magnetotaxis has advanced considerably during the past 27 years, the biology of other magnetic effects has remained largely on a phenomenological level, a fact that is partly due to a lack of model organisms and model responses; and in great part also to the circumstance that the biological community at large takes little notice of the field, and in particular of the available physical theories. We review the known magnetobiological effects for bacteria, protists and fungi, and try to show how the variegated empirical material could be approached in the framework of the available physical models. c© Versita Warsaw and Springer-Verlag Berlin Heidelberg. All rights reserved.

Keywords: magnetic field, magnetoreception, ion-cyclotron resonance, magnetosomes, quantum coherence, radical-pair mechanism, ecology, climate change

∗ E-mail: † E-mail:

598 A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659


B magnetic flux density (magnetic induction) BAC alternating magnetic field (generated by alternating current) BDC static magnetic field (generated by directed current) CD coherent domain ELF extremely low frequency (i.e. magnetic field, ∼3-300 Hz) EMF electromagnetic field H magnetic field strength IIM ion interference mechanism ISC intersystem crossing ICR ion cyclotron resonance IPR ion parametric resonance LF low frequency (i.e. magnetic field) MF magnetic field

1 Introduction

The earliest studies on the influence of electromagnetism on organisms date back to the late 19th century, probably beginning in St. Petersburg [1]. A larger, more general interest arose only some decades later, coinciding with worldwide electrification and telecommunication. Although microorganisms play a major role in the global ecosystem, the number of publications covering magnetoreception in fungi, protists and non-magnetotactic bacteria is small compared to similar reports on humans and animals [1–6]; and is perhaps comparable to the state of knowledge in plants [7]. The magnetoorientation of magnetotactic bacteria [8], as well as that of migrating birds and insects, belongs to the best understood and most intensely studied phenomena of magnetoreception [6, 9, 10]. Recently Ritz et al. [1, 12] suggested a light-driven, radical-pair mechanism for the magnetoreception of birds mediated by cryptochrome [13, 14]. There is evidence that even plant cryptochromes are involved in the magnetoreception of Arabidopsis [15]. Bacterial magnetotaxis is based on the magnetoorientation of magnetite crystals; thus representing the only magnetoreception mechanism completely elucidated up to now [8, 16–18].

The two central questions in this context: (i) whether or not microorganisms are able to perceive geomagnetic fields, and (i) whether or not magnetoreception is an essential and vital environmental factor for survival, have remained largely unanswered, even though magnetoreception must be regarded as an established fact. Furthermore, we contend that the recent discussion regarding the mechanisms of climate change and global warming should consider other, non-anthropogenic contributions, e.g. the altered gas exchange of microorganisms as a consequence of the steadily changing geomagnetic field.

Despite the numerous magnetobiological effects that have been described in the pertinent literature, there is an apparent lack of model organisms, model responses and genetic approaches; tools that are typical for modern research strategies commonly found

A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 599 in other biological fields. The problem is compounded by the fact that magnetic effects are observed for a huge range of magnetic flux densities, which cover more than 10 orders of magnitude. To come to grips with such a huge dynamic range, that is similar to that of human vision, one would expect the study of dose-response relationships to be of paramount importance. It thus comes as a surprise that there exists only one doseresponse curve for a biomagnetic effect in DC fields [19], and only a limited number of dose-response studies for AC-fields. Despite the limited information it is nevertheless apparent that magnetobiological dose-response relationships differ drastically from the ones usually found in physiology, where one typically finds exponential rise or decay curves, and in some cases optimum curves. Magnetoresponses, in contrast, show certain “windows” of magnetic flux densities, or, in the case of AC-fields, windows of frequencies for which a response is obtained. Applying a higher field strength may thus not necessarily guarantee a stronger response. As a consequence, some of the apparent contradictory results in the magnetism literature could be explained by the fact that different authors often used different magnetic flux densities that were within or outside these windows. Research on biological magnetoresponses can be roughly divided into experiments that employ static magnetic fields (BDC), or alternating fields (BAC), or as in most cases, a combination of both (BAC+B DC). The body of literature on AC fields and their concomitant effects dominates by far that of DC fields. This appears surprising in view of the fact that DC experiment is required to find out how geomagnetic fields influence life.

Most experiments with AC fields are done with frequencies near 50 or 60 Hz, i.e. frequencies akin to that of ubiquitous electric appliances. Much of this type of research was historically motivated by the wish to find out whether or not our electric environment influences life, and specifically, human health. Even though this line of research may not directly contribute to the understanding of how static geomagnetic fields influence life, it nevertheless represents a powerful technical tool to investigate the involvement of specific ions in a given biomagnetic response. It had earlier been noticed that AC fields elicit responses most prominently at the cyclotron resonance frequencies (including their harmonics and subharmonics) of biologically important ions, in particular Ca2+.T his pattern gives rise to dose-response curves with several minima and maxima. Therefore it is not difficult to understand why explaining this type of dose-response relationship is the subject of several theories (ion-cyclotron resonance, ion-parametric resonance, ioninterference mechanism, coherence mechanism; see below).

One of the reasons that magnetobiological responses frequently meet with reservations is based on the fact that the energy content of biologically actinic magnetic fields can be several orders of magnitude below their thermal energy content (kT problem). We will show how the problem can be addressed within the framework of modern theories. Also the hunt for “the” magnetoreceptor remains presently an unresolved task (with the exception of magnetite in bacterial magnetotaxis, see below). As function of fact, it is contested as to whether or not there exists only one type of magnetoreceptor; the requisite literature rather indicates that in prokaryotic and eukaryotic cells several

600 A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 magnetosensitive molecules, and physically distinct mechanisms, exist that could mediate magnetoreception. Because cell membranes do not constitute barriers for magnetic fields, magnetoreception could in theory occur on many different levels. For example, DNA itself, and the transcription and translation machineries, have been proposed as targets. It is thus noteworthy that even cell-free systems of protein biosynthesis are receptive to magnetic fields (see below).

2 Magnetic effects on solutes

Magnetic fields can affect organisms not only directly, but also indirectly, by changing the physical properties of solutes and growth media. For example, a 1 minute magnetic pretreatment of culture media stimulated the subsequent growth of Escherichia coli in a geomagnetic field [20]. Magnetically treated water inhibits the germination of the microfungus Alternaria alternata [21], and treatment of nutritional media affects the subsequent growth of Saccharomyces fragilis, Brevibacterium and Bacillus mucilaginosus [2]. Such effects are possible because magnetic treatments alters solutes, for example the formation of calcium carbonate [23], water vaporization [24], ion hydration and resin absorption [25]. Indeed effects on Ca2+ hydration after short treatment with a weak magnetic field or pulses, applicable, for example, to organismal growth stimulation, is reported by Goldsworthy et al. [26]. As these effects were usually achieved with very strong magnetic fields in the mT to T range, they may not be pertinent for experiments done in very weak or geomagnetic fields.

3 Bacterial magnetotaxis

3.1 Magnetosomes and their role in magnetoreception

Apart from the general effects of all types of magnetic fields on growth and morphogenesis, some organisms have succeeded in employing the directional qualities of magnetic fields for orientation purposes. Animals using geomagnetic fields for navigation are either long-distance travellers, such as migratory birds, whales, sharks, turtles and butterflies, or depend for other reasons on the ability for exact orientation, e.g. honeybees. Clearly, microorgansims do not fall into a category where magnetoorientation would be expected. Nonetheless this behaviour, known as magnetotaxis, is the best studied of the magnetoresponses [27–29]. This reaction has been globally observed in a number of marine [30, 31] and freshwater bacteria [32–34], as well as in several types of unicellular eukaryotic microorganisms. The latter are rarely observed, probably because they are easily overlooked in samples teeming with bacteria. Due to their overall high fragility and sensitivity, eukaryotic laboratory strains usable for detailed analyses have not yet been established; yet the occurrence of magnetoperception in eukaryotes may be rather widely distributed, as magnetotactic species have been detected in a number of major groups, such as dinoflagellates, ciliates, cryptophytes [35, 36].

A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 601

3.2 Magnetotactic microorganisms

Research into magnetoresponsive prokaryotes began with the discovery that certain bacteria consistently preferred one geomagnetic pole over the other, and therefore always swam to the same side of a water droplet on microscopic slides [16]. Magnetotaxis, although biased for fast-swimming organisms, provided a handy tool for the isolation of several similar species [4, 37–39, 41–45]; and moreover also allows for the enrichment and studies of unculturable strains and communities [46, 47]. Magnetotactic bacteria are flagellated chemolithoautotrophs exhibiting various morphotypes; to date cocci, spirilla, rod-shaped, and vibroid or helical forms have been described [47, 48]. They inhabit the oxygen-anoxygen transition zone of marine or freshwater sediments or chemically stratified water columns, where they occasionally occur in high cell densities [31, 49]a nd are either obligate anaerobes [30, 45] or facultatively anaerobic microaerophils [50–53]. Although likely to be of polyphyletic origin [54], most of the characterized morphotypes have been grouped into the Proteobacteria, with a distinct subcluster present in its alpha subgroup [32, 42, 46, 5, 56]( Table 1). Genome data are available from Magnetospirillum magnetotacticum MS-1 (GenBank accession AAAP00000000, Magnetospirillum gryphiswaldense MSR-1A [57]; GenBank acc. BX571797) and Magnetospirillum magneticum AMB-1 (GenBank acc. AP007626). Highly organized aggregates of magnetotactic cells have also been described from a variety of other locations [28, 58–61].

Magnetotaxis is defined as movement parallel to the field lines of an external magnetic field. Nevertheless it is not a taxis in the strictest sense as the organisms are not following the direction of the magnetic field itself, but rather ultilize the directional information to support other orientation mechanisms. It was generally assumed that they navigate along the inclination of the magnetic field lines to locate a suitable environment within an oxygen (magneto-aerotaxis), or other chemical, gradient; thus reducing search movement in turbid surroundings to just one dimension - up and down [62]. The key benefit of magnetotaxis in this process is the enhancement of the bacterium’s ability to detect oxygen, not an increase in average speed of reaction [29]. Movement along a straight path allows for earlier detection of an existing oxigen gradient, and thus enhances the flight from oxygen. One study suggests a role for magnetosome formation in mediating the response to gravity, as magnetosomes and magnetotaxis were shown to be completely absent in prolonged microgravity [63]. In magnetotaxis, polar and axial magnetotactic strains can be discriminated between. Bipolar flagellated cells display axial behavior by swimming back and forth within a local applied magnetic field. In polar magnetotaxis, the cells follow a preferential direction and swim away when the local field is reversed [64]. This classification apparently results from cellular morphology, and has no impact on orientation efficiencies in natural environments. The observation that polar magnetotactic cells in the southern hemisphere predominantly exhibited a south-seeking behavior in

602 A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 laboratory tests was taken as support for the importance of the magnetic field line for magnetotaxis [65]. The recent discovery of seasonally occurring, predominantly southseeking polar bacteria, in populations from the northern hemisphere call this explanation into question. Instead, the oxidation-reduction potential at any given position of a water column seems to influence the polarity of movement [31, 49].

3.4 Magnetosomes

All magnetotactic cells contain magnetosomes. These organelles consist of a ferrimagnetic crystal surrounded by a specialized membrane. In prokaryotes, the magnetosome crys- tals result from the controlled biomineralization of either magnetite (Fe3O4), or greigite (Fe3S4)[ 6–69]. Additionally one single strain has been found that contains iron pyrite

(FeS2)[ 67]. A few morphotypes mineralize magnetite and greigite within the same cell, and even within the same crystal aggregate [70, 71].

In both magnetite and greigite, crystal structures follow the spinel type, consisting of two interlocking grid systems with different numbers of grid coordinates (nodes). Magnetite, as well as greigite, contains a mixture of two- and three-valent iron, with each form occupying specific nodes. This leads to the complete extinction of the atomic magnetic dipol moments of Fe3+. The magnetic properties, therefore, are solely attributed to Fe2+. Each morphotype is usually associated with a particular crystalline habit of magnetite, whereas greigite crystals of different shapes may occur simultaneously [48, 72–74]. Cuboid, bullet-, tooth- and drop-shaped crystals have been described [45, 73, 75, 76].

Besides eukaryotic microorganisms, magnetite crystals that are similar in appearance and structure to those of bacteria were also found in animal cells [7]; however no information exists on their origin and biosynthesis. Ferrimagnetic crystals interact in excess of a million times more strongly with magnetic fields than do diamagnetic or paramagnetic materials. If a ferrimagnetic nanocrystal were fixed to an ion channel - an assumption that has not been verified yet - it would generate torque in a weak geomagnetic field that would suffice to alter ion movement across a membrane. Such considerations show that magnetites hold, at least in theory, the potential to directly influence ion transport [7]. It has also been pointed out that trace amounts of magnetite may be ubiquitous, and that a single 100-nm magnetite crystal, exposed to a 60 Hz, 0.1 mT magnetic field, could absorb sufficient energy to supersede several times the thermal background noise [78]. Magnetite particles can have dramatic effects on the dynamics of photogenerated free radicals [79]. It is thus pertinent to reckon with a modulating effect of magnetites if present, particularly in context of the radical pair mechanism (see below).

Fossil records of magnetosome crystals date back to the Precambrian time; while similar crystals have been detected in 4 billion year-old carbonate blebs of martian meteorite fragments [80–82]. Although controversely discussed [83, 84], it appears possible that the martian magnetites are of a biogenic origin. This would also imply that these martian minerals constitute the oldest fossils on Earth, and at the same time provide evidence for the possibility of panspermia [85].

A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 603

The number of magnetosomes within a single cell ranges from a few large structures in cryptophyte cells [35] to several hundred [73, 86], with 10 − 20 the average number in magnetotactic spirilla [8, 16, 72]. Crystal sizes range from 35 to 200 nm, which indicates their single-domain status [86–8]. The size of a magnetic domain depends on the material, and can be roughly calculated. According to such estimations, a domain of magnetite corresponds to a size between 35 and 75 nm, and in elongated crystals up to 120 nm [89–91]. As single domain crystals, the magnetosome cystals are especially susceptible to efficient magnetization and alignment, and they produce stable magnetic fields. Exceptions have been published by Farina et al. [92] and Spring et al. [56], who demonstrated that at least two strains isolated from the Itaipu lagoon in Brazil contained magnetosomes with dimensions up to, and even exceeding, 200 nm, a size that could easily harbour two magnetic domains. In such large crystals a metastable, single-domain state is only possible when the crystals are aligned within a chain [93]. The extracellular formation of single-domain magnetite for biotechnological applications has also been performed by a biologically-induced, biomineralization process of non-magnetotactic bacteria [94, 95]; and by the aerobic fungi Fusarium oxysporum and Verticillium sp. [96].

3.5 Magnetosome organization and synthesis

In some morphotypes magnetosomes form loose aggregates within the cell [60, 97], however in the majority of strains studied they are arranged in one or more chains spanning the cytoplasm. The magnetosomes within a given chain are separated by a gap containing no particulate structures, as observed in transmission electron micrographs [98]. In Magnetospirillum species, the single magnetosome chain is usually located close to the inner membrane [98, 9]. The combination of disposition in chains, and size control, results in a high magnetic to thermal energy ratio. The total magnetic moment of a magnetosome chain equals the sum of the individual particle moments [100], and substantially surpasses thermal noise [73, 101].

Organization into chains implicates the crystals in magnetizing each other, and aligning their magnetic dipole moments with each other. These processes start at synthesis, thus each newly formed magnetosome crystal is influenced by the pre-existing chain. Biologically controlled biomineralization is a highly precise process, and is necessarily subject to very exacting control. Therefore the organism first creates a matrix, delimiting the space within which the mineral will grow. The form and size of the nascent crystals depend on the interactions between organic and inorganic phases, and are influenced by parameters such as pH, redox conditions, ionic strength, lattice geometry, polarity, stereochemistry and topography. The biomineralization of greigite is less well studied than that of magnetite. It seems to be less organized, and to require considerably more time [102]. Similar to magnetite biomineralization, it requires several mineralization steps, leading from the non-magnetic precursors, mackinawite and cubis FeS [103], to the final product over a transition period of several days or weeks. During this latter period, iron atoms

604 A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 are rearranged between adjacent sulfur layers, and some of the iron is lost and likely deposited as amorphous iron oxide aggregates. The processes leading towards the formation of magnetite have been reviewed by

Schuler [64, 72, 76, 104–106]. At the onset, a low oxygen potential is likely to be a regulatory signal for metabolic induction of biomineralization, as in Magnetospirillum gryphiswaldense, M. magnetotacticum and Magnetospirillum sp. AMB-1. Thus biominer- alization only occurs at pO2 values below a threshold of 20 mbar [107]. Biomineralization occurs inside a specialized organella, the magnetosome, providing a scaffold organized by membrane proteins which ensures the spatial and temporal accuracy of the process. The scaffold need not be proteinaceous: it can also be thought of as a matrix of amporphous mineral precursors [108]. Indeed amorphous iron oxide has been found to form a layer surrounding maturing crystal [102]. Magnetosomes are enmeshed in a network of cytoskeletal filaments [9], and provides a surrounding for the precise coordination of events involved in magnetite biomineralization [109]. The whole complex consists of several structural entities: the magnetite crystal, magnetosome membrane, surrounding matrix, and, as described for Magnetospirillum magnetotacticum MS-1, an interparticle connection [110]. It has been assumed that magnetosomes are invaginations of the cell membrane; indeed proteins probably involved in such an invagination process have in fact been identified in the magnetosome membrane [1]. Recently electron cryotomography revealed that the membrane surrounding magnetite crystals is continuous with the inner membrane [9]. Nevertheless, some questions remain. The process of iron acquisition and biomineralization would require a closed compartment. Also, in the electron cryotomography picture series, the connection of magnetosomes with the inner membrane was only visible for the innermost structures; the largest magnetosomes seemingly already contained finished crystals. The small, incomplete magnetosomes at the chain ends were completely inside the cytoplasm, with no apparent contact with the inner membrane. Moreover, the lipid and protein composition of the magnetosome membrane differed from all of the other membrane systems of the cell [112]. If it really does originates from the inner membrane, then it is at the very least least subject to extensive modifications. The hitherto identified proteins are apparently involved in iron import, iron conversion and in magnetite synthesis [57, 113, 114].

These membrane vesicles precede magnetite biomineralization and may exist independently [109]. When cells grown under iron limitation are changed to iron sufficiency, biomineralization occurs simultaneously in many pre-formed vesicles, and from the same location within each vesicle. In cells with sufficient iron supply, new magnetite crystals are formed in vesicles at the end of the fully developed magnetosome chain [109]. Usually, the magnetosome chain is distributed to daugther cells at the point of cell division, and during cellular growth. However complete de novo synthesis is also possible [71].

As a prerequisite for biomineralization, iron needs to be imported into the cell. This is a very fast process, for example in iron depleted cells of Magnetospirillum sp. AMB-1 iron uptake is complete within 10 minutes [115], and in Magnetospirillum gryphiswaldense an increase in magnetite measured as intracellular insoluble iron can also be found within

A. Pazur et al. / Central European Journal of Biology 2(4) 2007 597–659 605

10 minutes [116]. Indeed a recent microarray analysis of iron-inducible genes demonstrated the up-regulation of Fe2+ transporters in iron-rich conditions [115].

(Parte 1 de 7)