Вы находитесь на странице: 1из 12

REVIEWS

Types and origins of bacterial


membrane vesicles
Masanori Toyofuku1*, Nobuhiko Nomura1 and Leo Eberl 2
*
Abstract | Most bacteria release membrane vesicles (MVs) that contain specific cargo molecules
and have diverse functions, including the transport of virulence factors, DNA transfer,
interception of bacteriophages, antibiotics and eukaryotic host defence factors, cell
detoxification and bacterial communication. MVs not only are abundant in nature but also show
great promise for applications in biomedicine and nanotechnology. MVs were first discovered to
originate from controlled blebbing of the outer membrane of Gram-negative bacteria and are
therefore often called outer-membrane vesicles (OMVs). However, recent work has shown that
Gram-positive bacteria can produce MVs, that different types of MVs besides OMVs exist and
that, in addition to membrane blebbing, MVs can also be formed by endolysin-triggered cell lysis.
In this Review, we provide an overview of the structures and compositions of the various vesicle
types and discuss novel formation routes, which may lead to distinct vesicle types that serve
particular functions.

Bacteria release membrane vesicles (MVs) with sizes their structure and composition reflect the routes by
ranging from 20 to 400 nm in diameter that affect which they were formed. We summarize the published
diverse biological processes, including virulence, hori- evidence showing that MVs that originate from cell lysis
zontal gene transfer, export of cellular metabolites, phage are different from MVs originating from membrane
infection and cell-to-cell communication1–3. Microbial blebbing of living cells and that this has consequences
MVs also have immunomodulatory activities and are for their composition and content and possibly for
therefore used as vaccines and show great potential for their function.
the development of anticancer drugs and for applica-
tions in nanotechnology4,5. Bacterial MVs are abundant Types of membrane vesicles
in coastal and open-ocean sea water, which implies that Recent work has shown that MVs can differ in their
these structures are important for carbon cycling in the structure and composition, and some of these differ-
marine ecosystem6. MVs are also abundant in micro- ences might be explained by the different routes by
bial biofilms, where they are an integral constituent of which the MVs were formed (FIGS 1,2; TABLE 1).
the biofilm matrix and can protect biofilm cells from
certain antibiotics7–10. Outer-membrane vesicles. OMVs are the archetypal
MVs were first found to be produced through bacterial MV. They are spherical particles that consist
controlled blebbing of the outer membrane of Gram- of an outer leaflet of lipopolysaccharide (LPS) and an
negative bacteria and are therefore often referred to as inner leaflet of phospholipid, which is derived from the
outer- membrane vesicles (OMVs). The composition outer membrane of Gram-negative bacteria1,11–13. Classic
of OMVs and their biological functions, as well as vari- OMVs originate from blebbing of the outer mem-
1
Department of Life and
ous OMV formation mechanisms and the underlying brane and are therefore enriched for outer-membrane
Environmental Sciences, environmental triggers, have been summarized in a proteins, show specific lipid compositions and differ in
University of Tsukuba, number of excellent articles1,3,11–13. This Review focuses the amount and content of cargo molecules depending
Tsukuba, Japan. on recent work that has demonstrated that other types of on growth conditions3,16–19. Many studies have shown
2
Department of Plant and MVs exist in addition to OMVs (FIG. 1), including outer- that they contain periplasmatic and cytosolic proteins,
Microbial Biology, University inner membrane vesicles (OIMVs), cytoplasmic mem- DNA and RNA and that they transport virulence fac-
of Zurich, Zurich, Switzerland.
brane vesicles (CMVs) and tube-shaped membranous tors14,20–24. OMVs are therefore considered to represent
*e-mail: toyofuku.masanori.
structures (TSMSs), and that MVs can also be formed as a specialized bacterial secretion pathway25.
gf@u.tsukuba.ac.jp;
leberl@botinst.uzh.ch a consequence of phage endolysin-triggered cell lysis14,15. However, despite many years of research, the sorting
https://doi.org/10.1038/ On the basis of these studies, we suggest that different mechanism required for cargo selection is unknown,
s41579-018-0112-2 formation routes lead to distinct types of MVs and that and the presence of cytoplasmic contents within

NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | JANUARY 2019 | 13


REVIEWS

a c e

CMV

Phage tail

Phage head

b Peptidoglycan d f

Inner membrane Outer membrane

Fig. 1 | Types of vesicles. a | Outer-membrane vesicles (OMVs) are produced by blebbing of the outer membrane
and thus consist of a single membrane bilayer (green), as shown in an electron cryotomographic (ECT) image (left) and
a reconstructed 3D model (right) of an OMV isolated from an Acinetobacter baumannii culture (scale bar = 50 nm).
b | By contrast, outer-inner membrane vesicles (OIMVs) have two membrane bilayers derived from the outer membrane
(green) and the inner membrane (purple), and in between the membranes there is peptidoglycan (OIMV from
A. baumannii; scale bar = 100 nm). c | Gram-positive bacteria can produce cytoplasmic membrane vesicles (CMVs),
as exemplified here by an ECT image of a CMV produced by Bacillus subtilis. Notably, the vesicle carries PBSX phage
particles. PBSX encodes endolysin, and its expression leads to the formation of holes in the peptidoglycan layer through
which the cytoplasmic membrane protrudes, forming CMVs (scale bar = 100 nm). d | Tube-shaped membranous
structures are considered a specialized type of membrane vesicle. B. subtilis forms intercellular nanotubes that connect
neighbouring cells (scale bar = 1 µm). e | Myxococcus xanthus forms OMV chains, which can connect cells in biofilms (scale
bar = 100 nm). f | Vibrio vulnificus forms segmented tubes, from which OMVs can be pinched off (scale bar = 200 nm).
Parts a and b reproduced from Koning, R. I. et al. Cryo-electron tomography analysis of membrane vesicles from
Acinetobacter baumannii ATCC19606 T. Research in Microbiology 2013; 164(5):397–405. Copyright © 2013 Elsevier
Masson SAS. All rights reserved. Part c adapted from REF.15, CC-BY-4.0. Part d adapted with permission from REF.50,
Elsevier. Part e reproduced with permission from REF.41, Wiley-VCH. Part f adapted from REF.48, CC-BY-4.0.

OMVs has remained unexplained26. It is possible that electron cryomicroscopy study of Shewanella vesiculosa
previous reports, which found nucleic acids or cyto- M7 supernatant32, which unambiguously demonstrated
solic proteins in OMVs, might not have analysed bona the production of double bilayered MVs, which were
fide OMVs, which are formed through the blebbing of originally named O-IMVs (referred to as OIMVs in the
the outer membrane and therefore do not have direct following sections). The study also showed that although
access to cytoplasmic contents26,27. It is worth noting, this bacterium produces both OMVs and OIMVs, most
however, that OMVs that are formed through cell of the DNA is contained in OIMVs. Subsequent work
lysis of a small subpopulation of the cultured bacteria demonstrated that diverse bacteria produce OIMVs
can contain RNA, chromosomal DNA and endolysins14 and that DNA is specifically packed into this type of
(see below). MV32–34. The fraction of OIMVs produced seems to
be highly variable, ranging from 0.1% of total MVs in
Outer-inner membrane vesicles. Kadurugamuwa and S. vesiculosa M7 to 49% in Pseudoalteromonas
Beveridge reported in 1995 that MVs from Pseudomonas marina32,35. Although the proposed blebbing model
aeruginosa contain DNA, confirming earlier stud- could explain how plasmid DNA enters OIMVs, it is
ies demonstrating the presence of DNA and RNA in difficult to understand how DNA fragments of the
Endolysins
Gram-negative MVs28–31. As none of the blebbing models bacterial chromosome are packaged into MVs with-
Hydrolytic enzymes that are at the time could convincingly explain how DNA out assuming that the producing cell dies. Indeed, the
produced by bacteriophages passes through the inner membrane and enters OMVs, presence of chromosomal DNA in MVs is considered
to degrade the cell wall of the they proposed a new formation mechanism. In their to be indicative of cell lysis26,27. ‘Explosive cell lysis’ is
bacterial host during the final
stage of the lytic cycle.
model, the peptidoglycan layer of the bacterial cell is another possible route for the formation of MVs con-
weakened by autolysins such that the inner membrane taining chromosomal DNA, as cell lysis is triggered by
Autolysins protrudes into the periplasm, allowing cytoplasmic con- DNA damage and results in cell envelope fragments that
Peptidoglycan hydrolases that tents such as DNA to enter the vesicle, which is even- re-circularize and thereby enclose the released DNA (see
are produced by bacteria for
tually pinched off from the cell surface together with below). Notably, this would also explain why vesicles
peptidoglycan turnover and to
complete cell division by
a surrounding outer membrane. The first experimen- can be enriched for endolysins, as peptidoglycan deg-
separating the daughter cell tal evidence for MVs consisting of both the outer and radation is the key process in explosive cell lysis14,36,37.
from the mother cell. the inner membranes was provided by a transmission In support of this hypothesis, a recent study showed

14 | JANUARY 2019 | VOLUME 17 www.nature.com/nrmicro


REVIEWS

Gram-negative bacteria
Outer membrane Peptidoglycan Gram-positive bacteria
Endolysin
Inner membrane Cytoplasmic molecule Peptidoglycan Cytoplasmic membrane

Membrane DNA
molecule

Hydrophobic Phage
molecule Endolysin

Imbalanced
cell wall Explosive Bubbling
Intercalation cell lysis cell death
biosynthesis

OMV OIMV EOMV CMV


Fig. 2 | Different routes lead to the formation of distinct membrane vesicle types. Gram-negative bacteria have two
main routes for vesicle formation: blebbing of the outer membrane and explosive cell lysis. Blebbing leads to the production
of outer-membrane vesicles (OMVs) and occurs as a result of cell envelope disturbances, which can be caused by either
imbalance of peptidoglycan biosynthesis or intercalation of hydrophobic molecules into the outer membrane. As the inner
membrane remains intact, cytoplasmic components have no direct access to these OMVs. Explosive cell lysis is triggered by
phage-derived endolysin that degrades the peptidoglycan cell wall. Once the peptidoglycan is degraded, the cell rounds
up and explodes, and the shattered membrane fragments round up and self-assemble into outer-inner membrane vesicles
(OIMVs) and explosive outer-membrane vesicles (EOMVs). In contrast to OMVs formed by blebbing, EOMVs randomly
contain cytoplasmic components. Endolysin also triggers ‘bubbling cell death’ in Gram-positive bacteria, in which it gives
rise to cytoplasmic membrane vesicles (CMVs). CMVs can contain membrane and cytoplasmic components. Other
peptidoglycan-damaging enzymes and treatments may have similar consequences as endolysin. In Gram-negative bacteria,
CMVs, arising from the inner membrane, have also been reported, but the underlying mechanism is not understood.

that treatment of Stenotrophomonas maltophilia with Tube- shaped membranous structures. Several dif-
ciprofloxacin not only induces the SOS response and ferent bacteria produce TSMSs, often referred to as
consequently cell lysis (BOX 1) but also stimulates the nanotubes, nanowires or nanopods. These structures
production of OIMVs38. are tube-like protrusions of the cytoplasmic membrane
of Gram-positive bacteria or of the outer membrane of
Cytoplasmic membrane vesicles. Given that MVs were Gram-negative bacteria and are considered to be spe-
originally thought to originate solely from the pinching- cialized types of MVs. They often decorate the surface
off of outer-membrane material from Gram-negative of the producing cell and form bridges between cells that
bacteria, it came as a surprise when vesicle formation enable the exchange of various cellular components.
was observed in Gram- positive bacteria. Over the The simplest version of these structures is a chain of
past two decades, vesicle formation has been demon- OMVs, which was first observed in Myxococcus xan-
strated in numerous Gram-positive bacteria (reviewed thus41,42. Using various imaging techniques, including
previously2). Given the lack of an outer membrane, we 3D focused ion beam scanning electron microscopy, a
suggest that MVs produced by Gram-positive bacteria study41 showed that these structures range between 30
be named CMVs. Although it has been demonstrated and 60 nm in width and up to 5 μm in length and form
that CMVs can arise from dying cells15,39 (see below), it is an extensive membrane-enclosed network that seems
conceivable that conservative blebbing mechanisms exist to connect cells within biofilms at the level of the peri-
that are independent of cell death. Large intracellular plasmic space. It has been suggested that these vesicular
vesicles have also been observed in the Gram-negative connections constitute intercellular bridges that enable
bacterium Acidiphilium cryptum JF-5 when grown under cytoplasmic molecular trade between biofilm cells to
anaerobic conditions40. The strain also formed blebs at coordinate social activities43,44.
the cell surface and tube-shaped projections, similar to The metal- reducing microorganism Shewanella
the nanotubes discussed below. Electron micrographs oneidensis MR-1 produces OMV chains that can elongate
of thin sections showed morphological continuity and turn into smooth filaments45. The most remarka-
between the inner membrane (or cytoplasmic membrane), ble property of these tubular structures is their electri-
the blebs and the projections, indicating that the mem- cal conductivity and their ability to connect separate
branes of the blebs and projections originated from the S. oneidensis MR-1 cells. They are therefore considered
inner membrane. a special type of microbial nanowire, which is normally

NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | JANUARY 2019 | 15


REVIEWS

formed by electrically conductive proteins46. Whether Nanotubes that originate from extensions of the
redox-functionalized membrane extensions represent outer membrane may enable the intercellular transfer of
a general strategy for long-range extracellular electron periplasmic metabolites, membrane proteins and lipids
transport between bacteria remains to be shown, but it but not of cytoplasmic contents. This is in contrast to
is worth noting that, in addition to OMV chains, nano- nanotubes formed by the Gram- positive bacterium
tubes that connect M. xanthus cells in biofilms have been Bacillus subtilis, which the bacterium uses to exchange
observed41,42. Francisella tularensis, the causative agent of cytoplasmic contents such as proteins and plasmid
the zoonotic disease tularaemia, produces OMVs and DNA49,50. Electron microscopic analyses showed that
TSMSs in liquid media, which contribute to cell–cell these multilayered structures consist of cell wall material,
communication and the exchange of molecules without membrane components and cytoplasmic contents43.
forming bridges between cells47. A recent study48 nicely The texture of the nanotubes resembles the bacterial cell
illustrated the interplay between OMVs and nanotubes. surface, suggesting similar exterior composition and
Cryo-electron microscopy showed that in cultures of surface continuity.
capsulated Vibrio vulnificus strains, individual OMVs
were pinched off from nanotubes, creating an intrigu- Routes of vesicle formation
ing ‘beads-on-a-string’ pattern. This suggests that, in Evidence has accumulated that suggests that two prin-
addition to strings of OMVs fusing to form nanotubes, cipal routes for the formation of MVs exist: blebbing
as observed for M. xanthus, nanotubes, in turn, can also of membrane material of living cells that gives rise
disintegrate into OMVs. to classic OMVs, and endolysin- triggered cell lysis

Table 1 | Proposed vesicle types, their routes of formation and their compositions
Formation routes Outer- Cytoplasmic (or Plasmids RNA and Endolysinsa Virulence Hydrophobic Phagesa
membrane inner) membrane chromosomal factors molecules
proteins proteins DNA
Outer-membrane vesicle (Gram-negative)
Outer-membrane blebbing + – (+)b + – (+)c – + – –
(imbalanced peptidoglycan
turnover, phospholipid
accumulation or deacylation
of lipid A)
Intercalation of hydrophobic + – (+)b ? – (+)c – + +d –
molecules (such as bacterial
signalling molecules,
antibiotics and hydrophobic
substrates) into the outer
membrane
Shedding of + – ? – – ? – –
lipopolysaccharide
from flagella
Explosive outer-membrane vesicle (Gram-negative)
OMVs generated through + + + + + + + +
explosive cell lysis (either
directly or through
rearrangement of OIMVs)
Outer-inner membrane vesicle (Gram-negative)
Explosive cell lysis and + + + + + + + +
cell budding
Cytoplasmic membrane vesicle (Gram-positivee)
‘Bubbling cell death’ and NA + + + + + – +
bacterial autolysins
Tube-shaped membranous structure (Gram-negative)
Mechanism unknown + – – – ? – – –
Tube-shaped membranous structure (Gram-positive)
Mechanism unknown NA – + + ? ? – –
+, present; –, absent; ?, unknown; CMV, cytoplasmic membrane vesicle; EOMV, explosive outer-membrane vesicle; NA , not applicable; OIMV, outer-inner membrane
vesicle; OMV, outer-membrane vesicle; TSMS, tube-shaped membranous structure. aMembrane vesicles (MVs) originating from cell lysis are intrinsically associated
with endolysins and sometimes entire phages, whereas they are expected to be absent from MVs originating through blebbing mechanisms. OMVs can adsorb
phages on their surface but would not carry phages inside. bOccasionally , cytoplasmic proteins have been observed in OMVs, but it is not known how these proteins
are packaged into OMVs. cAlthough several reports have demonstrated the presence of plasmid DNA in OMVs, the presence of chromosomal DNA and RNA is
considered an indication of cell lysis. We therefore speculate that these nucleic acids are mainly associated with OIMVs and EOMVs. dHydrophobic molecules that
can induce vesicle formation through blebbing will stay associated with the vesicles. eVesicles originating from the inner (or cytoplasmic) membrane have also been
reported in Gram-negative bacteria, but the underlying formation route is unclear.

16 | JANUARY 2019 | VOLUME 17 www.nature.com/nrmicro


REVIEWS

Box 1 | SOS response in bacteria formation in the mutant. A recent study showed that
phospholipid accumulation in the outer leaflet of the
The SOS response is a global regulatory system that allows bacteria to respond to DNA outer membrane of Haemophilus influenzae promotes
damage. It was first identified in Escherichia coli, in which it was shown to control the release of OMVs without compromising mem-
prophage induction, cell filamentation and mutation upon exposure to ultraviolet brane integrity61. Phospholipids accumulated when
radiation125. Since then, it has become a paradigm of a general stress response system
the VacJ (also known as MlaA) and Yrb ATP-binding
that is conserved and widespread among bacteria. Two regulatory proteins modulate
the expression of genes involved in the SOS response126: the LexA repressor, which cassette (ABC) transport system, a proposed phospho-
inhibits expression of the SOS genes during normal cell growth by binding to specific lipid transporter, is either mutated or its expression is
DNA sequences (SOS boxes) upstream of target genes and thus blocking their downregulated under iron-limiting conditions. Given
transcription127; and the RecA protein, which binds to single-stranded DNA generated that inactivation of this transporter increased vesicu-
by DNA-damaging agents or inhibition of replication. The binding of RecA to single- lation in phylogenetically diverse bacteria, it has been
stranded DNA, which is assisted by the components of the RecFOR and RecBCD suggested that this OMV biogenesis pathway could be
proteins of the two major recombinational repair pathways in E.coli128, activates RecA universal in Gram-negative bacteria16. In Salmonella
and stimulates autocatalytic cleavage of LexA129. Cleaved LexA is no longer able to bind enterica subsp. enterica serovar Typhimurium, deacyl-
to SOS boxes, and thus, expression of SOS genes is derepressed. In E. coli and many ation of LPS was shown to affect the level of OMV prod-
other bacteria, the SOS regulon comprises genes of three DNA repair pathways,
uction: expression of the lipid A deacylase PagL led to
namely, homologous recombination, nucleotide excision repair and translesion DNA
synthesis130. Translesion DNA synthesis often introduces mutations, as this pathway overvesiculation with deacylated lipid A accumulation
involves error-prone DNA polymerases, and thus increases the occurrence of antibiotic- exclusively in OMVs18.
resistant strains131,132. The SOS response also affects cell division and blocks septum Another mechanism for OMV formation through
formation, leading to filamentation133. Expression of lexA is negatively autoregulated, blebbing is the accumulation of peptidoglycan fragments
and thus, SOS gene expression is repressed once the DNA damage has been or misfolded proteins in the periplasmic space59,62,63.
repaired134,135. Importantly, RecA also activates self-cleavage of other transcriptional This is thought to create a turgor stress on the outer
repressor proteins, most notably phage repressors, and therefore, phage production is membrane that eventually leads to OMV release.
typically induced upon DNA damage. Another route for OMV formation through bleb-
bing involves integration of molecules that induce
that leads to the formation of OIMVs, explosive outer- membrane curvature into the cell envelope of bacteria.
membrane vesicles (EOMVs) (discussed below) and The best investigated example for this ‘bilayer-couple’
CMVs (FIG. 2; TABLE 1). model of OMV formation comes from P. aeruginosa:
the quorum-sensing molecule Pseudomonas quinolone
Vesicle formation through membrane blebbing. signal (PQS) mediates its own packaging and transport
Membrane blebbing results from cell envelope distur- by stimulating OMV formation through intercalation
bances, which can be caused by either the unbalanced into the outer membrane64–66. The accumulation of PQS
biosynthesis of cell wall components or the intercalation in the outer leaflet is thought to change the membrane
Intercalation of hydrophobic molecules into the outer membrane of curvature, providing the initial driving force for OMV
The reversible insertion of Gram-negative bacteria. Although the release of OMVs formation. PQS was shown to increase anionic repul-
molecules into materials with occurs during normal cell growth without affecting sions between LPS molecules by sequestering divalent
layered structures such as the
cell viability, growth conditions can have a profound cations, which are important to stabilize the negatively
cellular membrane.
effect on vesiculation51. Several blebbing models have charged B- band LPS molecules67. The bilayer- couple
Turgor pressure been proposed, including localized membrane remod- model may not be limited to bacterial signal molecules,
Turgor is the force that pushes elling, altered turgor pressure of the periplasmic space, as previous studies have demonstrated that antibiotics
the cytoplasmic membrane anionic charge repulsion between LPS molecules and that cause membrane perturbations, such as gentamicin
against the cell wall as a result
of the osmotic flow of water.
loss of linkage between the peptidoglycan and the and the cationic antimicrobial peptides polymyxin and
outer membrane. colistin, induce OMV formation8,31,68,69 (BOX 2).
Quorum sensing The common assumption of these models is that Another mechanism for OMV formation has been
A cell-to-cell communication OMVs protrude from regions of the outer membrane, demonstrated in Aliivibrio fischeri. This bacterium
mechanism in bacteria by
in which disrupted crosslinking between peptide gly- assembles flagella that are surrounded by a sheath
which gene regulation is
controlled in a population- can and the outer membrane leads to dissociation derived from the outer membrane. Membrane blebs
dependent manner through of the outer membrane from the peptidoglycan layer, are commonly found along the sheathed flagella, and
the production and perception for example, at the site of cell division52–54. Mutants these are released when the flagella rotate 70,71. The
of signal molecules. with defects in crosslinking produce more OMVs LPS that is associated with these OMVs is an impor-
B-band LPS
than wild-type strains55–59. However, as some of these tant trigger for morphogenesis of the light organ of its
Pseudomonas aeruginosa mutants show substantial leakage of intracellular com- symbiotic host, the Hawaiian bobtail squid, Euprymna
synthesizes two types of ponents, it cannot be ruled out that at least some of scolopes71,72. Likewise, Vibrio cholerae was shown to use
lipopolysaccharide (LPS) the increase in vesiculation is due to cell lysis. More its sheathed flagella to release OMVs carrying LPS70,
referred to as A-band and
recently, an Escherichia coli nlpI mutant with a subtle which can trigger inflammation and an immune
B-band LPS. The A-band LPS
contains a conserved defect in crosslinking (a reduction of approximately response in the host. Flagellar motility is an important
O-polysaccharide region 40% compared with wild- type cells) was shown to virulence trait of the pathogens Brucella melitensis and
composed of d-rhamnose hypervesiculate without detectable cell lysis59,60. NlpI Helicobacter pylori, whose sheathed flagella have also
(homopolymer), whereas the is an outer-membrane lipoprotein that is involved in been shown to carry membrane blebs73,74, suggesting
B-band O-antigen
(heteropolymer) structure
modulating peptidoglycan dynamics, and it was pos- that the phenomenon of flagellar-mediated LPS release
varies among different tulated that an imbalance in peptidoglycan turnover through OMVs may be widespread among bacteria
serotypes. prevented crosslinking and thus stimulated OMV with sheathed flagella.

NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | JANUARY 2019 | 17


REVIEWS

Box 2 | Antibiotics and bacterial vesicles peptidoglycan layer 14. As a consequence, the cells
round up and explode, which is why this phenomenon
Treatment of bacteria with sublethal concentrations of certain antibiotics is a well- was named explosive cell lysis. The remaining shattered
established trigger of bacterial vesicle formation. At least three mechanisms by which membrane fragments round up and self-assemble into
antibiotics stimulate membrane vesicle (MV) formation can be distinguished: cell MVs. Explosive cell lysis was shown to be stimulated
envelope stress caused by antibiotics, induction of the SOS response and inhibition of
in biofilms and under anoxic conditions14,75. Another
cell wall biosynthesis (see the figure).
study showed that, similar to P. aeruginosa, expression
Polymyxin Ciprofloxacin β-lactams of an endolysin encoded by a defective prophage trig-
gers vesicle formation in the Gram-positive bacterium
Outer-membrane DNA Cell wall B. subtilis15. Although the enzymatic activities of the
stress damage biosynthesis endolysins weaken peptidoglycan in both organisms,
SOS the consequences are different: whereas P. aeruginosa
response cells round up and explode, B. subtilis cells pro-
trude cytoplasmic membrane material through holes
in the peptidoglycan, which is released as CMVs.
OMVs OIMVs EOMVs CMVs
Furthermore, whereas P. aeruginosa cells completely
disintegrate, the thick Gram- positive cell wall of
B. subtilis is not entirely hydrolysed, although most cells
die owing to the loss of membrane integrity, as evi-
denced by the formation of ghost cells and intracellular
CMVs. This phenomenon has therefore been named
‘bubbling cell death’. The endolysins released from
Antibiotics that weaken the cell wall, such as β-lactams, stimulate cytoplasmic dying B. subtilis cells were shown to trigger CMV for-
membrane vesicle (CMV) formation in Gram-positive bacteria15,136,137. This treatment mation in neighbouring cells, as they hydrolyse the cell
will generate holes in the peptidoglycan layer through which cytoplasmic membrane wall from outside15. Similarly, MVs originating from
material can protrude into the extracellular space and be released as CMVs. It has been explosive cell lysis of Gram-negative bacteria also carry
well established that some antibiotics, particularly quinolones such as ciprofloxacin,
endolysins and therefore may lyse other cells37,68, which
induce the cellular SOS response (BOX 1), which in turn has been shown to stimulate
vesicle production14,75,138,139. As the SOS response triggers expression of endolysins
in turn could generate MVs. These studies suggest that
encoded by prophages, these antibiotics stimulate vesicle formation through lysis. phage-triggered cell death may be an underestimated
This pathway seems to be the main route for outer-inner membrane vesicle route for vesicle formation in nature. In addition to
(OIMV) production38. phage- derived endolysins, bacterial autolysins have
Antibiotics that cause cell envelope stress, such as polymyxin or gentamicin, will cause been reported to be involved in vesicle formation in
outer-membrane vesicle (OMV) formation through blebbing of the outer membrane8,140. some bacteria76,77.
A particularly interesting case is the biosynthesis of linearmycin by Streptomyces sp. Mg1 Conceivably, explosive cell lysis can generate
(REF.141). This hydrophobic compound, which has antifungal and antibacterial properties, OIMVs, as supported by the finding that ciprofloxacin
is released and transported by OMVs, and it induces OMV formation, possibly by treatment of S. maltophilia triggered phage expression,
affecting membrane curvature.
cell lysis and OIMV formation38. However, it cannot be
The fact that some antibiotics stimulate vesicle formation has important implications
ruled out that during cell explosion the inner and outer
for the choice of antibiotics for the treatment of infections. For example, Escherichia coli
strains O157:H7 and O104:H4 secrete Shiga toxin (Stx), the major virulence factor membranes of the cell envelope fragments become sep-
of these bacteria, through MVs118,119. The synthesis of Stx is associated with the arated and spontaneously re-circularize to form OMVs
induction of the Stx prophage that carries the toxin gene115. Hence, antibiotics that or possibly CMVs. Incidentally, Koning et al.78 reported
induce the SOS response not only induce Stx production but also trigger explosive that in some OIMVs of Acinetobacter baumannii the
cell lysis and thus dispersal of the toxin through MVs. For these reasons, the use of inner membrane was partially detached from the pep-
ciprofloxacin for the treatment of patients infected with enterohaemorrhagic E. coli has tidoglycan layer, which remained associated with the
been discouraged142. outer membrane. Moreover, the authors also observed
As a consequence of their formation routes, OIMVs, explosive outer-membrane OIMVs with partial inner membranes and OMV for-
vesicles (EOMVs) and CMVs carry endolysins, which can lyse and kill bacteria. They may
mation from OIMVs. Hence, explosive cell lysis will
also carry the antibiotics that triggered the SOS response, and these may also contribute
to the toxicity of these MVs. Antibiotics that intercalate into the outer membrane induce
also lead to the formation of OMVs (FIG. 2). However,
OMV formation through blebbing. These OMVs carry the inducing antibiotics and as these vesicles likely differ from classic OMVs with
therefore can kill susceptible bacteria68,96,143. Classic OMVs that are produced owing to respect to their cargo, we refer to them as EOMVs.
outer-membrane blebbing can function as decoys for envelope-targeting antibiotics
and thereby protect the bacteria8. Tube- shaped membranous structure formation.
Very little is known about the mechanisms of TSMS
Vesicle formation through endolysin-triggered cell formation. It was suggested that local lysis of the
death. Recent work has unravelled novel vesicle for- Gram- positive cell wall causes bulging of the cyto-
mation routes that are based on the enzymatic action of plasmic membrane, which, in turn, initiates nanotube
endolysins, which are typically used by double-stranded formation49. This is reminiscent of the early stage of
DNA phages to lyse their hosts so that the phage prog- endolysin- triggered CMV formation, during which
eny can be released. In the Gram-negative bacterium material from the cytoplasmic membrane protrudes
P. aeruginosa, DNA-damaging stress induces expres- through holes in the peptidoglycan layer15. However,
sion of an endolysin, which is part of a pyocin bio- whereas membrane blebs are released during CMV
synthesis gene cluster, leading to degradation of the formation, these membrane extrusions elongate during

18 | JANUARY 2019 | VOLUME 17 www.nature.com/nrmicro


REVIEWS

fluidity3,13,16. Recent work has shown that deacylation


of lipid A and the accompanying outer- membrane
Iron limitation Outer- remodelling in S. Typhimurium and phospholipid
membrane accumulation in the outer leaflet of the outer mem-
blebbing
Signalling molecules brane of H. influenzae and V. cholerae lead to hyper-
(PQS) vesiculation18,61. Iron limitation has been suggested
to control OMV formation by affecting expression of
phospholipid transporter genes in V. cholera, E. coli and
Hydrophobic compounds
H. influenzae61, but very little is known about the envi-
Explosive ronmental stimuli that affect expression of the other
Antibiotics cell lysis genes involved in OMV formation.
Growth of bacteria on highly hydrophobic carbon
sources, such as hexadecane or phenanthrene, which
DNA-damaging agents likely interact with the outer membrane, was also
reported to stimulate vesiculation 79,80. OMVs can
Peptidoglycan- Bubbling also detoxify cells from membrane- targeting com-
degrading enzymes cell death
pounds, as has been shown for Pseudomonas putida
IH-2000 (REF.81). In this example, toluene, a solvent, inter-
calates into hydrophobic regions of the outer membrane,
Fig. 3 | Different triggers that induce membrane vesicle formation. Certain growth
conditions, such as iron limitation, and antibiotics can influence the lipid homeostasis in causing a change in membrane curvature and the pro-
Gram-negative bacteria and trigger the blebbing of the outer membrane. Blebbing is duction of OMVs. Toluene was found to be enriched in
also triggered by intercalation of hydrophobic molecules including signalling molecules, OMVs compared to the cell envelope, suggesting that
such as the Pseudomonas quinolone signal (PQS), and by membrane-targeting the OMVs are used to dispose of the toxic compound
antibiotics. Antibiotics and DNA-damaging agents can induce the expression of phage- and thus protect the cell.
derived endolysin through the SOS response, and this can trigger vesicle formation
through explosive cell lysis in Gram-negative bacteria and ‘bubbling cell death’ in Gram- Factors triggering vesicle formation through
positive bacteria. Other peptidoglycan-degrading enzymes might also trigger explosive endolysin- triggered cell death. Vesicle formation
cell lysis and bubbling cell death. β-lactam antibiotics affect the integrity of the cell wall
through explosive cell lysis or bubbling cell death
and thus can induce the formation of membrane vesicles, which is similar to what
depends on the expression of phage-derived endolysins.
ultimately happens in bubbling cell death. See also BOX 2 for the effects of antibiotics
on vesicle formation. Double-stranded DNA phages use these peptidoglycan-
hydrolysing enzymes to ensure the release of newly
packaged phage particles82. To reach the peptidoglycan,
nanotube formation. It is possible that the size of the endolysins depend on small hydrophobic proteins called
hole in the peptidoglycan layer determines the fate of holins. These proteins oligomerize in the cytoplasmic
the membrane protrusion. The enzymatic activity of membrane and form pores that allow the endolysins to
endolysins may create large and more quickly expand- access the peptidoglycan83.
ing holes, which will lead to cell lysis, whereas the holes Whereas lytic phages immediately enter a produc-
for nanotube formation may instead be confined to tive cycle, temperate phages can integrate into the
small punctures in the peptidoglycan layer. bacterial host genome and are then transmitted as
prophages to daughter cells at each cell division. In this
Triggers of vesicle formation lysogenic state, the prophage does not promote cell
Certain stimuli trigger specific outer- membrane death or the production of phage particles. The pro-
vesicle formation routes. Evidence has accumulated phage can enter the lytic cycle again when the cell is
that vesicle production not only depends on the genetic stressed, typically upon exposure to DNA-damaging
background of the producing strain but also is strongly agents or ultraviolet radiation, which are conditions
influenced by the growth conditions. A plethora of envi- that activate the SOS response (BOX 1). The prophage
ronmental factors affect the rate of vesicle formation, then induces expression of lytic genes that promote
including media composition, growth phase, temper- DNA replication, phage particle assembly, DNA
ature, iron and oxygen availability, exposure to anti- packaging, host DNA degradation and bacterial lysis.
biotics and genotoxic stress3,11. In the following sections, Even in the absence of an obvious stressor, a low rate
we summarize the evidence that suggests that some of spontaneous induction of lytic production occurs.
conditions and stimuli trigger certain routes of vesicle In P. aeruginosa and E. coli, less than 1% of the popu-
formation (FIG. 3). lation induces the SOS response when grown under
optimal conditions in liquid medium14,84. However, a
Factors triggering vesicle formation through bleb- considerably higher proportion of the population was
bing. Forward genetic screens in several bacterial found to be induced when the cells were growing in
species have identified several genes that affect OMV biofilms85,86. This is in agreement with previous work
production through blebbing. In most cases, these genes that has shown that MVs are major components of
either affect envelope crosslinking, the accumulation of biofilm matrices9,10. It is noteworthy that cell lysis will
peptidoglycan fragments or misfolded proteins in the also lead to the release of extracellular DNA, which
periplasmic space, or the lipid composition of the outer is known to be an important structural component of
membrane, which determines membrane curvature and bacterial biofilms87.

NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | JANUARY 2019 | 19


REVIEWS

Peptidoglycan Cytoplasmic membrane


Susceptible cell
Decoy
Infection MV formation
Resistant cell

Phage

Phage receptor Endolysin Receptor


transfer

Fig. 4 | Roles of membrane vesicles during phage infection. Phages specifically infect cells that carry matching phage
receptors. When the phage expresses endolysin in an infected cell, the resulting membrane vesicles (MVs) also carry the
phage receptors. On the one hand, these MVs can adsorb extracellular phages and function as a decoy that protects other
cells from infection. On the other hand, when such MVs fuse with a previously resistant cell, they can transfer phage
receptors to make the cells susceptible to the phage.

Endolysin- triggered cell lysis is required for the increased both the amount of DNA associated with
release of not only phages but also phage tail- like the MVs and the transfer efficiency of plasmid DNA69.
bacteriocins in Gram-negative bacteria, as has been most MVs of several bacterial species were shown to lyse
intensively investigated for the R-type and F-type pyo- other bacteria37, and it has been suggested that these
cins of P. aeruginosa88. The initial report on explosive ‘predatory MVs’ provide a fitness benefit to the vesicle-
cell lysis as a novel route for vesicle formation was based producing population. Moreover, some bacterivorous
on the observation that SOS-dependent expression of bacteria that thrive on the degradation of macromol-
the holin (PA0614) and endolysin (PA0629) genes not ecules secrete a cocktail of antibiotics and bacteriolytic
only caused cell death but also greatly increased the enzymes to lyse other microbial cells99,100. A particularly
number of MVs produced14,75. interesting case is Lysobacter sp. XL1, which secretes
In conclusion, enzymes that degrade peptidoglycan five bacteriolytic enzymes into its surroundings to
are widespread among bacteria and can serve a wide kill prey bacteria101. At least one of these enzymes, the
variety of functions, including bacterial killing, release endopeptidase L5, has been shown to be exclusively
of phages and nutrient acquisition. Independently of secreted through OMVs and was suggested to also be
their eventual purpose, these enzymes will lyse bacteria involved in OMV formation. However, the mecha-
and thereby trigger vesicle formation. It is of note that nism by which L5 affects OMV formation remains to
these processes are induced under certain environmen- be elucidated102,103.
tal conditions, and thus, the cargo of the vesicle reflects Phage- triggered cell lysis releases not only phage
the physiological state of the cell at the time of death progeny but also OIMVs from Gram-negative bacteria
rather than being the consequence of a specific sort- and CMVs from Gram-positive bacteria. Recent work
ing mechanism. As the SOS response is normally acti- has unravelled a complex interplay between phages, bac-
vated in only a physiologically distinct subpopulation, teria and their MVs (FIG. 4). For example, B. subtilis cells
it may, at least partly, explain why MVs have a different that are sensitive to the phage SPP1 release CMVs that
protein and lipid profile from the outer membrane of carry phage receptors104. When these CMVs fuse with
living cells. phage-resistant cells, they render the cells sensitive to
the phage. This mechanism allows phages to infect non-
Functional consequences host species and thus promotes phage spread and inva-
Given their different compositions, it is possible that sion in mixed natural communities. As this facilitates
some MV types have certain functions. For example, transduction and consequently horizontal gene transfer
two traits that are typically associated with MVs, namely, among phylogenetically diverse bacteria, this phenom-
DNA transfer 30,69,89–96 and bacterial killing 37,68,96–98, enon will accelerate bacterial evolution 104. Another
depend on the presence of DNA and enzymes that aspect of this interplay is the finding that OMVs can
hydrolyse peptidoglycan in MVs. As discussed above, function as decoys that can neutralize phages8,104–107.
Bacteriocins these constituents seem to be mainly associated with Similar to phages, some antibiotics stimulate OMV for-
Protein or peptide toxins MVs originating from cell lysis, that is, EOMVs mation, although by different routes (BOX 2), and these
produced by bacteria to kill or
inhibit growth of bacteria that
and OIMVs and in Gram- positive bacteria CMVs. OMVs can neutralize environmental agents that target
are closely related to the Exposure of Acinetobacter baylyi to sublethal concen- the outer membrane of Gram-negative bacteria, such as
producer. trations of gentamicin stimulated MV production and antimicrobial peptides and bacteriophages8.

20 | JANUARY 2019 | VOLUME 17 www.nature.com/nrmicro


REVIEWS

As mentioned above, the signal molecule PQS stim- and also show specific protein expression patterns. For
ulates OMV production through intercalation into the example, it is well documented that SOS induction trig-
outer membrane. The resulting OMVs transport this gers virulence in some bacteria by inducing the produc-
molecule within a population65. Recent reports suggest tion of phage-encoded toxins, such as the cholera toxin
a general role of MVs for the release and dispersal of in V. cholera and the Shiga toxin in pathogenic E. coli
hydrophobic signal molecules, such as long-chain N- strains115–117. Intriguingly, these toxins were shown to
acylhomoserine lactones (AHLs)34,108. For Paracoccus be secreted through MVs, which deliver them to the
denitrificans, it was shown that the amount of AHLs car- eukaryotic host cells118,119. Hence, the SOS response not
ried by a single vesicle suffices to trigger the quorum- only induces production of the toxins but also triggers
sensing response of a target cell. In contrast to the explosive cell lysis and thus ensures that the toxins are
classic quorum- sensing model, which assumes free secreted and delivered by MVs.
diffusion of signal molecules to synchronize bacterial As some stimuli can trigger certain vesicle for-
activities, vesicle-mediated signalling is binary (that is, mation routes (FIG. 3), it is likely that environmental
cells either receive a vesicle or not) and would lead to niches select for particular vesicle types. Recent DNA
physiologically heterogeneous populations108. At pres- and RNA-sequencing analyses have provided evidence
ent, it is not known whether long-chain AHL signal that a considerable proportion of the MVs present
molecules can stimulate vesicle formation through in natural habitats likely originate from cell lysis by
intercalation into the outer membrane or whether they demonstrating that MVs can harbour complete viral
are just packed into MVs for their release and pro- genomes120, that DNA associated with MVs isolated
tection. It is also possible that they constitute specific from open-ocean samples is highly enriched for viral
signals that trigger a regulatory pathway controlling sequences6,121,122 and that ultraviolet radiation of water
vesiculation, as has recently been suggested for the samples, a trigger for prophage induction, greatly stim-
diffusible signal factor (DSF) of S. maltophilia and ulates vesicle formation123. Given that phages are the
Xylella fastidiosa109,110. most abundant form of life on Earth, exceeding bacte-
ria in number by tenfold124, phage-triggered cell lysis
Conclusions appears to be an important and so far underestimated
We hope that this Review will raise awareness that dif- route for vesicle formation in nature. However, the
ferent types of MVs exist and provide a guide for their sequencing of environmental MVs may overestimate
identification. We have summarized evidence that sug- the role of endolysin-triggered cell death for vesicle for-
gests that two principal vesicle formation routes exist, mation in the environment, as RNA and DNA appear
namely, membrane blebbing and endolysin-triggered to be mostly associated with OIMVs and CMVs, which
cell lysis. These two routes give rise to various types of are mainly produced through cell lysis routes. In this
MVs that differ in their compositions and possibly in context, it will be important to unravel to what extent
their functions (FIG. 2; TABLE 1). This information may explosive cell lysis triggers the formation of OMVs
stimulate future research that aims at investigating the (which we refer to as EOMVs), and whether this occurs
abundance and ecological role of different vesicle types directly or from OIMVs78, as we predict that they will
in the environment. Recent work has shown that MVs differ from OMVs formed through blebbing mecha-
have important physiological roles in coral reefs34 and nisms (TABLE 1). EOMVs can carry cytosolic contents,
a so far underestimated effect on nutrient cycles across including chromosomal DNA and endolysins that can
the biosphere6,111. Even so, many aspects of vesicle func- kill bacteria. By contrast, classic OMVs formed by bleb-
tions, including cell detoxification, killing of bacteria, bing through conventional mechanisms are unlikely to
DNA transfer and signal molecule release, in natural carry chromosomal DNA or cytosolic proteins27 but
environments remain to be explored and appear to can be enriched for hydrophobic molecules, including
be highly valuable fields for future research. It would be bacterial signals and membrane-targeting antibiotics,
particularly interesting to elucidate which function is and may show specific lipid compositions and carry
associated with a particular vesicle type. particular cargo molecules, which are often virulence
Although endolysin- triggered vesicle formation factors. In bacteria carrying phage endolysins, cell
requires the death of a small subpopulation of cells, this lysis of a subpopulation of the cells is inevitable, even
may provide a benefit to the surviving population. For under optimal exponential growth conditions. It can
example, some antibiotics induce the SOS response, therefore not be ruled out that previous studies report-
cell lysis and vesicle formation in lysogenic bacteria. ing the presence of RNA and DNA in OMVs have been
The remaining bacterial population, however, is now biased by the presence of small amounts of EOMVs
protected by the produced OMVs, as they can neutral- and OIMVs.
ize environmental agents that target the outer mem- The isolation and biochemical and structural charac-
brane of Gram- negative bacteria, including phages, terization of MVs from environmental samples will be
membrane- targeting antibiotics and eukaryotic host required to determine which types of vesicle prevail in
defence factors8,105,112–114. each particular niche. This in turn may shed light on the
It is also worth noting that expression of phage endo- environmental conditions that the cells encounter and
lysins is often controlled by the SOS response, which is may allow conclusions to be made about the possible
a genetically hardwired regulatory cascade (BOX 1). This functions of the different types of vesicles.
means that cells, in which the SOS response has been
induced before lysis, are in a special physiological state Published online 5 November 2018

NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | JANUARY 2019 | 21


REVIEWS

1. Schwechheimer, C. & Kuehn, M. J. Outer-membrane Proc. Natl Acad. Sci. USA 111, E1548–E1556 43. Baidya, A. K., Bhattacharya, S., Dubey, G. P.,
vesicles from Gram-negative bacteria: biogenesis and (2014). Mamou, G. & Ben-Yehuda, S. Bacterial nanotubes:
functions. Nat. Rev. Microbiol. 13, 605–619 (2015). 24. Wai, S. N. et al. Vesicle-mediated export and assembly a conduit for intercellular molecular trade. Curr. Opin.
This comprehensive Review presents classic models of pore-forming oligomers of the enterobacterial ClyA Microbiol. 42, 1–6 (2017).
of OMV formation and function. cytotoxin. Cell 115, 25–35 (2003). 44. Cao, P., Dey, A., Vassallo, C. N. & Wall, D.
2. Brown, L., Wolf, J. M., Prados-Rosales, R. & 25. Guerrero-Mandujano, A., Hernandez-Cortez, C., How Myxobacteria cooperate. J. Mol. Biol. 427,
Casadevall, A. Through the wall: extracellular vesicles Ibarra, J. A. & Castro-Escarpulli, G. The outer 3709–3721 (2015).
in Gram-positive bacteria, mycobacteria and fungi. membrane vesicles: secretion system type zero. Traffic 45. Pirbadian, S. et al. Shewanella oneidensis MR-1
Nat. Rev. Microbiol. 13, 620–630 (2015). 18, 425–432 (2017). nanowires are outer membrane and periplasmic
This Review discusses vesicle production by 26. Renelli, M., Matias, V., Lo, R. Y. & Beveridge, T. J. extensions of the extracellular electron transport
organisms not belonging to the Gram-negative DNA-containing membrane vesicles of Pseudomonas components. Proc. Natl Acad. Sci. USA 111,
bacteria. aeruginosa PAO1 and their genetic transformation 12883–12888 (2014).
3. Orench-Rivera, N. & Kuehn, M. J. Environmentally potential. Microbiology 150, 2161–2169 (2004). 46. Sure, S. K., Ackland, L. M., Torriero, A. A., Adholeya, A.
controlled bacterial vesicle-mediated export. 27. Zhou, L., Srisatjaluk, R., Justus, D. E. & Doyle, R. J. & Kochar, M. Microbial nanowires: an electrifying
Cell. Microbiol. 18, 1525–1536 (2016). On the origin of membrane vesicles in gram-negative tale. Microbiology 162, 2017–2028 (2016).
4. Gujrati, V. et al. Bioengineered bacterial outer bacteria. FEMS Microbiol. Lett. 163, 223–228 47. McCaig, W. D., Koller, A. & Thanassi, D. G. Production
membrane vesicles as cell-specific drug-delivery (1998). of outer membrane vesicles and outer membrane
vehicles for cancer therapy. ACS Nano 8, 1525–1537 28. Dorward, D. W. & Garon, C. F. DNA-binding proteins tubes by Francisella novicida. J. Bacteriol. 195,
(2014). in cells and membrane blebs of Neisseria 1120–1132 (2013).
5. Kaparakis-Liaskos, M. & Ferrero, R. L. Immune gonorrhoeae. J. Bacteriol. 171, 4196–4201 (1989). 48. Hampton, C. M. et al. The opportunistic pathogen
modulation by bacterial outer membrane vesicles. 29. Dorward, D. W. & Garon, C. F. DNA is packaged within Vibrio vulnificus produces outer membrane vesicles in
Nat. Rev. Immunol. 15, 375–387 (2015). membrane-derived vesicles of Gram-negative but not a spatially distinct manner related to capsular
6. Biller, S. J. et al. Bacterial vesicles in marine Gram-positive bacteria. Appl. Environ. Microbiol. 56, polysaccharide. Front. Microbiol. 8, 2177 (2017).
ecosystems. Science 343, 183–186 (2014). 1960–1962 (1990). 49. Dubey, G. P. et al. Architecture and characteristics
This study presents evidence that MVs are 30. Dorward, D. W., Garon, C. F. & Judd, R. C. Export of bacterial nanotubes. Dev. Cell 36, 453–461
abundant in open-ocean samples and and intercellular transfer of DNA via membrane (2016).
demonstrates that the DNA associated with MVs blebs of Neisseria gonorrhoeae. J. Bacteriol. 171, 50. Dubey, G. P. & Ben-Yehuda, S. Intercellular nanotubes
is highly enriched for viral sequences. 2499–2505 (1989). mediate bacterial communication. Cell 144,
7. Grande, R. et al. Helicobacter pylori ATCC 43629/ 31. Kadurugamuwa, J. L. & Beveridge, T. J. Virulence 590–600 (2011).
NCTC 11639 outer membrane vesicles (OMVs) from factors are released from Pseudomonas aeruginosa 51. Sutterlin, H. A. et al. Disruption of lipid homeostasis
biofilm and planktonic phase associated with in association with membrane vesicles during normal in the Gram-negative cell envelope activates a novel
extracellular DNA (eDNA). Front. Microbiol. 6, 1369 growth and exposure to gentamicin: a novel cell death pathway. Proc. Natl Acad. Sci. USA 113,
(2015). mechanism of enzyme secretion. J. Bacteriol. 177, E1565–E1574 (2016).
8. Manning, A. J. & Kuehn, M. J. Contribution of 3998–4008 (1995). 52. Hoekstra, D., van der Laan, J. W., de Leij, L. &
bacterial outer membrane vesicles to innate bacterial 32. Pérez-Cruz, C. et al. New type of outer membrane Witholt, B. Release of outer membrane fragments
defense. BMC Microbiol. 11, 258 (2011). vesicle produced by the Gram-negative bacterium from normally growing Escherichia coli. Biochim.
This paper demonstrates that OMVs can function Shewanella vesiculosa M7T: implications for DNA Biophys. Acta 455, 889–899 (1976).
as decoys that neutralize phages and membrane- content. Appl. Environ. Microbiol. 79, 1874–1881 53. Burdett, I. D. & Murray, R. G. Electron microscope
targeting antibiotics. (2013). study of septum formation in Escherichia coli strains B
9. Toyofuku, M., Roschitzki, B., Riedel, K. & Eberl, L. 33. Pérez-Cruz, C., Delgado, L., López-Iglesias, C. & and B/r during synchronous growth. J. Bacteriol. 119,
Identification of proteins associated with the Mercade, E. Outer-inner membrane vesicles naturally 1039–1056 (1974).
Pseudomonas aeruginosa biofilm extracellular matrix. secreted by gram-negative pathogenic bacteria. 54. Deatherage, B. L. et al. Biogenesis of bacterial
J. Proteome Res. 11, 4906–4915 (2012). PLOS ONE 10, e0116896 (2015). membrane vesicles. Mol. Microbiol. 72, 1395–1407
10. Schooling, S. R. & Beveridge, T. J. Membrane vesicles: 34. Li, J., Azam, F. & Zhang, S. Outer membrane vesicles (2009).
an overlooked component of the matrices of biofilms. containing signalling molecules and active hydrolytic 55. Bernadac, A., Gavioli, M., Lazzaroni, J. C., Raina, S. &
J. Bacteriol. 188, 5945–5957 (2006). enzymes released by a coral pathogen Vibrio shilonii Lloubés, R. Escherichia coli tol-pal mutants form outer
11. Kulp, A. & Kuehn, M. J. Biological functions and AK1. Environ. Microbiol. 18, 3850–3866 (2016). membrane vesicles. J. Bacteriol. 180, 4872–4878
biogenesis of secreted bacterial outer membrane 35. Hagemann, S. et al. DNA-bearing membrane vesicles (1998).
vesicles. Annu. Rev. Microbiol. 64, 163–184 (2010). produced by Ahrensia kielensis and 56. Sonntag, I., Schwarz, H., Hirota, Y. & Henning, U.
12. Beveridge, T. J. Structures of gram-negative cell walls Pseudoalteromonas marina. J. Basic Microbiol. 54, Cell envelope and shape of Escherichia coli: multiple
and their derived membrane vesicles. J. Bacteriol. 1062–1072 (2014). mutants missing the outer membrane lipoprotein and
181, 4725–4733 (1999). 36. Kadurugamuwa, J. L. & Beveridge, T. J. Natural other major outer membrane proteins. J. Bacteriol.
13. Mashburn-Warren, L. M. & Whiteley, M. Special release of virulence factors in membrane vesicles 136, 280–285 (1978).
delivery: vesicle trafficking in prokaryotes. Mol. by Pseudomonas aeruginosa and the effect of 57. Yem, D. W. & Wu, H. C. Physiological characterization
Microbiol. 61, 839–846 (2006). aminoglycoside antibiotics on their release. of an Escherichia coli mutant altered in the structure
14. Turnbull, L. et al. Explosive cell lysis as a mechanism J. Antimicrob. Chemother. 40, 615–621 (1997). of murein lipoprotein. J. Bacteriol. 133, 1419–1426
for the biogenesis of bacterial membrane vesicles and 37. Li, Z., Clarke, A. J. & Beveridge, T. J. Gram-negative (1978).
biofilms. Nat. Commun. 7, 11220 (2016). bacteria produce membrane vesicles which are 58. Murata, M., Noor, R., Nagamitsu, H., Tanaka, S. &
This paper demonstrates for the first time that capable of killing other bacteria. J. Bacteriol. 180, Yamada, M. Novel pathway directed by sigma E to
MVs can be formed as a consequence of explosive 5478–5483 (1998). cause cell lysis in Escherichia coli. Genes Cells 17,
cell lysis, which is induced by the expression of a 38. Devos, S. et al. Membrane vesicle secretion and 234–247 (2012).
phage endolysin. prophage induction in multidrug-resistant 59. McBroom, A. J., Johnson, A. P., Vemulapalli, S.
15. Toyofuku, M. et al. Prophage-triggered membrane Stenotrophomonas maltophilia in response to & Kuehn, M. J. Outer membrane vesicle production
vesicle formation through peptidoglycan damage in ciprofloxacin stress. Environ. Microbiol. 19, by Escherichia coli is independent of membrane
Bacillus subtilis. Nat. Commun. 8, 481 (2017). 3930–3937 (2017). instability. J. Bacteriol. 188, 5385–5392 (2006).
This study shows that CMV production in the Gram- This study shows that treatment of S. maltophilia 60. Schwechheimer, C., Rodriguez, D. L. & Kuehn, M. J.
positive bacterium B. subtilis is triggered by the with ciprofloxacin, which is known to induce the NlpI-mediated modulation of outer membrane vesicle
expression of a phage endolysin, a phenomenon SOS response owing to DNA damage, stimulates production through peptidoglycan dynamics in
named ‘bubbling cell death’. the production of not only OIMVs but also large Escherichia coli. MicrobiologyOpen 4, 375–389
16. Roier, S., Zingl, F. G., Cakar, F. & Schild, S. Bacterial amounts of phages. (2015).
outer membrane vesicle biogenesis: a new mechanism 39. Shingaki, R., Kasahara, Y., Inoue, T., Kokeguchi, S. 61. Roier, S. et al. A novel mechanism for the biogenesis
and its implications. Microb. Cell 3, 257–259 (2016). & Fukui, K. Chromosome DNA fragmentation and of outer membrane vesicles in Gram-negative
17. Kulp, A. J. et al. Genome-wide assessment of outer excretion caused by defective prophage gene bacteria. Nat. Commun. 7, 10515 (2016).
membrane vesicle production in Escherichia coli. expression in the early-exponential-phase culture of This study demonstrates that iron limitation
PLOS ONE 10, e0139200 (2015). Bacillus subtilis. Can. J. Microbiol. 49, 313–325 controls OMV formation by affecting expression of
18. Elhenawy, W. et al. LPS remodeling triggers formation (2003). phospholipid transporter genes in H. influenzae.
of outer membrane vesicles in Salmonella. mBio 7, 40. Küsel, K., Dorsch, T., Acker, G. & Stackebrandt, E. 62. McBroom, A. J. & Kuehn, M. J. Release of outer
e00940-16 (2016). Microbial reduction of Fe(III) in acidic sediments: membrane vesicles by Gram-negative bacteria is a
19. Bager, R. J. et al. Outer membrane vesicles reflect isolation of Acidiphilium cryptum JF-5 capable of novel envelope stress response. Mol. Microbiol. 63,
environmental cues in Gallibacterium anatis. coupling the reduction of Fe(III) to the oxidation of 545–558 (2007).
Vet. Microbiol. 167, 565–572 (2013). glucose. Appl. Environ. Microbiol. 65, 3633–3640 63. Tashiro, Y. et al. Outer membrane machinery and
20. Koeppen, K. et al. A novel mechanism of host-pathogen (1999). alginate synthesis regulators control membrane
interaction through sRNA in bacterial outer membrane 41. Remis, J. P. et al. Bacterial social networks: structure vesicle production in Pseudomonas aeruginosa.
vesicles. PLOS Pathog. 12, e1005672 (2016). and composition of Myxococcus xanthus outer J. Bacteriol. 191, 7509–7519 (2009).
21. Sjöström, A. E., Sandblad, L., Uhlin, B. E. & Wai, S. N. membrane vesicle chains. Environ. Microbiol. 16, 64. Florez, C., Raab, J. E., Cooke, A. C. & Schertzer, J. W.
Membrane vesicle-mediated release of bacterial RNA. 598–610 (2014). Membrane distribution of the Pseudomonas
Sci. Rep. 5, 15329 (2015). This study uses several imaging techniques to quinolone signal modulates outer membrane vesicle
22. Bitto, N. J. et al. Bacterial membrane vesicles visualize chains of OMVs in M. xanthus biofilms. production in Pseudomonas aeruginosa. mBio 8,
transport their DNA cargo into host cells. Sci. Rep. 7, 42. Wei, X., Vassallo, C. N., Pathak, D. T. & Wall, D. e01034-17 (2017).
7072 (2017). Myxobacteria produce outer membrane-enclosed 65. Mashburn, L. M. & Whiteley, M. Membrane vesicles
23. Altindis, E., Fu, Y. & Mekalanos, J. J. Proteomic tubes in unstructured environments. J. Bacteriol. traffic signals and facilitate group activities in a
analysis of Vibrio cholerae outer membrane vesicles. 196, 1807–1814 (2014). prokaryote. Nature 437, 422–425 (2005).

22 | JANUARY 2019 | VOLUME 17 www.nature.com/nrmicro


REVIEWS

66. Schertzer, J. W. & Whiteley, M. A bilayer-couple model 89. Domingues, S. & Nielsen, K. M. Membrane vesicles bacteremia by recognition of the outer membrane
of bacterial outer membrane vesicle biogenesis. mBio and horizontal gene transfer in prokaryotes. protein U. J. Innate Immun. 8, 269–283 (2016).
3, e00297-11 (2012). Curr. Opin. Microbiol. 38, 16–21 (2017). 113. Duperthuy, M. et al. Role of the Vibrio cholerae matrix
67. Mashburn-Warren, L. et al. Interaction of quorum 90. Ho, M. H., Chen, C. H., Goodwin, J. S., Wang, B. Y. & protein Bap1 in cross-resistance to antimicrobial
signals with outer membrane lipids: insights into Xie, H. Functional advantages of Porphyromonas peptides. PLOS Pathog. 9, e1003620 (2013).
prokaryotic membrane vesicle formation. gingivalis vesicles. PLOS ONE 10, e0123448 (2015). 114. Codemo, M. et al. Immunomodulatory effects of
Mol. Microbiol. 69, 491–502 (2008). 91. Kolling, G. L. & Matthews, K. R. Export of virulence pneumococcal extracellular vesicles on cellular and
68. Kadurugamuwa, J. L. & Beveridge, T. J. Bacteriolytic genes and Shiga toxin by membrane vesicles of humoral host defenses. mBio 9, e00559-18 (2018).
effect of membrane vesicles from Pseudomonas Escherichia coli O157:H7. Appl. Environ. Microbiol. 115. Kimmitt, P. T., Harwood, C. R. & Barer, M. R. Toxin
aeruginosa on other bacteria including pathogens: 65, 1843–1848 (1999). gene expression by shiga toxin-producing Escherichia
conceptually new antibiotics. J. Bacteriol. 178, 92. Yaron, S., Kolling, G. L., Simon, L. & Matthews, K. R. coli: the role of antibiotics and the bacterial SOS
2767–2774 (1996). Vesicle-mediated transfer of virulence genes from response. Emerg. Infect. Dis. 6, 458–465 (2000).
69. Fulsundar, S. et al. Gene transfer potential of outer Escherichia coli O157:H7 to other enteric bacteria. 116. Quinones, M., Kimsey, H. H. & Waldor, M. K.
membrane vesicles of Acinetobacter baylyi and effects Appl. Environ. Microbiol. 66, 4414–4420 (2000). LexA cleavage is required for CTX prophage induction.
of stress on vesiculation. Appl. Environ. Microbiol. 80, 93. Rumbo, C. et al. Horizontal transfer of the OXA-24 Mol. Cell 17, 291–300 (2005).
3469–3483 (2014). carbapenemase gene via outer membrane vesicles: 117. Chatterjee, D. & Chaudhuri, K. Association of cholera
70. Brennan, C. A. et al. A model symbiosis reveals a role a new mechanism of dissemination of carbapenem toxin with Vibrio cholerae outer membrane vesicles
for sheathed-flagellum rotation in the release of resistance genes in Acinetobacter baumannii. which are internalized by human intestinal epithelial
immunogenic lipopolysaccharide. eLife 3, e01579 Antimicrob. Agents Chemother. 55, 3084–3090 (2011). cells. FEBS Lett. 585, 1357–1362 (2011).
(2014). 94. Blesa, A. & Berenguer, J. Contribution of vesicle- 118. Bielaszewska, M. et al. Host cell interactions of outer
71. Aschtgen, M. S. et al. Rotation of Vibrio fischeri protected extracellular DNA to horizontal gene transfer membrane vesicle-associated virulence factors of
flagella produces outer membrane vesicles that induce in Thermus spp. Int. Microbiol. 18, 177–187 (2015). enterohemorrhagic Escherichia coli O157:
host development. J. Bacteriol. 198, 2156–2165 95. Klieve, A. V. et al. Naturally occurring DNA transfer intracellular delivery, trafficking and mechanisms of
(2016). system associated with membrane vesicles in cell injury. PLOS Pathog. 13, e1006159 (2017).
This study identifies a unique mechanism of OMV cellulolytic Ruminococcus spp. of ruminal origin. 119. Kunsmann, L. et al. Virulence from vesicles: Novel
formation that is based on the rotation of Appl. Environ. Microbiol. 71, 4248–4253 (2005). mechanisms of host cell injury by Escherichia coli
membrane-sheathed flagella. 96. Tashiro, Y. et al. Interaction of bacterial membrane O104:H4 outbreak strain. Sci. Rep. 5, 13252 (2015).
72. Aschtgen, M. S., Wetzel, K., Goldman, W., vesicles with specific species and their potential for 120. Gaudin, M. et al. Extracellular membrane vesicles
McFall-Ngai, M. & Ruby, E. Vibrio fischeri-derived delivery to target cells. Front. Microbiol. 8, 571 (2017). harbouring viral genomes. Environ. Microbiol. 16,
outer membrane vesicles trigger host development. 97. Li, Z., Clarke, A. J. & Beveridge, T. J. A major autolysin 1167–1175 (2014).
Cell. Microbiol. 18, 488–499 (2016). of Pseudomonas aeruginosa: subcellular distribution, This study shows that MVs can harbour viral
73. Geis, G., Suerbaum, S., Forsthoff, B., Leying, H. & potential role in cell growth and division and secretion genomes, suggesting a link between phage release
Opferkuch, W. Ultrastructure and biochemical studies in surface membrane vesicles. J. Bacteriol. 178, and vesicle formation.
of the flagellar sheath of Helicobacter pylori. J. Med. 2479–2488 (1996). 121. Biller, S. J. et al. Membrane vesicles in sea water:
Microbiol. 38, 371–377 (1993). 98. Prangishvili, D. et al. Sulfolobicins, specific heterogeneous DNA content and implications for viral
74. Qin, Z., Lin, W. T., Zhu, S., Franco, A. T. & Liu, J. proteinaceous toxins produced by strains of the abundance estimates. ISME J. 11, 394–404 (2017).
Imaging the motility and chemotaxis machineries extremely thermophilic archaeal genus Sulfolobus. 122. Soler, N., Krupovic, M., Marguet, E. & Forterre, P.
in Helicobacter pylori by cryo-electron tomography. J. Bacteriol. 182, 2985–2988 (2000). Membrane vesicles in natural environments: a major
J. Bacteriol. 199, e00695–16 (2016). 99. Seccareccia, I., Kost, C. & Nett, M. Quantitative challenge in viral ecology. ISME J. 9, 793–796 (2015).
75. Toyofuku, M. et al. Membrane vesicle formation is analysis of Lysobacter predation. Appl. Environ. 123. Gamalier, J. P., Silva, T. P., Zarantonello, V., Dias, F. F.
associated with pyocin production under denitrifying Microbiol. 81, 7098–7105 (2015). & Melo, R. C. Increased production of outer
conditions in Pseudomonas aeruginosa PAO1. 100. Casida, L. E. Minireview: Nonobligate bacterial membrane vesicles by cultured freshwater bacteria in
Environ. Microbiol. 16, 2927–2938 (2014). predation of bacteria in soil. Microb. Ecol. 15, 1–8 response to ultraviolet radiation. Microbiol. Res. 194,
76. Wang, X., Thompson, C. D., Weidenmaier, C. & Lee, J. C. (1988). 38–46 (2017).
Release of Staphylococcus aureus extracellular 101. Vasilyeva, N. V., Tsfasman, I. M., Suzina, N. E., 124. Breitbart, M. & Rohwer, F. Here a virus, there a virus,
vesicles and their application as a vaccine platform. Stepnaya, O. A. & Kulaev, I. S. Secretion of everywhere the same virus? Trends Microbiol. 13,
Nat. Commun. 9, 1379 (2018). bacteriolytic endopeptidase L5 of Lysobacter sp. XL1 278–284 (2005).
77. Hayashi, J., Hamada, N. & Kuramitsu, H. K. into the medium by means of outer membrane 125. Radman, M. SOS repair hypothesis: phenomenology
The autolysin of Porphyromonas gingivalis is involved vesicles. FEBS J. 275, 3827–3835 (2008). of an inducible DNA repair which is accompanied by
in outer membrane vesicle release. FEMS Microbiol. 102. Kudryakova, I. V., Suzina, N. E., Vinokurova, N. G., mutagenesis. Basic Life Sci. 5A, 355–367 (1975).
Lett. 216, 217–222 (2002). Shishkova, N. A. & Vasilyeva, N. V. Studying factors 126. Kenyon, C. J. & Walker, G. C. DNA-damaging agents
78. Koning, R. I. et al. Cryo-electron tomography analysis involved in biogenesis of Lysobacter sp. XL1 outer stimulate gene-expression at specific loci in Escherichia
of membrane vesicles from Acinetobacter baumannii membrane vesicles. Biochemistry (Mosc.) 82, coli. Proc. Natl Acad. Sci. USA 77, 2819–2823 (1980).
ATCC19606 T. Res. Microbiol. 164, 397–405 (2013). 501–509 (2017). 127. Fernandez De Henestrosa, A. R. et al. Identification of
79. Shetty, A. & Hickey, W. J. Effects of outer membrane 103. Kudryakova, I. V., Suzina, N. E. & Vasilyeva, N. V. additional genes belonging to the LexA regulon in
vesicle formation, surface-layer production and nanopod Biogenesis of Lysobacter sp. XL1 vesicles. FEMS Escherichia coli. Mol. Microbiol. 35, 1560–1572
development on the metabolism of phenanthrene by Microbiol. Lett. 362, fnv137 (2015). (2000).
Delftia acidovorans Cs1-4. PLOS ONE 9, e92143 104. Tzipilevich, E., Habusha, M. & Ben-Yehuda, S. 128. McPartland, A., Green, L. & Echols, H. Control of recA
(2014). Acquisition of phage sensitivity by bacteria through gene RNA in E. coli: regulatory and signal genes. Cell
80. Borneleit, P., Hermsdorf, T., Claus, R., Walther, P. & exchange of phage receptors. Cell 168, 186–199 20, 731–737 (1980).
Kleber, H. P. Effect of hexadecane-induced vesiculation (2017). 129. Little, J. W., Edmiston, S. H., Pacelli, L. Z.
on the outer membrane of Acinetobacter The study demonstrates that CMVs can transmit & Mount, D. W. Cleavage of the Escherichia coli LexA
calcoaceticus. J. Gen. Microbiol. 134, 1983–1992 phage receptors to phage-resistant cells, which protein by the RecA protease. Proc. Natl Acad. Sci.
(1988). then become phage sensitive. USA 77, 3225–3229 (1980).
81. Kobayashi, H., Uematsu, K., Hirayama, H. & 105. Kharina, A. et al. Temperate bacteriophages collected 130. Michel, B. After 30 years of study, the bacterial SOS
Horikoshi, K. Novel toluene elimination system in by outer membrane vesicles in Komagataeibacter response still surprises us. PLOS Biol. 3, e255 (2005).
a toluene-tolerant microorganism. J. Bacteriol. 182, intermedius. J. Basic Microbiol. 55, 509–513 (2015). 131. Tippin, B., Pham, P. & Goodman, M. F. Error-prone
6451–6455 (2000). 106. Manning, A. J. & Kuehn, M. J. Functional advantages replication for better or worse. Trends Microbiol. 12,
82. Feiner, R. et al. A new perspective on lysogeny: conferred by extracellular prokaryotic membrane 288–295 (2004).
prophages as active regulatory switches of bacteria. vesicles. J. Mol. Microbiol. Biotechnol. 23, 131–141 132. Fuchs, R. P. & Fujii, S. Translesion DNA synthesis and
Nat. Rev. Microbiol. 13, 641–650 (2015). (2013). mutagenesis in prokaryotes. Cold Spring Harb.
83. Catalão, M. J., Gil, F., Moniz-Pereira, J., São-José, C. 107. Reyes-Robles, T. et al. Vibrio cholerae outer membrane Perspect. Biol. 5, a012682 (2013).
& Pimentel, M. Diversity in bacterial lysis systems: vesicles inhibit bacteriophage infection. J. Bacteriol. 133. Schoemaker, J. M., Gayda, R. C. & Markovitz, A.
bacteriophages show the way. FEMS Microbiol. Rev. https://doi.org/10.1128/JB.00792-17 (2018). Regulation of cell division in Escherichia coli: SOS
37, 554–571 (2013). 108. Toyofuku, M. et al. Membrane vesicle-mediated induction and cellular location of the SulA protein,
84. Pennington, J. M. & Rosenberg, S. M. Spontaneous bacterial communication. ISME J. 11, 1504–1509 a key to lon-associated filamentation and death.
DNA breakage in single living Escherichia coli cells. (2017). J. Bacteriol. 158, 551–561 (1984).
Nat. Genet. 39, 797–802 (2007). 109. Devos, S. et al. The effect of imipenem and diffusible 134. Little, J. W. & Harper, J. E. Identification of the lexA
85. Taddei, F., Matic, I. & Radman, M. cAMP-dependent signaling factors on the secretion of outer membrane gene product of Escherichia coli K-12. Proc. Natl
SOS induction and mutagenesis in resting bacterial vesicles and associated Ax21 proteins in Acad. Sci. USA 76, 6147–6151 (1979).
populations. Proc. Natl Acad. Sci. USA 92, Stenotrophomonas maltophilia. Front. Microbiol. 6, 135. Craig, N. L. & Roberts, J. W. E. coli RecA protein-
11736–11740 (1995). 298 (2015). directed cleavage of phage lambda repressor requires
86. Bernier, S. P. et al. Starvation, together with the SOS 110. Ionescu, M. et al. Xylella fastidiosa outer membrane polynucleotide. Nature 283, 26–30 (1980).
response, mediates high biofilm-specific tolerance vesicles modulate plant colonization by blocking 136. Biagini, M. et al. The human pathogen Streptococcus
to the fluoroquinolone ofloxacin. PLOS Genet. 9, attachment to surfaces. Proc. Natl Acad. Sci. USA 111, pyogenes releases lipoproteins as lipoprotein-rich
e1003144 (2013). E3910–E3918 (2014). membrane vesicles. Mol. Cell. Proteomics 14,
87. Okshevsky, M. & Meyer, R. L. The role of extracellular 111. Lynch, J. B. & Alegado, R. A. Spheres of hope, packets 2138–2149 (2015).
DNA in the establishment, maintenance and of doom: the good and bad of outer membrane 137. Wichgers Schreur, P. J., Rebel, J. M., Smits, M. A.,
perpetuation of bacterial biofilms. Crit. Rev. Microbiol. vesicles in interspecies and ecological dynamics. van Putten, J. P. & Smith, H. E. Lgt processing is an
41, 341–352 (2015). J. Bacteriol. 199, e00012-17 (2017). essential step in Streptococcus suis lipoprotein
88. Scholl, D. Phage tail-like bacteriocins. Annu. Rev. Virol. 112. Aung, K. M. et al. Naturally occurring IgG antibodies mediated innate immune activation. PLOS ONE 6,
4, 453–467 (2017). provide innate protection against Vibrio cholerae e22299 (2011).

NATURE REVIEWS | MICROBIOLOGY VOLUME 17 | JANUARY 2019 | 23


REVIEWS

138. Maredia, R. et al. Vesiculation from Pseudomonas 142. Freedman, S. B. et al. Shiga toxin–producing Author contributions
aeruginosa under SOS. Sci. World J. 2012, 402919 Escherichia coli infection, antibiotics, and risk of M.T. and L.E. researched data for the article and contributed
(2012). developing hemolytic uremic syndrome: a meta- substantially to discussion of the content. All authors wrote
139. Bauwens, A., Kunsmann, L., Karch, H., Mellmann, A. analysis. Clin. Infect. Dis. 62, 1251–1258 (2016). the article and reviewed and edited the manuscript before
& Bielaszewska, M. Antibiotic-mediated modulations 143. MacDonald, K. L. & Beveridge, T. J. Bactericidal effect submission.
of outer membrane vesicles in enterohemorrhagic of gentamicin-induced membrane vesicles derived from
Escherichia coli O104:H4 and O157:H7. Pseudomonas aeruginosa PAO1 on Gram-positive Competing interest
Antimicrob. Agents Chemother. 61, e00937-17 bacteria. Can. J. Microbiol. 48, 810–820 (2002). The authors declare no competing interests.
(2017).
140. Kadurugamuwa, J. L., Clarke, A. J. & Beveridge, T. J. Acknowledgements
Surface action of gentamicin on Pseudomonas M.T. was supported by the Ministry of Education, Culture, Publisher’s note
aeruginosa. J. Bacteriol. 175, 5798–5805 Sports, Science and Technology of Japan (MEXT) (project Springer Nature remains neutral with regard to jurisdictional
(1993). 16H06189), N.N. was supported by the Japan Science and claims in published maps and institutional affiliations.
141. Hoefler, B. C. et al. A link between linearmycin Technology Agency (JST) (ERATO project JPMJER1502),
biosynthesis and extracellular vesicle genesis connects and L.E. was supported by the Swiss National Science Reviewer information
specialized metabolism and bacterial membrane Foundation (SNSF) (Project 31003A_169307). The authors Nature Reviews Microbiology thanks S. Schild, S. N. Wai and
physiology. Cell Chem. Biol. 24, 1238–1249.e7 acknowledge K. Agnoli-Antkowiak for valuable comments on the other anonymous reviewer(s) for their contribution to the
(2017). the manuscript. peer review of this work.

24 | JANUARY 2019 | VOLUME 17 www.nature.com/nrmicro

Вам также может понравиться