om bio genesis

8/11/2019 Om Bio Genesis

http://slidepdf.com/reader/full/om-bio-genesis 1/27

Biogenesis of theGram-Negative BacterialOuter Membrane

Martine P. Bos, Viviane Robert,and Jan Tommassen

Department of Molecular Microbiology and Institute of Biomembranes,Utrecht University, 3584 CH Utrecht, The Netherlands; email: [email protected],

[email protected], [email protected]

Annu. Rev. Microbiol. 2007. 61:191–214

First published online as a Review in Advance on June 7, 2007

The Annual Review of Microbiology is online at micro.annualreviews.org

This article’s doi:10.1146/annurev.micro.61.080706.093245

Copyright c 2007 by Annual Reviews. All rights reserved

0066-4227/07/1013-0191$20.00

Key Words

lipopolysaccharide, outer membrane proteins, lipoproteins,phospholipids

Abstract

The cell envelope of gram-negative bacteria consists of two membranes, the inner and the outer membrane, that are separated by

the periplasm. The outer membrane consists of phospholipidslipopolysaccharides, integral membrane proteins, and lipoproteins

These components are synthesized in the cytoplasm or at the inne

leaflet of the inner membrane and have to be transported across theinner membrane and through the periplasm to assemble eventually

in the correct membrane. Recent studies in Neisseria meningitidis and Escherichia coli have led to the identification of several machineries

implicated in these transport and assembly processes.

191



IM: innermembrane

OM: outermembrane

Periplasm:compartment between the innerand outer

membranesPeptidoglycan:bacterial cell wallconsisting of sugarpolymers covalently connected viaoligopeptides

Lipopolysaccharide(LPS): glycolipidconstituting theouter layer of thebacterial outermembrane

β-barrel: commonstructure of outermembrane proteins,consisting of amphipathicantiparallelβ-strands

Contents

INTRODUCTION... . . . . . . . . . . . . . . 192INTEGRAL OUTER MEMBRANE

PROTEINS . . . . . . . . . . . . . . . . . . . . . 193Passage Across the IM and

Through the Periplasm . . . . . . . . 193

Assembly into the OM . . . . . . . . . . . . 194Role of LPS in OMP Biogenesis . . 199 Model for OMP Biogenesis . . . . . . . 200

LIPOPROTEINS . . . . . . . . . . . . . . . . . . . 200LIPOPOLYSACCHARIDE . . . . . . . . . 202

Structure and Biosynthesis . . . . . . . . 202

Transport to the Cell Surface . . . . . 202 Models for LPS Transport . . . . . . . . 205

PHOSPHOLIPIDS . . . . . . . . . . . . . . . . . 206PERS PECT IV ES . . . . . . . . . . . . . . . . . . . 2 0 6

INTRODUCTION

The cell envelope of gram-negative bacteriaconsists of two membranes, the inner mem-

brane (IM) and the outer membrane (OM),that are separated by the periplasm containing

the peptidoglycan layer. The two membraneshave an entirely different structure and com-

position. Whereas the IM is a phospholipid

bilayer, the OM is an asymmetrical bilayer,consisting of phospholipids and lipopolysac-charides (LPS) in the inner and outer leaflet,

respectively. Additionally, these membranes

differ with respect to the structure of theintegral membrane proteins. Whereas inte-

gral IM proteins typically span the membranein the form of hydrophobic α -helices, inte-

gral OM proteins (OMPs) generally consist of antiparallelamphipathicβ-strandsthatfold

into cylindrical β-barrels with a hydrophilicinterior and hydrophobic residues pointing

outward to face the membrane lipids (49).Both membranes also contain lipoproteins,

which are anchored to the membranes viaan N-terminal N -acyl-diacylglycerylcysteine,

with the protein moiety usually facing theperiplasm in the case of Escherichia coli . In

other gram-negative bacteria, however, the

protein moiety of OM lipoproteins may alsoextend into the extracellular medium; ex-

amples are the LbpB and TbpB compo-nents of the lactoferrin and transferrin re-

ceptor, respectively, of Neisseria meningitidis

(70).

The OM functions as a selective barrierthat protects the bacteria from harmful com-pounds, such as antibiotics, in the environ-

ment. Unlike the IM, the OM is not energizedby a proton gradient and ATP is not avail-

able in the periplasm. In the absence of read-ily available energy sources, nutrients usually

pass the OM by passive diffusion via an abun-dant class of trimeric OMPs called porins(66)

Porins form water-filled channels that allowthe passage of small hydrophilic solutes with

molecular weights up to ∼

600 Da. Never-theless, energy-requiring transport processes

in the OM have also been described. Suchprocesses are dependent on complex energy-

coupling systems, such as the TonB system

which couples the proton-motive force of theIM to receptor-mediated uptake processes in

the OM (110). Whereas the composition, structure, and

function of the OM have been known fordecades, its assembly in the absence of en-

ergy sources has remained largely enigmatic All the components of the OM are syn-

thesized in the cytoplasm or at the cyto-plasmic face of the IM, and they have to

be transported across the IM and throughthe periplasm to reach their destination and

to assemble into the OM. Recently, many

new components involved in these processeshave been described, and in this review we

focus on these recent developments. Apartfrom studies in the classical model organisms

E. coli and Salmonella enterica, much progressin this field was reached by studying N. menin-

gitidis , and a comparison between these sys-tems reveals important differences in these

fundamental processes. Hence, what is truefor E. coli is not necessarily true for other

bacteria.

192 Bos · Robert · Tommassen



INTEGRAL OUTER MEMBRANEPROTEINS

Passage Across the IM and Through the Periplasm

Integral OMPs are synthesized in the cyto-

plasm as precursors with an N-terminal signal

sequence, which is required for translocationacross the IM. Two translocation machinerieshave been identified: the Sec system for the

translocation of unfolded proteins (23) and

the Tat system, which transports proteinsfolded in the cytoplasm (57). However, to

our knowledge, all OMPs studied to date aretransported via the Sec system, indicating that

they reach the periplasm in an unfolded state. After transport across the IM, the nascent

OMPs are accessible to periplasmic chaper-

ones (30). These chaperones have been stud-ied extensively, mainly in E. coli . One of these chaperones, Skp (seventeen-kilodalton-

protein), immediately interacts with theOMPs as soon as they emerge from the

Sec channel (37). The crystal structure of

this homotrimeric protein resembles a three-pronged grasping forceps that could bind a

nonnative OMP between the prongs and pro-tect it and prevent it from aggregating dur-

ing passage through the periplasm (52, 109).

An skp mutant of E. coli is viable but containsdecreased amounts of OMPs (16). Anotherperiplasmic chaperone, SurA, was initially

identified as a protein required for survivalduring the stationary phase (102). Mutants in

surA display an OMP assembly defect (56),and specifically the conversion of unfolded

monomers of OMPs into folded monomers

appears affected in such mutants (77). Theprotein has peptidyl-prolyl cis / trans isomerase

(PPIase) activity (56, 62), which nevertheless

appears to be dispensable for the chaperonefunction (5). Unlike most cytoplasmic chaper-ones, SurA is selective andpreferentially binds

nonnative OMPs over other proteins (5). By screening peptide libraries, Hennecke et al.

(38) demonstrated that SurA binds peptidesrich in aromatic residues and preferentially

those containing Ar-Ar or Ar-X-Ar motifs

Lipoprotein:protein attached tothe bacterial inner orouter membrane viaan N-terminal lipidmoiety

Porin: protein that forms water-filledchannels in the outermembrane

Sec system: generaprotein export apparatus

Chaperone: proteinthat guides anothermolecule to itsdestination, beingthe folded state orthe right cellularcompartment

(where Ar is an aromatic residue and X is any

residue). Transmembraneβ-strands of OMPsare particularly enriched with such motifs.

Like skp mutants, surA mutants are viable, but the combination of an skp and a surA muta-

tion results in a synthetically lethal phenotype(75). Therefore, it was suggested that Skp and

SurA are functionally redundant and that they operate in parallel pathways for chaperone ac-tivity (75). However, this is not the only pos-

sible explanation for the synthetic phenotype. We favor the possibility that Skp and SurA act

sequentially in the same pathway (Figure 1).Skp may act as a holding chaperone, prevent-

ing aggregation of nonnative OMPs in theperiplasm, andSurA mayact as a folding chap-

erone. The demand for the holding chaper-one may be limited when the subsequent fold-

ing and assembly steps are efficient. Underthose conditions, the OMPs have only little

chance to aggregate and an skp null mutation

Periplasm

OM

IMSec

Omp85

Skp

SurA

C

PorinOMP

YfgL YfiO NlpB

C

Figure 1

Model for OMP biogenesis. As soon as an OMP emerges from the Sec

translocon, it is bound by the chaperone Skp, which may prevent aggregation in the periplasm. The C-terminal signature sequence of theOMP functions as a targeting signal and binds the periplasmic domain of Omp85. Other chaperones and folding catalysts, such as SurA, may act onthe OMP. After folding, the OMP inserts into the lipid bilayer, possibly inbetween the Omp85 subunits. Oligomerization of certain OMPs, such asporins, may occur after insertion. The exact function of the accessory lipoproteins YfgL, YfiO, and NlpB is not known. Another accessory lipoprotein recently identified, SmpA (84), is not indicated here.

www.annualreviews.org • Outer Membrane Biogenesis 193



PPIase:peptidyl-prolylcis / trans isomerase

σE: alternativesigma factor that

guides RNA polymerase to thepromoters of genesinvolved in relievingperiplasmic stressconditions

has only a limited effect. The demand for such

achaperonewillincreasewhenthesubsequent folding and assembly processes are hampered,

e.g., in a surA mutant. Hence, it is not possibleto construct the skp surA double mutant.

Apart from SurA, the periplasm containsthree other PPIases, PpiA (also known as

RotA), PpiD, and FkpA. PpiA is by far themost active of them, because its inactivationleads to barely detectable residual PPIase ac-

tivity (47). However, a ppiA null mutation hadnodetectableeffectonOMPassembly(47).In

contrast, a ppiD null mutation led to an overallreduction in the level and folding of OMPs,

and a combination of ppiD and surA mutations was lethal (19), suggesting functional redun-

dancy. However, PpiD is anchored in the IM, whereas OMP folding is presumably initiated

after targeting to the OM,with which, indeed,a proportion of the SurA molecules cofrac-

tionated (38). Later, it was reported that ppiD

mutants, like ppiA and fkpA mutants and evena ppiD ppiA fkpA triple mutant, did not show

different overall OMP patterns or OM per-meability compared with the wild type and

that a ppiD surA double mutant could be con-structed, which had the same phenotype as a

surA single mutant (44). Overall, there is littleevidence for a direct role of PpiA, PpiD, or

FkpA in OMP biogenesis.Other proteins that may play a role at

the periplasmic stage of OMPs are DsbA andDsbC, two enzymes involved in the forma-

tion and isomerization, respectively, of disul-

fide bonds. DsbA catalyzed the formation of a disulfide bond between two cysteines engi-

neered in OM porin PhoE at positions not accessible from the periplasm once the porin

is inserted into the OM. This observationshowed that the disulfide bond was formed

duringperiplasmictransitandthatatleastpar-tial folding occurs prior to OM insertion (30).

DegP is a protease that degrades unfolded ormisfolded proteins in the periplasm but also

has chaperone activity, as was shown in vitroon the soluble substrates MalS andcitrate syn-

thase (87). A combination of surA and degP

mutations was synthetically lethal, suggesting

a role for DegP in OMP assembly (75). How-

ever, the role of DegP as a protease, remov-ing misfolded or unfolded OMPs from the

periplasm, may be more important than itsfunction as a chaperone in this respect (13).

Assembly into the OM

The insertion of OMPs into the OM has longremained enigmatic and has been considered

a spontaneous process (98). However, recent work has identified a proteinaceous machin-

ery that is essential for this process.

Omp85, an essential component of the

OMP assembly machinery. Recently, we

(105) identified a first component requiredfor OMP insertion, i.e., a protein designated

Omp85 in N. meningitidis . Omp85 was foundto be essential for the viability of the bacteriaand homologues of the omp85 gene were

found in all gram-negative bacteria for whichthe genome sequence was available (106)

suggesting its involvement in an importantprocess. Moreover, in many of these genomes

including those of N. meningitidis and E. coli

the omp85 gene is flanked by the skp gene

which encodes the periplasmic chaperone involved in OMP biogenesis, and rseP (formally

designated yaeL), which encodes a proteaseinvolved in the σE-dependent stress response

of E. coli that is induced upon accumulation of

unfolded OMPs in the periplasm (see below) All these features are consistent with a vital

role of Omp85 in OMP assembly. Upon de-pletion of Omp85 in a genetically engineered

strain, unfolded forms of all OMPs examinedincluding porins, a siderophore receptor, an

enzyme, and a secretin involved in type IVpili assembly, accumulated as aggregates in

the periplasm (105, 106). The role of Omp85in OMP assembly was confirmed in E. coli

in which the corresponding gene, designated

yaeT , was also an essential gene. Either deple-

tion of Omp85 or growth of a temperature-

sensitive mutant at the restrictive temperatureresulted in severe OMP assembly defects

(27, 112, 114). Even more interestingly, a




homolog of Omp85 was essential for the

assembly of β-barrel proteins into the outermembrane of mitochondria, a eukaryotic cell

organelle of endosymbiont origin (33, 54,69). Apparently, the OMP assembly pathway

is highly conserved during evolution. A difference was observed with respect to

the fate of OMPs in Omp85-depleted cellsof either N. meningitidis or E. coli . Whereasunfolded OMPs accumulated as aggregates

in such cells of N. meningitidis (105), Omp85depletion in E. coli primarily led to severely

reduced amounts of detectable OMPs (112,114). The difference is presumably caused by

the absence of the σE-dependent periplasmic

stress response in N. meningitidis . In E. coli ,this stress response is induced upon accumu-

lation of unfolded OMPs in the periplasm (fora review, see Reference 79). Such OMPs are

sensed by the PDZ domain of the membrane-bound protease DegS, upon which a prote-

olytic cascade is initiated that involves theprotease domain of DegS and RseP, result-

ing in the cleavage of the antisigma factor

RseA and the release of σE in the cytoplasm(Figure 2). The released σE then binds the

RNA polymerase core enzyme, resulting inthe transcription of the genes for periplasmic

D e g S

P D Z

P D Z

R s e A

R s e

B

R s e P

D e g S

R s e A

R s e

B

R s e P

Periplasm

Periplasmic

stress response

Stress

OM

IM

C

RNA

polymerase

sRNAsskp

surA

degP

σE

σE

P r o t e a s

e

P r o t e a s e

OMPs

OMP

Figure 2

Periplasmic stressresponse. Uponaccumulation of unfolded OMPs inthe periplasm, astress response isinitiated that starts with the recognitionof the C termini of the OMPs by the

PDZ domain of theprotease DegS. Thisrecognition initiatesa proteolytic cascaderesulting in therelease of thealternative sigmafactor σE, whichbinds RNA polymerase and leadsto enhancedtranscription of genes encoding

periplasmicchaperones, such asSkp and SurA, andthe protease DegP. The σE response alsoleads to productionof sRNAs that negatively regulateOMP expression.




sRNA: smallregulatory RNA

chaperones, such as Skp and SurA, and for

the potent protease DegP. Together thesechaperones and protease relieve the periplas-

mic stress. Moreover, the σE response resultsin the production of small regulatory RNAs

(sRNAs) in Enterobacteriaceae that prevent

OMP expression at the translational level (31,

43, 68). Thus, in E. coli , misfolded OMPsare rescued or degraded in the periplasm andOMP synthesis is inhibited during the stress

period. Although chaperones such as Skp andSurA are present in N. meningitidis , many es-

sential components of the signal transductionpathway are lacking. Whereas E. coli contains

three genes for the related proteases DegP,DegS, and DegQ (107), only a homologue of

DegQ can be found in N. meningitidis (en-coded by the NMB0532 locus in the N. menin-

gitidis strain MC58). Moreover, homologuesof RseA and RseB appear to be absent. Al-

though there is an alternativeσ factor belong-

ing to the ECF (extracytoplasmic factor) fam-ily (to which σE also belongs) encoded on the

meningococcal chromosome (i.e., NMB2144in strain MC58), it seems to have an entirely

different function. Its inactivation in the re-lated bacterium Neisseria gonorrhoeae did not

affect global gene expression as analyzed by microarray analysis, whereas its overexpres-

sion affected only very few genes, includingthose for methionine sulfoxide reductase (35).

Thus, theσE-dependent periplasmicstressre-

sponse appears to be absent in the pathogenic Neisseriaceae. As a result, when there is an

OMP assembly defect, OMPs will continue tobe synthesized and they will accumulate in the

periplasm, where they will form aggregates.

Omp85 is part of a multisubunit complex.

Sodium dodecyl sulfate polyacrylamide gel

electrophoresis (PAGE) analysis under non-denaturing conditions revealed that Omp85

of N. meningitidis is present in a high-molecular-weight complex (105). The only

other protein identified in this complex was

the RmpM protein ( J. Geurtsen, R. Voulhoux& J. Tommassen, unpublished results), which

has a peptidoglycan-binding motif and largely

resides in the periplasm but is firmly associ-

ated with many different integral OMPs viaan N-terminal peptide (71). RmpM proba-

bly anchors the OM to the underlying pep-tidoglycan layer (88), and it has no particular

function in OMP biogenesis. Cross-linkingexperiments revealed that the Omp85 ho-

molog of E. coli forms a complex with threelipoproteins, YfgL, YfiO, and NlpB (114). Asimilar complex was identified upon a pro-

teomic analysis of OMP complexes resolvedby blue-native PAGE (93). Copurification ex-

periments using strains with mutations in thegenes for various components of the complex

indicated that YfgL and YfiO directly interact with Omp85, whereas NlpB is associated with

the complex via YfiO (58). The yfgL gene was originally identified

in a screen for suppressors that restored theOM permeability defect of a partial loss-of-

function mutation in the imp gene (29, 78)

which encodes an OMP involved in LPSbiogenesis (see below). Although yfgL is not

an essential gene, null mutations create anOM permeability defect and result in reduced

amounts of OMPs (67, 78, 114), consistent with a role in OMP assembly. Remarkably, al-

though the yfgL gene is widely disseminatedamong gram-negative bacteria, we could not

identify a homolog in the sequenced genomesof N. meningitidis and N. gonorrhoeae.

The yfiO gene is essential in E. coli (67)but a mutant with a transposon insertion near

the 3 end of the gene was viable (114). This

mutant showed increased OM permeabilityand reduced amounts of OMPs, and a yfiO

depletion strain showed a phenotype similarto that of the yaeT / omp85 depletion strain of

E. coli (58). In N. gonorrhoeae, a transposoninsertion in the middle of the yfiO homolog

was described (32). This mutant was viableshowed a reduced cell size, and was transfor-

mation deficient; therefore, the gene was des-ignated comL. However, attempts to introduce

other comL truncations into the chromosome

failed (32), and we have not yet succeededto generate a complete comL deletion in the

chromosome of N. meningitidis (E. Volokhina




M.P. Bos & J. Tommassen, unpublished re-

sults). Hence, also in the Neisseriaceae, YfiO(ComL) appears essential for viability and the

N-terminal half of the protein appears suffi-cient for partial functionality. TheComL pro-

tein is covalently linked to the peptidoglycan,both in N. gonorrhoeae and when expressed

in E. coli (32). Furthermore, yfgL mutationsaffected peptidoglycan synthesis, possibly by regulating the activity of lytic transglycosy-

lases (29). Thus, both YfgL and YfiO might have a role in modulating the peptidoglycan

to facilitate the passage of OMPs through thislayer.

Whereas yfiO, like yfgL and omp85 , is widely disseminated among gram-negative

bacteria, this is less so for nlpB. Null mutantsin nlpB of E. coli are viable; they show only

moderate OM permeability defects and noobvious defects in OMP assembly (67, 114).

However, because an nlpB surA double knock-

out mutant has a synthetic lethal phenotype(67), it is clear that NlpB also has a direct role

in OMP assembly, possibly redundant to that of the periplasmic chaperone SurA.

Structure of Omp85. On the basis of

the sequence, we (105) have proposedthat Omp85 consists of two domains, a

C-terminal β

-barrel embedded in the OMand an N-terminal domain extending into

the periplasm. Support for this model wasobtained in proteolytic digestion experi-

ments, which resulted in the degradation of

the N-terminal domain and left the predictedmembrane-embeddedβ-barrel domain intact

(76, 91). The periplasmic extension containsrepeats of a conserved domain, POTRA

(polypeptide transport associated), suggestedto have chaperone-like qualities (81). Such

POTRA domains were further identified inother members of the Omp85 superfamily,

which includes the Omp85 homologs of mi-tochondria, the Toc75 OM component of the

chloroplast protein import machinery, andthe OM-localized TpsB component of the

two-partner secretion (TPS) systems of gram-

negative bacteria, and in members of the

POTRA:polypeptidetransport associated

FtsQ/DivIB family of IM proteins involved in

cell division. The bacterial Omp85 homologsare unique in having five of these POTRA do-

mains, whereas the number of such domainsin the other proteins with known functions

is restricted to one or two. Another memberof the Omp85 superfamily that is widely

disseminated among gram-negative bacteria(encoded by the ytfM gene in E. coli and theNMB2134 locus in N. meningitidis strain

MC58) contains three POTRA domains. The function of this protein is unknown (92).

To study its structure and function inmore detail, we produced the Omp85 protein

of E. coli in inclusion bodies, which wereisolated, and refolded the protein into the

native conformation in vitro (76). The re-folded protein formed oligomers, presumably

tetramers, similarly as reported for anothermember of the Omp85 superfamily, i.e.,

HMW1B, the TpsB component of a TPS

system of Haemophilus influenzae (95). Of note, the interactions between the subunits in

this homo-oligomeric complex are not stable;blue-native PAGE indicated equilibrium

between monomeric and oligomeric forms(76). Consistently, the hetero-oligomeric

Omp85/YfiO/YfgL/NlpB complex identifiedin vivo appeared to consist of one copy of

each subunit (93).

Interaction of Omp85 with its substrate

OMPs. When reconstituted in planar lipid

bilayers, both in vitro refolded Omp85 and

Omp85 extracted from E. coli cell envelopesformed narrow ion-conductive channels (76,

91). This property was used to study the in-teraction between Omp85 and its substrate

OMPs, emanating from the idea that such aninteraction would affect the channel activity.

Indeed, nonnative OMPs drastically enhanceOmp85 channel activities (76), showing that

these substratesinteract directly withOmp85.Furthermore, using mutant OMPs and syn-

thetic oligopeptides in this assay, we demon-strated that Omp85 specifically recognizes a

C-terminal motif in its substrates(76) that was

shown to be required for efficient assembly




of these proteins into the OM in vivo (94).

This C-terminal signature motif, which con-sists of a phenylalanine (or tryptophan) at the

C-terminal position, a tyrosine or hydropho-bic residue at position 3 from the C terminus,

and also hydrophobic residues at positions 5,7, and 9 from the C terminus, is present in

most bacterial OMPs, including porins, re-ceptors, enzymes, and autotransporters. It isinteresting to note that the same signature

sequence is recognized by the PDZ domainof DegS when unfolded OMPs accumulate

in the periplasm (108), thereby initiating the

σE-dependentperiplasmicstressresponse (see

above).In the planar lipid bilayer experiments an

OMP of N. meningitidis , in contrast to E. coli

OMPs, did not enhance the activity of the E.

coli Omp85 channels (76). Consistently, high-level expression of neisserial OMPs in E. coli

is often lethal and leads to their misassembly,

suggesting that the E. coli OMP assembly ma-chinery cannot deal efficiently with neisserial

proteins. Indeed, expression of E. coli omp85

cannot complement an omp85 mutation in N.

meningitidis and vice versa (E. Volokhina, V.Robert, M.P. Bos & J. Tommassen, unpub-

lished observations). Although the C terminiof neisserial OMPs do contain the signature

sequence with the features outlined above,they are further characterized by the presence

of a positively charged residue at the penulti-

mate position, which could inhibit the inter-action with E. coli Omp85. Indeed substitu-

tionof this positively charged residue in the N.

meningitidis porin PorA drastically improved

PorA assembly into the E. coli OMinvivo(76). Thus, although the OMP assembly machin-

ery is highly conserved among gram-negativebacteria, species-specific adaptations seem to

have occurred during evolution.Of note, the C-terminal signature se-

quence of OMPs is not absolutely essentialfor assembly into the OM. Thus, whereas

the high-level expression of a mutant form of

the E. coli porin PhoE lacking the C-terminalphenylalanine was lethal and resulted in its

periplasmic aggregation (21, 94), its low-level

expression was tolerated and allowed for its

assembly into the OM (21). Similarly, severalstudies reporting the assembly of neisserial

OMPs into the E. coli OM in vivo (113) maybe explained by low expression levels. Pulse-

chase experiments in E. coli overexpressingPhoE revealed the existence of two assem-

bly pathways: approximately half of the PhoEmolecules assembled within the 30-s pulse

period into their native conformation in the

OM, whereas the other half of the moleculesfollowed much slower kinetics and allowed

for the detection of several assembly inter-mediates (42). In these assays, the mutant

PhoE lacking the C-terminal phenylalaninefollowedonlytheslowerkineticpathway.This

study underscores the role of the C-terminalsignature sequence as a targeting signal and

indicates that there must be alternative, lessefficient targeting signals in OMPs. Further-

more,kineticpartitioningbetweenOMincor-poration and aggregation of periplasmic in-

termediates may explain the observation that

OMPs with defective signature sequences arestill assembled into the OM at low expression

levels, when the kinetics of aggregation is lowand hence the time span for assembly into the

OM is elongated (76). What could be the nature of the al-

ternative targeting sequences in OMPs? Ofnote, the C-terminal signature sequence con-

tributes an amphipathic β-strand to the β-barrel in the folded protein. In the absence

of the C-terminal phenylalanine, Omp85may recognize, though less efficiently, in-

ternal β-strands, many of which also end

with an aromatic residue. In some classesof OMPs, the C-terminal signature motif

could not be discerned (94). Perhaps, inthese cases also, Omp85 recognizes an in-

ternal β-strand. One of the major OMPs of

E. coli , OmpA, consists of two domains, an

N-terminal OM-embeddedβ-barrel and a C-terminal periplasmicextension. The signature

motifwasfoundinthiscaseattheendofthe β-barrel domain (94) and its importance in OM

targeting wasdemonstrated in a deletion analysis (48). Although this example demonstrates




that the targeting motif can be located inter-

nally in the primary structure of an OMP, thisappears to be a rather exceptional case be-

cause the C-terminal extension of an OMPsequence with, for example, a His tag usually

results in severe assembly defects (91). Another class of OMPs that lack the C-

terminal signature motif is constituted by TolC and its homologs, which are involved intype I protein secretion and drugs extrusion.

In its native trimeric structure, TolC forms a

β-barrel in the OM, to which each monomer

contributes four β-strands, whereas the major portion of the protein extends as long

α -helices into the periplasm (53). TolC is de-pendent on Omp85 for its assembly (112);

hence, possibly also in this case, an inter-nal β-strand is recognized. In contrast, Wza,

an OMP involved in the export of capsu-lar polysaccharides, has an entirely different

structure; the membrane-embedded portion

of this octameric protein consists of a barrelof eight amphipathic α -helices, to which each

monomer contributes one helix (28). It is con-ceivable that this protein assembles entirely

independently of the Omp85 machinery. Thesecretins form a class of OMPs involved in di-

verse processes, including type II and type IIIprotein secretion, type IV pili biogenesis, and

the extrusion of filamentous phages (7). Thestructure of these large multimeric proteins is

not known at atomic resolution, but they ap-pear to be rather poor inβ-sheet content (14),

which could indicate that they also follow a

different assembly pathway. However, at least one secretin, PilQ of N. meningitidis , depends

on Omp85 for its assembly (105). Perhaps, thebinding of these proteins and other OMPs

that lack the C-terminal signature motif isindirect and requires accessory factors, such

as the lipoproteins YfiO, YfgL, and NlpB, orspecific chaperones, such as the pilotins in the

case of the secretins (3).

Role of LPS in OMP Biogenesis

Since the observation that the amounts of sev-

eral OMPs are severely decreased in deep-

rough mutants of E. coli and Salmonella ty-

phimurium, which contain truncated LPSmolecules (2, 51), a role of LPS in OMP

assembly has been suggested. Consistently,the OM porin PhoE of E. coli could be con-

verted in vitro into a folded monomeric formin the presence of LPS, detergent, and diva-

lent cations. This folded monomer appearedto be an assembly intermediate because it could subsequently be converted into its na-

tive trimeric OM-inserted form (22). Alsofor other OMPs, LPS-dependent folding in

vitro has been reported (83). However, LPS-independent in vitro folding conditions were

later established for many OMPs, includingPhoE (42), arguing against a specific role of

LPS in OMP folding. Moreover, an lpxA mu-tant of N. meningitidis , which is completely

defective in LPS biosynthesis, appeared to be viable (90), and all OMPs examined, including

porins, were correctly assembled in vivo into

the LPS-free OM of this mutant (89). Nev-ertheless, species-specific differences with re-

spect to the LPS dependency of OMP assem-bly cannot yet be excluded at this stage.

The crystal structure of E. coli Skp revealed a putative LPS-binding site (109), con-

sistent with the earlier description of the pro-teinin Salmonella minnesota as an LPS-binding

protein (34). Thus, LPS may exert its rolein OMP biogenesis via Skp. Furthermore, it

was demonstrated in protease-digestion as-says that binding of LPS and phospholipids

could inversely modulate the structure of Skp

in vitro (20). Thus, after binding nonnativeOMPs at the IM, a conformational change

triggered by LPS binding may release thecargo at the OM. Note that LPS is present

only in the outer leaflet of the OM. Hence,Skp should bind to LPS molecules that have

not yet reached their destination. Consis-tently, OMP biogenesis is heavily affected by

cerulinin, a drug that inhibits lipid synthe-sis (8). However, the putative LPS-binding

site in E. coli Skp is largely conserved in

N. meningitidis Skp, where OMP biogene-

sis is independent of LPS. Hence, the role

of this putative LPS-binding site remains to




Imp: increasedmembranepermeability

be determined in site-directed mutagenesis

experiments. We speculate that the role of LPS in OMP

biogenesis is restricted to late stages afterthe insertion into the OM, such as the sta-

bilization of porin trimers (55), which pre-sumably requires some LPS-mediated rear-

rangements in the cell surface-exposed loopsof theproteins. Theseverely reduced amountsof OMPs in deep-rough mutants may be ex-

plained by the induction of the σE responsein such mutants, resulting in the production

of sRNAs that inhibit the synthesis of many OMPs (see above). Changes in LPS structure

induce the σE response in E. coli by a hithertounknown mechanism (97). At least one OMP

of E. coli , TolC, appears totally unaffected by LPS structure (111). Presumably, the expres-

sion of TolC is unaffected by the sRNAs. Theassembly of this lipid-independent OMP is in-dependent of the accessory component YfgL

of the Omp85 assembly machinery (15). Of note, a homolog of the yfgL gene is lacking in

N. meningitidis , in which the assembly of allOMPs is independent of LPS. Hence, there

may be an additional role for LPS in the as-sembly of lipid-dependent OMPs in E. coli ,

and YfgL may play a role specifically in theassembly of this class of OMPs. In this re-

spect, it may indeed be relevant that the yfgLmutants were initially picked up as suppres-

sors of a partial loss-of-function imp mutant

(29), and the imp gene product is specifically involved in LPS biogenesis (9) (see below).

Model for OMP Biogenesis

We propose the following model for OMP

biogenesis (Figure 1). After their transport through the Sec translocon, nascent OMPs

are immediately bound by Skp, which may as-sisttheirreleasefromtheIMandpreventtheir

aggregation in the periplasm. The Skp/OMPcomplex is targeted to the Omp85 complex in

the OM, whereby the C-terminal signature

motif of the OMPs functions as the primary targeting signal that binds directly to Omp85,

presumably to its N-terminal POTRA do-

mains. Binding initiates folding, which results

in the release of Skp (20) and may be assistedby the presumed chaperone activities of the

POTRA domains and by periplasmic proteinssuch as SurA and DsbA. Binding of an OMP

also results in a conformational change in theC-terminal domain of Omp85, which is re-

flected by the increased pore activity observedin the planar lipid bilayer experiments (76) This conformational change allows the OMPs

to insert into the OM, possibly in between theOmp85 subunits. Dissociation of the Omp85

subunits releases the assembled OMPs intothe OM, where final conformation changes in

the cell surface-exposed loops may be inducedupon interaction with LPS.

The specific role of the accessory lipopro-teins YfgL, YfiO, and NlpB has not been ad-

dressed experimentally so far. Of note, no ho-mologs of these proteins have been implicated

in the assembly of β-barrel proteins into the

mitochondrial OM, suggesting that their roleis less crucial, although YfiO is definitely es-

sential, at least in E. coli (67). These proteinsmay play a role in the recognition of OMPs

that do not display the C-terminal signaturemotif, which includes the essential protein

Imp (increased membrane permeability), inmodulating the peptidoglycan layer, and/or

they may function as chaperones in the fold-ing of OMPs. YfgL may have a role in coor-

dinating the assembly of a subclass of OMPsthat require LPS for their assembly, but such a

role is difficult to assess in E. coli , in which as-

sembly defects of OMPs are associated witha feedback inhibition on their synthesis via

σE-induced sRNAs, whereas N. meningitidis

which does not display such a feedback inhibi-

tion, does not have a YfgL homolog. The ab-sence of this feedback inhibition mechanism

makes N. meningitidis a favorable organism tostudy the role of other assembly factors.

LIPOPROTEINS

Bacterial lipoproteins are membraneattached via an N-terminal N -acyl-

diacylglycerylcysteine. Lipidation and folding




of these proteins take place after their translo-

cation over the IM via the Sec machinery.Prior to cleavage of the signal sequence, a

diacylglycerol group is transferred by theenzyme Lgt from phosphatidylglycerol to

the sulfhydryl group of the cysteine that isinvariably present at the +1 position relative

to the processing site (82). Subsequently, thediacylglycerylprolipoprotein is processed by adedicated signal peptidase, signal peptidase II

(24), after which the free α -amino group of the cysteine is acylated by the enzyme Lnt

(36), yielding the mature lipoprotein.Lipoproteins are sorted to the IM or OM

according to a sorting signal that comprisesthe amino acids flanking the lipidated cysteine

in the mature protein (101). Lipoproteinslacking an IM retention signal, usually an

aspartate at the +

2 position of the matureprotein, are transported to the OM by the

Lol system (Figure 3). The first component

of the Lol system was identified in an in vitrosystem, whereby lipoproteins were synthe-

sized in a radioactive form in spheroplasts of E. coli (60). The mature forms of the newly

synthesized lipoproteins were retained at the surface of the spheroplasts, from which

they could be released upon addition of aperiplasmic extract. The active component

of the periplasmic extract was identified anddesignated LolA. Furthermore, a complex of

the three proteins LolC, LolD, and LolE, which constitutes an ATP-binding cassette

(ABC) transporter in the IM, is required

for lipoprotein transport (65). The integralmembrane components of this transporter,

LolC and LolE, show considerable sequencesimilarity and may function as a heterodimer.

Of note, N. meningitidis contains only one of these proteins (encoded by locus NMB1235

in strain MC58), which may form a homod-imer. The current model (Figure 3) states

that lipoproteins destined for the OM are first bound by LolC/E in an ATP-independent

manner. This binding results in an increase inthe affinity of LolD for ATP. Subsequently,

ATP binding to LolD causes a conformational

change in the LolC/E moiety that results in a

Periplasm

OM

IM

C - X C

- X

C -X

C-X

C

- X

C - X

C - X

C - D

C - D / X

Sec

Sorting

LolA

LolA

LolB

LolC

LolD

LolE

LolD

?

ATP ADP ATP ADP

Figure 3

Lipoprotein transport through the bacterial cell envelope. After theirtransport via the Sec system and subsequent lipidation at the N-terminalcysteine (C), lipoproteins bind to the ABC-transporter LolCDE, providedthey do not possess a Lol-avoidance motif, which usually is an aspartate(D) flanking the N-terminal C (C-D). Lipoproteins with another aminoacid at the +2 position (X) are destined for the OM. Energy from ATPhydrolysis by LolD is transferred to LolC and LolE and utilized to openthe hydrophobic LolA cavity to accommodate the lipoprotein. When theLolA-lipoprotein complex interacts with the OM receptor LolB, thelipoprotein is transferred to LolB and inserted into the OM. Furthertransport to the outer leaflet of the OM, if applicable, occurs throughunknown mechanisms.

Spheroplasts:bacterial cells of which the outermembrane and thepeptidoglycan layerhave been disruptedby treatment withEDTA and lysozyme

ABC: ATP-bindingcassette

decrease in lipoprotein binding affinity. ATPhydrolysis is then required for transfer of the

lipoprotein from LolC/E to the periplasmic

chaperone LolA (41). The LolA-lipoproteincomplex crosses the periplasm and interacts

with an OM receptor, LolB (61). The lipopro-tein is then transferred to LolB according

to the affinity difference between LolA andLolB. LolA and LolB are structurally similar;

both contain a hydrophobic cavity. TheLolA cavity is composed mostly of aromatic

residues, whereas the cavity in LolB is mademostly of leucine and isoleucine residues,

which is likely to explain the difference inlipoprotein-binding affinity (64, 96).

In E. coli , all known lipoproteins face the

periplasm, but in other bacteria, most no-tably in members of the spirochetes (12), cell

surface-exposed lipoproteins also are present.




Whether lipoprotein transport over the OM

occurs through an extension of the Lol sys-tem or by an unrelated transport system

is presently unclear. The only cell surface-exposed lipoprotein studied in this respect is

the starch-disbranching enzyme pullulanaseof Klebsiella oxytoca (72). Pullulanase contains

an aspartate at position +

2, indicating that it is not a substrate for the Lol system. In-deed the transport of pullulanase to the cell

surface requires a type II protein secretionapparatus. In the absence of such an appa-

ratus, pullulanase is retained in the IM (72). Thus, in K. oxytoca, sorting between periplas-

mically and cell surface-exposed lipoproteinstakes place at the IM level. However, this is

not necessarily always the case. N. meningitidis

contains several cell surface-exposed lipopro-

teins, including LbpB and TbpB, but the se-quenced genomes do not reveal the presence

of a type II secretion apparatus (104). Further-

more, LbpB and TbpB contain an isoleucineand a leucine, respectively, at the +2 posi-

tion, suggesting they are substrates for theLol system. Thus, these lipoproteins may be

transported to the cell surface via an exten-sion of the Lol system, which remains to be

identified.

LIPOPOLYSACCHARIDE

Structure and Biosynthesis

LPS is a complex glycolipid exclusively

presentintheouterleafletoftheOMofgram-negative bacteria. It consists of a hydrophobic

membrane anchor, lipid A, substituted withan oligosaccharide core region, which in some

bacteria (e.g., in most E. coli strains, but not inthe laboratory strain E. coli K-12 and also not

in N. meningitidis ) is extended with a repeatingoligosaccharide, the O-antigen. These differ-

entLPSconstituentsaresynthesized at thecy-toplasmic leaflet of the IM. The lipid A moi-

ety of LPS is rather well conserved among

gram-negative bacteria. It usually consists of a

β-1,6-linkedd-glucosamine disaccharide car-

rying ester- and amide-linked 3-hydroxy fatty

acids at the 2, 3, 2, and 3 positions and phos-

phate groups at the 1 and 4 positions. Theprimary 3-hydroxy fatty acids may be substi-

tuted with secondary fatty acids. The lipid Abiosynthetic pathway, also known as the Raetz

pathway (74), has been characterized in detailmostly in E. coli and Salmonella typhimurium

but it appears to be highly conserved amongother gram-negative bacteria (74). Remark-ably, although LPS has so far only been de-

tected in gram-negative bacteria, homologs ofthe genes encoding the lipid A biosynthetic

enzymes have also been found in sequencedplant genomes (74).

The core oligosaccharide is much more variable between bacterial species. In addi-

tion, a huge amount of LPS heterogeneity iscreated by numerous modifications of both

the lipid A part and the core part of LPS(73). The modifying enzymes, which are lo-

calized in different cellular compartments, are

not generally conserved; so different speciescan express uniquely modified types of LPS

(103). The O-antigen, if present, is the most variable part of LPS and shows even a high

degree of variability between different strainsof the same species.

Transport to the Cell Surface

In contrast to the understanding of its biosyn-

thesis, the mechanism of the transport of LPS

from its site of synthesis to its final destina-tion forms a much less complete picture (25).

It is clear that the lipid A–core moiety andthe O-antigen subunits, if present, are trans-

ported separately overthe IM. The O-antigensubunits are transferred over the IM by one

of three different routes: the Wzy-, ABC-transporter-, or synthase-dependent pathway

(74).After polymerization, the subsequentlig-ation of the O-antigen to the lipid A–core

moiety at the periplasmic side of the IM isan incompletely understood process that in-

volves at least the WaaL ligase (45).

The translocation of the lipid A–core moi-ety over the IM is mediated by an ABC fam-

ily transporter called MsbA, as inferred from




the accumulation of LPS in the IM of a

temperature-sensitive E. coli msbA mutant at the restrictive temperature (116). The LPS

accumulated was not modified by periplas-mic enzymes, demonstrating that it was not

transported to the periplasmic leaflet of theIM (26). More evidence for a role of MsbA

in LPS transport came from a study of anmsbA mutant of N. meningitidis . The viabil-ity of LPS-deficient N. meningitidis mutants

makes this organism well suited for the study of LPS biogenesis, because clean knockouts

of genes involved in this process can be con-structed. Indeed, MsbA is a nonessential pro-

tein in N. meningitidis (99). A neisserial msbA

mutant produced only low amounts of LPS, a

feature indicative of a defect in LPS transport in this species. In N. meningitidis , biosynthe-

sis and transport of LPS are coupled in such a way that synthesis is reduced under conditions

in which transport is halted (9, 99).

The subsequent steps in LPS transport to the exterior of the bacterium have long

remained obscure. However, an OM compo-nent required for the appearance of LPS at the

bacterial cell surface was identified recently. This component is an OMP known as Imp

or OstA (organic solvent tolerance), because

E. coli strains expressing mutant versions of

this protein showed altered membrane per-meability (1, 80). Imp is an essential proteinin

E. coli . In a conditional imp mutant, correctly folded OMPs accumulate in aberrant mem-

branes with an increased density, indicative

of an altered lipid/protein ratio (11). Theprecise role of Imp was demonstrated in

N. meningitidis . Imp was not essential in thisbacterial species, allowing the construction

of an imp deletion mutant. The phenotypeof this mutant demonstrated a role for Imp

in LPS biogenesis: It produced less than10% of wild-type levels of LPS, which were

not accessible to LPS-modifying enzymesrecombinantly expressed in the OM or added

to the extracellular medium (9). Therefore,Imp appears to function in the transport of

LPS over the OM to the cell surface. This

role of Imp was confirmed in an E. coli imp

depletion strain. Upon depletion of Imp,

newly synthesized LPS was not accessible toLPS-modifying enzymes in the OM, showing

that it did not reach the outer leaflet of theOM (115). In E. coli , another OM component

was identified to play a role in LPS transport,i.e., the essential lipoprotein RlpB. Depletion

of this lipoprotein resulted in a phenotypesimilar to that expressed upon depletion of

Imp. Moreover, Imp and RlpB exist in a com-

plex. Depletion of Imp or RlpB resulted inan increased total cellular LPS content (115),

indicating that in E. coli , in contrast to N.

meningitidis , defective transport does not lead

to feedback inhibition on LPS biosynthesis.Imp and RlpB homologs are widely dis-

seminated among gram-negative bacteria,suggesting that the mechanism of LPS trans-

port is highly conserved. Imp is predicted tocontain a β-barrel domain embedded in the

OM, with a long N-terminal domain and ashort C-terminal domain extending into the

periplasm (Figure 4). In the conserved do-

main database at the National Center forBiotechnology Information (59), the β-barrel

domain is recognized as a conserved domain,designated OstA-C, in all Imp homologs. An-

other conserved domain, COG1934, is recog-nized in theN-terminal periplasmic extension

of many but not all Imp homologs (Figure 4).Nevertheless, all Imp homologs show se-

quence similarity over the entire length, andalso when the COG1934 domain is not

recognized. Additional components putatively in-

volved in LPS translocation were identified

by Sperandeo et al. (86), who discovered sev-eral new essential genes in E. coli, some of

which appeared to play a role in cell enve-lope biogenesis. Depletion of recombinant

bacteria for the proteins encoded by two of these genes resulted in similar phenotypes as

described for Imp- and RlpB-depleted cells: The bacteria exhibited an altered OM density

and an increased cellular LPS content. Thegenes, which form an operon, were designated

lptA and lptB (LPS transport) (85). Unfortu-nately, we cannot use the same designations in




Imp

YrbK

102

#

195

319

214

1

1

1

LptA/

NMB0355

Acidobacteria bacterium

Ellin345

Nostoc sp. PCC712

Acinetobacter sp. ADP1

COG1934

COG1934

COG1934

C O G 1 9 3 4

COG1934 COG1934

COG3117

COG3117

COG3117

COG3117

OstA_C

OstA_C

OstA_C

Figure 4Conserved protein domains present in LPS transport components. The Conserved Domain Database at the National Center for Biotechnology Information was searched in December 2006 with the proteinsshown at the left. The type and number (#) of retrieved architectures are shown. The species expressingthe most unusual architectures are indicated at the right. A topology model for Imp, based on manualinspection for putative β-strands, is shown at the top right-hand corner.

N. meningitidis , for which the acronym lptA is

used for a gene encoding an LPS phospho-ethanolamine transferase (17).

Also the neisserial homolog of E. coli

LptA, encoded by the NMB0355 locus in

N. meningitidis strain MC58, plays a role inLPS transport: In these bacteria, the cor-

responding gene is not essential, and itsdeletion results in severely decreased lev-

els of LPS (10). The LPS is accessible toperiplasmic LPS-modifying enzymes in this

mutant, indicating that the LptA/NMB0355protein acts at a step after translocation by

MsbA (10). The LptA protein of E. coli wasfound in the soluble periplasmic fraction (85),

but we found the majority of the corre-

sponding neisserial protein (NMB0355) inthe membrane fraction, although it has no

obvious membrane-spanning segments (M.P.Bos & J. Tommassen, unpublished obser-

vations). LptA largely consists of the con-served domain COG1934, the same domain

found in the N-terminal periplasmic domain

of Imp (Figure 4), indicative of a commonfunction.

LptB is a 27-kDa protein present in a

140-kDa IM complex; unfortunately, no in-teracting partners were identified (93). The

protein possesses the typical features of an ABC protein but has no obvious membrane-

spanning segments. Thus, the current data suggest the involve-

mentofanovelABCtransporterinLPStrans-port. LptB is the ABC component of this

transporter, but the cognate integral mem-brane component remains to be identified

The protein encoded by the yrbK gene, whichis located immediately upstream of the lptAB

operon and which is also essential for viability

in E. coli (86), may be a part of the transporter The genetic organization of the yrbK-lptA-

lptB locus is highly conserved among gram-negative bacteria, and the observation that a

conserved protein domain present in YrbKis sometimes present in one polypeptide to-

gether with conserved domains from Imp or




LptA (Figure 4) is suggestive for a role of

YrbK in LPS transport. However, secondary structure predictions of YrbK show only one

putative transmembrane helix, making it un-likelythatthisproteinfunctionsastheintegral

membrane component of the ABC trans-porter, as such components of ABC trans-

porters usually contain multiple transmem-brane helices (6).

Models for LPS Transport

Two models for LPS transport through thecell envelope have been considered, and

with the current situation, no model can beconsidered definitively proven or discounted

(Figure 5). One possibility is that LPS passes

through the periplasm in a soluble complex

with a chaperone that shields its hydrophobicmoiety, similar to the Lol system for lipopro-

tein transport (Figure 3). Indeed, the recent identification of a novel ABC transporter in-

volved in LPS transport may suggest sim-ilarities with the lipoprotein transport sys-

tem. The LptA protein, which is a solubleperiplasmic protein in E. coli (86), may func-tion as the LPS chaperone, as LolA does for

lipoproteins. Furthermore, while LolA passesits cargo to the structurally related OM recep-

tor LolB, LptA may pass the LPS moleculesto theperiplasmic N-terminal domainof Imp,

which shows sequence similarity to LptA. The

β-barrel of Imp may form a channel for fur-

ther transport to the cell surface. Secondary

Periplasm

OM

Membrane contact sites Lol system-like

IM

Imp

MsbA MsbA

Imp

RlpB

LptA

LptA

LptB LptB

YrbK YrbK

YrbK

RlpB

? ?

ATP ADP

ATP ADP ATP ADP

ATP ADP

Figure 5

Models for LPS transport through the bacterial cell envelope. LPS is synthesized at the inner leaflet of the IM and transported over the IM by the ABC transporter MsbA. LPS then travels through theperiplasm by an unknown mechanism that could resemble the Lol system (right ). LptB together with anunknown (?) membrane protein and possibly YrbK may form an ABC transporter that delivers LPS tothe periplasmic chaperone LptA. LPS may then be transferred from LptA to the periplasmic domain of Imp, which may have a structure similar to LptA. Alternatively, LPS transport may take place at contact sites between the two membranes (left ). YrbK, LptA, LptB, and an unknown transmembrane domain may function in the formation of these contact sites. Finally, Imp and RlpB are required to transfer LPS to thecell surface, perhaps by acting as a flippase complex.




structure predictions for LptA show many

β-strands, possibly forming a soluble beta-barrel, resembling LolA (10, 96). The LptB

protein could be the functional LolD ho-molog of the LPS transport system. As ex-

plained above, the putative LolC/E compo-nents of such an LPS transport system remain

to be found. The other model postulates that LPS ac-tually never leaves its membranous environ-

ment and that it is transported at contact sites between the IM and OM (Figure 5).

The first indication for the existence of suchsites, known as zones of adhesion or Bayer

junctions, came from electron microscopy studies (4), although fixation procedures may

have affected the results (46). Later, M ¨ uhlradt et al. (63) reported that newly synthesized

LPS appears in patches in the OM, closeto the membrane contact sites. Membrane

fractionation studies showed the existence

of a minor fraction, designated OML, that contains IM, peptidoglycan, and OM. This

membrane fraction, which contained pepti-doglycan biosynthesis activity, may represent

membrane contact sites. Pulse-chase exper-iments combined with fractionation proce-

dures showed that newly synthesized LPStransiently passed through this fraction on its

way to the OM (40). Furthermore, when asimilar approach was used that led to the iden-

tification of LolA (see above), newly synthe-

sized LPS could not be released from sphero-plasts upon addition of periplasmic extracts.

Rather, LPS transport from IM to OM con-tinued in the spheroplasts,suggesting that this

process does not involve a soluble periplas-mic component and proceeds via contact sites

(100). In this model, the LptA, LptB, and YrbK components may have a role in the for-

mation of these contact sites (Figure 5).

PHOSPHOLIPIDS

The major OM phospholipids of E. coli are

phosphatidylethanolamine and phosphatidyl-glycerol. Phospholipids are synthesized at the

cytoplasmic side of the IM (18, 39). Then, in

order to reach the OM, they first need to ro-

tate (flip-flop) over the membrane. It is notclear whether a dedicated flippase is neces-

sary for this process. The LPS transporter MsbA was also implicated in phospholipid

transport because the conditional E. coli msbA

mutant accumulated both LPS and phospho-

lipids in the IM under restrictive conditions(116). However, an msbA mutant of N. menin-

gitidis appeared viable and still made a dou-

ble membrane, showing that at least in thisbacterium MsbA is not required for phospho-

lipid transport (99). Another distant msbA ho-molog in N. meningitidis , i.e., the NMB0264

locus in strain MC58, could be disrupted without causing any obvious phenotype (M.P

Bos, unpublished observations). Moreover various α -helical membrane-spanning pep-

tides, but curiously not MsbA, induced phos-pholipid translocation in synthetic lipid bi-

layers (50). Thus, flip-flop of phospholipids

may not require a specific transporter butmerely the presence of the typical α -helical

membrane-spanning segments of some IMproteins. The next steps in phospholipid bio-

genesis, i.e., transport through the periplasmand incorporation into the inner leaflet of the

OM, remain obscure. Unlike LPS transportthe transport of phospholipids was halted in

spheroplasts, and unlike lipoproteins, newlysynthesized phospholipids could not be re-

leased from the spheroplasts upon addition ofa periplasmic extract (100). Thus, the trans-

port mechanism appears different from those

of LPS and lipoproteins. Any components involved in phospholipid transport remain to be

identified.

PERSPECTIVES

In the past few years, much progress hasbeen made in the field of OM biogenesis

with the identification of many new compo-nents involved in the process. The field will

rapidly move forward, gaining mechanistic

insights to which structural analysis of thenewlyidentified componentsby X-raycrystal-

lography will make important contributions




The major players required for OMP assem-

bly have likely been identified. Provided that no energy-coupling system is required and

that protein folding and partitioning into thehydrophobic environment of the membrane

are the driving forces, it may be possible toset up an in vitro system for OMP assembly

with purified components. For LPS, a majorissue remains how it is transported throughthe periplasm. Studying the binding of LPS

to the components together with immuno-gold electron microscopy studies to deter-

mine whether these components are associ-ated with the contact sites between IM and

OM will help to address these questions. Forlipoproteins, an important issue is how such

molecules are transported to the cell surface.Research on the transport of phospholipids

to the OM has to start more or less from thebeginning.

Importantly, much progress in the fieldhas

been reached by studies in two model organ-isms, E. coli and N. meningitidis . These stud-

ies have revealed similarities as well as differ-

ences. For example, whereas an LPS transport defect in N. meningitidis results in feedback in-

hibition of its synthesis, this is not the case in

E. coli . For OMP assembly, the reverse is true:

An OMP assembly defect leads to feedback inhibition in E. coli , but not in N. meningitidis .

Such differences make it attractive to study specific aspects of OM biogenesis in different organisms. In addition, considering the dif-

ferences already observed between these twomodel organisms, it is likely that studies in

other bacteria will uncover new, unanticipatedfeatures.

Further studies in this field will remainimportant, because they will uncover funda-

mental biological processes. In addition, theknowledge gained from these studies may be

useful for medical applications: The essentialnature of the bacterial machineries involved

and their surface localization make them at-

tractive targets for the development of new antimicrobial drugs and vaccines.

SUMMARY POINTS

1. OMPs and lipoproteins are transported across the IM via the Sec system.

2. Assembly of bacterial outer membrane proteins requires the outer membrane proteinOmp85, which is evolutionary conserved and found even in the OM of mitochondria.

3. Omp85 recognizes its substrate OMPs by virtue of their C-terminal signaturesequences.

4. Other proteins involved in OMP transport and assembly are the periplasmic chap-erones Skp and SurA, and the OM-associated lipoproteins YfiO, YfgL, NlpB, and

SmpA, the function of which remains to be determined.

5. Transport of lipoproteins to the OM depends on the Lol system, which consists of

an ABC transporter in the IM, a soluble periplasmic chaperone, and an OM-attachedreceptor.

6. MsbA is an ABC transporter required for the transport of LPS across the IM.7. Further transport of LPS to the cell surface requires, at least, the ABC protein LptB,

the periplasmic protein LptA, the OM-attached lipoprotein RlpB, and the integralOMP Imp.

8. Nothing is known regarding the transport of phospholipids to the OM.




DISCLOSURE STATEMENT

The authors are not aware of any biases that might be perceived as affecting the objectivity ofthis review.

ACKNOWLEDGMENTS

The work in our laboratory is supported by grants from the Netherlands Research Council for

Earth and Life Sciences (ALW) and the Netherlands Research Council for Chemical Sciences(CW) with financial aid from the Netherlands Organization for Scientific Research (NWO).

LITERATURE CITED

1. Abe S, Okutsu T, Nakajima H, Kakuda N, Ohtsu I, et al. 2003. n-Hexane sensitivity

of Escherichia coli due to low expression of imp / ostA encoding an 87 kDa minor protein

associated with the outer membrane. Microbiology 149:1265–732. Ames GFL, Spudich EN, Nikaido H. 1974. Protein composition of the outer membrane

of Salmonella typhimurium: effect of lipopolysaccharide mutations. J. Bacteriol. 117:406–163. Bayan N, Guilvout I, Pugsley AP. 2006. Secretins take shape. Mol. Microbiol. 60:1–4

4. Bayer ME. 1968. Areas of adhesion between wall and membrane of Escherichia coli . J. Gen Microbiol. 53:395–404

5. Behrens S, Maier R, de Cock H, Schmid FX, Gross CA. 2001. The SurA periplasmicPPIase lacking its parvulin domains functions in vivo and has chaperone activity. EMBO

J. 20:285–946. Biemans-Oldehinkel E, Doeven MK, Poolman B. 2006. ABC transporter architecture

and regulatory roles of accessory domains. FEBS Lett. 580:1023–357. Bitter W. 2003. Secretins of Pseudomonas aeruginosa: large holes in the outer membrane

Arch. Microbiol. 179:307–148. Bolla JM, Lazdunski C, Pag es JM. 1988. The assembly of the major outer membrane

protein OmpF of Escherichia coli depends on lipid synthesis. EMBO J . 7:3595–99

9. First demonstration that an integral OMP,Imp, is required for the transport of LPS to the cellsurface.

9. Bos MP, Tefsen B, Geurtsen J, Tommassen J. 2004. Identification of an outermembrane protein required for lipopolysaccharide transport to the bacterial cell

surface. Proc. Natl. Acad. Sci. USA 101:9417–22

10. Bos MP, Tommassen J. 2006. Identification of LPS transport components in Neisseria

meningitidis . In Fifteenth Int. Pathogenic Neisseria Conf., ed. J Davies, M Jennings, pp. 33

West Leederville, Australia: Cambridge Publ.11. Braun M, Silhavy TJ. 2002. Imp/OstA is required for cell envelope biogenesis in Es-

cherichia coli. Mol. Microbiol. 45:1289–30212. Casjens S. 2000. Borrelia genomes in the year 2000. J. Mol. Microbiol. Biotechnol. 2:401–1013. CastilloKeller M, Misra R. 2003. Protease-deficient DegP suppresses lethal effects of a

mutant OmpC protein by its capture. J. Bacteriol. 185:148–54

14. Chami M, Guilvout I, Gregorini M, Remigy HW, M ¨ uller SA, et al. 2005. Structuralinsights into the secretin PulD and its trypsin-resistant core. J. Biol. Chem. 280:37732–4115. Charlson ES, Werner JN, Misra R. 2006. Differential effects of yfgL mutation on Es-

cherichia coli outer membrane proteins and lipopolysaccharide. J. Bacteriol. 188:7186–9416. Chen R, Henning U. 1996. A periplasmic protein (Skp) of Escherichia coli binds a class of

outer membrane proteins. Mol. Microbiol. 19:1287–9417. Cox AD, Wright JC, Li J, Hood DW, Moxon ER, Richards JC. 2003. Phosphorylation

of the lipid A region of meningococcal lipopolysaccharide: identification of a family of




transferases that add phosphoethanolamine to lipopolysaccharide. J. Bacteriol. 185:3270–

77

18. Cronan JE. 2003. Bacterial membrane lipids: Where do we stand? Annu. Rev. Microbiol.

57:203–24

19. Dartigalongue C, Raina S. 1998. A new heat-shock gene, ppiD, encodes a peptidyl-prolyl

isomerase required for folding of outer membrane proteins in Escherichia coli . EMBO J.

17:3968–80

20. De Cock H, Sch ¨ afer U, Potgeter M, Demel R, M ¨ uller M, et al. 1999. Affinity of theperiplasmic chaperone Skp of Escherichia coli for phospholipids, lipopolysaccharides andnon-native outer membrane proteins. Role of Skp in the biogenesis of outer membrane

protein. Eur. J. Biochem. 259:96–103

21. De Cock H, Struyv e M, Kleerebezem M, van der Krift T, Tommassen J. 1997. Role of

the carboxy-terminal phenylalanine in the biogenesis of outer membrane protein PhoEof Escherichia coli K-12. J. Mol. Biol. 269:473–78

22. De Cock H, Tommassen J. 1996. Lipopolysaccharides and divalent cations are involvedin the formation of an assembly-competent intermediate of outer membrane protein

PhoE of E. coli . EMBO J. 15:5567–73

23. De Keyzer J, van der Does C, Driessen AJM. 2003. The bacterial translocase: a dynamic

protein channel complex. Cell Mol. Life Sci. 60:2034–5224. Dev IK, Ray PH. 1984. Rapid assay and purification of a unique signal peptidase that

processes the prolipoprotein from Escherichia coli B. J. Biol. Chem. 259:11114–20

25. Doerrler WT. 2006. Lipid trafficking to the outer membrane of gram-negative bacteria.

Mol. Microbiol. 60:542–52

26. Doerrler WT, Gibbons HS, Raetz CRH. 2004. MsbA-dependent translocation of lipids

across the inner membrane of Escherichia coli . J. Biol. Chem. 276:45102–9

27. Doerrler WT, Raetz CRH. 2005. Loss of outer membrane proteins without inhibition

of lipid export in an Escherichia coli YaeT mutant. J. Biol. Chem. 280:27679–87

28. Dong C, Beis K, Nesper J, Brunkan-Lamontagne AL, Clarke BR, et al. 2006. Wza thetranslocon for E. coli capsular polysaccharides defines a new class of membrane protein.

Nature 444:226–2929. Eggert US, Ruiz N, Falcone BV, Branstrom AA, Goldman RC, et al. 2001. Genetic basis

for activity differences between vancomycin and glycolipid derivatives of vancomycin.

Science 294:361–64

30. Eppens EF, Nouwen N, Tommassen J. 1997. Folding of a bacterial outer membraneprotein during passage through the periplasm. EMBO J. 16:4295–301

31. Figueroa-Bossi N, Lemire S, Maloriol D, Balbontin R, Casadesus J, et al. 2006. Lossof Hfq activates the σE-dependent envelope stress response in Salmonella enterica. Mol.

Microbiol. 62:838–52

32. Fussenegger M, Facius D, Meier J, Meyer TF. 1996. A novel peptidoglycan-linked

lipoprotein (ComL) that functions in natural transformation competence of Neisseria

gonorrhoeae. Mol. Microbiol. 19:1095–105

33. Gentle I, Gabriel K, Beech P, Waller R, Lithgow T. 2004. The Omp85 family of proteins

is essential for outer membrane biogenesis in mitochondria and bacteria. J. Cell Biol.

164:19–24

34. Geyer R, Galanos C, Westphal O, Golecki JR. 1979. A lipopolysaccharide-binding cell-surface protein from Salmonella minnesota. Isolation, partial characterization and occur-

rence in different Enterobacteriaceae. Eur. J. Biochem. 98:27–38




35. Gunesekere IC, Kahler CM, Ryan CS, Snyder LA, Saunders NJ, et al. 2006. Ecf, an

alternative sigma factor from Neisseria gonorrhoeae, controls expression of msrAB, whichencodes methionine sulfoxide reductase. J. Bacteriol. 188:3463–69

36. Gupta SD, Gan K, Schmid MB, Wu HC. 1993. Characterization of a temperature-sensitive mutant of Salmonella typhimurium defective in apolipoprotein N -acyltransferase

J. Biol. Chem. 268:16551–5637. Harms N, Koningstein G, Dontje W, Muller M, Oudega B, et al. 2001. The early inter-

action of the outer membrane protein PhoE with the periplasmic chaperone Skp occursat the cytoplasmic membrane. J. Biol. Chem. 276:18804–1138. Hennecke G, Nolte J, Volkmer-Engert R, Schneider-Mergener J, Behrens S. 2005. The

periplasmic chaperone SurA exploits two features characteristic of integral outer mem-brane proteins for selective substrate recognition. J. Biol. Chem. 280:23540–48

39. Huijbrechts RPH, de Kroon AIPM, de Kruijff B. 2000. Topology and transport of mem-brane lipids in bacteria. Biochim. Biophys. Acta 1469:43–61

40. Ishidate K, Creeger E, Zrike J, Deb S, Glauner B, et al. 1986. Isolation of differentiatedmembrane domains from Escherichia coli and Salmonella typhimurium, including a fraction

containing attachment sites between the inner and outer membrane and the mureinskeleton of the cell envelope. J. Biol. Chem. 261:428–43

41. Ito Y, Kanamaru K, Taniguchi N, Miyamoto S, Tokuda H. 2006. A novel ligand bound ABC transporter, LolCDE, provides insights into the molecular mechanisms underlying

membrane detachment of bacterial lipoproteins. Mol. Microbiol. 62:1064–75

42. Jansen C, Heutink M, Tommassen J, de Cock H. 2000. The assembly pathway of outermembrane protein PhoE of Escherichia coli . Eur. J. Biochem. 267:3792–800

43. Johansen J, Rasmussen AA, Overgaard M, Valentin-Hansen P. 2006. Conserved smalnoncoding RNAs that belong to the σE regulon: role in down-regulation of outer mem-

brane proteins. J. Mol. Biol. 364:1–844. Justice SS, Hunstad DA, Harper JR, Duguay AR, Pinkner JS, et al. 2005. Periplasmic

peptidyl prolyl cis -trans isomerases are not essential for viability, but SurA is required forpilus biogenesis in Escherichia coli . J. Bacteriol. 187:7680–86

45. Kaniuk NA, Vinogradov E, Whitfield C. 2004. Investigation of the structural require-ments in the lipopolysaccharide core acceptor for ligation of O antigens in the genus

Salmonella: WaaL “ligase” is not the sole determinant of acceptor specificity. J. Biol

Chem. 279:36470–8046. Kellenberger E. 1990. The ‘Bayer bridges’ confronted with results from improved elec-

tron microscopy methods. Mol. Microbiol. 4:697–70547. Kleerebezem M, Heutink M, Tommassen J. 1995. Characterization of an Escherichia coli

rotA mutant, affected in periplasmic peptidyl-prolyl cis / trans isomerase. Mol. Microbiol

18:313–20

48. Klose M, Schwarz H, MacIntyre S, Freudl R, Eschbach ML, et al. 1988. Internal deletionsinthegeneforan Escherichia coli outer membrane protein define an area possibly important

for recognition of the outer membrane by this polypeptide. J. Biol. Chem. 263:13291–9649. Koebnik R, Locher KP, Van Gelder P. 2000. Structure and function of bacterial outer

membrane proteins: barrels in a nutshell. Mol. Microbiol . 37:239–5350. Kol M, van Dalen A, de Kroon AIPM, de Kruijff B. 2003. Translocation of phospholipids

is facilitated by a subset of membrane-spanning proteins of the bacterial cytoplasmic

membrane. J. Biol. Chem. 278:24586–9351. Koplow J, Goldfine H. 1974. Alterations in the outer membrane of the cell envelope of

heptose-deficient mutants of Escherichia coli . J. Bacteriol . 181:527–43




52. Kornd ¨ orfer IP, Dommel MK, Skerra A. 2004. Structure of the periplasmic chaperone

Skp suggests functionalsimilarity with cytosolic chaperonesdespite different architecture.

Nat. Struct. Mol. Biol. 11:1015–20

53. Koronakis V, Sharff A, Koronakis E, Luisi B, Hughes C. 2000. Crystal structure of thebacterial membrane protein TolC central to multidrug efflux and protein export. Nature

405:914–1954. Kozjak V, Wiedemann N, Milenkovic D, Lohaus C, Meyer HE, et al. 2003. An essential

role of Sam50 in the protein sorting and assembly machinery of the mitochondrial outermembrane. J. Biol. Chem. 278:48520–2355. Laird MW, Kloser AW, Misra R. 1994. Assembly of LamB and OmpF in deep rough

lipopolysaccharide mutants of Escherichia coli K-12. J. Bacteriol. 176:2259–6456. Lazar SW, Kolter R. 1996. SurA assists the folding of Escherichia coli outer membrane

proteins. J. Bacteriol. 178:1770–7357. Lee PA, Tullman-Ercek D, Georgiou G. 2006. The bacterial twin-arginine translocation

pathway. Annu. Rev. Microbiol. 60:373–9558. Malinverni JC, Werner J, Kim S, Sklar JG, Kahne D, et al. 2006. YfiO stabilizes the YaeT

complex and is essential for outer membrane protein assembly in Escherichia coli . Mol.

Microbiol. 61:151–64

59. Marchler-Bauer A, Anderson JB, Cherukuri PF, DeWeese-Scott C, Geer LY, et al.2005. CDD: a Conserved Domain Database for protein classification. Nucleic Acids Res .

33:D192–96

60. Matsuyama S, Tajima T, Tokuda H. 1995. A novel carrier protein involved in the sortingand transport of Escherichia coli lipoproteins destined for the outer membrane. EMBO J.

14:3365–7261. Matsuyama S, Yokota N, Tokuda H. 1997. A novel outer membrane lipoprotein, LolB

(HemM), involved in the LolA (p20)-dependent localization of lipoproteins to the outermembrane of Escherichia coli . EMBO J. 16:6947–55

62. Missiakas D, Betton JM, Raina S. 1996. New components of protein folding in extracy-toplasmic compartments of Escherichia coli SurA, FkpA and Skp/OmpH. Mol. Microbiol.

21:871–8463. M ¨ uhlradt PF, Menzel J, Golecki JR, Speth V. 1973. Outer membrane of Salmonella. Sites

of export of newly synthesised lipopolysaccharide on the bacterial surface. Eur. J. Biochem.

35:471–81

64. Narita S, Matsuyama S, Tokuda H. 2004. Lipoprotein trafficking in Escherichia coli . Arch.

Microbiol. 182:1–665. Narita S, Tokuda H. 2006. An ABC transporter mediating the membrane detachment of

bacterial lipoproteins depending on their sorting signals. FEBS Lett . 580:1164–7066. Nikaido H. 2003. Molecular basis of bacterial outer membrane permeability revisited.

Microbiol. Mol. Biol. Rev. 67:593–65667. Onufryk C, Crouch ML, Fang FC, Gross CA. 2005. Characterization of six lipoproteins

in the σE regulon. J. Bacteriol. 187:4552–6168. Papenfort K, Pfeiffer V, Mika F, Lucchini S, Hinton JCD, et al. 2006. σE-dependent

small RNAs of Salmonella respond to membrane stress by accelerating global omp mRNA decay. Mol. Microbiol. 62:1674–88

69. Paschen SA, Waizenegger T, Stan T, Preuss M, Cyrklaff M, et al. 2003. Evolutionary

conservation of biogenesis of β-barrel membrane proteins. Nature 426:862–6670. Pettersson A, Poolman JT, van der Ley P, Tommassen J. 1997. Response of Neisseria

meningitidis to iron limitation. Antonie van Leeuwenhoek 71:129–36




71. Prinz T, Tommassen J. 2000. Association of iron-regulated outer membrane proteins of

Neisseria meningitidis with the RmpM (class 4) protein. FEMS Microbiol. Lett. 183:49–5372. Pugsley AP. 1993. The complete general secretory pathway in gram-negative bacteria

Microbiol. Rev. 57:50–10873. Raetz CRH, Reynolds CM, Trent MS, Bishop RE. 2007. Lipid A modification systems

in gram-negative bacteria. Annu. Rev. Biochem. 76:295–32974. Raetz CRH, Whitfield C. 2002. Lipopolysaccharide endotoxins. Annu. Rev. Biochem

71:635–70075. Rizzitello AE, Harper JR, Silhavy TJ. 2001. Genetic evidence for parallel pathways of

chaperone activity in the periplasm of Escherichia coli . J. Bacteriol. 183:6794–800

76. Demonstratesthat the C-terminalsignature sequenceof OMPs functionsas a targeting factor, recognizedby assembly factor Omp85.

76. Robert V, Volokhina EB, Senf F, Bos MP, Van Gelder P, et al. 2006. Assembly

factor Omp85 recognizes its outer membrane protein substrates by a species-

specific C-terminal motif. PLoS Biol. 4:1984–9577. Rouvi ere PE, Gross CA. 1996. SurA, a periplasmic protein with peptidyl-prolyl isomerase

activity, participates in the assembly of outer membrane porins. Genes Dev. 10:3170–8778. Ruiz N, Falcone B, Kahne D, Silhavy TJ.2005.Chemicalconditionality: a genetic strategy

to probe organelle assembly. Cell 121:307–1779. Ruiz N, Silhavy TJ. 2005. Sensing external stress: watchdogs of the Escherichia coli cell

envelope. Curr. Opin. Microbiol. 8:122–26

80. Sampson BA, Misra R, Benson SA. 1989. Identification andcharacterization of a newgeneof Escherichia coli K-12 involved in outer membrane permeability. Genetics 122:491–501

81. Sanchez-Pulido L, Devos D, Genevrois S, Vincente M, Valencia A. 2003. POTRA: a

conserved domain in the FtsQ family and a class of β-barrel outer membrane proteinsTrends Biochem. Sci. 28:523–26

82. Sankaran K, Wu HC. 1994. Lipid modification of bacterial prolipoprotein. Transfer of

diacylglyceryl moiety from phosphatidylglycerol. J. Biol. Chem. 269:19701–683. Sen K, Nikaido H. 1990. In vivo trimerization of OmpF porin secreted by spheroplasts

of Escherichia coli . Proc. Natl. Acad. Sci. USA 87:743–4784. Sklar JG, Wu T, Gronenberg LS, Malinverni JC, Kahne D, et al. 2007. Lipoprotein

SmpA is a component of the YaeT complex that assembles outer membrane proteins in

Escherichia coli . Proc. Natl. Acad. Sci. USA 104:6400–5

85. Identifies an ABC protein and aperiplasmic protein as new factorsrequired for thetransport of LPS tothe cell surface.

85. Sperandeo P, Cescutti R, Villa R, Di Benedetto C, Candia D, et al. 2007. Char-

acterization of lptA and lptB , two essential genes implicated in lipopolysaccharide

transport to the outer membrane of Escherichia coli . J. Bacteriol. 189:244–5386. Sperandeo P, Pozzi C, Deho G, Polissi A. 2006. Non-essential KDO biosynthesis and

new essential cell envelope biogenesis genes in the Escherichia coli yrbG- yhbG locus. Res

Microbiol. 157:547–5887. Spies C, Beil A, Ehrmann M. 1999. A temperature-dependent switch from chaperone to

protease in a widely conserved heat shock protein. Cell 97:339–4788. Steeghs L, Berns M, ten Hove J, de Jong A, Roholl P, et al. 2002. Expression of foreign

LpxA acyltransferases in Neisseria meningitidis results in modified lipid A with reduced

toxicity and retained adjuvant activity. Cell. Microbiol. 4:599–61189. Steeghs L, de Cock H, Evers E, Zomer B, Tommassen J, et al. 2001. Outer membranecomposition of a lipopolysaccharide-deficient Neisseria meningitidis mutant. EMBO J

20:6937–45

90. First demonstration that

N. meningitidis is viable without LPS,making thisorganism suitablefor further studieson LPS transport.

90. Steeghs L, den Hartog R, den Boer A, Zomer B, Roholl P, et al. 1998. Meningitis

bacterium is viable without endotoxin. Nature 392:449–5091. Stegmeier JF, Andersen C. 2006. Characterization of pores formed by YaeT (Omp85)

from Escherichia coli . J. Biochem. 140:275–83




92. Stegmeier JF, Gl ¨ uck A, Sukumaran S, M ¨ antele W, Andersen C. 2007. Characterization

of YtfM, a second member of the Omp85 family in Escherichia coli . Biol. Chem. 388:37–4693. Stenberg F, Chovanec P, Maslen SL, Robinson CV, Ilag LL, et al. 2005. Protein com-

plexes of the Escherichia coli cell envelope. J. Biol. Chem. 280:34409–1994. Struyv e M, Moons M, Tommassen J. 1991. Carboxy-terminal phenylalanine is essential

for the correct assembly of a bacterial outer membrane protein. J. Mol. Biol. 218:141–4895. Surana NK, Grass S, Hardy GG, Li H, Thanassi DG, et al. 2004. Evidence for conserva-

tionof architecture andphysical properties of Omp85-like proteins throughout evolution.

Proc. Natl. Acad. Sci. USA 101:14497–503

96. Provideshigh-resolution structures of LolA and LolB,providing insightsinto the lipoproteintransport mechanism.

96. Takeda K, Miyatake H, Yokota N, Matsuyama S, Tokuda H, et al. 2003. Crystal

structures of bacterial lipoprotein localization factors, LolA and LolB. EMBO J.

22:3199–20997. Tam C, Missiakas D. 2005. Changes in lipopolysaccharide structure induce the σE-

dependent response of Escherichia coli . Mol. Microbiol. 55:1403–1298. Tamm LK, Arora A, Kleinschmidt JH. 2001. Structure and assembly of β-barrel mem-

brane proteins. J. Biol. Chem. 276:32399–40299. Tefsen B, Bos MP, Beckers F, Tommassen J, de Cock H. 2005. MsbA is not required for

phospholipid transport in Neisseria meningitidis . J. Biol. Chem. 280:35961–66100. Tefsen B, Geurtsen J, Beckers F, Tommassen J, de Cock H. 2005. Lipopolysaccharide

transport to the bacterial outer membrane in spheroplasts. J. Biol. Chem. 280:4504–9101. Terada M, Kuroda T, Matsuyama S, Tokuda H. 2001. Lipoprotein-sorting signals eval-

uated as the LolA-dependent release of lipoproteins from the cytoplasmic membrane of

Escherichia coli . J. Biol. Chem. 276:47690–94102. Tormo A, Almir ´ on M, Kolter R. 1990. surA, an Escherichia coli gene essential for survival

in stationary phase. J. Bacteriol. 172:4339–47103. Trent MS, Stead CM, Tran AX, Hankins JV. 2006. Diversity of endotoxin and its impact

on pathogenesis. J. Endotoxin Res. 12:205–23104. Van Ulsen P, Tommassen J. 2006. Protein secretion and secreted proteins in pathogenic

Neisseriaceae. FEMS Microbiol. Rev. 30:292–319

105. First

demonstration thatan integral OMP,Omp85, is requiredfor the assembly ofOMPs.

105. Voulhoux R, Bos MP, Geurtsen J, Mols M, Tommassen J. 2003. Role of a highly

conserved bacterial protein in outer membrane protein assembly. Science 299:262–65

106. Voulhoux R, Tommassen J. 2004. Omp85, an evolutionarily conserved bacterial proteininvolved in outer-membrane-protein assembly. Res. Microbiol. 155:129–35

107. Waller PR, Sauer RT. 1996. Characterization of degQ and degS , Escherichia coli genesencoding homologs of the DegP protease. J. Bacteriol. 178:1146–53

108. Demonstratesthat the C-terminasignature sequenceof OMPs triggersthe σE-dependent periplasmic stress

response when unfolded OMPsaccumulate in theperiplasm.

108. Walsh NP, Alba BM, Bose B, Gross CA, Sauer RT. 2003. OMP peptide signals

initiate the envelope-stress response by activating DegS protease via relief of in-

hibition mediated by its PDZ domain. Cell 113:61–71109. Walton TA, Sousa MC. 2004. Crystal structure of Skp, a prefoldin-like chaperone that

protects soluble and membrane proteins from aggregation. Mol. Cell 15:367–74

110. Wandersman C, Delepelaire P. 2004. Bacterial iron sources: from siderophores tohemophores. Annu. Rev. Microbiol. 58:611–47111. Werner J, Augustus AM, Misra R. 2003. Assembly of TolC, a structurally unique and

multifunctional outer membrane protein of Escherichia coli K-12. J. Bacteriol. 185:6540–47

112. Werner J, Misra R. 2005. YaeT (Omp85) affects the assembly of lipid-dependent andlipid-independent outer membrane proteins of Escherichia coli . Mol. Microbiol. 57:1450–

59




Annual Review of

Microbiology

Volume 61, 2007 Contents

Frontispiece

Margarita Salas

40 Years with Bacteriophage Ø29

Margarita Salas

The Last Word: Books as a Statistical Metaphor for Microbial Communities

Patrick D. Schloss and Jo Handelsman

The Mechanism of Isoniazid Killing: Clarity Through the Scope

of Genetics

Catherine Vilchèze and William R. Jacobs, Jr.

Development of a Combined Biological and Chemical Process for

Production of Industrial Aromatics from Renewable Resources

F. Sima Sariaslani

The RNA Degradosome of Escherichia coli : An mRNA-Degrading

Machine Assembled on RNase E Agamemnon J. Carpousis

Protein Secretion in Gram-Negative Bacteria via the Autotransporter

Pathway

Nathalie Dautin and Harris D. Bernstein

Chlorophyll Biosynthesis in Bacteria: The Origins of Structural and

Functional Diversity

Aline Gomez Maqueo Chew and Donald A. Bryant

Roles of Cyclic Diguanylate in the Regulation of Bacterial Pathogenesis

Rita Tamayo, Jason T. Pratt, and Andrew Camilli Aggresomes and Pericentriolar Sites of Virus Assembly:

Cellular Defense or Viral Design?

Thomas Wileman

As the Worm Turns: The Earthworm Gut as a Transient Habitat for

Soil Microbial Biomes

Harold L. Drake and Marcus A. Horn

vi



Biogenesis of the Gram-Negative Bacterial Outer Membrane

Martine P. Bos, Viviane Robert, and Jan Tommassen 191

SigB-Dependent General Stress Response in Bacillus subtilis and

Related Gram-Positive Bacteria

Michael Hecker, Jan Pané-Farré, and Uwe Völker 215

Ecology and Biotechnology of the Genus Shewanella Heidi H. Hau and Jeffrey A. Gralnick 237

Nonhomologous End-Joining in Bacteria: A Microbial Perspective

Robert S. Pitcher, Nigel C. Brissett, and Aidan J. Doherty 259

Postgenomic Adventures with Rhodobacter sphaeroides Chris Mackenzie, Jesus M. Eraso, Madhusudan Choudhary, Jung Hyeob Roh,

Xiaohua Zeng, Patrice Bruscella, Ágnes Puskás, and Samuel Kaplan 283

Toward a Hyperstructure Taxonomy

Vic Norris, Tanneke den Blaauwen, Roy H. Doi, Rasika M. Harshey, Laurent Janniere, Alfonso Jim´ enez-S´ anchez, Ding Jun Jin,Petra Anne Levin, Eugenia Mileykovskaya, Abraham Minsky,Gradimir Misevic, Camille Ripoll, Milton Saier, Jr., Kirsten Skarstad,and Michel Thellier 309

Endolithic Microbial Ecosystems

Jeffrey J. Walker and Norman R. Pace 331

Nitrogen Regulation in Bacteria and Archaea

John A. Leigh and Jeremy A. Dodsworth 349

Microbial Metabolism of Reduced Phosphorus Compounds Andrea K. White and William W. Metcalf 379

Biofilm Formation by Plant-Associated Bacteria

Thomas Danhorn and Clay Fuqua 401

Heterotrimeric G Protein Signaling in Filamentous Fungi

Liande Li, Sara J. Wright, Svetlana Krystofova, Gyungsoon Park,and Katherine A. Borkovich 423

Comparative Genomics of Protists: New Insights into the Evolution

of Eukaryotic Signal Transduction and Gene Regulation

Vivek Anantharaman, Lakshminarayan M. Iyer, and L. Aravind 453

Lantibiotics: Peptides of Diverse Structure and Function

Joanne M. Willey and Wilfred A. van der Donk 477

The Impact of Genome Analyses on Our Understanding of

Ammonia-Oxidizing Bacteria

Daniel J. Arp, Patrick S.G. Chain, and Martin G. Klotz 503

Co nt en ts v ii



Morphogenesis in Candida albicans Malcolm Whiteway and Catherine Bachewich

Structure, Assembly, and Function of the Spore Surface Layers

Adriano O. Henriques and Charles P. Moran, Jr.

Cytoskeletal Elements in Bacteria

Peter L. Graumann

Indexes

Cumulative Index of Contributing Authors, Volumes 57–61

Cumulative Index of Chapter Titles, Volumes 57–61

Errata

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

om bio genesis

Documents