Genomics, Genetics, and Cell Biology of Magnetosome Formation · • Genomics, Genetics, and Cell Biology of Magnetosome Formation 503 Annu. Rev. Microbiol. 2009.63:501-521. Downloaded

ANRV387-MI63-24 ARI 14 August 2009 16:26

Genomics, Genetics,and Cell Biology ofMagnetosome FormationChristian Jogler and Dirk SchulerDepartment of Biology I, LMU Biozentrum, Ludwig-Maximilians-Universitat Munchen,82152 Planegg-Martinsried, Germany; email: [email protected]

Annu. Rev. Microbiol. 2009. 63:501–21

The Annual Review of Microbiology is online atmicro.annualreviews.org

This article’s doi:10.1146/annurev.micro.62.081307.162908

Copyright c© 2009 by Annual Reviews.All rights reserved

0066-4227/09/1013-0501$20.00

Key Words

magnetite biomineralization, magnetosome island, magnetotacticbacteria, horizontal gene transfer, bacterial organelles, evolution ofmagnetotaxis

AbstractMagnetosomes are specialized organelles for magnetic navigation thatcomprise membrane-enveloped, nano-sized crystals of a magnetic ironmineral; they are formed by a diverse group of magnetotactic bacteria(MTB). The synthesis of magnetosomes involves strict genetic controlover intracellular differentiation, biomineralization, and their assemblyinto highly ordered chains. Physicochemical control over biomineral-ization is achieved by compartmentalization within vesicles of the mag-netosome membrane, which is a phospholipid bilayer associated with aspecific set of proteins that have known or suspected functions in vesicleformation, iron transport, control of crystallization, and arrangementof magnetite particles. Magnetosome formation is genetically complex,and relevant genes are predominantly located in several operons withina conserved genomic magnetosome island that has been likely trans-ferred horizontally and subsequently adapted between diverse MTBduring evolution. This review summarizes the recent progress in ourunderstanding of magnetobacterial cell biology, genomics, and the ge-netic control of magnetosome formation and magnetotaxis.

501

Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


Magnetotacticbacteria (MTB):phylogenetically andphysiologically diversegroup of aquaticmicroorganisms thatorient and swim alongmagnetic field lines bymagnetotaxis owing tothe presence ofmagnetosomes

Magnetosomes:intracellular organellesfor magneticnavigation thatcomprise crystals of amagnetic mineralenveloped by aphospholipidmembrane thatcontains specificproteins

Contents

INTRODUCTION . . . . . . . . . . . . . . . . . . 502STRUCTURE AND

COMPOSITION OFMAGNETOSOMES . . . . . . . . . . . . . . 503Biomineralization of the

Inorganic Core . . . . . . . . . . . . . . . . . 503Biochemical Composition

of the MM. . . . . . . . . . . . . . . . . . . . . . 504GENETICS AND GENOMICS

OF MAGNETOSOMEFORMATION . . . . . . . . . . . . . . . . . . . . 505General Genome Characteristics . . . 505Identification of Magnetosome

Genes . . . . . . . . . . . . . . . . . . . . . . . . . . 506Molecular Organization

of Magnetosome Genes . . . . . . . . . 508Functions of Individual Genes

and Proteins . . . . . . . . . . . . . . . . . . . . 509Evolution of Magnetotaxis . . . . . . . . . . 514

DIRECTIONS FOR FUTURERESEARCH . . . . . . . . . . . . . . . . . . . . . . 516

INTRODUCTION

Magnetotactic bacteria (MTB) synthesize mag-netosomes, which are specialized intracellularstructures for magnetic navigation that com-prise membrane-enveloped crystals of a mag-netic iron mineral (4). It has recently becomeapparent that magnetosomes display charac-teristics of prokaryotic organelles, exhibitinga comparable degree of complexity as theireukaryotic counterparts (39, 41). Formation ofmagnetosomes occurs in different phylogeneticgroups of eubacteria, and MTB are morpholog-ically and physiologically diverse (2). Althoughthe function of the magnetosome chain is nottotally understood, it is widely accepted thatmagnetotaxis enables navigation of cells basedon an interaction with other sensory mecha-nisms. Briefly, cells use chemotaxis, aerotaxis,and perhaps phototaxis to locate and main-tain their growth-favoring position in verticalchemical gradients within their aquatic habitats

while they are passively oriented along geo-magnetic field lines by the permanent magneticdipole moment imparted by magnetosomes(Figure 1a). Because the magnetic alignmentof the cell reduces a three-dimensional searchproblem to a one-dimensional search problem,magnetotaxis is thought to make the cellularchemotactic response more efficient. However,the signal transduction mechanisms involvedin magnetotaxis are poorly understood, andthe recent observation of MTB with reversedor variable magnetic polarity suggests that themechanism of magnetotaxis might be morecomplex than originally assumed and thatthe magnetosomes might fulfill different oradditional functions (20, 23, 33, 77).

The synthesis of bacterial magnetosomes in-volves genetic control over the biomineraliza-tion of perfectly shaped and sized magneticcrystals and their assembly into highly orderedchains to serve most efficiently as a magneticfield sensor. Magnetosome formation is geneti-cally complex, and a rather large set of candidategenes predominantly located within a genomicmagnetosome island (MAI) has been identifiedso far (70).

Understanding the formation of mag-netosomes is of interest to diverse disci-plines such as microbiology, geobiology, andbiotechnology (70, 72, 85). In addition, MTBrecently emerged as a model for study-ing prokaryotic organelle formation and cellbiology (39).

Although mud bacteria, whose motility ismagnetically directed, were first documentedby Salvatore Bellini as early as 1963 (10), itwas only their rediscovery and the descriptionof magnetotaxis by Richard Blakemore in the1970s (11) that stimulated a wealth of diverseresearch activities. Since Blakemore’s seminalreview of early discoveries, which was publishedin this journal in 1982 (12), the growing knowl-edge on MTB and magnetosome formationhas been frequently summarized and recentlyhas been the subject of several comprehensivereview articles (7, 20, 39) and a monograph(69). This review summarizes the enormousprogress recently made on the understanding

502 Jogler · Schuler

Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


of the unique cell biology of MTB as well asthe genetic control of magnetosome formationand magnetotaxis, which has become possibleby the availability of genomic information aswell as novel genetic and ultrastructural toolsfor analysis and manipulation of MTB.

STRUCTURE ANDCOMPOSITION OFMAGNETOSOMES

Biomineralization of theInorganic Core

In most MTB the mineral core of magnetosomeconsists of the magnetic iron oxide magnetite(Fe3O4). Several uncultivated MTB from ma-rine environments synthesize magnetosomesconsisting of the magnetic iron sulfide greigite(Fe3S4), the biomineralization of which, how-ever, is poorly understood owing to the inabilityto cultivate greigite producer in the laboratory.The size and morphology of magnetite crystalsare subject to species-specific genetic control,and a considerable diversity of magnetosomemorphologies is found in different MTB, whichare mostly unknown from magnetite particlesformed by chemical synthesis (20) (Figure 1b).

The biomineralization of magnetite crys-tals requires the accumulation of large amounts(>4% per dry weight) of iron, which can betaken up in either its ferric or ferrous form frommicromolar extracellular concentrations (18,71). The intracellular pathway for uptake andsequestration then has to be strictly controlledbecause of the potentially harmful effect offree intracellular iron levels. Genomic analysisand preliminary experimental data suggest thatcommon constituents of the iron metabolism,such as uptake systems for ferrous and ferriciron, iron storage, as well as iron-regulatory ele-ments, and possibly siderophores, are present inMTB, although their significance for magnetitebiomineralization is poorly understood (9, 16,48, 91). There are indications that uptake andintracellular processing of iron for magnetitesynthesis proceeds through a distinct pathwayinvolving membrane-associated precursors that

Magnetosome chain

Bulletshaped

Cubo-octohedral Elongated-prismatic

Single chain Multiple chains

N

S

M

a

b

Figure 1(a) Transmission electron micrograph of a magnetotactic bacterium (MTB).The chain of magnetosomes (arrow) imparts a magnetic dipole moment thatorients the cell in ambient magnetic fields. N and S stand for the north andsouth pole of the magnetic dipole, respectively. (b) Diversity of magnetosomemorphologies found in different MTB.

Magnetotaxis: abilityof certain motile,aquaticmicroorganisms toorient in a magneticfield

Magnetosome island(MAI): large (80–150 kb) instablegenomic regionharboring manymagnetosome genesthat is found in allanalyzed MTB anddisplays conservedstructural andcompositionalcharacteristics,presumably transferredhorizontally betweendifferent species

include a ferrous high-spin and a ferritin-likecompound, but no mineral precursor other thanmagnetite (18). According to this study, a pro-cess has to be assumed by which nucleation ofmagnetite crystals predominantly occurs at thecytoplasmic membrane (CM) level, and sub-sequent crystal growth may proceed spatiallyseparated from the CM after detachment ofmagnetosome vesicles. However, the proposedmechanism has not yet been fully characterizedat the genetic and biochemical levels. Accumu-lation of supersaturating quantities of iron intothe magnetosomes is assumed to proceed viadistinct routes involving specific transport pro-teins. In Magnetospirillum gryphiswaldense themagnetosome-associated MamB and MamMproteins, which are members of the cation dif-fusion facilitator (CDF) family of metal trans-porters, are involved in magnetosomal irontransport (36).

Magnetite synthesis requires strict controlof the iron supersaturation, pH, and Eh levels

www.annualreviews.org • Genomics, Genetics, and Cell Biology of Magnetosome Formation 503

Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


MM: magnetosomemembrane

(20). This is achieved by compartmentalizationof crystallization by the magnetosome mem-brane (MM), which forms a distinct vesicularcompartment that serves as a nanoreactor inwhich conditions for magnetite precipitationcan be strictly controlled. Despite the strictbiological control exerted on magnetosomeformation, magnetite crystallization is affectedby environmental parameters. Specifically,the morphology of magnetite crystals is alsodetermined by the rates of Fe uptake, andfast-growing crystallites lack the structural per-fection of those formed slowly (19). Synthesisof magnetite crystals further depends onmicrooxic or anoxic conditions, whereas higheroxygen concentrations suppress magnetitebiomineralization or result in the formation ofsmaller and aberrantly shaped crystals (32).

Biochemical Composition of the MM

Similar to eukaryotic organelles, magnetosomecrystals are individually enveloped by a lipidbilayer membrane, the MM (27, 39, 67). TheMM is formed prior to magnetite formationand originates as an invagination of the CM(39, 40), but it may become pinched off laterin the cell cycle (18, 61). The MM has beenstudied at the structural and biochemical lev-els so far only in strains of Magnetospirillum,but it can be assumed that similar structuresare also present in all other MTB. By electron

microscopy the MM is visible as vesicu-lar structures intimately attached to the CM(Figure 2) (40). Komeili et al. (41) havesuggested that magnetosome vesicles requireactivation, possibly by the MM-associatedtetratricopeptide repeat (TPR) MamA protein.

Biochemical analysis of isolated magneto-some revealed that they contain phospholipidsand are associated with specific proteins(27, 29, 30, 66, 81). In M. gryphiswaldensethe phospholipid composition of the MMresembles that of the CM and includes phos-phatidylethanolamine, phosphatidylglycerol,ornithinamid lipid, and an unidentified aminolipid (29). Its protein composition, however, israther distinct from that of other subcellularcompartments. The MM of M. gryphiswaldenseis associated with a specific set of more than20 proteins present in various amounts (29, 30,58). Similar biochemical studies and genomicanalyses of other strains suggested that manymagnetosome proteins are conserved amongMTB (48, 52, 81, 82). MM proteins thathave been named either Mam (magnetosomemembrane), Mms (magnetic particle mem-brane specific), Mtx (magnetotaxis), or Mme(magnetosome membrane) belong to char-acteristic protein families, including TPRproteins (MamA), CDF transporters (MamBand MamM), HtrA-like serine proteases(MamE, MamP, and MamO), actin-like pro-teins (MamK), generic transporters (MamH

100 nm

a b

Figure 2(a) Electron micrograph of a Magnetospirillum gryphiswaldense cell. The inset shows a micrograph of isolatedmagnetosomes enveloped by the magnetosome membrane (MM). (b) Electron micrographs of an ultrathinsection of an iron-starved M. gryphiswaldense cell showing empty and partially filled vesicles of the MM(arrows). (Micrograph by G. Wanner; modified from Reference 60).


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


and MamN), and MTB-specific proteins withno homology to other proteins in nonmagneticorganisms (MamG, F, D, C, J, W, X, and Y,Mms6, MmeA, and MtxA) (29, 30, 35, 58).As shown by 1D and 2D electrophoresis, theprotein composition of the MM is distinct, andit has been demonstrated by GFP fusions andimmunodetection that several proteins arespecifically targeted to the MM. The molecularmechanisms controlling the sorting of MMproteins to the magnetosome compartmentare currently unknown, and no commonsequence motifs or targeting signals have beenidentified so far (39, 68). Whereas many MMproteins display the characteristics of typicalmembrane proteins and are tightly bound tothe magnetosome crystals, others appear to berather hydrophilic and seem loosely attachedso that they can be selectively solubilized bymild detergents (68).

Recruitment of magnetosome proteinscould be by direct binding to the surface ofmagnetite crystals or by specific protein-protein interactions. The latter could be me-diated for instance by the PDZ and TPR do-mains found in several MM proteins, whichgenerally mediate protein-protein interactions,act as scaffolding proteins, and typically coordi-nate the assembly of proteins into multisubunit

complexes at particular subcellular locations(7, 35). Although the role of many MM proteinshas remained obscure, several proteins were al-ready assigned functions in magnetosomal irontransport, magnetite nucleation and crystalliza-tion, and assembly of magnetosome chains.

GENETICS AND GENOMICS OFMAGNETOSOME FORMATION

General Genome Characteristics

Understanding the genetic basis of magneto-some formation has enormously benefited fromthe development of tools for the genetic manip-ulation as well as the genome analysis of sev-eral MTB. As of January 2009 two completeand two nearly complete genome sequencesfrom Alphaproteobacteria-MTB are availableto the public, which include the freshwaterM. gryphiswaldense (58), M. magneticum (48),and M. magnetotacticum, as well as the ma-rine magnetic coccus strain MC-1 (63) (forgeneral genome characteristics see Table 1).In addition, there is substantial genome infor-mation available, including large contigs har-boring the MAI for the marine vibrio strainMV-1 (33) and two uncultivated MTB thathave been collected from freshwater habitats by

Table 1 General genome features of the four magnetotactic bacteria with publicly availablegenomic sequencea,b

Feature MSR-1 AMB-1 MS-1 MC-1Nucleotide statisticsGenome size (bp) 4,264,908 4,967,148 4,503,280 4,719,581GC content (%) 62.8 65.1 64.0 54.8No. of contigs 373 1 316 1Gene predictionsNo. of predicted ORFs 4,268 4,559 4,925 3,716No. of 16S/23S/5S rRNA genes 2/2/2 2/2/NA 2/2/2 3/3/3No. of tRNA genes 47 50 44 45Genome featuresNo. of putative prophagesb 6 12 ND 14No. of chemotaxis genes ND 162b 326b 65b

aModified after Reference 58.bReference 64.ND, not determined; ORFs, open reading frames; NA, not annotated in the original publication.


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


a metagenomic approach (34). Genome anal-ysis of the deltaproteobacterium Desulfovibriomagneticus RS-1 was recently completed aswell (NCBI: NC 012796). Genome compar-isons and phylogenetic analysis of magnetospir-illa and the magnetic coccus MC-1 have in-dicated that the magnetotactic phenotype isonly weakly reflected by the overall gene con-tent of the genome. For instance, Richter et al.(58) identified a magnetobacterial core genomeshared by all four sequenced strains of 891genes, and even the three closely related mag-netospirilla have only 1932 of their nearly 5000(4268–4925) genes in common. However, inaddition to the magnetosome genes discussedbelow, all MTB seem to share a notable num-ber of common characteristics: (a) genomessizes between 4.2 and 4.9 Mb, (b) G+C con-tent between 54.8 and 65.1%, and (c) unusuallyhigh numbers of putative chemotaxis transduc-ers found in bacterial genomes (65 in M. mag-neticum and M. magnetotacticum, 5 in Escherichiacoli ) (1, 25) as well as 65 signal transductionrelated genes in MC-1 (64). Another character-istic is that the highest number of hemerythringenes (32 in M. magneticum) among prokary-otic genomes (compare with 5 in Rhodospirillumrubrum) are implicated in oxygen sensing andsignal transduction and might play a central rolein magneto-aerotaxis (23, 24). Despite their dif-ferent habitats, all analyzed MTB have genesassociated with an autotrophic lifestyle includ-ing those encoding putative enzymes for carbonfixation. While magnetospirilla seem to fix CO2

via the Calvin-Benson-Bassham cycle (6, 48),MC-1 lacks RuBisCO genes but fixes CO2 viathe reverse tricarboxylic acid cycle (6, 64, 90).Magnetospirillum species and MV-1 are capableof anaerobic growth with nitrate or N2O as ter-minal electron acceptor, respectively, while themagnetic coccus MC-1 depends on oxygen res-piration (5, 8).

Among MTB the magnetic coccus MC-1exhibits the most conspicuous genome, andits phylogenetic position was controversiallydiscussed because of its mosaic-like genomiccomposition (17, 37). In addition, its genomecontains 14 predicted prophages, which is the

highest number found in a prokaryotic genomeso far (13).

Identification of Magnetosome Genes

Because of the limited amount of MTB speciesavailable in pure culture, genetic analysishas focused so far on the alphaproteobacte-rial freshwater strains M. gryphiswaldense andM. magneticum as model systems, which hasbeen facilitated by the availability of genomedata. Requirements for genetic manipulations,such as clonal selection on agar plates, suitableselection markers, gene transfer via conjugationor electroporation, site-directed and randommutagenesis, and the expression of GFP fu-sions, as well as the establishment of expressionsystems, are fulfilled only by these two MTB(41, 42, 44, 46, 51, 73–75).

The trait of magnetotaxis is genetically com-plex and requires a range of diverse gene func-tions, which are still not known completely.However, by using different approaches, thegenetic basis of magnetosome formation waspartially unveiled, and three main strategiesidentified a set of candidate genes with knownor suspected functions in magnetosome forma-tion, which are discussed below.

Reverse genetics by proteomic analysis. Onthe basis of the assumption that magnetosome-associated proteins might have specific rolesin their biogenesis, several studies attemptedto identify the corresponding genes in sev-eral MTB. For example, the first of thesestudies identified the mam22 gene (equivalentto mamA ) in M. magnetotacticum encodinga conserved TPR protein (52, 53). InM. gryphiswaldense, proteomic analysis and re-verse genetics so far led to the identifica-tion of 23 genes encoding specific bona fidemagnetosome-associated proteins (29, 30, 58).The proteins were assigned to various proteinfamilies and named MamA, B, C, D, E, F, G, J,M, N, O, Q, R, S, T, W, Y, and X; MmsF; Mms6;as well as MM22, MmeA, or MtxA. With the ex-ception of mmeA, mtxA, mms16, and MM22, allidentified genes were found colocalized within a


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


single genomic region that was recognized lateras an MAI.

In a similar study in M. magneticum, 78proteins copurified with isolated magnetosomeparticles. However, only 14 of them (Mms16,Mms6, and MamC, F, E, J, M, A, R, B, D,K, S, and O) overlapped with the magneto-some subproteome of M. gryphiswaldense (29,81). Because a large proportion of the identifiedproteins are highly abundant cellular proteinsinvolved in general housekeeping metabolism,most proteins identified by Tanaka et al. (81)probably represent contaminations from othercellular compartments that became adsorbed tothe magnetosome particles during cell disrup-tion. For example, the phasin Mms16 (ApdaA),which was originally suggested to be involvedin the formation of MM vesicles in M. mag-neticum because of its detection in the MM frac-tion (50), is not associated with magnetosomesin vivo, and its function seems entirely unre-lated to magnetosome formation but instead isassociated with the formation of polyhydrox-ybutyrate granules (31, 73). In summary, al-though the proteomic approach is powerful forthe identification of genes encoding abundantmagnetosome-associated proteins, it is prone tofalse detection of contaminations and bears therisk of missing less abundant proteins, or thosethat are not or only temporarily localized at theMM but nevertheless have relevant functionsfor magnetosome formation.

Forward genetics by mutant analysis. Spon-taneous mutants were observed at high fre-quency (up to 10−2) in M. gryphiswaldense duringprolonged cold storage and stationary growth,or during exposure to oxidative stress. Mutantswere recognized on plates by their pale appear-ance, compared with the dark brown color ofwild-type colonies (63, 86), and displayed vari-ous phenotypes including smaller, fewer, or nomagnetosomes. All nonmagnetic mutants har-bored large and sometimes multiple deletionsthat mapped within a single genomic regioncorresponding to the MAI. Excision was likelycaused by homologous recombination betweenthe numerous direct repeats present in this area

(33, 86). In M. magneticum the spontaneous lossof the MAI, or of large parts of it, was also ob-served and resulted in nonmagnetic or weaklymagnetic phenotypes (25, 49). The tendency offrequent, spontaneous loss of the magneticphenotype and in particular of high geneticinstability of the genomic MAI under certainconditions requires extreme caution because itmight obscure the results of genetic analysis, aswe discuss below.

Although transposon mutagenesis hasthe potential for an unbiased identificationof all genes involved in magnetosome for-mation, the polar character of transposoninsertions precludes higher-resolution analysisbecause of the operonal organization of mostmagnetosome genes. Nevertheless, severalstudies with partially controversial resultswere described. For example, a nonmagnetic,complementable transposon mutant (NM4)from M. gryphiswaldense had an insertionwithin a gene encoding a CheIII-like protein.The gene mapped within the neighborhoodof mamW within the MAI (45). A mini-Tn5mutagenesis in M. magneticum was correlatedwith the isolation of 62 nonmagnetic clones(46, 87). Curiously, neither of the previouslyidentified mam and mms genes, nor other geneslocated within these operons, were identifiedin the studies mentioned above, which is instriking contrast to an independent study inwhich all transposon mutants obtained in thesame organism resulted from insertions in themamAB cluster (41). In addition, as none of thetransposon insertants was trans-complementedso far, the conclusions of the aforementionedstudies must be treated with caution, and manyof the supposed mutants are likely to representsecond-site mutations (46, 87, 88). In fact,Nash (49) demonstrated in a comprehensivemutagenesis study in the same organism that,by screening 5809 transposon insertants, 173of 192 with various nonmagnetic or weaklymagnetic phenotypes had lost one or severalgenes from the mamAB operon by spontaneousdeletion. Fourteen of the remaining 19 hadtransposon insertion sites inside the mamABcluster in six different genes, whereas only 5 of


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


them had insertions in previously unsuspectedgenes outside the MAI clusters, although itcannot be excluded that these carry second-sitemutations outside the screened mamAB cluster.

Comparative genome analyses. Cross-comparison of the available MTB genomeswas motivated by the assumption that bacteriathat share the common feature of magnetotaxiscan be expected to have a distinct set of genesin common (58). By comparing the genomesof M. gryphiswaldense, M. magneticum, M. mag-netotacticum, and the magnetic coccus MC-1against each other and genome databases,152 genes that are shared exclusively be-tween various magnetospirilla were identified.Twenty-eight genes were group specific andthus exclusively found in all four analyzedgenomes of MTB, but not in any nonmagneticorganism (MTB specific), or displayed only re-mote similarity to nonmagnetobacterial genes(MTB related). Eighteen of these genes arelocated within the MAI of M. gryphiswaldense(mamD, T, F, S, C, I, X, H, K, O, P, Q, A, B, M,E and mamH-like as well as mmsF ), and 10 arelocated outside.

In conclusion, the three strategies resultedin the identification of a partially redundantset of candidate genes. For example, both pro-teomic analysis of the MM and bioinformaticprediction revealed an extensive overlap in theidentified candidate genes. For some of them,including mamA, J, K, G, F, D, C, M, B, theirfunction related to magnetosome formationwas already experimentally confirmed by tar-geted deletion (36, 40, 41, 60, 61). On the otherhand, the comparative approach also led tothe identification of 14 previously unsuspectedgenes. If all genes that are colocated togetherwith the predicted or verified candidate genesin several large operons within and outsidethe MAI were taken in account, the tentativenumber of genes potentially involved inmagnetotaxis might be as high as 50. However,most genes identified by reverse genetics andbioinformatics are confined to the MAI, whichis in agreement with the fact that all the reliablemutants with a magnetosome phenotype map

to this region. In addition, it has been arguedthat the number of candidate genes identifiedoutside the MAI in fact encode functionsinvolved in signal transduction and motilityrather than biomineralization and chain assem-bly (58). A proportion of the predicted genesseem redundant, as they provide accessoryfunctions, or are present in multiple copies.Thus, the actual number of genes required formagnetosome formation might be as low as15–25, and this number might shrink furtherif only genes essential for magnetite biominer-alization, not chain assembly, are considered.In summary, the most probable estimation ofthe minimum specific gene set currently seemsto be between 20 and 50, a number that is suf-ficiently small for experimental analysis. Thatmeans that less than 1% of the genome is likelyto be dedicated to magnetotaxis. Thus, thenumber of predicted genes approaches that im-plicated in other complex metabolic traits, suchas, for example, bacterial photosynthesis (57).

Molecular Organizationof Magnetosome Genes

Characteristics of the genomic MAI. A re-verse genetic approach in M. gryphiswaldense(30) led to the first indication that magnetosomegenes are not scattered throughout the chro-mosome but are apparently confined to a singlegenomic region that was later recognized as anMAI (30), in analogy to other genomic islandsdescribed previously (15). Additional studiesprovided evidence that homologous MAIs arepresent in the genomes of M. magneticum (25),MV-1 (33), MC-1 (64), and two uncultivatedMTB (34). The MAI region displays the follow-ing characteristics: (a) It harbors most identifiedmagnetosome (mam/mms) genes, (b) it containsa high proportion of transposase genes (up to20% of the coding region), and (c) it is geneti-cally unstable owing to frequent homologousrecombination between numerous direct andinverted repeats (except for MC-1). In addi-tion, a large fraction of hypothetical genes arepresent, as well as hemerythrins or other genesputatively involved in signal transduction.


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


(d ) The MAI region extends over 60–150 kb(25, 33, 86), and (e) it displays a bias in G+Ccontent and tetranucleotide frequency (33).

Diversity and organization of magnetosomegenes within the MAI. The operon orga-nization of magnetosome genes is conservedamong different Magnetospirillum strains andalso, to a lesser extent, strains MC-1, MV-1,and two metagenomic clones (Fos001 andFos002). This finding provides evidence thatmany magnetosome genes are universal amongcultivated and uncultivated MTB (Figure 3).Encoded proteins display substantial sequencedivergence, which is in the range of 30–45%identity for most magnetosome proteins. Be-sides some conspicuous organizational patterns(for instance, mamK, M, O, P, A, Q, and B arealways arranged within a single mamAB-likeoperon, and mamO, P, A are always syntenic),several genes universally present in the MAIare shuffled between different magnetosomegene clusters in different MTB. The sequenceand organization of magnetosome genes areless conserved in strains MC-1 and MV-1 thanin magnetospirilla (Figure 3). In the MAI ofMV-1, magnetosome genes are organized intofive putative operons, along with a number ofyet unknown hypothetical and conserved hy-pothetical genes. For example, the largest genecluster comprises 17 genes, 14 of which (mamI,E, K, M, N, O, P, A, Q, R, B, S, T, and Y ) areorthologous to magnetosome genes found in allMagnetospirillum species and that are, with theexception of mamN, R, and Y, also present inMC-1.

Despite the general conservation, there arealso some notable differences between vari-ous MTB. For instance, several magnetosomegenes are present as duplicates in the MAIof some analyzed MTB. For example, ho-mologs of mamE, O, Q, R, and B are presentwithin an additional operon in M. magneto-tacticum and M. magneticum but are missing inM. gryphiswaldense. Whereas only one mreB-likegene (mamK ) is present within the mamAB-like cluster of MC-1 and the magnetospirilla,a second mamK gene is located within an ad-

ditional operon in MV-1 (mamDFHK cluster),and two divergent copies of mamK are presentin the metagenomic clone Fos001 as well. Onthe other hand, the presence of the mamJ gene,which has a function in magnetosome chain as-sembly in M. gryphiswaldense, seems to be con-fined to the magnetospirilla, but it is absentfrom the genome of other MTB.

In summary, a set of genes including mamH,I, E, K, M, O, P, A, Q, B, S, T, C, D, Z, and X, aswell as mms6 and mmsF, is shared by the MAIof all five cultivated MTB, while only mamE,K, M, O, P, A, Q, B, and S are present in thelimited available genome sequence of metage-nomic clones. That only a fraction of mag-netosome genes is conserved with respect tosequence homology and MAI localization sug-gests that the minimal set of genes requiredfor magnetosome biomineralization might besmaller than previously suggested (58) and thatsome genetic redundancy is present in the mag-netospirilla. Alternatively, other functions es-sential for magnetosome formation might beperformed by unrelated genes by nonorthol-ogous displacement in other MTB. In gen-eral, the considerable genetic variability foundin the MAI may account for the observed di-versity in magnetosome morphologies and ar-rangements found in many MTB other thanmagnetospirilla.

Functions of Individual Genesand Proteins

Although a large set of candidate genes has beenidentified, only a few examples of genes andproteins have been characterized in detail withrespect to their function.

Mms6 affects magnetite crystallization invitro. The small Mms6 protein, which is atightly bound constituent of the MM, may playa role in the magnetite crystallization process(47). Mms6 bears a Leu-Gly-rich motive thatis also conserved in the magnetosome proteinsMamG and MamD (29) and that is reminiscentof self-aggregating proteins of other biomin-eralization systems (3, 68). The overproducedMms6 protein from M. magneticum exhibited


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


Mag

neto

spiri

llum

gry

phis

wal

dens

e M

SR-1

mm

s6

mam

GFD

C

mam

GFD

C

mm

s6

mm

s6

mam

AB

mam

AB

mam

AB li

ke

mam

AB li

ke

mam

AB li

ke

mam

AB li

kem

amD

FHK

mm

s6 li

kem

amCX

Z

mam

XY li

ke mam

D li

keM

mc1

_2

24

0M

mc1

_2

25

7M

mc1

_2

26

6M

mc1

_2

27

0M

mc1

_2

27

2M

mc1

_2

27

1

mv

1g

07

0m

v1

g0

68

mv

1g

05

7m

v1

g0

52

mv

1g

05

0m

v1

g0

45

mm

sFm

v1

g0

46

mv

1g

04

2m

v1

g0

41

mv

1g

03

2m

v1

g0

31

mv

1g

02

5m

v1

g0

30

mv

1g

02

6

mv

1g

04

7

A li

ke

mam

HIE

Cm

ms

mam

GFD

mam

XY

mam

XY

Fts

Z li

ke

Fts

Z li

ke

mm

sF

mm

sF

mm

sFfu

r

6W

GF

mgr

40

74

mm

sFm

gr4

07

0

mgr

40

71

D

DF

6H

IE

JK

LM

NO

PA

QR

BS

TU

V

HI

EJ

KL

MN

O

E

6F

HI

E

FH

IE

K

IE

KM

NO

PA

QR

BS

TY

K-I

IH

mm

sFm

amD

like

6C

XZ

R li

ke

KM

NO

PA

QR

BS

CK

FM

OP

AQ

BS

TZ

X

KL

MN

O

PA

QR

BS

T

PA

QR

BS

TG

FD

UV

WE

OQ

RB

ZX

YF

EO

QR

BZ

XY

FW

C

DF

6C

CH

IE

JK

LM

NO

PA

QR

BS

TU

Z

Fts

Z li

ke

B li

ke

XY

mam

GFD

Cm

amAB

mam

XY

M. m

agne

ticum

AM

B-1

M. m

agne

tota

ctic

um M

S-1

Mag

neto

cocc

us m

arin

us M

C-1

Fos0

02

Fos0

01

Mag

netic

vib

rio

MV-

1

TP

R p

rote

in

Ge

ne

ric

tra

nsp

ort

er

Aci

dic

re

pe

titi

ve p

rote

in

Lem

A-l

ike

pro

tein

Hyp

oth

eti

cal p

rote

in

Ge

ne

ral m

ag

ne

toso

me

ge

ne

Act

in-l

ike

Mre

B p

rote

inM

TB

-sp

eci

fic

ge

ne

PD

Z d

om

ain

(H

trA

-lik

e

seri

ne

pro

tea

se)

1 k

bC

DF

tra

nsp

ort

er


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


iron-binding activity and had a striking effecton the morphology of growing magnetite crys-tals in vitro by catalyzing the formation of uni-form 30-nm-sized, single-domain particles insolution (3, 56). However, the significance ofMms6 for magnetosomal magnetite synthesisin vivo remains to be shown.

MamGFDC control the size of magneto-somes. Mature magnetite crystals producedby MTB typically are within the stable singlemagnetic domain size range (between 30and 120 nm) (7). Smaller, superparamagneticparticles would not efficiently contribute tothe cellular magnetic moment, whereas largercrystals tend to reduce their total magneticmoment by developing multiple magneticdomains. Thus, biological control of crystalsizes within a narrow range is crucial for theirefficient use in magnetic navigation.

The four small, hydrophobic magnetosomeproteins MamG, MamF, MamD, and MamCare specifically involved in the size controlof magnetite crystals in M. gryphiswaldense(Figure 4a) (60). Except for mamG, which ispresent only in Magnetospirillum strains, themamD, mamF, and mamC genes are part ofthe MTB-specific set of 28 signature genes de-scribed above (58). The MamGFDC proteinsaccount for approximately 35% of all MM pro-teins and are tightly bound to the MM (29),and their intracellular localization seems to beconfined to the magnetosomes (44). The mostabundant MM protein, MamC, has been uti-lized as an MM anchor in translational fusionsfor the magnetosome-specific display of het-erologous hybrid proteins (44, 92). The sec-ond most abundant MM protein, MamF, tendsto form stable oligomers even in the presenceof sodium dodecyl sulfate (29). The MamD

and MamG proteins share a conspicuous mo-tif containing a Leu-Gly repeat also found inthe Mms6 protein (Figure 4a). MamGFDCare not essential for magnetosome formation,as mutants deficient in either mamC or theentire mamGFDC operon still synthesize mag-netite crystals and form intracellular MM vesi-cles (60). However, crystals formed by thesemutants were significantly smaller and less reg-ular with respect to morphology and chain-like organization (60) (Figure 4a). Apparently,growth of mutant crystals was not only spa-tially constrained by the size of MM vesicles,but the absence of MamGFDC likely also in-hibited the growth of magnetite crystals di-rectly. Formation of wild-type-sized magnetitecrystals could be gradually restored by trans-complementation with any combination of one,two, or three genes of the mamGFDC operon,whereas the expression of all four genes resultedin large crystals, even exceeding wild-type size.These experiments have demonstrated that theMamGFDC proteins have partially redundantfunctions and, in a cumulative manner, controlthe growth of magnetite crystals (60).

MamJ and MamK: cytoskeletal elementsare involved in the assembly of magne-tosome chains. For maximum sensitivity ofthe magnetic orientation, magnetosomes arearranged in single or multiple linear chainsin which the cellular magnetic dipole is thesum of the permanent magnetic dipole mo-ments of the individual single-domain mag-netosome particles (22). However, a string ofmagnetic dipoles has an inherent tendency ofcollapsing to lower its magnetostatic energywithout mechanical stabilization (38). Dipolarattractions between particles are not sufficientfor maintaining straight chains, as isolated

←−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−Figure 3Molecular organization of magnetosome genes in the magnetosome island (MAI) of all sequenced magnetotactic bacteria (MTB) areshown. Different colored arrows indicate characteristic features of the encoded proteins. The red outline indicates MTB-specific genes,which are shared by all MTB but have no homologs in nonmagnetic organisms (57). Diagrams of the representative morphologies areshown. In the case of the anonymous metagenomic clones Fos001 and Fos002, the assignment of a morphotype is tentative and basedon a phylogenetic analysis of encoded genes, which revealed that the MAI sequences shown most likely originate from an organismresembling a magnetic coccus or a magnetic vibrio, respectively (32).


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


CAR domain

Wild typeΔmamJ mutant

Electron micrograph of ΔmamJ mutant

Alanine rich Glycine richTransmembrane domain

Molecularweight(kDa)

Isoelectricpoint

30.2

12.4 4.9

9.7

9.612.3

9.37.7

Relativeabundancein MM (%)

16.3

3.1

14.8

>1.0

- 125MamC 1 -

- 112MamF 1 -

MamG - 881 -

LGLGLGLGAWGP

LGLGLGLGAWGP

MamD - 3141 -

ΔmamGFDC mutant

a Domain organization of the MamGFDC proteins

b Domain organization of the MamJ protein

N- -C1 426

81 256Tandem repeat

unit 1

Tandem repeat unit 1



Gly richAla rich

Acidic

Amino acids:

Basic

HydrophobicHydrophilic (except Glu and Asp)

MM vesicles

Magnetite crystals

Prolin

Cell membrane

Magnetosome filament

Figure 4Characteristics and domain organization of the (a) MamGFDC and (b) MamJ proteins, and phenotypes of their respective deletionmutants. The MamJ protein contains several domains with highly biased amino acid composition. The central acidic repeat (CAR)domain has a variable length (Reference 62) and contains two identical stretches arranged as tandem repeats that are rich in acidicamino acid residues. Color pictures show cryoelectron tomograms of wild-type cells and a mamJ deletion mutant. Electronmicrographs and diagrams by A. Scheffel (black-and-white pictures), cryoelectron tomograms (color pictures) by M. Gruska andJ. Plitzko. Schematics and micrographs modified from References 60 (panel a) and 61 (panel b). MM, magnetosome membrane.

magnetosomes tend to form structures such asflux-closure rings and folded chains that repre-sent lower magnetostatic energy states for in-plane dipoles (54). The assembly and mainte-nance of well-ordered chains in vivo are highlycontrolled at the genetic and structural levels.Apparently, constituents of the MM mediatecoherence within the chain (61, 82), as isolatedmagnetosomes are interconnected by adherent

organic material that appears to form junc-tions between individual particles. Recently,two complementary cryoelectron tomographystudies demonstrated a network of filaments,3–4 nm in diameter, that traverses cells ofM. gryphiswaldense and M. magneticum closelyadjacent to the CM. Magnetosomes wereclosely arranged along this cytoskeletal struc-ture, which has been tentatively called a


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


magnetosome filament (MF) and is presum-ably formed by the MamK protein (21)(Figure 4b). MamK has homology to the cy-toskeletal actin-like MreB protein, which is in-volved in a number of essential cellular pro-cesses in bacteria, such as cell shape determina-tion, establishment of cell polarity, and chromo-some segregation (26, 28). Because of this in-triguing similarity it was speculated that MamKmight have a role in aligning and stabilizingthe magnetosome chain (68). In fact, the MFwas absent in a mutant of M. magneticum fromwhich the mamK gene was deleted (40). Themagnetosome chains were less regular and dis-persed throughout the cell, which led to theconclusion that the MF is involved in the sta-bilization and anchoring of the magnetosomechain within the cell (39, 40). MamK of M. mag-neticum formed straight filaments, structurallyand functionally distinct from the known MreBand ParM filaments, rather than helical struc-tures (55). The formation of MamK straight fil-aments in M. gryphiswaldense was independentof the presence of other magnetosome genes,as revealed by a GFP-based localization studyin a deletion mutant lacking large parts of theMAI (62). Recombinant MamK of M. magne-totacticum expressed in E. coli in vitro polymer-ized into long straight filamentous bundles, theformation of which was dependent on the pres-ence of a nonhydrolyzable ATP analogue (83).In MTB other than magnetospirilla and MC-1, multiple divergent mamK genes were spec-ulated to have slightly distinct functions andmight also be correlated to the more complexintracellular magnetosome architectures, suchas the organization of multiple chains as foundin several uncultivated MTB (34).

In addition to MamK, the acidic MamJ pro-tein identified in the MM of M. gryphiswaldense(29) is involved in the control of magnetosomechain assembly. The mamJ gene immediatelyprecedes mamK within the mamAB operon,which is transcribed from a single promoter (63,65). The most conspicuous sequence feature ofthe MamJ protein is the central acidic repetitive(CAR) domain (Figure 4b), which, however,does not seem absolutely essential for its func-

MF: magnetosomefilament

tion in chain assembly (62). A deletion mutantof mamJ no longer produces straight magne-tosome chains, but magnetite crystals are ar-ranged in compact clusters (61, 62) (Figure 4b),whereas empty vesicles and immature crystalsare scattered throughout the cytoplasm and de-tached from the MFs. Apparently, in cells lack-ing MamJ, mature magnetosome crystals arefree to agglomerate once they are in close prox-imity to each other (39, 61). Localization studieswith GFP suggest that MamJ interacts with theMF, and the direct interaction between MamKand MamJ was experimentally confirmed bytwo-hybrid experiments (62). One obviousmodel derived from these results is that MamJconnects magnetosomes to the cytoskeletal MFformed by MamK, which stabilizes the magne-tosome chain and prevents it from collapsing(62). However, so far it is unknown if additionalproteins are involved in chain formation, andseveral conflicting observations in two differ-ent Magnetospirillum species seem to indicatethat MamJ and MamK proteins could performdifferent or additional functions (39, 70).

This additional function could be, for in-stance, the sorting and assembly of newlybiomineralized crystals into the nascent chains(21, 80). In cells of M. gryphiswaldense that weregrown at steadily high iron concentrations, im-mature crystals were preferentially found atthe ends of the chain, whereas empty vesicleswere dispersed along the entire length of thecell in iron-starved cells devoid of magnetitecrystals (61). In contrast, immature crystalliteswere formed simultaneously at multiple sitesalong the entire length of the cell after in-duction of magnetite synthesis by the additionof iron to iron-starved, resting cells (18, 61).These growing crystals, subsequently concen-trated at midcell, first assembling into imper-fect, loosely spaced chains, gradually developedinto straight, tightly spaced chains of matureparticles. It is not yet known how this dynamiccellular localization is controlled or what causesnew magnetosomes to form at the ends of theinherited chain (21). MamK, similar to somerelated MreB-like proteins, might function as amolecular motor (28, 84), thereby establishing


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


Horizontal genetransfer:incorporation ofgenetic material by anorganism from anotherorganism withoutbeing the offspring ofthat organism

the chain by transporting newly formed magne-tosomes with growing crystals to the ends of thenascent preexisting magnetosome chains. Thisnotion is supported by the observation that theassembly of MamK filaments in E. coli is a highlydynamic and kinetically asymmetrical process(55), and polymerization and depolymerizationprocesses might generate forces sufficient forthe transport and relocation of magnetosomes.Furthermore, the position and polarity of themagnetosome chain might relate to other cel-lular structures relevant to magnetotaxis, suchas the flagella motor (80), and it was suggestedthat the inherent molecular polarity of MamKis translated into a mechanism for controllingglobal cell polarity (55).

This also relates to the problem of howmagnetosomes are correctly positioned at mid-cell, and how they are properly segregated todaughter cells during cell division (21). Lossof MamJ resulted in an uneven segregationof magnetosome particles during cell division(62), and there are further indications for acontrolled mechanism of magnetosome posi-tioning. It was suggested that spatial informa-tion for localization and distribution of themagnetosome chains could be provided by in-teraction with other positional determinantscontrolling cell division in bacteria, such asthe tubulin-like FtsZ protein involved in di-visome formation (26, 28, 84). In addition tothe generic ftsZ gene present elsewhere in thegenome, an ftsZ-like gene is colocalized withother putative magnetosome genes within themamXY operon of M. gryphiswaldense, whichled to speculations whether tubulin-like cy-toskeletal structures other than the actin-likeMF are involved in magnetosome assembly (58,70). In one of the two mamK-like genes froman uncultivated MTB, an N-terminal FtsZ-like domain is linked to a C-terminal MamK-like actin domain protein. The fusion of atubulin- and an actin-like domain within a sin-gle chimeric polypeptide lends further supportto the hypothesis that both proteins may inter-act and that, in addition to MamK, FtsZ-likecytoskeletal elements might be involved inmagnetosome assembly (34, 58, 70).

Evolution of Magnetotaxis

Magnetotaxis and magnetosome biomineral-ization are widespread among Proteobacteriaand the different species produce crystalsof different morphology (2, 7, 14). As il-lustrated in Figure 5, most of the knownMTB belong to the Alphaproteobacteria,while Desulfovibrio magneticus and the many-celled magnetotactic prokaryote belong tothe Deltaproteobacteria (14, 59, 89). The“Candidatus Magnetobacterium bavaricum”branches close to the Nitrospira phylum (79),and there is some evidence that uncultivatedrod-shaped MTB produce greigite crystals inthe Gammaproteobacteria (78).

Alphaproteobacteria-MTB do not form acoherent phylogenetic group, but based on 16SrRNA gene analysis some MTB-containingbranches of the alphaproteobacterial phyloge-netic tree are interspersed with nonmagneto-tactic representatives (2). For example, non-magnetotactic spirilla were repeatedly isolatedthat seem phylogenetically closely related tomagnetotactic Magnetospirillum species (76). Asshown in Figure 5, Magnetospirillum speciesand their nonmagnetotactic relatives branchclosely together even below the genus level(33), whereas the closest relative outside thegenus Magnetospirillum is the photosyntheticPhaeospirillum molischianum, and strains MV-1and MC-1 are even more distantly related tothe magnetospirilla.

Two possible evolutionary scenarios havebeen hypothesized to explain the phylogeneticsituation of MTB. In the first scenario, themagnetotactic trait has evolved independentlyseveral times and was also lost in severallineages (14). The second scenario implieshorizontal transfer of genes required to raisethe magnetotactic phenotype among differentphyla and species (33, 63, 86). The strict con-gruence of 16S rRNA gene and magnetosomegene phylogenies of all shared mamAB genes, aswell as the occurrence of magnetosome genesoutside the supposedly transferred MAI, seemsto argue against a recent event of horizontalgene transfer in magnetospirilla (33, 58).


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


Barbell-shaped MTB

MMP

Desulfovibrio burkinabensis

Desulfovibrio magneticus

Magnetobacterium bavaricum

Magnetic coccus MC-1

Metagenomic clone 1

Phaeospirillum molischianum

M. magneticum

M. magnetotacticum

Magnetospirillumgryphiswaldense

Aquaspirillum polymorphum

Agrobacterium tumefaciens

Rhodospirillum rubrum

Metagenomic clone 2

Magnetic vibrio MV-1

MTB?

HGT?

HGT?

HGT?

HGT

HGT

HGT

HGT

HGT

HGT ?

δ γ α

Nitrospira

Figure 5Putative evolution of magnetotaxis. The branching order of the schematic tree is according to Reference 2. Branching length is not toscale. Putative events of HGT are indicated based on sequence data; hypothetical HGT events are indicated with question marks.Representative morphotpypes of displayed branches are shown. According to the model, the genomic MAI has been horizontallytransferred among different phyla as indicated. The origin of magnetosome genes remains unknown. Abbreviations: HGT, horizontalgene transfer; MAI, magnetosome island; MMP, magnetotactic prokaryote; MTB, magnetotactic bacteria.

However, this finding might be explained bya period of coevolution of the transferred MAIand the genome of Magnetospirillum speciesafter an ancient event of transfer. It is possi-ble that magnetosome genes were acquiredby the ancestors of recent magnetospirilla,which may have subsequently separated intodifferent species such as M. magneticum, M.gryphiswaldense, and M. magnetotacticum (33).

During further evolution, the ancestral MAIthen underwent extensive rearrangements,such as duplication, deletion, insertion, andphage integration, leading to distinctive com-positional features. In the absence of selectivepressure for magnetotaxis, nonmagnetotactic

representatives branching within the MTB,such as A. polymorphum, may then haveoriginated from secondary loss of the MAI,for example, by large deletions caused byhomologous recombination between directrepeats within the MAI, a process that occursfrequently during lab cultivation of variousmagnetospirilla (25, 33, 86). In addition,this model suggests the independent hori-zontal transfer of magnetosome genes froman unknown ancestor to strains MC-1 andMV-1 occurred a longer time ago. The closephylogenetic affinity of a photosynthetic bac-terium (P. molischianum) to the magnetospirillasuggests that their ancestor might have been a


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


phototrophic organism, with whom MTB alsoshare the ability to form intracellular mem-branes. This hypothesis is further supported bythe presence of genes in M. magnetotacticum,which might represent remnants of a previousphotosynthetic lifestyle (9). For example,a gene encoding a putative photosyntheticreaction center M subunit displays similarity tohomologs in Methylobacterium chloromethanicumand Rhodospirillum centenum (97% and 76%identity, respectively). Taken together, theseresults suggest that a model of MTB evolutioninvolving horizontal gene transfer is muchmore likely than independent evolution of thiscomplex trait.

DIRECTIONS FOR FUTURERESEARCH

Despite the considerable progress in our under-standing of the molecular mechanisms govern-ing magnetotaxis and biomineralization, a num-ber of remaining questions need to be solved byfuture approaches. Although the largest part ofcandidate genes involved in magnetosome for-mation has likely been identified, elucidation oftheir individual functions by genetic and bio-chemical analysis will remain a challenge forthe next few years. On the other hand, onlya few magnetotactic species, mostly from theAlphaproteobacteria, have been analyzed withrespect to the genetic basis of magnetosomeformation. It will be of interest to discoverwhether the same genetic mechanisms do ap-ply in more remotely related MTB from theDelta-, Gamma-, or Nitrospira Proteobacteriaphyla. However, only a small minority of MTBfrom natural populations have been isolatedin pure culture, whereas the great diversity of

naturally occurring species has been largelyunexplored, including, for example, giant ormulticellular forms of MTB, as well as thoseproducing crystals of different chemical com-position (i.e., greigite). However, unlike otheruncultivated bacteria, MTB can be collectedfrom environmental samples to virtual homo-geneity by utilizing their magnetically directedmotility. It has been already demonstrated thatthis can be used for selective cloning of largegenomic fragments harboring the MAI fromuncultivated MTB (34).

Not only will metagenomic analysis providea census of genetic diversity in magnetosomeformation, but by use of single-cell approachesthis might also be extended to the analysis ofcomplete genomes from single or a few individ-ual cells, even from underrepresented speciesof natural populations. The increasing genomeinformation will also bring us closer to the iden-tification of the minimal set of genes requiredfor magnetosome formation. This might in turnfacilitate the heterologous expression of singlegenes, entire operons, or ultimately, the ge-netic reconstitution of the complete pathwayfrom uncultivated MTB in heterologous hosts.In the future this could provide accessto the biomineralization of the spectacularhost of crystal morphologies and arrange-ments, which are expected to display unprece-dented magnetic and crystalline characteristics.A complete understanding of magnetosomeformation at the molecular level will also beuseful for the engineering and functionaliza-tion of magnetosomes for their use as mag-netic nanomaterials with tailored magneticand crystalline properties that could be uti-lized in various biotechnological applications(43).

SUMMARY POINTS

1. Magnetosomes are specialized organelles for magnetic navigation comprisingmembrane-enveloped, nano-sized crystals of a magnetic iron mineral.

2. Physicochemical control over biomineralization is achieved by its compartmentalizationwithin vesicles of the MM, that is a phospholipid bilayer originating from the CM byinvagination.


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


3. The MM is associated with a specific set of Mam, Mms, Mtx, and Mme proteins thatbelong to characteristic families including TPR proteins, CDF transporters, HtrA-likeserine proteases, actin-like proteins, generic transporters, and MTB-specific proteinswith no homology to other proteins.

4. The magnetotactic phenotype is only weakly reflected by the overall gene content ofmagnetobacterial genomes, and MTB analyzed thus far share a magnetotactic core set of891 genes and have a set of 28 conserved signature genes in common that is specificallyimplicated in magnetotaxis.

5. By various approaches, including forward and reverse genetics, as well as bioinformaticprediction, a partially redundant set of 15–50 candidate genes implicated in magnetosomeformation was identified.

6. Most magnetosome genes are located in several operons within a conserved genomicMAI, which shares common structural and compositional characteristics between variousMTB. The MAI also displays differences with respect to gene content and organizationthat may account for the diversity in magnetosome morphologies and arrangements inmany MTB.

7. Examples with known in vivo functions are the MamGFDC proteins, which regulate thesize of magnetite crystals, and the MamK and MamJ proteins, which control the assemblyof magnetosome chains along dedicated cytoskeletal structures, presumably formed bythe actin-like MamK protein.

8. Magnetotaxis is widespread within different lineages of eubacteria and most likely evolvedby horizontal transfer of the MAI to various phylogenetic groups.

DISCLOSURE STATEMENT

The authors are not aware of any affiliations, memberships, funding, or financial holdings thatmight be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENT

Research in the author’s lab has been funded by the Deutsche Forschungsgemeinschaft, theGerman BMBF, and the European Union. We wish to greatly knowledge the contributions ofall our previous and current students, coworkers, and collaborators.

LITERATURE CITED

1. Alexandre G, Greer-Phillips S, Zhulin IB. 2004. Ecological role of energy taxis in microorganisms. FEMSMicrobiol. Rev. 28:113–26

2. Amann R, Peplies J, Schuler D. 2006. Diversity and taxonomy of magnetotactic bacteria. In Magnetore-ception and Magnetosomes in Bacteria, ed. D Schuler, pp. 25–36. Heidelberg, Ger.: Springer

3. Arakaki A, Webb J, Matsunaga T. 2003. A novel protein tightly bound to bacterial magnetic particles inMagnetospirillum magneticum strain AMB-1. J. Biol. Chem. 278:8745–50

4. Balkwill D, Maratea D, Blakemore R. 1980. Ultrastructure of a magnetotactic spirillum. J. Bacteriol.141:1399–408

5. Bazylinski D, Blakemore RP. 1983. Denitrification and assimilatory nitrate reduction in Aquaspirillummagnetotacticum. Appl. Environ. Microbiol. 46:1118–24


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


6. Bazylinski DA, Dean AJ, Williams TJ, Long LK, Middleton SL, Dubbels BL. 2004. Chemolithoautotro-phy in the marine, magnetotactic bacterial strains MV-1 and MV-2. Arch. Microbiol. 182:373–87

7. Bazylinski DA, Frankel RB. 2004. Magnetosome formation in prokaryotes. Nat. Rev. Microbiol. 2:217–308. Bazylinski DA, Schubbe S. 2007. Controlled biomineralization by and application of magnetotactic bac-

teria. Adv. Appl. Microbiol. 62:1–629. Bazylinski DA, Williams TJ. 2006. Ecophysiology of magnetotactic bacteria. In Magnetosomes and Mag-

netoreception in Bacteria, ed. D Schuler, pp. 37–75. Heidelberg, Ger.: Springer10. Bellini S. 1963. Su di un particolare comportamento di batteri d’acqua dolce (About a particular behavior

of freshwater bacteria) and Ulteriori studi sui “batteri magnetosensibili” (Further studies on “Magnetosen-sitive Bacteria.”) J. Inst. Microbiol., Univ. Pavia, Italy. http://www.calpoly.edu/∼rfrankel/SBellini1.pdf

11. Blakemore R. 1975. Magnetotactic bacteria. Science 190:377–7912. Blakemore R. 1982. Magnetotactic bacteria. Annu. Rev. Microbiol. 36:217–3813. Bose M, Barber RD. 2006. Prophage Finder: a prophage loci prediction tool for prokaryotic genome

sequences. In Silico Biol. 6:223–2714. DeLong EF, Frankel RB, Bazylinski DA. 1993. Multiple evolutionary origins of magnetotaxis in bacteria.

Science 259:803–615. Dobrindt U, Hochhut B, Hentschel U, Hacker J. 2004. Genomic islands in pathogenic and environmental

microorganisms. Nat. Rev. Microbiol. 2:414–2416. Dubbels BL, DiSpirito AA, Morton JD, Semrau JD, Neto JN, Bazylinski DA. 2004. Evidence for a

copper-dependent iron transport system in the marine, magnetotactic bacterium strain MV-1. Microbiology150:2931–45

17. Esser C, Martin W, Dagan T. 2007. The origin of mitochondria in light of a fluid prokaryotic chromosomemodel. Biol. Lett. 3:180–84

18. Faivre D, Bottger L, Matzanke B, Schuler D. 2007. Intracellular magnetite biomineralization in bacteriaproceeds via a distinct pathway involving membrane-bound ferritin and ferrous iron species. Angew. Chem.Int. Ed. 46:8647–52

19. Faivre D, Menguy N, Posfai M, Schuler D. 2008. Environmental parameters affect the physical propertiesof fast-growing magnetosomes. Am. Mineral. 93:463–69

20. Faivre D, Schuler D. 2008. Magnetotactic bacteria and magnetosomes. Chem. Rev. 108:4875–9821. Frankel RB, Bazylinski DA. 2006. How magnetotactic bacteria make magnetosomes queue up. Trends

Microbiol. 14:329–3122. Frankel RB, Williams TJ, Bazylinski DA. 2006. Magneto-aerotaxis. In Magnetosomes and Magnetoreception

in Bacteria, ed. D Schuler, pp. 1–24. Heidelberg, Ger.: Springer23. Frankel RB, Williams TJ, Bazlinski DA. 2007. Magneto-aerotaxis. In Handbook of Biomineralization, ed.

E Baeuerlein, pp. 1–24. Weinheim, Ger.: Wiley-VCH24. French CE, Bell JM, Ward FB. 2008. Diversity and distribution of hemerythrin-like proteins in prokary-

otes. FEMS Microbiol. Lett. 279:131–4525. Fukuda Y, Okamura Y, Takeyama H, Matsunaga T. 2006. Dynamic analysis of a genomic island in

Magnetospirillum sp. strain AMB-1 reveals how magnetosome synthesis developed. FEBS Lett. 580:801–12

26. Gitai Z. 2007. Diversification and specialization of the bacterial cytoskeleton. Curr. Opin. Cell Biol. 19:5–12

27. First biochemicalanalysis of the MM andobservation of severalspecific bands in gelelectrophoresis ofproteins associated withpurified magnetosomes.

27. Gorby YA, Beveridge TJ, Blakemore RP. 1988. Characterization of the bacterial magnetosomemembrane. J. Bacteriol. 170:834–41

28. Graumann PL. 2007. Cytoskeletal elements in bacteria. Annu. Rev. Microbiol. 61:589–61829. Describes theproteomic analysis andthe identification of abona fide set ofmagnetosome proteinsand their sequenceanalysis.

29. Grunberg K, Muller EC, Otto A, Reszka R, Linder D, et al. 2004. Biochemical and proteomic anal-ysis of the magnetosome membrane in Magnetospirillum gryphiswaldense. Appl. Eviron. Microbiol.

70:1040–5030. Grunberg K, Wawer C, Tebo BM, Schuler D. 2001. A large gene cluster encoding several magnetosome

proteins is conserved in different species of magnetotactic bacteria. Appl. Environ. Microbiol. 67:4573–8231. Handrick R, Reinhardt S, Schultheiss D, Reichart T, Schuler D, et al. 2004. Unraveling the function

of the Rhodospirillum rubrum activator of polyhydroxybutyrate (PHB) degradation: The activator is aPHB-granule-bound protein (phasin). J. Bacteriol. 186:2466–75


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


32. Heyen U, Schuler D. 2003. Growth and magnetosome formation by microaerophilic Magnetospirillumstrains in an oxygen-controlled fermentor. Appl. Microbiol. Biotechnol. 61:536–44

33. Jogler C, Kube M, Schubbe S, Ullrich S, Teeling H, et al. 2009. Comparative analysis of magnetosomegene clusters in magnetotactic bacteria provides further evidence for horizontal gene transfer. Environ.Microbiol. 11:1267–77

34. Describes a methodfor targetedmetagenomic analysis ofMTB from theenvironment andprovides the firstevidence that largeclusters ofmagnetosome genes arealso present inuncultivated MTB.

34. Jogler C, Lin W, Meyerdierks A, Kube M, Katzmann E, et al. 2009. Towards cloning the magne-totactic metagenome: identification of magnetosome island gene clusters in uncultivated magne-totactic bacteria from different aquatic sediments. Appl. Environ. Microbiol. 75:3972–79

35. Jogler C, Schuler D. 2007. Genetic and biochemical analysis of magnetosomes in Magnetospiril-lum gryphiswaldense. In Handbook of Biomineralization, ed. E Baeuerlein, pp. 145–62. Weinheim, Ger.:Wiley-VCH

36. Junge K. 2008. Die Funktion der CDF-Transporter MamB und MamM beim magnetosomalen Eisentransportin Magnetospirillum gryphiswaldense. Doctoral thesis. Univ. Bremen. Ger. 141 pp.

37. Kainth P, Gupta RS. 2005. Signature proteins that are distinctive of alpha proteobacteria. BMC Genomics6:94

38. Kirschvink JL. 1982. Paleomagnetic evidence for fossil biogenic magnetite in western Crete. Earth Plan.Sci. Lett. 59:388–92

39. Komeili A. 2007. Molecular mechanisms of magnetosome formation. Annu. Rev. Biochem. 76:351–66 40. Demonstrates thatthe MM derives fromthe CM by invagination,identifies bycryoelectrontomography a networkof cytoskeletal filamentsthat are absent fromcells in which theactin-like mamK gene isdeleted.

40. Komeili A, Li Z, Newman DK, Jensen GJ. 2006. Magnetosomes are cell membrane invaginationsorganized by the actin-like protein MamK. Science 311:242–45

41. Describes the firstunmarked allelicreplacement and theexpression of GPF in aMTB for functionalcharacterization of amagnetosome protein.

41. Komeili A, Vali H, Beveridge TJ, Newman DK. 2004. Magnetosome vesicles are present be-fore magnetite formation, and MamA is required for their activation. Proc. Natl. Acad. Sci. USA

101:3839–4442. Lang C, Pollithy A, Schuler D. 2009. Identification of promoters for efficient gene expression in Magne-

tospirillum gryphiswaldense. Appl. Environ. Microbiol. 75:4206–1043. Lang C, Schuler D. 2006. Biogenic nanoparticles: production, characterization, and application of bacterial

magnetosomes. J. Phys. Condens. Matter 18:2815–2844. Lang C, Schuler D. 2008. Expression of green fluorescent protein fused to magnetosome proteins in

microaerophilic magnetotactic bacteria. Appl. Environ. Microbiol. 74:4944–5345. Li F, Li Y, Jiang W, Wang Z, Li J. 2005. Cloning and functional analysis of the sequences flanking mini-

Tn5 in the magnetosomes deleted mutant NM4 of Magnetospirillum gryphiswaldense MSR-1. Sci China C.Life. Sci. 48:574–84

46. Matsunaga T, Nakamura C, Burgess J, Sode S. 1992. Gene transfer in magnetic bacteria: transposonmutagenesis and cloning of genomic DNA fragments required for magnetosome synthesis. J. Bacteriol.174:2748–53

47. Matsunaga T, Okamura Y. 2003. Genes and proteins involved in bacterial magnetic particle formation.Trends Microbiol. 11:536–41

48. First report of acomplete genomesequence of a MTB.

48. Matsunaga T, Okamura Y, Fukuda Y, Wahyudi AT, Murase Y, Takeyama H. 2005. Completegenome sequence of the facultative anaerobic magnetotactic bacterium Magnetospirillum sp. strainAMB-1. DNA Res. 12:157–66

49. Nash C. 2008. Mechanisms and Evolution of Magnetotactic Bacteria. PhD thesis. Calif. Inst. Technol.,Pasadena. 75 pp.

50. Okamura Y, Takeyama H, Matsunaga T. 2001. A magnetosome-specific GTPase from the magneticbacterium Magnetospirillum magneticum AMB-1. J. Biol. Chem. 276:48183–88

51. Okamura Y, Takeyama H, Sekine T, Sakaguchi T, Wahyudi AT, et al. 2003. Design and applicationof a new cryptic-plasmid-based shuttle vector for Magnetospirillum magneticum. Appl. Environ. Microbiol.69:4274–77

52. Okuda Y, Denda K, Fukumori Y. 1996. Cloning and sequencing of a gene encoding a new memberof the tetratricopeptide protein family from magnetosomes of Magnetospirillum magnetotacticum. Gene171:99–102

53. Okuda Y, Fukumori Y. 2001. Expression and characterization of a magnetosome-associated protein,TPR-containing MAM22, in Escherichia coli. FEBS Lett. 491:169–73


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


54. Philipse AP, Maas D. 2002. Magnetic colloids from magnetotactic bacteria: chain formation and colloidalstability. Langmuir 18:9977–84

55. Describes theexpression ofGFP-tagged MamKprotein in E. coli, whichdisplays a distinctlocalization patterndifferent from otherbacterial actin-likeproteins.

55. Pradel N, Santini CL, Bernadac A, Fukumori Y, Wu LF. 2006. Biogenesis of actin-like bacterialcytoskeletal filaments destined for positioning prokaryotic magnetic organelles. Proc. Natl. Acad.

Sci. USA 103:17485–8956. Prozorov T, Mallapragada SK, Narasimhan B, Wang L, Palo P, et al. 2007. Protein-mediated synthesis

of uniform superparamagnetic magnetite nanocrystals. Adv. Funct. Mater. 17:951–5757. Raymond J, Zhaxybayeva O, Gogarten JP, Gerdes SY, Blankenship RE. 2002. Whole-genome analysis of

photosynthetic prokaryotes. Science 298:1616–20

58. Describes themagnetotactic coregenome and theidentification of asignature set of 28genes that is specificallyassociated with themagnetotacticphenotype.

58. Richter M, Kube M, Bazylinski DA, Lombardot T, Glockner FO, et al. 2007. Comparative genomeanalysis of four magnetotactic bacteria reveals a complex set of group-specific genes implicatedin magnetosome biomineralization and function. J. Bacteriol. 189:4899–910

59. Sakaguchi T, Arakaki A, Matsunaga T. 2002. Desulfovibrio magneticus sp. nov., a novel sulfate-reducingbacterium that produces intracellular single-domain-sized magnetite particles. Int. J. Syst. Evol. Microbiol.52:215–21

60. Demonstrates bydeletion mutagenesisand complementationanalysis that theabundant MamGFDCmagnetosome proteinshave partially redundantfunctions andcollectively control thesize of magnetitecrystals.

60. Scheffel A, Gardes A, Grunberg K, Wanner G, Schuler D. 2008. The major magnetosome proteinsMamGFDC are not essential for magnetite biomineralization in Magnetospirillum gryphiswaldense,but regulate the size of magnetosome crystals. J. Bacteriol. 190:377–86

61. Demonstrates thatthe mutagenesis of themagnetosome proteinMamJ results incollapsed magnetosomechains andagglomeration ofmagnetosome inclusters.

61. Scheffel A, Gruska M, Faivre D, Linaroudis A, Graumann PL, et al. 2006. An acidic protein alignsmagnetosomes along a filamentous structure in magnetotactic bacteria. Nature 440:110–14

62. Scheffel A, Schuler D. 2007. The acidic repetitive domain of the Magnetospirillum gryphiswaldense MamJprotein displays hypervariability but is not required for magnetosome chain assembly. J. Bacteriol.189:6437–46

63. First description ofthe genomic MAI,which comprises mostmagnetosome genes andis genetically unstable.

63. Schubbe S, Kube M, Scheffel A, Wawer C, Heyen U, et al. 2003. Characterization of a spontaneousnonmagnetic mutant of Magnetospirillum gryphiswaldense reveals a large deletion comprising aputative magnetosome island. J. Bacteriol. 185:5779–90

64. Schubbe S, Williams TJ, Xie G, Kiss HE, Brettin TS, et al. 2009. Complete genome sequence of thechemolithoautotrophic marine magnetotactic coccus strain MC-1. Appl. Environ. Microbiol. 75:4835–52

65. Schubbe S, Wurdemann C, Heyen U, Wawer C, Peplies J, et al. 2006. Transcriptional organization andregulation of magnetosome operons in Magnetospirillum gryphiswaldense. Appl. Environ. Microbiol. 72:5757–65

66. Schuler D. 2000. Characterization of the magnetosome membrane in Magnetospirillum gryphiswaldense. InBiomineralization, ed. E Baeuerlein, pp. 109–18. Weinheim, Ger.: Wiley-VCH

67. Schuler D. 2004. Biochemical and genetic analysis of the magnetosome membrane in Magnetospirillumgryphiswaldense. In Biomineralization, ed. E Baeuerlein, pp. 61–74. Weinheim, Ger.: Wiley-VCH

68. Schuler D. 2004. Molecular analysis of a subcellular compartment: the magnetosome membrane inMagnetospirillum gryphiswaldense. Arch. Microbiol. 181:1–7

69. Schuler D, ed. 2006. Magnetoreception and Magnetosomes in Bacteria. Microbiology Monographs, Vol. 3.Heidelberg, Ger.: Springer

70. Schuler D. 2008. Genetics and cell biology of magnetosome formation in magnetotactic bacteria. FEMSMicrobiol. Reviews 32:654–57

71. Schuler D, Baeuerlein E. 1996. Iron-limited growth and kinetics of iron uptake in Magnetospirillumgryphiswaldense. Arch. Microbiol. 166:301–7

72. Schuler D, Bazylinski DA. 2007. Techniques for studying uncultured and cultured magnetotactic bacteria.In Manual of Environmental Microbiology, ed. CJ Hurst, RL Crawford, JL Garland, DA Lipson, AL Mills,LD Stetzenbach, pp. 1129–38. Washington, DC: ASM Press

73. Schultheiss D, Handrick R, Jendrossek D, Hanzlik M, Schuler D. 2005. The presumptive magnetosomeprotein Mms16 is a PHB-granule bound protein (phasin) in Magnetospirillum gryphiswaldense. J. Bacteriol.187:2416–25

74. Schultheiss D, Kube M, Schuler D. 2004. A system for gene replacement in Magnetospirillumgryphiswaldense: Inactivation of the flagellin gene flaA results in nonmagnetotactic mutants lacking flagellarfilaments. Appl. Environ. Microbiol. 70:3624–31


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


75. Schultheiss D, Schuler D. 2003. Development of a genetic system for Magnetospirillum gryphiswaldense.Arch. Microbiol. 179:89–94

76. Shinoda Y, Akagi J, Uchihashi Y, Hiraishi A, Yukawa H, et al. 2005. Anaerobic degradation of aro-matic compounds by Magnetospirillum strains: isolation and degradation genes. Biosci. Biotechnol. Biochem.69:1483–91

77. Simmons SL, Bazylinski DA, Edwards KJ. 2006. South-seeking magnetotactic bacteria in the NorthernHemisphere. Science 311:371–74

78. Simmons SL, Sievert SM, Frankel RB, Bazylinski DA, Edwards KJ. 2004. Spatiotemporal distributionof marine magnetotactic bacteria in a seasonally stratified coastal salt pond. Appl. Environ. Microbiol.70:6230–39

79. Spring S, Amann R, Ludwig W, Schleifer K-H, van Gemerden H, Petersen N. 1993. Dominating roleof an unusual magnetotactic bacterium in the microaerobic zone of a freshwater sediment. Appl. Environ.Microbiol. 50:2397–403

80. Stephens C. 2006. Bacterial cell biology: managing magnetosomes. Curr. Biol. 16:R363–6581. Tanaka M, Okamura Y, Arakaki A, Tanaka T, Takeyama H, Matsunaga T. 2006. Origin of magnetosome

membrane: proteomic analysis of magnetosome membrane and comparison with cytoplasmic membrane.Proteomics 6:5234–47

82. Taoka A, Asada R, Sasaki H, Anzawa K, Wu L-F, Fukumori Y. 2006. Spatial localizations of Mam22 andMam12 in the magnetosomes of Magnetospirillum magnetotacticum. J. Bacteriol. 188:3805–12

83. Taoka A, Asada R, Wu LF, Fukumori Y. 2007. Polymerization of the actin-like protein MamK, which isassociated with magnetosomes. J. Bacteriol. 189:8737–40

84. Thanbichler M, Shapiro L. 2008. Getting organized—how bacterial cells move proteins and DNA.Nat. Rev. Microbiol. 6:28–40

85. Thomas-Keprta KL, Clemett SJ, Bazylinski DA, Kirschvink JL, McKay DS, et al. 2002. Magnetofossilsfrom ancient Mars: a robust biosignature in the martian meteorite ALH84001. Appl. Environ. Microbiol.68:3663–72

86. Ullrich S, Kube M, Schubbe S, Reinhardt R, Schuler D. 2005. A hypervariable 130 kb genomic region ofMagnetospirillum gryphiswaldense comprises a magnetosome island, which undergoes frequent rearrange-ments during stationary growth. J. Bacteriol. 187:7176–84

87. Wahyudi AT, Takeyama H, Matsunaga T. 2001. Isolation of Magnetospirillum magneticum AMB-1 mutantsdefective in bacterial magnetic particle synthesis by transposon mutagenesis. Appl. Biochem. Biotechnol.91–93:147–54

88. Wahyudi AT, Takeyama H, Okamura Y, Fukuda Y, Matsunaga T. 2003. Characterization of aldehydeferredoxin oxidoreductase gene defective mutant in Magnetospirillum magneticum AMB-1. Biochem. Biophys.Res. Commun. 303:223–29

89. Wenter R, Wanner G, Schuler D, Overmann J. 2009. Ultrastructure, tactic behaviour and potential forsulfate reduction of a novel multicellular magnetotactic prokaryote from North Sea sediments. Environ.Microbiol. 11:1493–505

90. Williams TJ, Zhang CL, Scott JH, Bazylinski DA. 2006. Evidence for autotrophy via the reverse tricar-boxylic acid cycle in the marine magnetotactic coccus strain MC-1. Appl. Environ. Microbiol. 72:1322–29

91. Yijun H, Weijia Z, Wei J, Chengbo R, Ying L. 2007. Disruption of a fur-like gene inhibits magnetosomeformation in Magnetospirillum gryphiswaldense MSR-1. Biochemistry 72:1247–53

92. Yoshino T, Matsunaga T. 2006. Efficient and stable display of functional proteins on bacterial magneticparticles using mms13 as a novel anchor molecule. Appl. Environ. Microbiol. 72:465–71


Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.

AR387-FM ARI 5 August 2009 13:23

Annual Review ofMicrobiology

Volume 63, 2009Contents

FrontispieceLars G. Ljungdahl � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �xii

A Life with Acetogens, Thermophiles, and Cellulolytic AnaerobesLars G. Ljungdahl � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 1

Regulation of Translation Initiation by RNA Binding ProteinsPaul Babitzke, Carol S. Baker, and Tony Romeo � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �27

Chemotaxis-Like Regulatory Systems: Unique Rolesin Diverse BacteriaJohn R. Kirby � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �45

Aminoacyl-tRNA Synthesis and Translational Quality ControlJiqiang Ling, Noah Reynolds, and Michael Ibba � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �61

Resurrected Pandemic Influenza VirusesTerrence M. Tumpey and Jessica A. Belser � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �79

Interspecies Chemical Communication in Bacterial DevelopmentPaul D. Straight and Roberto Kolter � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �99

Lipid Signaling in Pathogenic FungiRyan Rhome and Maurizio Del Poeta � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 119

Biological Insights from Structures of Two-Component ProteinsRong Gao and Ann M. Stock � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 133

Role of GTPases in Bacterial Ribosome AssemblyRobert A. Britton � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 155

Gene Transfer and Diversification of Microbial EukaryotesJan O. Andersson � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 177

Malaria Parasite Development in the Mosquito and Infectionof the Mammalian HostAhmed S.I. Aly, Ashley M. Vaughan, and Stefan H.I. Kappe � � � � � � � � � � � � � � � � � � � � � � � � � � � � 195

How Sweet it is! Cell Wall Biogenesis and Polysaccharide CapsuleFormation in Cryptococcus neoformansTamara Lea Doering � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 223

v

Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


Mitochondrial Evolution and Functions in Malaria ParasitesAkhil B. Vaidya and Michael W. Mather � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 249

Probiotic and Gut Lactobacilli and Bifidobacteria: MolecularApproaches to Study Diversity and ActivityMichiel Kleerebezem and Elaine E. Vaughan � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 269

Global Emergence of Batrachochytrium dendrobatidis and AmphibianChytridiomycosis in Space, Time, and HostMatthew C. Fisher, Trenton W.J. Garner, and Susan F. Walker � � � � � � � � � � � � � � � � � � � � � � � � � 291

Anaerobic Oxidation of Methane: Progress with an Unknown ProcessKatrin Knittel and Antje Boetius � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 311

The Trypanosoma brucei Flagellum: Moving Parasitesin New DirectionsKatherine S. Ralston, Zakayi P. Kabututu, Jason H. Melehani,Michael Oberholzer, and Kent L. Hill � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 335

Plants, Mycorrhizal Fungi, and Bacteria: A Network of InteractionsPaola Bonfante and Iulia-Andra Anca � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 363

Evolutionary Roles of Upstream Open Reading Frames in MediatingGene Regulation in FungiHeather M. Hood, Daniel E. Neafsey, James Galagan, and Matthew S. Sachs � � � � � � � � � 385

Single-Cell Ecophysiology of Microbes as Revealed by RamanMicrospectroscopy or Secondary Ion Mass Spectrometry ImagingMichael Wagner � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 411

Microbiology of the Atmosphere-Rock Interface: How BiologicalInteractions and Physical Stresses Modulate a SophisticatedMicrobial EcosystemAnna A. Gorbushina and William J. Broughton � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 431

What Sets Bacillus anthracis Apart from Other Bacillus Species?Anne-Brit Kolst, Nicolas J. Tourasse, and Ole Andreas Økstad � � � � � � � � � � � � � � � � � � � � � � � � � � � 451

The Expanding World of Methylotrophic MetabolismLudmila Chistoserdova, Marina G. Kalyuzhnaya, and Mary E. Lidstrom � � � � � � � � � � � � � � 477

Genomics, Genetics, and Cell Biology of Magnetosome FormationChristian Jogler and Dirk Schüler � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 501

Predatory Lifestyle of Bdellovibrio bacteriovorusRenee Elizabeth Sockett � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 523

Plant-Growth-Promoting RhizobacteriaBen Lugtenberg and Faina Kamilova � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 541

vi Contents

Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.


Photorhabdus and a Host of HostsNick R. Waterfield, Todd Ciche, and David Clarke � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 557

Management of Oxidative Stress in BacillusPeter Zuber � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 575

Sociobiology of the MyxobacteriaGregory J. Velicer and Michiel Vos � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 599

Index

Cumulative Index of Contributing Authors, Volumes 59–63 � � � � � � � � � � � � � � � � � � � � � � � � � � � 625

Errata

An online log of corrections to Annual Review of Microbiology articles may be found athttp://micro.annualreviews.org/

Contents vii

Ann

u. R

ev. M

icro

biol

. 200

9.63

:501

-521

. Dow

nloa

ded

from

ww

w.a

nnua

lrev

iew

s.or

gby

WIB

6106

- U

NIV

ER

SIT

AE

TSB

IBL

IOT

HE

K M

UE

NC

HE

N o

n 09

/30/

10. F

or p

erso

nal u

se o

nly.

Documents

Genomics, Genetics, and Cell Biology of Magnetosome Formation · • Genomics, Genetics, and Cell Biology of Magnetosome Formation 503 Annu. Rev. Microbiol. 2009.63:501-521. Downloaded