On the origin of microbial magnetoreception

Abstract

A broad range of organisms, from prokaryotes to higher animals, have the ability to sense and utilize Earth's geomagnetic field—a behavior known as magnetoreception. Although our knowledge of the physiological mechanisms of magnetoreception has increased substantially over recent decades, the origin of this behavior remains a fundamental question in evolutionary biology. Despite this, there is growing evidence that magnetic iron mineral biosynthesis by prokaryotes may represent the earliest form of biogenic magnetic sensors on Earth. Here, we integrate new data from microbiology, geology and nanotechnology, and propose that initial biomineralization of intracellular iron nanoparticles in early life evolved as a mechanism for mitigating the toxicity of reactive oxygen species (ROS), as ultraviolet radiation and free-iron-generated ROS would have been a major environmental challenge for life on early Earth. This iron-based system could have later been co-opted as a magnetic sensor for magnetoreception in microorganisms, suggesting an origin of microbial magnetoreception as the result of the evolutionary process of exaptation.

Full text

REVIEW National Science Review
7: 472–479, 2020
doi: 10.1093/nsr/nwz065
Advance access publication 21 May 2019
EARTH SCIENCES
On the origin of microbial magnetoreception
Wei Lin1,2,3,∗, Joseph L. Kirschvink4,5, Greig A. Paterson1,6, Dennis A. Bazylinski7
and Yongxin Pan1,2,3,8,∗
1Key Laboratory of
Earth and Planetary
Physics, Institute of
Geology and
Geophysics, Chinese
Academy of Sciences,
Beijing 100029, China;
2Institutions of Earth
Science, Chinese
Academy of Sciences,
Beijing 100029, China;
3France-China Joint
Laboratory for
Evolution and
Development of
Magnetotactic
Multicellular
Organisms, Chinese
Academy of Sciences,
Beijing 100029, China;
4Division of
Geological &
Planetary Sciences,
California Institute of
Technology, Pasadena,
CA 91125, USA;
5Earth-Life Science
Institute, Tokyo
Institute of
Technology, Tokyo
152–8551, Japan;
6Department of Earth,
Ocean and Ecological
Sciences, University
of Liverpool, Liverpool,
L69 7ZE, UK; 7School
of Life Sciences,
University of Nevada
at Las Vegas, Las
Vegas, NV
89154-4004, USA and
8College of Earth and
Planetary Sciences,
University of Chinese
Academy of Sciences,
Beijing 100049, China
∗Corresponding
authors. E-mails:
weilin0408@gmail.com;
yxpan@mail.iggcas.
ac.cn
Received 7 March
2019; Revised 16
May 2019; Accepted
20 May 2019
ABSTRACT
A broad range of organisms, from prokaryotes to higher animals, have the ability to sense and utilize Earth’s
geomagnetic eld—a behavior known as magnetoreception. Although our knowledge of the physiological
mechanisms of magnetoreception has increased substantially over recent decades, the origin of this
behavior remains a fundamental question in evolutionary biology. Despite this, there is growing evidence
that magnetic iron mineral biosynthesis by prokaryotes may represent the earliest form of biogenic magnetic
sensors on Earth. Here, we integrate new data from microbiology, geology and nanotechnology, and
propose that initial biomineralization of intracellular iron nanoparticles in early life evolved as a mechanism
for mitigating the toxicity of reactive oxygen species (ROS), as ultraviolet radiation and free-iron-generated
ROS would have been a major environmental challenge for life on early Earth. is iron-based system could
have later been co-opted as a magnetic sensor for magnetoreception in microorganisms, suggesting an
origin of microbial magnetoreception as the result of the evolutionary process of exaptation.
Keywords: magnetoreception, biomineralization, magnetotactic bacteria, exaptation
INTRODUCTION
Earth’s magnetosphere protects the surface environ-
ment from solar wind and cosmic radiation, and has,
therefore, been an essential factor in the persistence
of life on Earth. It has also provided a natural global
positioning system that various organisms have
exploited for navigation and migration via the genet-
ically controlled biomineralization of ferrimagnetic
iron minerals [1–3]. is iron-based magnetore-
ception has been identied in microorganisms
(prokaryotes and some protists) and diverse an-
imals from sh to mammals, suggesting that it
was a primal sensory system of all living systems
[4–12]. However, the origin and early evolution
of magnetoreception remain major enigmas. It has
been proposed that magnetoreception evolved from
a pre-existing trait (i.e. biomineralization) through
the process of exaptation [13], while, more recently,
a non-genetically controlled photoferrotrophy-
driven hypothesis has been proposed [14]. How
and why biogenic magnetic sensors rst evolved
remain maers of debate, and resolving these ques-
tions is important for understanding the origin and
evolution of magnetoreception not only in prokary-
otes, but also in eukaryotes. Here, we integrate new
data from microbiology, geology and nanotechnol-
ogy that support an exaptation model for microbial
magnetoreception (also known as magnetotaxis)
from an initial iron-based system for scavenging
intracellular free radicals generated by ultraviolet
radiation (UVR) and/or ferrous iron on early Earth.
THE FIRST MAGNETORECEPTIVE
ORGANISMS ON EARTH
One of the most extensively studied magnetic-
sensing organisms are magnetotactic bacteria
(MTB)—a group of diverse prokaryotes that
synthesize intracellular chain-arranged, nano-sized,
membrane-bounded magnetic crystals of mag-
netite (Fe3O4) and/or greigite (Fe3S4) called
magnetosomes [2]. Magnetosome chains are the
magnetic sensors in MTB, which act as an internal
compass needle and cause cells to align passively
along the local geomagnetic eld (Fig. 1). MTB
are the most primitive magnetic-sensing organisms
known thus far, with no current evidence of this
ability in viruses or the Archaea. In addition to
the MTB, magnetosome-like structures have been
C
e Author(s) 2019. Published by Oxford University Press on behalf of China Science Publishing & Media Ltd. is is an Open Access article distributed under the terms of the Creative
Commons Aribution License (hp://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original
work is properly cited.
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

REVIEW Lin et al.473
Figure 1. A magnetotactic bacterium (∼2.2 μm in length) with a single chain of Fe3O4
magnetosomes (brown inclusions). A agellum is inserted schematically on the right
side of the cell. Magnetosomes impart a permanent magnetic dipole moment to the cell
and act as an internal compass needle, causing it to align passively along geomagnetic
eld lines as it swims.
discovered in eukaryotic algae, protozoans and
vertebrates [6,7], which led Vali and Kirschvink
[15] to propose that the rst eukaryotes may have
inherited the ability to biomineralize magnetosomes
from a magnetotactic alphaproteobacterium during
the endosymbiotic development of mitochondria,
with subsequent gene transfer to the nucleus.
MTB were discovered independently by Salva-
tore Bellini and Richard P. Blakemore in 1963 and
1974, respectively [4,16]. ese bacteria have a
global distribution in aquatic environments from
marine to freshwater ecosystems [17]. In addition,
they have been shown to be important in the global
biogeochemical cycling of Fe as well as other ele-
ments,suchasS,N,CandP[18–21]. In some envi-
ronments, magnetosomes from MTB are preserved
in sediments or rocks as fossils, referred to as mag-
netofossils [22,23]. Magnetofossils are important
contributions to the remanent magnetization of sed-
iments and have been suggested as biomarkers for
reconstructing paleoenvironmental conditions [24].
Magnetofossil records trace an evolutionary history
of MTB to the Cretaceous and, with less certainty, to
the Precambrian around ∼1.9 Ga [25].
Until a few years ago, all MTB were only as-
signed to one of two major bacterial phyla: the
Proteobacteria or the Nitrospirae [26]. Use of
cultivation-independent approaches (such as 16S
rRNA gene-targeting analyses, metagenomics and
single-cell genomics) has led to the discovery of
previously unidentied MTB lineages, which greatly
expands our knowledge of their diversity. MTB
have a patchy phylogenetic distribution and are
now known to lie within at least ve bacterial phyla,
including Proteobacteria,Nitrospirae,Planctomycetes
and the candidate phyla of Latescibacteria and
Omnitrophica, which suggests that the traits of
magnetotaxis and magnetosome biomineralization
occur widely in the domain Bacteria [17,27–29].
Molecular, genetic and genomic advances in
MTB have led to the identication of a large gene
cluster (referred to as a magnetosome gene clus-
ter or MGC) containing a group of genes involved
in magnetosome biomineralization and in construc-
tion of the magnetosome chain [30–35]. Because
of their essential roles in magnetotaxis, comparative
and phylogenetic analyses of MGCs from dierent
MTB taxonomies can shed light on the origin and
evolution of microbial magnetoreception in bacte-
ria. Recent expansion of MGCs has enabled the re-
construction of the evolutionary history of MTB,
which suggests a monophyletic origin of magne-
totaxis from a single common ancestor [33,36,37]
prior to or near the divergence between the Nitrospi-
rae and Proteobacteria phyla during the mid-Archean
Eon [38] or maybe even earlier, in the last com-
mon ancestor of the Proteobacteria,Nitrospirae,Om-
nitrophica,Latescibacteria and Planctomycetes phyla
(Fig. 2)[35]. Bacterial magnetotaxis, therefore, ap-
pears to be a primal physiological process and the
rst example of magnetoreception and the rst ex-
ample of controlled biomineralization on Earth.
THE FUNCTION OF MAGNETOSOMES IN
EXTANT MTB
Magnetotaxis is clearly the main function of magne-
tosomes in extant MTB. e presence of these iron
nanoparticles imparts a magnetic dipole moment on
MTB cells and enables the cells to orient passively,
which then allows them to swim actively along the
geomagnetic eld direction. In general, MTB also
appear to have a ‘polarity’—a preference to swim
in a particular direction under oxic conditions; that
is, they swim to the magnetic north in the north-
ern hemisphere and to the magnetic south in the
southern hemisphere [2], although several types of
MTB have the opposite polarity in each hemisphere
[39,40]. In conjunction with other tactic responses,
such as aerotaxis [41], phototaxis [42], chemotaxis
[43] or redox taxis [43], magnetotaxis allows MTB
to more eciently locate and maintain positions in
their preferred less-oxygenated microhabitats near
the oxic-anoxic transition zone in aquatic environ-
ments.
It has been estimated that, for a cell of a Mag-
netospirillum species, a magnetosome chain of 20
Fe3O4crystals would provide a sucient magnetic
dipole moment for magnetotaxis [44]. We note,
however, that as few as three to ve magnetosomes
per cell appear to be enough to provide a strong
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

474 Natl Sci Rev, 2020, Vol. 7, No. 2 REVIEW
Evolutionary time Evolutionary time
Ancestral MTB
b
Magnetite magnetosome
Greigite magnetosome
Unknown type of magnetosome
MGC duplication and divergence
Horizontal gene transfers
Loss events of magnetite-type MGCs
Loss events of greigite-type MGCs
Ancestral MTB
a
Figure 2. Proposed scenarios for the evolution of magnetotaxis in bacteria at or above
the class or phylum taxonomic levels [35]. The last common ancestor of magnetotac-
tic bacteria (MTB) was either (a) magnetite-producing or (b) a bacterium containing an
unknown magnetosome type. Both scenarios suggest a monophyletic origin of mag-
netosome gene clusters (MGCs) from a single common ancestor that existed early in
Earth history. Vertical inheritance followed by multiple independent gene losses is a
major force that drove the evolution of magnetotaxis in bacteria at or above the class
or phylum levels [35,36], while, within lower-level ranks, the evolutionary history of
magnetotaxis appears to be much more complicated (e.g. [81–83]).
magnetic dipole for orientation in some uncultured
environmental MTB (Fig. 3). Some MTB, includ-
ing ‘Candidatus Magnetobacterium bavaricum’ [45]
and ‘Candidatus Magnetobacterium casensis’ [46]
from the Nitrospirae phylum, synthesize hundreds
of magnetosomes in a single cell—far greater than
would be needed for magnetotaxis. e redundancy
or ‘overproduction’ of magnetic particles suggests
that magnetosomes in MTB may have other func-
tions in addition to magnetotaxis.
Several possible functions have been suggested
for magnetosomes, such as iron storage and seques-
tration, electrochemical baeries, gravity sensors
or providing locally strong magnetic elds for
enhancing and stabilizing magnetochemical
reaction pathways involving free-radical pairs
[15,25,47,48]. All of these, however, await conr-
mation by experimental studies. Recently, however,
it has been shown experimentally that Fe3O4mag-
netosomes in some MTB exhibit peroxidase-like
activity that can eliminate intracellular levels of
reactive oxygen species (ROS) [49]. Moreover, this
activity can be further enhanced under irradiation
by visible light [50]. ese ndings indicate strongly
the potential functions of magnetosome nanoparti-
cles in the detoxication of ROS or toxic free iron.
AN ORIGIN OF PROKARYOTIC
MAGNETOTAXIS THROUGH EXAPTATION
Exaptation—an evolutionary process by which a
biological entity is co-opted for a new role that is
unrelated to its initial function [51]—was likely
central in the evolution of magnetotaxis. Accumu-
lating evidence indicates that microbial life was
present at least since the Archean [52–54] and,
as noted above, MTB appear to have originated
in the mid-Archean Eon [38]. During the early
to late Archean, the primordial atmosphere was
anoxic, with ≤10−5of the present atmospheric
level of molecular O2[55,56]. Due to the lack of
an eective ozone layer on early Earth, harmful
ultraviolet radiation (UVR) was considerably higher
than in the present day and would have exerted
signicant environmental selection pressure on
microorganisms in the surface and shallow-water
conditions [57]. High UVR levels are detrimental
to living microorganisms by either directly causing
lesions on native DNA molecules or indirectly
through the accumulation of ROS inside cells.
Archean oceans were predominantly anoxic, with
abundant dissolved ferrous iron (>30 μm) supplied
from mid-ocean ridges, hydrothermal vents and
500 nm 200 nm
3 magnetosomes 4 magnetosomes
ab
200 nm
5 magnetosomes
c
Figure 3. Transmission electron microscope images of uncultured environmental magnetotactic bacteria with (a) three, (b)
four and (c) ve magnetosome particles per cell (white arrows point to each magnetosome), which indicates that three to
ve magnetosomes may provide a sufcient magnetic dipole moment for magnetotaxis in these bacteria.
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

REVIEW Lin et al.475
Figure 4. Exaptation model of microbial magnetoreception on early Earth. (a) Reactive
oxygen species (ROS) were a major challenge to which ancient life had to adapt. ROS
would have been generated and enhanced through ultraviolet radiation (UVR) (yellow),
accumulating free Fe(II) inside cells (purple) and/or mineral-induced formation (orange).
(b) The ancestral role of intracellular iron-oxide nanoparticles (initial magnetosomes)
formed through ancient biomineralization processes was to help early life cope with
oxidative stress because of their antioxidant enzyme-like activities and reducing intra-
cellular free iron. (c) Initial magnetosomes were later co-opted to serve an additional
new role of magnetoreception as a mineral magnetic sensor. (d) Modication of mag-
netosomes by natural selection, such as the increase in magnetosome particles and
formation of a chain arrangement, would impart a greater magnetic dipole moment to
the cell, leading to much more efcient magnetotaxis.
sediment diagenesis [58]. Ferrous iron likely could
diuse passively through the outer membrane of pri-
mordial organisms and would have stimulated toxic
intracellular ROS levels through the Fenton reac-
tion [59]. Furthermore, ROS might have also been
present in aqueous, atmospheric and rock environ-
ments on early Earth because of the formation of rad-
ical species on mineral surfaces induced by UVR, im-
pact shocks and mechanical grinding [60,61]. ROS
accumulation could damage genetic material, dete-
riorate proteins, cause lipid peroxidation and disturb
cellular homeostasis [62]; therefore, dealing with
ROS was a major survival challenge for early life on
Earth (Fig. 4a).
Extant organisms have evolved various antioxi-
dant systems to detoxify ROS, such as superoxide
dismutases, peroxiredoxins and catalases in aerobes
and superoxide reductases in anaerobes and mi-
croaerophiles [63]. e appearance of appreciable
O2concentrations would have led to signicant
oxidative stress, so it is generally accepted that major
antioxidant defense systems evolved prior to the
Great Oxygenation Event (GOE), which marked
a permanent molecular O2rise in the atmosphere
between 2.4 and 2.1 billion years ago [64]. An-
tioxidant defense systems then radiated massively
aer the GOE [65]. It remains unclear whether life
evolved primordial antioxidant enzymes at or prior
to the mid-Archean Eon, although some studies
suggest that the last universal common ancestor
might have possessed pathways to remove ROS
[63,66].
Discovery of intrinsic peroxidase- and catalase-
like activities of iron-oxide nanoparticles (IONPs,
including Fe3O4)[67–69] and of peroxidase-like
properties of magnetosomes [49,50] leads us to
propose that some ancient life forms might have
relied on the intracellular biomineralization of
IONPs (initial magnetosomes) as antioxidants to
cope with ROS stress on early Earth. IONPs have
been found to have pH-dependent dual enzyme-like
activities in intracellular microenvironments—that
is, they catalyse H2O2to generate hydroxyl radicals
under acidic conditions through peroxidase-like
activities and catalyse H2O2to H2O and O2at
neutral and basic pH through catalase-like activities
[69]. e median pH of the cytoplasm, periplasm
and lumen of the magnetosome vesicle are generally
neutral in Magnetospirillum magneticum strain
AMB-1 cells [70], while the cytoplasmic pH of
some uncultured MTB from acidic environments
is also close to neutral [29], which indicates that
Fe3O4magnetosomes may also have catalase-like
activity in vivo. Compared to traditional antioxidant
enzymes, IONPs have enhanced enzyme-like stabil-
ity under extreme conditions such as a wide range
of temperatures (4–90◦C) and pH (1–12) [71],
which could enable them to maintain antioxidant
function in harsh environments.
Microorganisms on early Earth with the ability
to mitigate ROS stress would have a competitive
advantage. Here, we argue that iron nanoparticle
formation (initial magnetosomes) in early primal
life had the function of mitigating intracellular ROS
toxicity, through their intrinsic antioxidant enzyme-
like activities and reducing intracellular toxic free
iron (Fig. 4b). With increasing magnetosome num-
bers, it appears that magnetosomes were co-opted
to provide the cell with a magnetic dipole moment
for orientation along the geomagnetic eld—a for-
mation that was likely established 3–4 billion years
ago (Fig. 4c). is primal magnetosensitive struc-
ture, which reduces a 3D search to an optimized 1D
search along geomagnetic eld lines, appears to have
further protected ancient life from lethal UVR by
allowing ecient directed swimming to deeper wa-
ter with less O2at or near the oxic-anoxic transition
zone either in the water column, the sediment–water
interface or deeper in the sediment. For this to occur,
natural selection would favor the biomineralization
of high-coercivity single-domain magnetic nanopar-
ticles arranged as a chain with dipoles aligned in the
same direction to maximize the net magnetic dipole
moment for the individual cell to optimize magnetic
orientation and navigation (Fig. 4d).
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

476 Natl Sci Rev, 2020, Vol. 7, No. 2 REVIEW
FUTURE PROSPECTS
An interesting yet unanswered question is: what
was the mineral phase of the rst magnetic sensor?
According to our model, the rst magnetosomes
should have had antioxidant activities for scaveng-
ing intracellular ROS. A growing number of iron
nanoparticles, such as Fe3O4,Fe
2O3and FeS, have
been shown to exhibit enzyme-like activity [72]. It
has been suggested that Fe3O4might have been
the mineral present in the rst magnetosomes [37]
(Fig. 2a). Alternately, the last common ancestor
of MTB could have synthesized an unknown iron-
containing biomineral with enzyme-like activity that
later, during evolution, perhaps through intracellu-
lar changes in enzymatic activity or redox, resulted
in the generation of Fe3O4and Fe3S4particles [35]
(Fig. 2b). Identication of this rst mineral magnetic
sensor remains to be elucidated and is an area of ac-
tive investigation. e search for putative magneto-
fossils in older rocks and the reconstruction of ances-
tral MGC proteins both have the potential to answer
this question.
e exaptation model of magnetotaxis imposes
an expected evolutionary sequence of magnetosome
genes. at is, genes that are involved in magneto-
some biosynthesis should have originated earlier
than those for magnetosome positioning and crystal
size, and for the number of magnetosomes per cell.
Genetic studies of MGCs reveal eight (mamIELM-
NOBQ) and six (mamELMOQB) magnetosome
genes that are essential for Fe3O4magnetosome
biosynthesis in Magnetospirillum magneticum strain
AMB-1 and M.gryphiswaldense strain MSR-1, re-
spectively [31,73]. Homologues of these genes have
been identied in MGCs of other MTB, thereby
emphasizing their important roles in magnetosome
biomineralization. Additional genomic, phyloge-
netic and evolutionary analyses are clearly necessary
to investigate whether these essential genes evolved
earlier than those that control magnetosome chain
construction (e.g. mamK [74]ormamJ [75]),
magnetosome crystal size (mms6,mmsF,etc.
[31,73]) and the number of magnetosomes per
cell. Moreover, studies of the linear organization
of magnetosomes and formation of magnetosome
membrane vesicles may also shed light on the
evolution of the cytoskeleton and vacuole formation
in both prokaryotes and eukaryotes [15,76].
It is also clear that further research is required
to characterize systematically any additional mag-
netosome functions beside magnetotaxis in extant
MTB. For example, determining whether magne-
tosome crystals play a role in storing cellular iron,
or as an electrochemical baery or gravity sensor,
or for promoting magnetochemistry awaits further
study. We propose here that Fe3O4magnetosome
crystals act as a type of iron-oxide nanozyme [69,71]
in MTB with neutral intracellular pH by exhibit-
ing catalase-like activity in addition to peroxidase-
like activity, although further experimental evidence
is required to support this hypothesis. Lastly, why
some MTB biomineralize Fe3S4magnetosomes as
opposed to Fe3O4remains unclear, especially con-
sidering the generally less perfect chain alignment
and poorer crystallinity of Fe3S4magnetosomes
compared with those of Fe3O4magnetosomes [77].
Chemically synthesized Fe3S4nanoparticles have
also been shown to have peroxidase-like activ-
ity [78]. us, any further studies, such as those
noted above, should also include Fe3S4-producing
MTB.
In space environments, UVR is one of the
most signicant hazards to living organisms. ere-
fore, the inferred adaptation of MTB to such
high-radiation environments makes them potential
model organisms in astrobiology research and may
provide an opportunity for studies on the responses
of organisms exposed to the near-space and low-
Earth-orbit space environments. Such studies could
in turn help to beer understand the origin and func-
tions of magnetosomes.
MTB are recognized as potentially signicant
contributors to present-day global iron cycling
[19,79]. Recent discovery of an Archean origin of
these magnetosensitive microorganisms further
suggests that they may have contributed to biogeo-
chemical cycling of iron throughout Earth’s history.
We suggest that the ROS-detoxication function
of magnetosomes and magnetotaxis capability
provided competitive advantages, which might
have helped ancient MTB to survive in diverse
aquatic environments on early Earth. Considering
their uptake of large amounts of environmental
iron and intracellular iron biomineralization, MTB
likely contributed to iron cycling on early Earth,
which further raises the question of whether these
microorganisms may have played as-yet-unknown
roles in the deposition of banded iron formations
that are distributed widely on the remnants of an-
cient cratons [80]. Future geochemical exploration
and magnetic characterization of both extant mag-
netosomes and magnetofossils will undoubtedly
provide new insights into this poorly understood,
yet geologically interesting, question.
CONCLUSIONS
e presence of precise biochemically controlled
biomineralization of ferrimagnetic minerals in two
domains of life provides strong evidence of Earth’s
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

REVIEW Lin et al.477
magnetic biosphere. However, the initial origin and
subsequent evolutionary history of magnetorecep-
tion have not been investigated to any signicant
degree. We posit that ancient magnetoreception in
prokaryotes might have originated via an exaptation
process from pre-existing intracellular iron nanopar-
ticles that initially decreased the toxicity of ROS in
early life forms. us, magnetosome particles in an-
cient life served a detoxication role and were later
co-opted for microbial magnetoreception or magne-
totaxis. is exaptation origin of magnetotaxis pro-
vides a conceptual model for study of the origin
and evolution of magnetoreception, as well as poten-
tially providing a genetic template for other biomin-
eralization systems and mechanisms. With the ever-
increasing genomic data from both cultivated and
uncultivated MTB as well as advancement of molec-
ular, genetic, chemical and evolutionary technolo-
gies, we anticipate great progress in understand-
ing microbial magnetoreception in the near future.
Shedding further light on the evolutionary origin of
this system will also provide additional constraints
on the paleoenvironments under which it evolved as
well as on the development of magnetoreception in
higher organisms.
FUNDING
W.L. and Y.P. acknowledge nancial support from the Strate-
gic Priority Research Program of Chinese Academy of Sciences
(XDA17010501) and the National Natural Science Foundation
of China (NSFC) (41621004). W.L. acknowledges support from
the NSFC (41822704) and the Youth Innovation Promotion
Association of the Chinese Academy of Sciences. J.L.K. is sup-
ported by the US National Aeronautics and Space Administra-
tion Exobiology (EXO14 2-0176). G.A.P. acknowledges sup-
port from the NSFC (41574063) and the Natural Environment
Research Council (NERC) Independent Research Fellowship
(NE/P017266/1). D.A.B. is supported by the US National Sci-
ence Foundation (NSF) (EAR-1423939).
REFERENCES
1. Kirschvink JL, Walker MM and Diebel CE. Magnetite-based
magnetoreception. Curr Opin Neurobiol 2001; 11: 462–7.
2. Bazylinski DA and Frankel RB. Magnetosome formation in
prokaryotes. Nat Rev Microbiol 2004; 2: 217–30.
3. Shaw J, Boyd A and House M et al. Magnetic particle-mediated
magnetoreception. J R Soc Interface 2015; 12: 499.
4. Blakemore RP. Magnetotactic bacteria. Science 1975; 190: 377–
9.
5. Walker MM, Kirschvink JL and Chang S-BR et al. A candidate
magnetic sense organ in the yellown tuna, Thunnus albacares.
Science 1984; 224: 751–3.
6. Dearaujo FFT, Pires MA and Frankel RB et al. Magnetite and
magnetotaxis in algae. Biophys J 1986; 50: 375–8.
7. Mann S, Sparks NHC and Walker MM et al. Ultrastructure, mor-
phology and organization of biogenic magnetite from sockeye
salmon, Oncorhynchus nerka: implications for magnetorecep-
tion. JExpBiol1988; 140: 35–49.
8. Tian L, Xiao B and Lin W et al. Testing for the presence of
magnetite in the upper-beak skin of homing pigeons. Biometals
2007; 20: 197–203.
9. Bauer GB, Fuller M and Perry A et al. Magnetoreception and
biomineralization of magnetite in Cetaceans. In: Kirschvink JL,
Jones DS and MacFadden BJ (eds). Magnetite Biomineraliza-
tion and Magnetoreception in Organisms: a New Biomagnetism.
Boston: Springer US, 1985, 489–507.
10. Bazylinski DA, Lef `
evre CT and Frankel RB et al. Magnetotactic
protists at the oxic-anoxic transition zones of coastal aquatic
environments. In: Altenbach AV, Bernhard JM and Seckbach J
(eds). Anoxia,Vol.21. Dordrecht: Springer Netherlands, 2012,
131–43.
11. Holland RA, Kirschvink JL and Doak TG et al. Bats use mag-
netite to detect the earth’s magnetic eld. PLoS One 2008; 3:
e1676.
12. Tian L, Lin W and Zhang S et al. Bat head contains soft magnetic
particles: evidence from magnetism. Bioelectromagnetics 2010;
31: 499–503.
13. Kirschvink JL and Hagadorn JW. A grand unied theory of
biomineralization. In: B¨
auerlein E (ed). The Biomineralisation
of Nano- and Micro-Structures. Weinheim: Wiley-VCH Verlag
GmbH, 2000, 139–50.
14. Strbak O and Dobrota D. Archean iron-based metabolism anal-
ysis and the photoferrotrophy-driven hypothesis of microbial
magnetotaxis origin. Geomicrobiol J 2019; 36: 278–90.
15. Vali H and Kirschvink JL. Observations of magnetosome orga-
nization, surface structure, and iron biomineralization of unde-
scribed magnetotactic bacteria: evolutionary speculations. In:
Frankel RB and Blakemore RP (eds). Iron Biominerals. New York:
Plenum Press, 1990, 278–90.
16. Bellini S. On a unique behavior of freshwater bacteria. Chin J
Ocean Limnol 2009; 27:3–5.
17. Lin W, Pan Y and Bazylinski DA. Diversity and ecology of and
biomineralization by magnetotactic bacteria. Env Microbiol Rep
2017; 9: 345–56.
18. Cox BL, Popa R and Bazylinski DA et al. Organization and ele-
mental analysis of P-, S-, and Fe-rich inclusions in a population
of freshwater magnetococci. Geomicrobiol J 2002; 19: 387–406.
19. Lin W, Bazylinski DA and Xiao T et al. Life with compass: diver-
sity and biogeography of magnetotactic bacteria. Environ Micro-
biol 2014; 16: 2646–58.
20. Rivas-Lamelo S, Benzerara K and Lef`
evre CT et al. Magnetotac-
tic bacteria as a new model for P sequestration in the ferrugi-
nous Lake Pavin. Geochem Persp Let 2017: 35–41.
21. Schulz-Vogt HN, Pollehne F and J ¨
urgens K et al. Effect of large
magnetotactic bacteria with polyphosphate inclusions on the
phosphate prole of the suboxic zone in the Black Sea. ISME
J2019; 13 : 1198–208.
22. Chang SBR and Kirschvink JL. Magnetofossils, the magnetiza-
tion of sediments, and the evolution of magnetite biomineral-
ization. Annu Rev Earth Planet Sci 1989; 17: 169–95.
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

478 Natl Sci Rev, 2020, Vol. 7, No. 2 REVIEW
23. Vasiliev I, Franke C and Meeldijk JD et al. Putative greigite magnetofossils from
the Pliocene epoch. Nat Geosci 2008; 1: 782–6.
24. Pan YX, Deng CL and Liu QS et al. Biomineralization and magnetism of bacterial
magnetosomes. Chin Sci Bull 2004; 49: 2563–8.
25. Kopp RE and Kirschvink JL. The identication and biogeochemical interpreta-
tion of fossil magnetotactic bacteria. Earth-Sci Rev 2008; 86: 42–61.
26. Jogler C and Sch¨
uler D. Genomics, genetics, and cell biology of magnetosome
formation. Annu Rev Microbiol 2009; 63: 501–21.
27. Kolinko S, Jogler C and Katzmann E et al. Single-cell analysis reveals a novel
uncultivated magnetotactic bacterium within the candidate division OP3. Env-
iron Microbiol 2012; 14: 1709–21.
28. Lin W and Pan Y. A putative greigite type magnetosome gene cluster from
the candidate phylum Latescibacteria. Env Microbiol Rep 2015; 7: 237–
42.
29. Abreu F, Le ˜
ao P and Vargas G et al. Culture-independent characterization of a
novel uncultivated magnetotactic member of the Betaproteobacteria class of
the Proteobacteria phylum from an acidic lagoon. Environ Microbiol 2018; 20:
2615–24.
30. Gr¨
unberg K, Wawer C and Tebo BM et al. A large gene cluster encoding sev-
eral magnetosome proteins is conserved in different species of magnetotactic
bacteria. Appl Environ Microb 2001; 67: 4573–82.
31. Murat D, Quinlan A and Vali H et al. Comprehensive genetic dissection of the
magnetosome gene island reveals the step-wise assembly of a prokaryotic or-
ganelle. Proc Natl Acad Sci USA 2010; 107: 5593–8.
32. Lohße A, Ullrich S and Katzmann E et al. Functional analysis of the magneto-
some island in Magnetospirillum gryphiswaldense: the mamAB operon is suf-
cient for magnetite biomineralization. PLoS One 2011; 6: e25561.
33. Abreu F, Cantao ME and Nicolas MF et al. Common ancestry of iron oxide- and
iron-sulde-based biomineralization in magnetotactic bacteria. ISME J 2011;
5: 1634–40.
34. Lef`
evre CT, Trubitsyn D and Abreu F et al. Comparative genomic analysis of
magnetotactic bacteria from the Deltaproteobacteria provides new insights
into magnetite and greigite magnetosome genes required for magnetotaxis.
Environ Microbiol 2013; 15: 2712–35.
35. Lin W, Zhang W and Zhao X et al. Genomic expansion of magnetotactic bacte-
ria reveals an early common origin of magnetotaxis with lineage-specic evo-
lution. ISME J 2018; 12: 1508–19.
36. Lef`
evre CT and Bazylinski DA. Ecology, diversity, and evolution of magnetotactic
bacteria. Microbiol Mol Biol R 2013; 77: 497–526.
37. Lef`
evre CT, Trubitsyn D and Abreu F et al. Monophyletic origin of magnetotaxis
and the rst magnetosomes. Environ Microbiol 2013; 15: 2267–74.
38. Lin W, Paterson GA and Zhu Q et al. Origin of microbial biomineralization and
magnetotaxis during the Archean. Proc Natl Acad Sci USA 2017; 114: 2171–
6.
39. Simmons SL, Bazylinski DA and Edwards KJ. South-seeking magnetotactic bac-
teria in the Northern Hemisphere. Science 2006; 311: 371–4.
40. Le˜
ao P, Teixeira LCRS and Cypriano J et al. North-seeking magnetotactic
Gammaproteobacteria in the Southern Hemisphere. Appl Environ Microbiol
2016; 82: 5595–602.
41. Frankel RB, Bazylinski DA and Johnson MS et al. Magneto-aerotaxis in marine
coccoid bacteria. Biophys J 1997; 73: 994–1000.
42. Shapiro OH, Hatzenpichler R and Buckley DH et al. Multicellular photo-
magnetotactic bacteria. Environ Microbiol Rep 2011; 3: 233–8.
43. Spring S and Bazylinski DA. Magnetotactic bacteria. In: Dworkin M (ed). The
Prokaryotes: An Evolving Electronic Resource for the Microbiological Commu-
nity. New York: Springer Verlag, 2006, 842–62.
44. Frankel RB, Zhang J-P and Bazylinski DA. Single magnetic domains in magne-
totactic bacteria. J Geophys Res 1998; 103: 30601–4.
45. Spring S, Amann R and Ludwig W et al. Dominating role of an unusual mag-
netotactic bacterium in the microaerobic zone of a freshwater sediment. Appl
Environ Microbiol 1993; 59: 2397–403.
46. Lin W, Deng A and Wang Z et al. Genomic insights into the uncultured genus
‘Candidatus Magnetobacterium’ in the phylum Nitrospirae. ISME J 2014; 8:
2463–77.
47. Kirschvink JL. Rock magnetism linked to human brain magnetite. Eos TransAGU
1994; 75: 178–9.
48. Uebe R and Sch¨
uler D. Magnetosome biogenesis in magnetotactic bacteria.
Nat Rev Microbiol 2016; 14: 621–37.
49. Guo FF, Yang W and Jiang W et al. Magnetosomes eliminate intracellular re-
active oxygen species in Magnetospirillum gryphiswaldense MSR-1. Environ
Microbiol 2012; 14: 1722–9.
50. Li K, Wang P and Chen C et al. Light irradiation helps magnetotactic bacte-
ria eliminate intracellular reactive oxygen species. Environ Microbiol 2017; 19:
3638–48.
51. Gould SJ and Vrba ES. Exaptation—a missing term in the science of form.
Paleobiology 1982; 8: 4–15.
52. Allwood AC, Walter MR and Burch IW et al. 3.43 billion-year-old stromatolite
reef from the Pilbara Craton of Western Australia: ecosystem-scale insights to
early life on Earth. Precambrian Res 2007; 158: 198–227.
53. Allwood AC, Grotzinger JP and Knoll AH et al. Controls on development and
diversity of Early Archean stromatolites. Proc Natl Acad Sci USA 2009; 106:
9548–55.
54. Sugitani K, Lepot K and Nagaoka T et al. Biogenicity of morphologically diverse
carbonaceous microstructures from the ca. 3400 Ma Strelley pool formation, in
the Pilbara Craton, Western Australia. Astrobiology 2010; 10: 899–920.
55. Poulton SW and Caneld DE. Ferruginous conditions: a dominant feature of the
ocean through Earth’s history. Elements 2011; 7: 107–12.
56. Johnson JE, Gerpheide A and Lamb MP et al. O2constraints from Paleopro-
terozoic detrital pyrite and uraninite. Geol Soc Am Bull 2014; 126: 813–30.
57. Cnossen I, Sanz-Forcada J and Favata F et al. Habitat of early life: Solar X-
ray and UV radiation at Earth’s surface 4–3.5 billion years ago. J Geophys Res
2007; 112: E02008.
58. Kendall B, Anbar AD and Kappler A et al. The global iron cycle. In: Knoll AH,
Caneld DE and Konhauser KO (eds). Fundamentals of Geobiology West Sus-
sex : John Wiley & Sons, Ltd, 2012, 65–92.
59. Winterbourn CC. Toxicity of iron and hydrogen peroxide: the Fenton reaction.
Toxicol Lett 1995; 82–83: 969–74.
60. Schoonen M, Smirnov A and Cohn C. A perspective on the role of minerals in
prebiotic synthesis. AMBIO: A Journal of the Human Environment 2004; 33:
539–51.
61. Xu J, Sahai N and Eggleston CM et al. Reactive oxygen species at the
oxide/water interface: formation mechanisms and implications for prebiotic
chemistry and the origin of life. Earth Planet Sci Lett 2013; 363: 156–67.
62. TouatiD. Iron and oxidative stress in bacteria. Arch Biochem Biophys 2000; 373:
1–6.
63. ´
Slesak I, ´
Slesak H and Zimak-Piekarczyk P et al. Enzymatic antioxidant systems
in early anaerobes: theoretical considerations. Astrobiology 2016; 16: 348–58.
64. Lyons TW, Reinhard CT and Planavsky NJ. The rise of oxygen in Earth’s early
ocean and atmosphere. Nature 2014; 506: 307–15.
65. Kirschvink JL, Gaidos EJ and Bertani LE et al. Paleoproterozoic snowball Earth:
extreme climatic and geochemical global change and its biological conse-
quences. Proc Natl Acad Sci USA 2000; 97: 1400–5.
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020

REVIEW Lin et al.479
66. ´
Slesak I, ´
Slesak H and Kruk J. Oxygen and hydrogen peroxide in the early evo-
lution of life on Earth: in silico comparative analysis of biochemical pathways.
Astrobiology 2012; 12: 775–84.
67. Gao L, Zhuang J and Nie L et al. Intrinsic peroxidase-like activity of ferromag-
netic nanoparticles. Nat Nanotech 2007; 2: 577–83.
68. Ragg R, Tahir MN and Tremel W. Solids go bio: inorganic nanoparticles as en-
zyme mimics. Eur J Inorg Chem 2016; 2016: 1906–15.
69. Chen Z, Yin J-J and Zhou Y-T et al. Dual enzyme-like activities of iron ox-
ide nanoparticles and their implication for diminishing cytotoxicity. ACS Nano
2012; 6: 4001–12.
70. Eguchi Y, Fukumori Y and Taoka A. Measuring magnetosomal pH of the mag-
netotactic bacterium Magnetospirillum magneticum AMB-1 using pH-sensitive
uorescent proteins. Biosci Biotechnol Biochem 2018; 8451:1–9.
71. Gao L, Fan K and Yan X. Iron oxide nanozyme: a multifunctional enzyme mimetic
for biomedical applications. Theranostics 2017; 7: 3207–27.
72. Wei H and Wang E. Nanomaterials with enzyme-like characteristics
(nanozymes): next-generation articial enzymes. Chem Soc Rev 2013; 42:
6060.
73. Lohße A, Borg S and Raschdorf O et al. Genetic dissection of the mamAB
and mms6 operons reveals a gene set essential for magnetosome biogen-
esis in Magnetospirillum gryphiswaldense. J Bacteriol 2014; 196: 2658–
69.
74. Komeili A, Li Z and Newman DK et al. Magnetosomes are cell membrane
invaginations organized by the actin-like protein MamK. Science 2006; 311:
242–5.
75. Scheffel A, Gruska M and Faivre D et al. An acidic protein aligns magneto-
somes along a lamentous structure in magnetotactic bacteria. Nature 2006;
440: 110–4.
76. Grant CR, Wan J and Komeili A. Organelle formation in Bacteria and Archaea.
Annu Rev Cell Dev Biol 2018; 34: 217–38.
77. P´
osfai M, Buseck PR and Bazylinski DA et al. Iron suldes from magnetotactic
bacteria; structure, composition, and phase transitions. Am Mineral 1998; 83:
1469–81.
78. Ding C, Yan Y and Xiang D et al. Magnetic Fe3S4 nanoparticles with
peroxidase-like activity, and their use in a photometric enzymatic glucose as-
say. Microchim Acta 2016; 183: 625–31.
79. Chen AP, Berounsky VM and Chan MK et al. Magnetic properties of unculti-
vated magnetotactic bacteria and their contribution to a stratied estuary iron
cycle. Nat Commun 2014; 5: 4797.
80. Frankel RB. Fossil record: magnetic skeletons in Davy Jones’ locker. Nature
1986; 320: 575.
81. Rioux J-B, Philippe N and Pereira S et al. A second actin-like MamK protein in
Magnetospirillum magneticum AMB-1 encoded outside the genomic magneto-
some island. PLoS One 2010; 5: e9151.
82. Ji B, Zhang SD and Zhang WJ et al. The chimeric nature of the
genomes of marine magnetotactic coccoid-ovoid bacteria denes a
novel group of Proteobacteria. Environ Microbiol 2017; 19: 1103–19.
83. Monteil CL, Perri `
ere G and Menguy N et al. Genomic study of a novel magne-
totactic Alphaproteobacteria uncovers the multiple ancestry of magnetotaxis.
Environ Microbiol 2018; 20: 4415–30.
Downloaded from https://academic.oup.com/nsr/article-abstract/7/2/472/5493124 by California Institute of Technology user on 28 May 2020