You are on page 1of 18


Daniel Acosta-Avalos, Centro Brasileiro de Pesquisas Físicas - CBPF, Rio de Janeiro, Brazil
Pedro Leão and Fernanda Abreu, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil
Dennis A Bazylinski, University of Nevada Las Vegas, Las Vegas, NV, United States
© 2018 Elsevier Inc. All rights reserved.

In memory of Ulysses Lins (1969–2017): Ulysses Lins, a professor at the Microbiology Institute of the Federal University of Rio de Janeiro (UFRJ) Brazil,
unexpectedly passed away at age 48. Lins’ passion in microbiology was the magnetotactic bacteria, in particular, a unique, uncultured group of these
organisms referred to as the “multicellular magnetotactic prokaryotes.” The impact of Lins’ outstanding work in biomineralization and magnetotactic bacteria
cannot be overstated based on numerous important discoveries in his field. For his dedication to microbiology and his sincerity and generosity, his students and
colleagues will always remember Professor Ulysses Lins with admiration and respect.

Introduction 2
General Features of Magnetotactic Bacteria 2
Characterization of Magnetotactic Bacteria in Pure Culture 4
Characterization of Uncultured Magnetotactic Bacteria 5
Magnetosomes 10
Biophysics of Magnetotaxis 10
Magnetotaxis 13
Magneto-Aerotaxis 13
Magnetotactic Eukaryotes 15
Geochemical, Geophysical, and Astrobiological Aspects of Magnetotactic Bacteria 15
Biotechnological Applications of Magnetotactic Bacteria, Magnetosomes, and Magnetosome Crystals 17
Perspectives 17
Further Reading 18

Biologically controlled mineralization Highly regulated process in organisms that produces relatively complex materials with
specific biological functions and well-organized structures.
Biologically induced mineralization The precipitation and deposition of minerals resulting from secondary chemical
interactions between metabolic processes of organisms and the surrounding medium.
Biomineralization The formation of inorganic minerals by biological organisms.
Electron holography Electron interference technique that enables the extraction of amplitude and phase information for an
electron wave scattered by a thin electron transparent specimen, in a transmission electron microscope.
Magnetosome Hallmark organelle of magnetotactic bacteria. A magnetosome consists of a nano-sized iron oxide or sulfide
magnetic crystal enveloped by a lipid-bilayer membrane containing numerous proteins.
Magnetotactic bacteria Motile bacteria propelled by flagella that are capable of aligning in a magnetic field and navigating
along magnetic field lines.
Magnetotaxis The term used to describe the magnetic response observed in magnetotactic bacteria and protists. It consists of
the passive orientation and the active motility of the cell along magnetic field lines.

BCM Biologically controlled mineralization
BIM Biologically induced mineralization
MMA Magnetotactic multicellular aggregates
MMO Magnetotactic multicellular organisms
MMP Many-celled magnetotactic prokaryotes
PCR Polymerase chain reaction

Change History: May 2018. D Acosta-Avalos, Pedro Leão, Fernanda Abreu and Dennis A Bazylinski updated the text and further readings to this entire article.
This article is an update of U. Lins and D.A. Bazylinski, Magnetotaxis, In Encyclopedia of Microbiology (Third Edition), edited by Moselio Schaechter, Academic
Press, Oxford, 2009, Pages 229–241.

Encyclopedia of Microbiology, 4th Edition 1

2 Magnetotaxis


Many bacteria are motile and swim using flagella and as they do, their swimming pattern changes from being random to being
directed toward a more favorable environment. This more favorable environment, which varies from species to species, is tracked on
the basis of sensing changes in a range of stimuli that includes both chemicals like nutrients and toxins and physical parameters like
pH, light, and temperature. This behavioral response to directional stimuli is called taxis; more specifically, if the organism swims
toward the stimulus, it is called positive taxis; and if away from the stimulus, it is called negative taxis. Among the most-studied
stimuli and tactic responses in bacteria are: chemicals and chemotaxis; and light and phototaxis. This article deals with an unusual
bacterial behavior in magnetic fields called magnetotaxis. It should be pointed out, however, that magnetotaxis is very different than
other types of taxis in that magnetotactic bacteria do not swim toward or away from a magnetic field and only passively align along
magnetic field lines like a compass needle.
The geomagnetic field is considered to be one of several important environmental cues for navigation by many different
biological organisms. The Earth’s geomagnetic field is important for life on this planet for several reasons including the fact that it
creates a planetary shield against the solar wind and that it appears to be responsible for the orientation and migration by a relative
large number of biological species. Different organisms that have been shown to respond in some way to the Earth’s geomagnetic
field include homing pigeons and other migratory birds, trout, sea turtles, social insects such as honeybees and ants, and the worm
Caenorhabditis elegans. Among the microbes, the motility of several types of bacteria and protists is affected by magnetic fields. This
behavior is referred to as magnetotaxis and the organisms as being magnetotactic. Magnetotactic microbes represent the best-
understood example of magnetically influenced, biological navigation on Earth.
The first report of magnetotactic bacteria was made by Salvatore Bellini in 1963 in a publication of the Microbiology Institute
(Instituto di Microbiologia) of the University of Pavia, Italy. His microscopic observations of freshwater sediments revealed large
numbers of bacteria that swam in a single, consistent direction. Bellini realized that the bacteria swam toward the North Pole and
hence called them magnetosensitive bacteria (batteri magnetosensibili). He presumed that some sort of biomagnetic compass was
present in the bacteria. His idea proved to be true. Richard P. Blakemore, then a Microbiology graduate student at the University of
Massachusetts at Amherst, reported in the journal Science in 1975, similar magnetosensitive behavior of bacteria in sediments
collected from a freshwater swamp in Woods Hole, Massachusetts. He discovered that these magnetosensitive bacteria contained
electron-dense, iron-rich crystalline inclusions. Blakemore also realized that these microorganisms were swimming toward the
geomagnetic north in the Earth’s magnetic field and named them magnetotactic bacteria and their behavior magnetotaxis. Since
then, many multidisciplinary studies have been done on these microorganisms, many of which include recent advances toward
understanding the molecular mechanisms underlying the biomineralization and ultrastructural organization of magnetosomes in
bacteria. The controlled synthesis and organization of lipid-bilayer membrane-bounded organelles, the magnetosomes, using
specific cytoskeleton elements has caused rethinking on some of the traditional boundaries used to distinguish eukaryotic cells
from their prokaryotic counterparts, as prokaryotes were thought to not contain lipid-bilayer membrane-bounded organelles or an
organized cytoskeleton. Magnetotactic bacteria are increasingly becoming the cell biology model for understanding the molecular
mechanisms underlying the basic rules that govern the intracellular ultrastructural organization of prokaryotes.
Despite the efforts of a number of different research groups since 1980, relatively few representatives of magnetotactic bacteria
have been isolated and cultivated in axenic culture. Due to a number of important improvements in developing growth media for
these organisms, approximately 25 pure cultures of magnetotactic bacterial species have been isolated although only a subset of
these have been validly named or available in microbial repositories for study. Difficulties in culturing members of this group are
related to their fastidiousness as to specific growth requirements and their overall diversity among species. As a consequence,
relatively little is known regarding their metabolic/physiological characteristics and versatility; many studies on their diverse
morphology and phylogeny as well as mineralogy of their magnetosomes have relied on culture-independent methods.

General Features of Magnetotactic Bacteria

The term “magnetotactic bacteria” has no taxonomic significance and represents a heterogeneous group of bacteria with various cell
morphologies/phylogenies/physiologies. The operational definition of magnetotactic bacteria is that they are a heterogeneous
collection of prokaryotes that share the traits of magnetotaxis and the biomineralization of magnetosomes. The magnetosome is
defined as an intracellular organelle consisting of a single-magnetic-domain crystal of a magnetic iron mineral enveloped by a lipid-
bilayer membrane that contains proteins that are unique to it. Despite their heterogeneity, there are other features shared by these
Magnetotactic bacteria are morphologically diverse, comprising cocci, rods, spirilla, vibrios, and multicellular forms. All possess
a Gram-negative cell wall and are phylogenetically associated with Gram-negative groups of prokaryotes (the Proteobacteria and the
Nitrospirae phyla; the candidate phylum “Omnitrophica” of the Planctomycetes–Verrucomicrobia–Chlamydiae (PVC) superphylum
and, possibly the candidate phylum “Latescibacteria”) in the domain Bacteria. Although magnetotactic bacteria appear to be
restricted to the domain Bacteria, there are no known specific reasons why there are no magnetotactic representatives phylogenet-
ically associated with the Archaea and Eukarya domains. Eukaryotic microbes (protists) that respond to an applied magnetic field
similarly to magnetotactic bacteria have been reported. However, this behavior appears to be associated with the ingestion of
Magnetotaxis 3

magnetotactic bacteria by these protists, leading to a temporary response to magnetic fields. Until now, there is no evidence to fully
support autonomous magnetosome biomineralization in eukaryotes.
All currently known magnetotactic bacteria are motile by means of flagella. Because most studies on uncultivated magnetotactic
bacteria relied on the use of a magnetic field for retrieving cells from water and sediment samples, it is possible that nonmotile
bacteria that synthesize magnetosomes exist that would not swim toward the magnet and thus would not be collected and observed.
By definition, the latter organisms would be magnetic rather than magnetotactic (the term tactic here implies motility). Flagellation
patterns in magnetotactic bacteria vary greatly and include polarly monotrichous, bipolar, or lophotrichous (possessing two tufts or
bundles of flagella) representatives. The number of flagella per cell ranges from one in the marine vibrio, Magnetovibrio blakemorei
strain MV-1, to over a thousand in the many-celled magnetotactic prokaryotes (MMPs). The swimming velocity also varies greatly
between species and ranges from about 40 mm per second in members of the genus Magnetospirillum to over 1000 mm per second in
some of the magnetotactic cocci. Although motility is a key factor in magnetotaxis, little is known about the structural and
molecular components of flagella in magnetotactic bacteria. In some cases, flagella in many magnetotactic bacteria (e.g., the
magnetococci) are shorter than those of many other bacteria (e.g., Escherichia coli) (Fig. 1). Tufts or bundles of flagella are typical of
the magnetotactic cocci and originate from a disk-shaped depression or pit on the cell. Targeted disruption of the flagellin gene flaA
was found to eliminate flagella formation and motility in Magnetospirillum gryphiswaldense. Large numbers of flagella are distributed
only on one side of each of the cells of the MMP; the part of the cell exposed to the environment and thus the entire organism can be
considered to be peritrichously flagellated.
Recently several studies with the magnetotactic coccus Candidatus Magnetococcus massalia, strain MO-1, isolated from the
Mediterranean Sea and capable of swimming 300 mm per second, revealed some interesting features of the flagella properties of
the magnetotactic cocci. In this strain, the flagella are powered by proton-motive force and a sodium ion gradient. The flagellar
bundle is enclosed by a sheath described as a six-part, right-handed helical tubular structure with a diameter of about 100 nm and a
helical pitch of about 260 nm. This sheath consists of a glycoprotein called a “sheath-associated protein” (Sap) and shows some
conserved amino acid sequence with some known bacterial adhesins and eukaryotic cadherin. Every bundle contains seven flagella,
with six individual flagella organized as a hexagon and the seventh in the middle. Strain MO-1 contains 14 homologs of flagellin
genes encoded in the genome. In general, the genomes of magnetotactic bacteria contain more flagella-associated genes than the
genomes of nonmagnetotactic bacteria. This characteristic might be associated with special swimming patterns or even the very high
swimming velocities observed in the magnetotactic cocci.
Most magnetotactic bacteria swim in helical tracks. The flagella force the cell body of the bacterium to rotate around its axis while
the cell is propelled by the flagella. Under the influence of the geomagnetic field, the consequence of the action of the short flagella
is a misalignment of the trajectories that enables the cells to change the trajectory direction when they, for example, hit an obstacle
in the sediment. The small misalignments during movement might be essential for certain magnetotactic bacteria to escape from
physical “traps” present in sediments where they thrive. Recently, high-speed imaging showed that the trajectory of a magnetococcus
was linear rather than helical as previously reported.
Magnetotactic bacteria are distributed worldwide in aquatic habitats that are neutral or close to neutral in pH. In general, they
inhabit many aquatic environments where there is strong chemical stratification, that is, where vertical chemical (e.g., O2)
concentration gradients exist. In general, magnetotactic bacteria are obligately microaerophilic and/or anaerobic organisms and
display a negative tactic and/or growth response to high concentrations of O2 but also show a positive aerotactic response to low

Fig. 1 Transmission electron micrograph of a freshwater magnetotactic bacterium. Three rows or chains (arrows) of bullet-shaped magnetosomes can be seen
aligned parallel to the main axis of the cell. A flagella bundle is present at one end of the cell (white arrow). Several intracellular (internal) granules, possibly sulfur-
rich inclusions, are also present (arrowheads).
4 Magnetotaxis

concentrations of O2 (these two responses are generally referred to as aerotaxis). In natural environments they appear to prefer to be
located at the oxic–anoxic interface where they are found in their highest numbers. This interface can occur in the water column or in
the uppermost millimeters to centimeters of unconsolidated sediments depending on the environment.
In general, most magnetotactic bacteria biomineralize a single type of mineral in their magnetosomes: either the iron oxide
magnetite (Fe3O4) or one or more iron sulfide, for example, greigite (Fe3S4) and its chemical precursors. The shape of the
magnetosome crystals is also considered to be species-specific. Currently described magnetosome crystal shapes include cubocta-
hedral (roughly cubic), elongated prismatic (appear rectangular in projection), bullet- or tooth-shaped, and pleomorphic (irregu-
larly shaped more typical for greigite magnetosomes). Magnetite producers are usually found at the oxic–anoxic interface proper or
slightly above it while some magnetotactic bacteria, particularly those that produce the iron sulfide greigite, are generally found
below the oxic–anoxic interface in the anaerobic zone where sulfide is also present. It was thought that all freshwater magnetotactic
bacteria and some marine, estuarine, and salt marsh morphotypes produce magnetite. However, greigite-producing magnetotactic
bacteria, only previously found in brackish to marine environments, have been recently reported in freshwater environments. Many
magnetotactic rod-shaped magnetotactic bacteria biomineralize both magnetite and greigite magnetosomes (sometimes in the
same chain) and have also been reported from freshwater environments. The recently formally named Desulfamplus magnetovalli-
mortis strain BW-1, isolated from a brackish spring at Badwater Basin, Death Valley National Park, California (United States) is an
excellent cultured example of a magnetotactic bacterium that biomineralizes both minerals.
All magnetotactic strains in pure culture produce magnetite magnetosomes except Desulfamplus magnetovallimortis strain BW-1
which biomineralizes both mineral types in culture depending on conditions. Metabolically, all cultivated strains of magnetotactic
bacteria have a respiratory metabolism and are unable to ferment except for Desulfovibrio magneticus strain RS-1 which has a solely
chemoorganoheterotrophic metabolism and ferments pyruvate to acetate, CO2, and H2 in the absence of a terminal electron
acceptor. Most known magnetotactic bacteria are mesophilic with respect to growth temperatures. Only recently have extremophilic
MTB been described, including alkaliphilic, psychrophilic, halophilic and moderate thermophilic species. Candidatus Thermomag-
netovibrio paiutensis strain HSMV-1 is an uncultured, moderately thermophilic magnetotactic bacterium found in Great Boiling
Springs located in Nevada, United States, at temperatures close to 63 C but not higher. Three different magnetotactic strains were
isolated from highly alkaline environments in California, United States. These alkaliphilic strains, designated as ML-1, ZZ-1 and
AV-1, show optimal growth at pH values between 9.0 and 9.5. Psychrophilic magnetotactic bacteria, predominantly cocci, have
been described in sediments from the Antarctic maritime region. The sequenced genomes of most cultivated magnetotactic bacteria
contain the nif genes responsible for the fixation of atmospheric N2 and almost all strains that have been tested exhibit nitrogenase
activity. Because N2 fixation is an O2-sensitive process, microaerophilic magnetotactic bacteria might be further restricted to the
oxic–anoxic interface in nitrogen-limited environments.

Characterization of Magnetotactic Bacteria in Pure Culture

All cultured magnetotactic bacteria phylogenetically belong to the Proteobacteria phylum (Table 1), most affiliated with the
Alphaproteobacteria class. Magnetospirillum species make up most of the currently cultured magnetotactic bacteria and are freshwater,

Table 1 Examples of cultivated magnetotactic bacteria

Class Species Cell morphology Magnetosome composition Magnetosome shape

Proteobacteria Magnetospirillum gryphiswaldense strain MRS-1 Spirillum Magnetite Cuboctahedral

Alphaproteobacteria Magnetospirillum magneticum strain AMB-1 Spirillum Magnetite Cuboctahedral
Magnetospirillum magnetotacticum strain MS-1 Spirillum Magnetite Cuboctahedral
Magnetospirillum caucaseum strain SO-1 Spirillum Magnetite Cuboctahedral
Magnetospirillum marisnigri strain SP-1 Spirillum Magnetite Cuboctahedral
Magnetospirillum moscoviense strain BB-1 Spirillum Magnetite Cuboctahedral
Magnetococcus marinus Coccus Magnetite Elongated, octahedral
strain MC-1
Magnetofaba australis strain IT-1 Coccus Magnetite Elongated, octahedral
Magnetococcus massalia strain MO-1 Coccus Magnetite Elongated, octahedral
Magnetovibrio blakemorei Vibrio Magnetite Elongated, octahedral
strain MV-1
Magnetospira thiophila Spirillum Magnetite Elongated, octahedral
strain MMS-1
Magnetospira sp. strain QH-2 Spirillum Magnetite Elongated, octahedral
Proteobacteria Desulfovibrio magneticus Vibrio Magnetite Bullet-shaped
Deltaproteobacteria strain RS-1
Desulfamplus magnetovallimortis strain BW-1 Rod Magnetite or greigite Bullet-shaped or irregular
Proteobacteria Strain SS-5 Rod Magnetite Elongated, octahedral
Gammaproteobacteria Strain BW-2 Rod Magnetite Cuboctahedral
Magnetotaxis 5

facultatively anaerobic microaerophiles that show varying tolerances to O2 and biomineralize cuboctahedral crystals of magnetite in
their magnetosomes. They are chemoorganoheterotrophs that use organic acids (e.g., succinic) as a source of electrons and carbon
but several have the potential for chemolithoautotrophy as at least one strain has been shown to grow autotrophically and as many
strains possess the gene for ribulose-1,5-bisphosphate carboxylase/oxygenase, a key enzyme of the Calvin-Benson-Bassham cycle for
autotrophy, in their genomes. In addition to oxygen, most strains can use nitrate as an alternate terminal electron acceptor for
anaerobic growth. Thus far, there are seven validly described species of Magnetospirillum: M. magnetotacticum strain MS-1,
M. gryphiswaldense strain MSR-1, M. magneticum strain AMB-1, M. caucaseum strain SO-1, M. marisnigri strain SP-1, M. moscoviense
strain BB-1, and M. bellicus strain VDY. Of these seven strains, cells of M. bellicus are unusual in that they are apparently unable to
biomineralize magnetosomes. A number of other strains of this genus have also been cultivated and partially characterized but not
validly named. Interestingly, several related nonmagnetotactic spirilla with  95% 16S rDNA sequence similarity to known species
of Magnetospirillum have also been isolated. Tractable genetic systems have been developed for M. gryphiswaldense and
M. magneticum.
Cultured marine species belonging to the Alphaproteobacteria include: the marine vibrio Magnetovibrio blakemorei strain MV-1; two
species of cocci, Magnetococcus marinus strain MC-1 and Magnetofaba australis strain IT-1; and the spirillum Magnetospira thiophila
strain MMS-1 and Magnetospira sp. strain QH-2. All of these strains are facultative chemolithoautotrophs also being able to grow
chemoorganoheterotrophically. Magnetovibrio blakemorei strain MV-1 and Magnetospira thiophila strain MMS-1 utilize the Calvin-
Benson-Bassham cycle for autotrophy, while Magnetococcus marinus strain MC-1 and Magnetofaba australis strain IT-1 utilize the
reverse or reductive tricarboxylic acid cycle for autotrophy. All four strains oxidize reduced sulfur compounds (e.g., sulfide and/or
thiosulfate) as a source of electrons and produce intracellular sulfur-rich globules. In marine, chemically stratified environments
where an oxygen: sulfide double inverse concentration gradient exists, oxygen and sulfide generally overlap at the oxic–anoxic
interface. Because these marine species require both compounds for energy, these organisms are restricted to the oxic–anoxic
interface. Interestingly, many uncultivated cells of magnetite-producing magnetotactic bacteria collected from these environments
also contain intracellular sulfur-rich globules suggesting that they also oxidize reduced sulfur compounds as a source of electrons.
Strain MV-1 is a facultatively anaerobic microaerophile, capable of anaerobic respiration on some nitrogen oxides (e.g., nitrate,
nitrous oxide) while strains MC-1, IT-1 and MMS-1 appear to be obligate microaerophiles. All four species biomineralize elongated
prismatic crystals of magnetite in their magnetosomes.
A few magnetotactic bacterial strains from proteobacterial classes other than the Alphaproteobacteria have also been isolated in
axenic culture. These include strains from the Gamma- and Deltaproteobacteria. Gammaproteobacterial strains include BW-2 and
SS-5. Both are rod-shaped and are chemolithoautotrophs that oxidize reduced sulfur compounds for electrons and use the Calvin-
Benson-Bassham cycle for autotrophy. Cells of strain BW-2 and SS-5 biomineralize cuboctahedral and elongated prismatic crystals
of magnetite, respectively. Cultivated magnetotactic strains from the Deltaproteobacteria are all chemoorganoheterotrophic and
dissimilatory sulfate reducing bacteria and include: Desulfovibrio magneticus strain RS-1 a curved rod-shaped sulfate-reducing
bacterium that can also reduce fumarate as a terminal electron acceptor and has recently been shown to grow microaerobically;
Desulfamplus magnetovallimortis strain BW-1, an obligately anaerobic bacterium; and several strains of obligately anaerobic, alkali-
philic bacteria isolated from strongly alkaline habitats all very closely related to the bacterium Desulfonatronum thiodismutans. All
deltaproteobacterial magnetotactic bacteria biomineralize bullet-shaped magnetite crystals with the exception of
D. magnetovallimortis strain BW-1 which biomineralizes greigite as well as bullet-shaped magnetites.

Characterization of Uncultured Magnetotactic Bacteria

The presence of magnetosomes and the trait of magnetotaxis allow for the magnetic manipulation and easy retrieval of uncultured
magnetotactic bacteria from environmental samples with the aid of a magnetic field. Our current knowledge regarding the diversity
of magnetotactic bacteria is based more on microorganisms harvested from the environment than on the isolation and character-
ization of pure cultures. Thanks to modern molecular techniques including the polymerase chain reaction (PCR), it is relatively easy
to obtain and analyze the sequence of their 16S rRNA genes and therefore be able to determine phylogenetic relationships of these
The most commonly found morphotype of magnetotactic bacteria found in marine and freshwater environments are the
magnetococci and thus numerous culture-independent studies have been focused on these organisms (Table 2). In addition, the
magnetotactic cocci, unlike many other magnetotactic morphotypes, persist for long periods in microcosm experiments. All known
magnetotactic cocci are phylogenetically affiliated with the Alphaproteobacteria class of the Proteobacteria phylum but are not closely
related to any other alphaproteobacterium and appear to form a unique lineage within the group. All magnetotactic cocci
biomineralize elongated prismatic crystals of magnetite although the arrangement of the crystals is variable including those with
multiple chains and those with a clump (rather than a chain) of magnetosomes at one side of the cell (Fig. 2).
The techniques of gold labeling in conjunction with in situ hybridization and transmission electron microscopy revealed the
presence of at least three different types of uncultured magnetococci in the Itaipu Lagoon in Brazil (Fig. 3). In this case,
polyribonucleotide probes were labeled with digoxigenin or fluorescein and detected by immunolabeling with antifluorescein
and antidigoxigenin antibodies bound to 10 and 15 nm gold particles. The use of this technique allowed for the correlation of
phylogenetic association with a specific magnetosome morphology and organization; that is, to determine which 16S rRNA gene
sequence belonged to which of the three cocci. The cocci were distributed as two clusters, referred to as Itaipu-1 and Itaipu-2, and
were closely related based on 16S rRNA gene sequences which also showed they belonged to the Alphaproteobacteria.
6 Magnetotaxis

Table 2 Uncultivated magnetotactic bacteria

Phylum Bacteria
Class Sampling site Cell morphology Magnetosome composition Magnetosome shape

Proteobacteria Uncultured coccus Itaipu-I Coccus Magnetite Elongated, prismatic

Alphaproteobacteria Brazil
Uncultured coccus Itaipu-III Coccus Magnetite Elongated, prismatic
Proteobacteria Candidatus Magnetoglobus multicellularis Spherical MMP Greigite Irregular
Deltaproteobacteria Brazil
Ca. Magnetomorum litorale Spherical MMP Magnetite Bullet-shaped
Germany Spherical MMP Magnetite or greigite Bullet-shaped or irregular
Ca. Magnetomorum rongchengroseum
Ca. Magnetananas tsingtaoensis Elliptical MMP Magnetite Bullet-shaped
Ca. Magnetananas rongchenensis Elliptical MMP Magnetite Bullet-shaped
Uncultured barbell-shaped bacterium Coccus Not determined Not determined
United States
Proteobacteria North-seeking bacterium strain NS-1 Vibrio Magnetite Elongated, prismatic
Gammaproteobacteria Brazil
Nitrospirae Ca. Magnetobacterium bavaricum Rod Magnetite Bullet-shaped
Ca. Magnetoovum mohavensis Ovoid Magnetite Bullet-shaped
United States
Ca. Thermomagnetovibrio paiutensis Vibrio Magnetite Bullet-shaped
strain HSMV-1
United States
Ca. Magnetobacterium bremense Rod Magnetite Bullet-shaped

Fig. 2 Transmission electron microscopy image of a magnetotactic coccus containing a “clump” of apparently disorganized magnetosomes at one side of the cell
rather than an organized chain. (A) Image of the magnetotactic coccus showing the presence of magnetosomes and granules. (B) Higher magnification image of

A number of as yet uncultured magnetotactic bacteria are phylogenetically affiliated with the Nitrospirae phylum. Morphotypes
of MTB phylogenetically associated with this phylum include: large rod- and ovoid-shaped cells and vibrios (Table 2).
Large numbers of a barbell-shaped magnetotactic bacterium were discovered in a chemically stratified coastal salt pond that
displayed unusual magnetotactic behavior (discussed in the Magneto-aerotaxis section). This microorganism appears to consist of a
chain of two to five cocci and belongs phylogenetically to the Deltaproteobacteria. Unfortunately, the composition of the magneto-
some mineral was not determined.
An uncultured, relatively large (5 mm long) slow-moving rod was retrieved from the chemically stratified Salt Pond, Woods
Hole, United States and was found to be a member of the Gammaproteobacteria class of the Proteobacteria although there is some
doubt as to the phylogenetic affiliation of this organism (Table 2).
Magnetotaxis 7

Fig. 3 Transmission electron microscope image of an ultra-thin section of a cell of a magnetic coccus collected from the Itaipu Lagoon in Brazil classified as Itaipu-
1. Section was hybridized with a 16S rRNA gene-digoxigenin probe and detected by antidigoxigenin antibodies conjugated with 10 nm gold particles (arrows).
Magnetosomes can be seen (arrowhead).

Although one magnetotactic bacterium capable of producing both magnetite- and greigite-containing magnetosomes has been
isolated in pure culture (discussed in the previous section), a number of uncultured magnetotactic bacteria have been described that
are capable of biomineralizing both minerals and partially characterized using culture-independent techniques, including geno-
mics. Evidence for the common ancestry of magnetite and greigite biomineralization in magnetotactic bacteria was revealed through
the partial sequencing of the genome of an uncultured greigite-producing multicellular magnetotactic bacterium from sulfide-rich
environment after magnetic concentration and separation. This approach allowed for the identification of specific genes considered
unique for magnetite magnetosome synthesis.
Morphologically conspicuous, multicellular magnetotactic bacteria have been reported from a number of different brackish,
marine, or hypersaline environments in both the Northern and the Southern Hemispheres, usually in sulfide-rich environments
very likely to be anaerobic (Table 2). These very unusual organisms are assemblage of Gram-negative cells that generally contain
chains of greigite-containing magnetosomes. These organisms are not a consortium of cells of different species, but an aggregate of
genetically identical cells. All those examined are phylogenetically affiliated with the Deltaproteobacteria class of the Proteobacteria
phylum (Table 2). For many years, because of their unique morphology, these multicellular bacteria were thought to represent a
single species. However, phylogenetic analyses of a collection of these organisms from a natural population present in a salt marsh
(Falmouth, MA, United States) showed that this group of multicellular prokaryotes had a number of different 16S rRNA gene
sequences separated by at least 5% divergence, suggesting that they are not a single species but represent a number of distinct species
within a single genus. An unaffiliated multicellular magnetotactic bacterium capable of producing both greigite and magnetite
magnetosomes was also reported (Fig. 4). Although several ellipsoidal-shaped types containing both magnetite and greigite have
been characterized (Fig. 5), most multicellular magnetotactic bacteria are spherical in shape (Fig. 6).
These organisms have not yet been validly named (largely because they have not been isolated and grown in pure culture) and
confusing and imprecise terminology has been used for them including magnetotactic multicellular aggregates (MMAs), many-
celled magnetotactic prokaryotes (MMPs), and magnetotactic multicellular organisms (MMOs).
Like other spherical multicellular magnetotactic prokaryotes, Candidatus Magnetoglobus multicellularis is a member of the
Deltaproteobacteria (Table 2). The most conspicuous structural feature of these microorganisms is the fact that its cells appear to be
precisely organized and must live as a “colony” or unit and are unable to live individually (Fig. 7). Each cell faces both the external
surrounding environment and the internal portion of the aggregate. If a cell leaves the assemblage, for example, when the organism
dies, it loses its ability to swim and to orientate along magnetic field lines.
Candidatus Magnetoglobus multicellularis exhibits a unique method of reproduction. This has been studied in some detail and
the following growth stages have been proposed: (1) individual cells grow in size thereby increasing their volume and their number
of magnetosomes; (2) the cells then divide doubling the number of cells in the microorganism; (3) individual cell within the
microorganism then reorganize; and (4) the microorganism divides into two new and identical units.
On the outer surface of the microorganism, cells of multicellular magnetotactic prokaryotes are covered with numerous flagella
that are responsible for its coordinated complex movement. Four different types of motility have been observed in Candidatus
Magnetoglobus multicellularis: free motion, rotation, walking, and escape motilities (Table 3). Coordination among cells in the
multicellular prokaryote are considered necessary for all types of motility. A negative photo response has been observed in a
8 Magnetotaxis

Fig. 4 Transmission electron microscopy image of a cell of a multicellular magnetotactic bacterium containing chains of bullet-shaped magnetite (arrowheads)
and rounded greigite magnetosomes (arrows).

Fig. 5 Electron microscopy of ellipsoidal MMPs. (A) Scanning electron microscopy image of the microorganism showing cells organized as an ovoid-shape
structure. (B) Transmission electron microscopy image of an ellipsoidal MMP showing cell arrangement inside the microorganism. Image in (A) was kindly given by
Prof. Tian Xiao and Prof. Long-Fei Wu.

multicellular magnetotactic prokaryote when illuminated with high intensity UV light (365 nm), violet-blue light (395–440 nm
filter) of about 80 W m2 of intensity and blue light (450–490 nm filter) of about 200 W m2 of intensity. For longer wavelengths
no photo response was observed. Photo kinesis has been observed in this multicellular magnetotactic prokaryote, decreasing their
velocity when illuminated with green light (517 nm, 0.46 W m2) and increasing their velocity when illuminated with red light
(628 nm, 0.16 W m2), both respectively to the velocity observed when illuminated with blue light (469 nm, 0.8 W m2). That
effect is related to the presence of a constant magnetic field and it can be canceled in the presence of radio-frequency electromagnetic
field at the Zeeman resonance frequency, showing the involvement of a radical pair mechanism, a very well-known magnetorecep-
tion mechanism used by migratory birds.
Unusual ellipsoidal-shaped multicellular magnetotactic prokaryotes were discovered in sediments and water from the intertidal
zone of the Yellow Sea in Qingdao (Tsingtao) City, China and were present along with spherical types (Table 2). Magnetosomes
Magnetotaxis 9

Fig. 6 Scanning electron microscopy image of Candidatus Magnetoglobus multicellularis showing its multicellular nature and its overall spherical morphology.

Fig. 7 Transmission electron microscopy image of an ultra-thin section of Candidatus Magnetoglobus multicellularis showing the general organization of cells and
the internal compartment (at asterisk). All cells have a cell wall surface structure of Gram-negative bacteria and face both the external environment and the internal
compartment. The numerous magnetosomes occur mostly at the periphery of the cells (arrows). Lipid or polyhydroxyalkanoate inclusions are marked (L).

were organized in chains close to parallel to the long axis of the organism. Motility in the ellipsoidal multicellular magnetotactic
prokaryotes was similar to that of spherical types, showing the characteristic “escape” motility and also negative phototaxis for UV,
violet and blue light.
Spherical multicellular magnetotactic prokaryotes change their behavior with magnetic field intensity in a way that cannot be
explained by the simple compass model used for magnetotaxis in unicellular bacteria. This unexpected behavior led to the
hypothesis that multicellular magnetotactic prokaryotes have a sort of magnetoreception mechanism in magnetic fields higher in
10 Magnetotaxis

Table 3 Motility of multicellular magnetotactic prokaryotes

Motility Description

Free The microorganism swims at speed up to 100 mm s1 in either straight or helical trajectories along magnetic field lines in a uniform magnetic field
Rotation At the edge of a drop under atmospheric conditions, the microorganism spins around an axis that passes through its center
Walking At the air–water interface on the top of the drop, the microorganism swims freely in a complex trajectory, maintaining the same sense of body rotation
Escape This movement consists of a backward movement (north-seeking in the Southern Hemisphere or south-seeking in the Northern Hemisphere) for tens
or even hundreds of micrometers, followed by a forward regular movement. In the backward movement, the microorganisms decelerate
continuously with time, whereas in the forward movement that follows, they display uniform acceleration. Also known as “ping-pong” response

intensity than that of the Earth’s geomagnetic field. This magnetoreception mechanism and the unusual “escape” motility might
imply transference of information from the magnetic interaction of the magnetosome chains to the flagellar motors. There is,
however, no scientific evidence for any type of magnetic “sensor” such as this in magnetotactic bacteria.


Many prokaryotic microorganisms facilitate the formation of a large number of various minerals, including magnetosome crystals,
through one of two biomineralization processes. These include biologically induced mineralization (BIM) and biologically
controlled mineralization (BCM). There are several important differences between BIM and BCM. The most important differences
relate to the control of the synthesis of the biominerals and the functionality of the biomineralized products. Generally, in BIM the
microorganism exerts little control over the biomineral, which has no known recognized function for the bacterium. In BCM, the
microorganism exerts strict crystallochemical control over the nucleation and growth of the biomineral, which in some cases has a
somewhat defined function. Because of these features, BCM processes are considered to be under specific biochemical and genetic
control. Examples of products of BCM in higher organisms are bones, teeth, or shells. In the microbial world, most known
biominerals are produced by bacteria through BIM. The mineral particles produced by bacteria in BCM are characterized as well-
ordered crystals with narrow size distributions, high chemical purity, and specific particle morphologies (Fig. 8). Generally,
biomineralized particles produce through BIM lack these qualities. The most characterized example of BCM in bacteria is the
Magnetosomes are iron oxide or sulfide magnetic crystals, each enveloped by a lipid bilayer membrane (i.e., the magnetosome
membrane). As previously stated, there are two compositional types of magnetic minerals in magnetosomes: magnetite (Fe3O4) and
greigite (Fe3S4). Additional nonmagnetic iron sulfide minerals, including mackinawite (tetragonal FeS) and a cubic form of FeS,
have been found in some greigite-producing magnetotactic bacteria and are assumed to represent precursors of greigite. In most
magnetotactic bacteria, magnetosomes are organized in a chain or chains within the cell. This has important physical significance
and is discussed later.
The nature of the magnetosome membrane has been studied in several magnetite-producing Magnetospirillum species. The
magnetosome membrane appears to originate as an invagination of the cell membrane and contains a number of proteins unique
to it that are thought to be responsible for regulating the size, the chemical purity, and the morphology of the magnetosome crystal.
The crystalline phase of the magnetosome has been extensively studied using a number of forms of electron microscopy.
Magnetotactic bacterial magnetite typically has cubic {100}, octahedral {111}, and dodecahedral {110} faces in crystals that have
quasi-rectangular projected shapes in the horizontal plane or are bullet-, tooth-, or arrowhead-shaped (Fig. 9). Morphologies of
greigite crystals are similar and include the same forms as magnetite but, in contrast, the crystals often appear “wrinkled” and have
an irregular border. The number of magnetosomes varies from a few to thousands per cell.
Regardless of mineral type, magnetosome crystals are generally in the size range of 35–120 nm, which corresponds to the single-
magnetic-domain size range for magnetite or greigite. However, magnetosomes up to 250 nm have been reported. The significance
of the size of the mineral crystals is discussed in the next section.

Biophysics of Magnetotaxis

As previously stated, in general, the sizes of magnetosome crystals in magnetotactic bacteria are within the permanent single-
magnetic-domain size range for magnetite ( 35–120 nm), meaning that they are uniformly magnetized with the maximum
magnetic dipole moment per unit volume. Magnetosomes larger than single magnetic domains are not uniformly magnetized
because of the formation of domain walls, resulting in multiple magnetic domains. The overall effect of being larger than the single-
magnetic-domain size range is the significant reduction of the magnetic dipole moment. On the other hand, very small particles
(<35 nm) are superparamagnetic, which means that they are still uniformly magnetized but their magnetic dipole moments are not
Magnetotaxis 11

Fig. 8 Transmission electron microscopy images of magnetosome magnetite crystal morphologies. (A) Cell of Magnetospirillum magnetotacticum containing a
chain of cuboctahedral magnetosomes (arrow), (B) an uncultivated marine coccus with two chains of elongated prismatic magnetosomes (arrows), (c) uncultured
freshwater magnetotactic bacterium with rows of bullet-shaped magnetosomes (arrows) and sulfur-rich inclusion (arrowheads).

constant because of thermally induced spontaneous reversals. This results in a net magnetic moment of zero over time. Thus, the
single-domain size range is the optimum particle size for the maximum magnetic dipole moment in a magnetosome.
Magnetosomes are anchored to the cell membrane in most, if not all, magnetotactic bacteria and are thus fixed in place within
the cell. In most magnetotactic bacteria, the magnetosomes are arranged as a chain or multiple chains. Experimental evidence
involving Magnetospirillum species shows that magnetic interactions between the crystals are not sufficient to align them in a chain.
The initial step in the biomineralization process is the invagination of the cytoplasmic membrane to form the magnetosome
membrane vesicle. Magnetite or greigite are biomineralized within these vesicles and may possibly nucleate even while the
magnetosome membrane is still an invagination with a connection to the periplasm. In most magnetotactic bacteria, at some
point, it is thought that the membrane invagination detaches from the cytoplasmic membrane, forming a separate magnetosome
vesicle. However, in cells of Magnetospirillum magneticum strain AMB-1, the magnetosome membrane may remain as cell membrane
invaginations. In both cases, cytoskeletal filaments composed by the protein MamK, an actin-like MreB homologue, are necessary to
organize the magnetosomes in chain in the cytoplasm and attach this structure to the cellular framework. This results in an efficient
transfer of torque to the cell body when the cell is in a magnetic field.
A number of the genes thought to be involved in the formation of the magnetosome membrane and biomineralization of
magnetite/greigite have been studied. Several different genes have been identified that encode magnetosome-associated proteins
using cultured Magnetospirillum species as a model (mainly Magnetospirillum gryphiswaldense strain MSR-1). Most genes associated
thought to be involved with magnetosome biogenesis are unique to magnetotactic bacteria. In Magnetospirillum gryphiswaldense
12 Magnetotaxis

Fig. 9 High-resolution transmission electron microscopy image of a bullet-shaped magnetite magnetosome biomineralized by magnetotactic bacteria.

these genes are clustered in a region of the genome that is flanked and interrupted by transposase and tRNA genes. These are features
typical of what is referred to as a genomic island and thus, this genomic region in Magnetospirillum gryphiswaldense is referred to as a
magnetosome island (MAI). In the magnetotactic Alphaproteobacteria, set of approximately 30 genes have been described and named
the mam (magnetosome membrane) and mms (magnetic particle membrane specific) genes. The following steps have been described
for the BCM biomineralization process of magnetosomes based on studies using Magnetospirillum species: (1) invagination of the
cytoplasmic membrane to form the magnetosome vesicle; (2) iron transport into the vesicle by the action of the cation transporters
MamM and MamB; (3) nucleation and controlled growth of the magnetic crystals;(4) formation of magnetosome chains through
the interaction of MamJ on the magnetosome vesicle with the cytoskeletal filament protein MamK and positioning the magneto-
some chain in the middle of the cell.
The magnetosome chain(s) within the cell impart a magnetic dipole moment to the cell that overcomes the thermal forces that
tend to randomize the position of the cell in the aqueous environment in which the bacteria live. The magnetosome chain causes
the cell to behave as a miniature magnetic compass needle, continuously causing the cell to passively align along local magnetic
field lines. This alignment is not perfect however and disturbed mostly by the thermal fluctuations of the aqueous medium. Thus
the alignment of magnetotactic bacteria along geomagnetic field lines depends on the ratio the magnetic to thermal energies, or MB/
kBT, where M is the magnetic dipole moment of the cell, B is the geomagnetic field strength, kB is the Boltzmann constant, and T is
the ambient temperature. When magnetosomes are arranged in a linear chain within the bacterial cell, the total magnetic dipole
moment of the cell is the sum of the magnetic dipole moments of the individual crystals. The organization of magnetosomes in
chains causes the individual magnetic dipole moments of the single-magnetic-domain magnetosome crystals to orient parallel to
each other along the chain. In this case, the permanent magnetic dipole moment for the entire chain is optimum, that is, it
approaches the maximum value possible. Electron holography in the electron microscope, magnetic force microscopy and pulsed-
magnetic-field remanence measurements on individual cells have demonstrated the permanent magnetic microstructure of the
magnetosome chain(s).
At ambient temperature, 10–20 magnetosomes (50 nm in diameter) in a chain would be sufficient for passive orientation of a
magnetotactic bacterium along the Earth’s magnetic field lines. The efficient orientation in the geomagnetic field results in the
migration of the cell along the magnetic field lines as it swims. If the magnetic field is increased (e.g., by exposing the cells to a bar
Magnetotaxis 13

magnet), the time-averaged orientation of the cell along the field is increased and the migration rate of the cell in the magnetic field
direction is increased, even though the swimming speed of the cell remains unchanged.
Intriguingly, many magnetotactic bacteria seem to produce more magnetosomes than is necessary for orientation suggesting that
magnetosomes might be responsible for other roles, perhaps physiological, than simply the magnetic orientation of the cell.
Unusually large magnetite-containing magnetosomes (up to 250 nm in length) were reported in an uncultured magnetotactic
coccus, named Itaipu-1, collected from the Itaipu Lagoon, near Rio de Janeiro, Brazil. Because of the large size of the magnetic
crystals, they would be expected to be outside the single-magnetic-domain range. However, electron holography measurements
demonstrated that the crystals in fact behave as single-magnetic-domains while they are organized in chains. Breaking or removal of
these magnetosome crystals from the chains results in a complicated magnetic microstructure with domain walls.
The functional significance of the unusually large volume of the magnetosomes is unknown. It has been speculated that the large
crystals in Itaipu-1 occur in order to compensate for the anomalous low magnetic field strength (0.23 gauss) of the South Atlantic
Magnetic Anomaly region, which includes Itaipu Lagoon. However, a calculation of the total magnetic dipole moment for a typical
Itaipu-1 cell results in MB/kBT ¼  250, which is over 10 times the minimum value required for efficient magnetotaxis, suggesting
there may be other functions of the magnetosomes in Itaipu-1.


It is important to understand that magnetotaxis is somewhat of a misnomer and does not represent a true taxis response as in
phototaxis or chemotaxis. Magnetotactic cells do not swim toward or away from a magnetic field source. Instead, magnetotaxis is
the combination of the passive orientation and active motility by flagella, of magnetotactic bacterial cells along magnetic field lines.
All known magnetotactic bacteria generally swim bidirectionally although some are known to “loop” while swimming. None
exhibit run-and-tumble motility as do some nonmagnetotactic bacteria (e.g., Escherichia coli). Most magnetotactic bacteria swim
following an helical-like trajectory whose characteristics depends on the relative inclination of the magnetosome chain and the
flagella, according to theoretical studies done on mono- and bi-flagellated bacteria.
Magnetotaxis is considered advantageous to bacteria in vertical chemical concentration gradients (e.g., O2) in locating and
maintaining an optimum position in the gradient because magnetotactic bacteria use the Earth’s magnetic field lines as a fixed
reference to swim, unlike other bacteria. Presumably, if a magnetotactic bacterium is displaced from its optimal position, it would
use geomagnetic field lines to more efficiently return to its optimal position. Magnetotactic bacteria, after aligning along geomag-
netic field lines, would have a one-dimensional search problem in locating their optimal position in a vertical chemical concen-
tration gradient whereas nonmagnetotactic bacteria would have a three-dimensional search problem, presumably expending more
time and more energy.
The original model of magnetotaxis stated that all magnetotactic bacteria had one of two possible swimming polarities: north-
seeking and south-seeking. The swimming polarity depends ultimately on the direction of motion of the cell under oxic conditions.
This was based on the observation that in the Northern Hemisphere, magnetotactic bacteria predominantly swim parallel to the
magnetic field toward the north, hence the term north-seeking. In the Southern Hemisphere, the opposite was observed and south-
seeking bacteria swim antiparallel to the field. Because, the geomagnetic field is inclined downward from the horizontal in the
Northern Hemisphere and upward in the Southern Hemisphere, with the magnitude of the inclination increasing from the equator
to the poles, the outcome of this situation is that in both hemispheres the bacteria swim downward toward the region of the
sediment with little oxygen or no oxygen. In the original model, once a bacterium reached a layer of the sediment with suitable
oxygen concentration, it would stop moving and possibly adhere to a sediment particle.
Two important observations were inconsistent with this magnetotaxis model: (1) the observation that north-seeking magneto-
tactic bacteria in pure culture (e.g., Magnetococcus marinus strain MC-1) form horizontal bands at certain positions below the
meniscus in culture tubes containing an O2 concentration gradient, instead of swimming persistently to the bottom of the culture
tube; and (2) some populations of uncultured magnetotactic bacteria were observed in the water column in chemically stratified
aquatic environments, not in the sediments. Should the original model be true, all magnetotactic bacteria would swim to the
bottom of culture tubes and to the sediment in natural aquatic environments. An expansion of the magnetotaxis model was then
proposed based on the observation that magnetotaxis appears to function in conjunction with aerotaxis. This is known as the
magneto-aerotaxis model.


Originally, two different magneto-aerotactic mechanisms were proposed: axial and polar magneto-aerotaxis. In axial magneto-
aerotaxis, cells do not show a polar preference and the magnetic field defines the axis but not the direction of swimming. This
mechanism is typical of cells of Magnetospirillum species grown in liquid media. Cells spontaneously reverse their swimming
direction in an applied magnetic field. Thus, the distinction between north-seeking and south-seeking does not apply to axial
magneto-aerotactic spirilla. Recent studies on cells of 12 strains of cultured magnetotactic bacteria of various morphotypes,
including spirilla, in microcapillary tubes containing an O2 concentration gradient and placed in a homogeneous magnetic field
showed six different behaviors for north-seeking magnetotactic bacteria when the O2 gradient and the magnetic field were parallel,
14 Magnetotaxis

and not simply polar and axial magneto-aerotaxis as described above. It was proposed from the results, that there are three general
magneto-aerotactic mechanisms: axial, dipolar (formerly called polar) and unipolar magneto-aerotaxis.
In dipolar magneto-aerotaxis, the most common mechanism observed in natural populations of magnetotactic bacteria, cells
show a polar preference in their swimming direction. The magnetic field provides both direction and axis for the magnetotactic
behavior. This is based on experimental findings of the cultured coccus Magnetococcus marinus strain MC-1. In the magneto-aerotaxis
model, magnetotactic bacteria can be either in an oxidized or in a reduced state. In the oxidized state, when in the oxic zone of a
vertical concentration gradient of O2, cells swim downward. They do this by rotating their flagella counterclockwise (Fig. 10).
However in the reduced state, their flagella would rotate clockwise and the bacteria would reverse direction without turning around
(they would still be aligned along magnetic field lines) and thus move upward, again toward the optimal O2 concentration. In this
model, magnetotactic bacteria still benefit from having a fixed axis for swimming.
The dipolar magneto-aerotaxis model can be extended to redoxtaxis in habitats in which spatially separated reservoirs of
reductants and oxidants coexist. In these cases, some magnetotactic bacteria would swim into anoxic zones of their habitat to
accumulate reduced compounds possibly as a source of electrons for growth (e.g., reduced sulfur compounds) and then swim back
to zones in which oxidized compounds (e.g., O2) are present and perhaps required for growth.
Unipolar magneto-aerotaxis is similar to the dipolar magneto-aerotaxis, although cells swim persistently in one direction
(toward the anoxic region or toward the oxic region) using the geomagnetic field as a guide. Once the persistent anoxic unipolar
magnetotactic bacteria reach an oxic region, they reverse their swimming direction, reaching the anoxic zone again and vice-versa.
There is evidence that magneto-aerotaxis does not explain the behavior of all magnetotactic bacteria particularly those in natural
environments. For example, a population of mainly south-seeking magnetotactic bacteria (the barbell-shaped bacterium discussed
in an earlier section) has recently been discovered in a chemically stratified environment in the Northern Hemisphere. In the
Southern Hemisphere, a population of north-seeking magnetotactic bacteria was also characterized in a lagoon in the South
Hemisphere. These microorganisms seem to represent a deviation from the magneto-aerotaxis model for magnetotactic bacteria as
they were found in the anoxic zone of their environment and putatively swam in the theoretically “wrong” direction, that is, they
had the “wrong” magnetic polarity. In these situations, if cells follow the polar magneto-aerotaxis principle, they would swim
persistently to a less favorable/optimal environment.
For cells of Magnetospirillum gryphiswaldense, magneto-aerotaxis experiments showed that an increase of the geomagnetic field
intensity 10-fold or a tilt of 45 degree of the magnetic field relative to the O2 concentration gradient did not significantly change the
magneto-aerotaxis efficiency, although it was drastically reduced when the magnetic field was null or when the tilt was 90 degree.
These results show that magneto-aerotaxis efficiency is not proportional to the cosine of the tilt angle, as assumed by classical
models of magneto-aerotaxis.
In both polar and dipolar magnetotactic bacteria, the swimming polarity can be switched from north-seeking to south-seeking
and vice-versa by exposing cells to demagnetizing magnetic fields of 50 Hz and above 0.1 T, or even at pulses of 1 to 5 ms at about
0.07 T. These observations were used to show that magnetosome magnetite crystals are single-magnetic-domains. There is now
evidence that redox gradients combined with manipulations of the magnetic field direction can also change the swimming polarity
of magnetotactic bacteria. In cultures of Magnetospirillum gryphiswaldense, it has been shown that the swimming behavior of cells can
be switched from axial to dipolar in the presence of O2 concentration gradients and magnetic fields oriented parallel or antiparallel
to the gradient. In addition, sudden increases in the local O2 concentration results in the conversion of cells to the opposite polarity.
Collecting and/or concentrating magnetotactic bacteria from mud and water samples using strong magnets can also reverse their
swimming polarity and thus caution must be used in determining the polarity of magnetotactic bacteria in environmental samples.

Fig. 10 Schematic drawing showing how magneto-aerotaxis may help magnetotactic bacteria to locate and maintain an optimal, preferred position in natural
habitats with vertical chemical concentration gradients (e.g., O2). See text for details.
Magnetotaxis 15

In addition, a high number of chemotactic receptor genes are present in many magnetotactic bacterial genomes (e.g., 56 genes
appear to encode for chemotactic receptors in Magnetospirillum gryphiswaldense strain MSR-1) in comparison to other bacterial
genomes (e.g., five genes encoding chemotactic receptors in Escherichia. coli), possibly indicating that chemotaxis is more complex in
magnetotactic bacteria possibly suggesting that the interaction between chemotaxis and magnetotaxis is also complex. Thus,
additional models of magnetotaxis, possibly associated with chemical concentration gradients other than O2 (e.g., sulfide), are
likely needed to explain the behavior of some magnetotactic bacteria.

Magnetotactic Eukaryotes

A magnetotactic response was reported in some eukaryotic microbes (protists) collected from coastal areas. Transmission electron
microscopy revealed intracellular iron-rich crystalline inclusions identical to bacterial magnetosome magnetite crystals in some of
these organisms. Intact magnetotactic bacteria were not observed in the cells. Two possibilities exist for the origin of these
magnetosomes. First, the protist itself might biomineralize magnetosomes. This seems to be the case for the magnetotactic
euglenoid alga Anisonema platysomum found to be present in a marine habitat in Brazil. This protist produces rows of chains of
bullet-shaped magnetosomes in an elaborate pattern along the long axis of the cell. The morphology of the crystals is uniform and
the magnetotactic response is consistent with the magnetic polarity of magnetotactic bacteria from the same environment as they
show a directional swimming preference in the magnetic field.
Second, bacteria-grazing protozoa may ingest magnetotactic bacteria and their magnetosomes, which may be stable or partially
or fully solubilized. They might be incorporated into the cell or present for a short time in food vacuoles, the likely site for
solubilization. This seems to be the case for certain biflagellates, dinoflagellates, and ciliates detected in coastal areas. The
significance and implications of this, as well as details of protozoal grazing of magnetotactic bacteria, is discussed later. Regardless
of the origin of magnetite or greigite crystals in protists, some of these eukaryotic microbes contain a significant amount of
intracellular iron.

Geochemical, Geophysical, and Astrobiological Aspects of Magnetotactic Bacteria

Cultured magnetotactic bacteria are known to facilitate a number of important geochemical processes and thus it is likely they play
important roles in these processes in natural environments. These include sulfide oxidation, sulfate reduction, denitrification,
nitrogen fixation, and iron redox reactions (iron cycling) through magnetosome biomineralization and possibly through iron
oxidation not related to magnetosome synthesis. In addition, all cultured marine strains are facultatively autotrophic through either
the Calvin-Benson-Bassham or the reverse (or reductive) tricarboxylic acid cycles. The freshwater magnetospirilla show potential for
autotrophy at least one species has been shown to grow autotrophically and as most appear to contain in their genomes an intact
ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) gene, which encodes for one of the key enzymes in the Calvin-
Benson-Bassham cycle. Thus magnetotactic bacteria have great potential in regulating important geochemical reactions and cycles
involving several important elements including sulfur, nitrogen, iron, and carbon.
Protozoa are known to be important components of the marine ecosystems as the primary consumers of bacteria and recyclers of
certain nutrients, including iron. Magnetotactic protists might play an important role in iron cycling either through direct
biomineralization or through grazing on magnetotactic bacteria and solubilizing magnetosomes. Iron is a limiting factor in primary
production in some oceanic and coastal areas. This element, which is often present in seawater in particulate and colloidal forms,
can be digested in the food vacuoles of protozoans during grazing of particulate and colloidal matter. This process might generate
more bioavailable iron for other species in the food chain. The chemical stability and thus the persistence of magnetosomes and
magnetosome crystals depend on its surrounding environment. Magnetite magnetosomes appear to be quite stable in the bacterial
cell possibly due to the magnetosome membrane and there is no evidence that magnetotactic bacteria can reuse iron in magneto-
somes for another purpose even when cells are starved for iron. Is it possible for protozoans grazing on magnetotactic bacteria to
make ingested magnetosome iron more bioavailable through solubilization? One possible mechanism of conversion of biologi-
cally unavailable forms of iron into more soluble forms is through the action of acidic food vacuoles on solid or colloidal forms of
iron in protozoa. This has been observed to occur with colloidal iron and with greigite magnetosomes from magnetotactic bacteria.
The dissolution of greigite magnetosomes by the ciliate Euplotes vannus has been demonstrated in in vitro experiments. The
ciliates ingested greigite-producing magnetotactic bacteria, the multicellular magnetotactic prokaryote discussed earlier, which
became enveloped in food vacuoles known to have an internal acidic pH. Transmission electron microscopy (Fig. 11) and X-ray
microanalysis of thin-sectioned ciliates showed that within the acidic vacuoles, the greigite crystals dissolved at different rates after
the magnetosome membrane was digested. The experiments with Euplotes suggest that protozoans that graze on magnetotactic
bacteria in natural environments are likely important in iron cycling as they may help to solubilize iron trapped in the magneto-
somes, making it bioavailable to other organisms. The impact of magnetite magnetosome digestion by protozoans remains
unknown, but, based on the experiments with greigite magnetosomes, there is potential for magnetosome magnetite to also be
solubilized by protozoans.
A simple calculation may help estimate the potential of the recycling of magnetosomes produced by magnetotactic bacteria.
Each multicellular magnetotactic prokaryote contains about 30 cells with approximately 100 magnetosomes per cell, which leads to
16 Magnetotaxis

Fig. 11 Transmission electron microscopy image of an ultra-thin section of a ciliated protozoan that has ingested a number of magnetotactic multicellular bacteria.
The bacterial cells (white arrows) are in vacuoles where they appear to be in the process of being digested. Magnetosome clusters can be seen (dark arrows). The
inset shows a high-magnification image of a cluster of magnetosomes in different degrees of dissolution. No magnetosome membrane is seen in heavily digested
crystals (arrow); m, mitochondria; L, Lipid inclusions.

each aggregate containing about 3000 magnetosomes. Using an average magnetosome size of 100 nm, which results in a cubic
volume of 106 nm3, and a greigite density of 4.1 g/cm3, each multicellular unit contains an estimated total amount of 1.2  1011 g
of greigite that could potentially be solubilized within the ciliates and eventually released to the environment. Marine magnetotactic
bacteria are known to reach population densities of 104 cells ml1 in some natural environments. Taking into consideration the
large amount of iron concentrated by magnetotactic bacteria in intracellular magnetosomes (1013 to 1015 g iron per cell) and
their estimated population density, in some habitats, nanomolar concentrations of iron may be sequestered by magnetotactic
bacteria and then solubilized by grazing protozoans that ingest them, thereby recycling significant amounts of iron.
Magnetotactic bacteria are considered to be a main source of the magnetization of sediments in a number of freshwater and
marine environments. In many of these environments, magnetosome crystals, particularly magnetite, is released from dead, lysing
cells and accumulate in sediments and persist relatively unaltered for long periods of time. In some cases, even magnetosome chains
persist. Because magnetosomes are single-magnetic-domain particles, they form ideal carriers of paleomagnetic information as they
cannot be demagnetized. Nanometer-sized magnetite grains have been recovered from a number of soils and lacustrine and modern
and ancient deep sea sediments. In many cases, these particles were identified as biogenic by their shape and size, which were similar
to magnetite crystals in magnetotactic bacteria and were therefore referred to as “magnetofossils.” Putative fossil magnetotactic
bacterial magnetite crystals have been found in stromatolitic and black cherts from the Precambrian, some dating as far back as 2
billion years. If these crystals are actually from magnetotactic bacteria, they represent some of the oldest bacterial fossils on Earth.
Many of the crystals, however, found in these rocks were partially degraded, showed indistinct edges, and underwent partial
oxidation to maghemite. The crystals have been used as supporting evidence that free O2 had begun to accumulate in the
atmosphere before 2000 Ma and that the present intensity of the Earth’s magnetic field strength had appeared by 2000 Ma. If
magnetosomes are biomineralized by protists or retained by protists after ingestion of magnetotactic bacteria, then these eukaryotes,
like the magnetotactic bacteria, might also directly contribute to the magnetization of sediments.
The cuboctahedral crystal habit is the equidimensional form of magnetite and it is commonly found in inorganically produced
magnetites. The elongated, nonequidimensional, pseudo-hexagonal prismatic and bullet-shaped forms of magnetite are very
unusual. Both have been found in magnetic separates of sediments and it would seem only the latter elongated forms of magnetite
might represent reliable magnetofossils of magnetotactic bacteria and could therefore be used as biomarkers of the past presence of
magnetotactic bacteria.
The presence of ultrafine-grained magnetite, pyrrhotite, and greigite in the rims of carbonate inclusions in the Martian meteorite
ALH84001 was considered one of the lines of evidence for life on ancient Mars. This meteorite, estimated now to be  4.1 billion
years old, contains magnetite crystals about 10–100 nm in size with cuboid, teardrop-shaped, and irregular morphologies. About
25% of the magnetite crystals present have an unusual morphology referred to as truncated hexa-octahedral, which is essentially the
same morphology of magnetite crystals found in the marine magnetotactic vibrio, Magnetovibrio blakemorei strain MV-1.
The elongated magnetite crystals from Magnetovibrio blakemorei strain MV-1 have been studied in great detail and show a number
of different distinctive properties putatively not found in inorganically produced magnetite crystals. These properties have been
considered by some to be suitable criteria for the determination of whether magnetite crystals are biogenic or not. These properties
Magnetotaxis 17

include (1) narrow size range (a non-log-normal size distribution with the mean centered in the single-magnetic-domain size
range); (2) restricted width-to-length ratios; (3) high chemical purity; (4) few crystallographic defects (crystals are generally defect-
free with the exception of occasional crystal twinning); (5) crystal morphology with unusual truncated hexa-octahedral geometry;
and (6) elongation along a specific crystallographic axis of the crystal. The 25% of the magnetite crystals in ALH84001 that resemble
those of Magnetovibrio blakemorei seem to fit the above criteria. However, there are several other hypotheses regarding the origin of
the magnetite crystals in meteorite ALH84001 that do not involve biology but only chemistry. Thus, some disagree that these criteria
are robust enough to distinguish between biogenically and abiogenically produced magnetite crystals and so the debate continues.

Biotechnological Applications of Magnetotactic Bacteria, Magnetosomes, and Magnetosome Crystals

The controlled biomineralization of magnetosomes and their single-magnetic-domain particles of relatively high structural
perfection and purity as well as consistent species-specific crystal morphologies suggest that there is great potential for biotechno-
logical applications. In general, the amount of magnetite and magnetosomes from magnetotactic bacteria is relatively low especially
considering the amount needed for specific applications. Thus, in order to produce enough cells, magnetosomes, and magnetite
crystals for these applications, cells must be grown in mass culture where the conditions for growth and magnetite synthesis must be
optimized. This was a major obstacle to research in this area for years. In the last decade, techniques including those involving
continuous culture methods have been devised for the successful mass culture of various Magnetospirillum species and Magnetovibrio
blakemorei strain MV-1. For example, for Magnetospirillum gryphiswaldense strain MSR-1, a chemostat culture technique based on
pH-stat feeding allowed the production of a high cell density in an auto-fermentor. Using this technique, it was possible a
magnetosome yield of about 83 mg L1 and a productivity of about 55 mg L1 day1.
Cells of magnetotactic bacteria have proven useful in many interesting applications. For example, living north-seeking cells of
polar magnetotactic bacteria have been used to determine south magnetic poles in meteorites and rocks containing fine-grained
(<1 mm) magnetic minerals. Cells of magnetotactic bacteria have also been used in a number of medical applications including the
magnetic separation of cells such as granulocytes and monocytes that phagocytized magnetotactic bacteria. Recently, the possibility
of using magnetotactic bacteria in the removal of heavy metals and radionuclides from wastewater was investigated. The advantage
here again is that if the cells can take up significant amounts of heavy metals they can be separated magnetically and the
contaminant can be eliminated from an aquatic system. Cells of the sulfate-reducing magnetotactic bacterium, Desulfovibrio
magneticus strain RS-1, have been used in cadmium recovery using magnetic separation. A recently described application is the
trapping of magnetotactic bacteria using a commercial magnetic recording head. This method may be useful in counting
magnetotactic bacteria cells in water samples or to detect magnetically labeled bacteria or magnetosomes. Magnetotactic bacteria
have been manipulated as nanorobots through a directional magnetic field that force them to push 3 mm beads along preplanned
paths and creating structures as pyramids.
Magnetosomes contain single-magnetic-domain crystals that have been shown to have useful magnetic and physical properties.
In addition, the organic phospholipid membrane that envelopes the crystals allows for the immobilization of biological molecules
such as nucleic acids or nonmagnetosome proteins on their surfaces. Magnetosomes have thus been used in the immobilization of
specific enzymes, some of which show higher enzymatic activity when attached to magnetosomes than when attached to
commercially synthesized magnetite particles. Antibodies linked to magnetosomes or magnetosome crystals have been developed
and have proven useful in various immunoassays involving the detection of allergens, certain cancer cells, and the quantification of
Magnetotactic bacterial magnetite crystals have been used to detect single nucleotide polymorphism based on a fluorescence
resonance energy transfer technique in which double-stranded labeled DNA is synthesized by PCR and immobilized to the
magnetite crystals hybridize to target DNA and a fluorescence signal is detected. Magnetosome membrane proteins on the surface
of bacterial magnetite particles have been used as anchor proteins for the assembly of foreign proteins on the surface of magnetite
magnetosomes. These so-called protein displays have been designed using the magnetosome membrane proteins MagA, MpsA,
Mms16, and Mms13 (MamC). Magnetotactic bacterial magnetite particles have been modified using compounds such as hyper-
branched polyamidoamine dendrimers or amino silanes for the extraction of DNA. Biotin attached to a monolayer-modified
substrate was detected by streptavidin immobilized to bacterial magnetite particles using a magnetic force microscope showing the
potential of magnetosomes in the detection of biomolecular interactions in medical and diagnostic analyses. In the case of
antitumor treatments, magnetosomes have been used as potential drug carries and as targets for hyperthermia. Finally magneto-
somes and/or magnetosome crystals are being currently used to attempt to enhance magnetic resonance imaging (MRI).


Magnetotactic bacteria and magnetosomes are a subject of an active and interdisciplinary field of research that includes topics such
as microbiology, cell biology, geomicrobiology, palaeomagnetism, microscopy, biomineralization, biophysics, and biochemistry to
name a few. New and exciting findings are constantly being reported on these microorganisms. New directions will come that will
inspire research groups in this highly multidisciplinary field of microbiological research.
18 Magnetotaxis

Further Reading
Abreu F, Martins JL, Silveira TS, et al. (2007) Candidatus Magnetoglobus multicellularis, gen. nov., sp. nov., a multicellular magnetotactic prokaryote from a hypersaline environment.
International Journal of Systematic and Evolutionary Microbiology 57: 1318–1322.
Abreu F, Araujo ACV, Leão P, et al. (2016) Culture-independent characterization of novel psychrophilic magnetotactic cocci from Antarctic marine sediments. Environmental
Microbiology 18: 4426–4441.
Bazylinski DA and Frankel RB (2004) Magnetosome formation in prokaryotes. Nature Reviews Microbiology 2: 217–230.
Bazylinski DA and Lefevre CT (2013) Magnetotactic bacteria from extreme environments. Life 3: 295–307.
Bazylinski DA, Schlezinger DR, Howes BH, Frankel RB, and Epstein SS (2000) Occurrence and distribution of diverse populations of magnetic protists in a chemically stratified coastal
salt pond. Chemical Geology 169: 319–328.
Dziuba M, Koziaeva V, Grouzdev D, et al. (2016) Magnetospirillum caucaseum sp. nov., Magnetospirillum marisnigri sp. nov. and Magnetospirillum moscoviense sp. nov., freshwater
magnetotactic bacteria isolated from three distinct geographical locations in European Russia. International Journal of Systematic and Evolutionary Microbiology 66: 2069–2077.
Frankel RB, Bazylinski DA, Johnson MS, and Taylor BL (1997) Magneto-aerotaxis in marine coccoid bacteria. Biophysical Journal 73: 994–1000.
Leão P, Teixeira LC, Cypriano J, Farina M, Abreu F, Bazylinski DA, and Lins U (2016) North-seeking magnetotactic Gammaproteobacteria in the Southern Hemisphere. Applied and
Environmental Microbiology 82(18): 5595–5602.
Leão P, Chen Y-R, Abreu F, Wang M, Zhang W-J, Zhou K, Xiao T, Wu L-F, and Lins U (2017) Ultrastructure of ellipsoidal magnetotactic multicellular prokaryotes depicts their complex
assemblage and cellular polarity in the context of magnetotaxis. Environmental Microbiology (Print) 19(6): 2151–2163.
Lefèvre CT and Bazylinski DA (2013) Ecology, diversity, and evolution of magnetotactic bacteria. Microbiology and Molecular Biology Reviews 77(3): 497–526.
Lefevre CT, Bennet M, Landau L, et al. (2014) Diversity of magneto-aerotactic behaviors and oxygen sensing mechanisms in cultured magnetotactic bacteria. Biophysical Journal
107: 527–538.
Lin W, Pan Y, and Bazylinski DA (2017) Diversity and ecology of and biomineralization by magnetotactic bacteria. Environmental Microbiology Reports 9: 345–356.
Lins U, McCartney MR, Farina M, Frankel RB, and Buseck PR (2005) Habits of magnetosome crystals in coccoid magnetotactic bacteria. Applied and Environmental Microbiology
71: 4902–4905.
Liu Y, Li GR, Guo FF, et al. (2010) Large-scale production of magnetosomes by chemostat culture of Magnetospirillum gryphiswaldense at high cell density. Microbial Cell Factories
9: 99.
Martins JL, Silveira TS, Abreu F, Silva KT, Silva-Neto ID, and Lins U (2007) Grazing protozoa and magnetosome dissolution in magnetotactic bacteria. Environmental Microbiology
9: 2775–2781.
de Melo RD and Acosta-Avalos D (2017) Light effects on the multicellular magnetotactic prokaryote ‘Candidatus Magnetoglobus multicellularis’ are cancelled by radiofrequency fields:
The involvement of radical pair mechanisms. Antonie Van Leeuwenhoek 110: 177–186.
Rahn-Lee L and Komeili A (2013) The magnetosome model: Insights into the mechanisms of bacterial biomineralization. Frontiers in Microbiology 4. p. 352.
Ruan J, Kato T, Santini CL, et al. (2012) Architecture of a flagellar apparatus in the fast-swimming magnetotactic bacteria MO-1. Proceedings of the National Academy of Sciences of
the USA 109: 20643–20648.
Schüler D (ed.) (2007) Magnetoreception and magnetosomes in bacteria. In: Microbiology monographs, vol. 3, p. 319. Berlin Heidelberg: Springer.
Simmons SL, Bazylinski DA, and Edwards KJ (2006) South-seeking magnetotactic bacteria in the Northern Hemisphere. Science 311: 371–374.
Uebe R and Schuler D (2016) Magnetosome biogenesis in magnetotactic bacteria. Nature Reviews Microbiology 14: 621–637.