Biofilm Review2004

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/8661019
Bacterial Bioﬁlms: From the Natural Environment to Infectious Diseases
Article in Nature Reviews Microbiology · March 2004

DOI: 10.1038/nrmicro821 · Source: PubMed
CITATIONS READS
6,152 14,588
3 authors, including:
Luanne Hall-Stoodley Paul Stoodley

The Ohio State University The Ohio State University
110 PUBLICATIONS 16,102 CITATIONS 403 PUBLICATIONS 33,873 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Biofilms in upper respiratory infections View project
Biofilm Conceptual Models View project
All content following this page was uploaded by Paul Stoodley on 16 May 2014.
The user has requested enhancement of the downloaded file.

REVIEWS
BACTERIAL BIOFILMS: FROM

THE NATURAL ENVIRONMENT
TO INFECTIOUS DISEASES
Luanne Hall-Stoodley*‡§, J. William Costerton§ and Paul Stoodley ||
Biofilms — matrix-enclosed microbial accretions that adhere to biological or non-biological surfaces
— represent a significant and incompletely understood mode of growth for bacteria. Biofilm
formation appears early in the fossil record (~3.25 billion years ago) and is common throughout a
diverse range of organisms in both the Archaea and Bacteria lineages, including the ‘living fossils’ in
the most deeply dividing branches of the phylogenetic tree. It is evident that biofilm formation is an
ancient and integral component of the prokaryotic life cycle, and is a key factor for survival in diverse
environments. Recent advances show that biofilms are structurally complex, dynamic systems with
attributes of both primordial multicellular organisms and multifaceted ecosystems. Biofilm formation
represents a protected mode of growth that allows cells to survive in hostile environments and also
disperse to colonize new niches. The implications of these survival and propagative mechanisms in
the context of both the natural environment and infectious diseases are discussed in this review.
What is of all things most yielding In this review, we will discuss the role of biofilm
Can overcome that which is most hard formation in the survival of prokaryotes in the natural
Lao Tzu
environment, and how these strategies are used for
*Department of Veterinary survival in the modern human world and in humans
Molecular Microbiology, Among the advances in microbiology that have taken themselves.
Biomedical Research
Building, 960 Technology place over the past 50 years, the realization of the
Park, Montana State extent to which microbial growth and development Biofilm formation in the environment
University, Bozeman, occurs on surfaces in complex communities has been Biofilms appear early in the fossil record. There is
Montana 59718, USA. one of the most subtle. Nevertheless, this behaviour evidence of biofilm formation early in the fossil
‡
Departments of
Microbiology and Civil
has profound consequences for how prokaryotic sur- record, particularly in hydrothermal environments.
Engineering, §Center for vival is viewed in both nature and disease. By the mid- Putative biofilm microcolonies have been identified
Biofilm Engineering, 366 twentieth century, Claude Zobell1 and others had by morphology in the 3.3–3.4-billion-year-old South
EPS Building, Montana noted that aquatic bacteria were more numerous on African Kornberg formation 4 and filamentous
State University, Bozeman, the solid surfaces of sample containers than as single, biofilms have been identified in the 3.2-billion-year-
Montana 59717, USA.
||
Center for Genomic suspended cells. Since then, the combination of high- old deep-sea hydrothermal rocks of the Pilbara
Sciences, Allegheny–Singer resolution three-dimensional imaging techniques, Craton, Australia5. Similar biofilm structures can be
Research Institute, 320 East specific molecular fluorescent stains, molecular-reporter found in modern hydrothermal environments, such
North Avenue, Pittsburgh, technology and BIOFILM-culturing apparatus2,3 has shown as hot springs 6 and deep-sea vents 7. Interestingly,
Pennsylvania 15212, USA.
Correspondence to P.S.
that biofilms are not simply passive assemblages of cells biofilm formation is also a characteristic of prokary-
e-mail: [email protected] that are stuck to surfaces, but are structurally and otic ‘living fossils’ in the most ancient lineages of the
doi:10.1038/nrmicro821 dynamically complex biological systems (FIG. 1). phylogenetic tree in both the Archaea and Bacteria —
NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 9 5

REVIEWS
7 Although antimicrobials damage

outer cell layers, the biofilm
community is resistant.
1 Starvation can induce bacteria

to shrink and adopt a spore-like 8 Propelled by shear forces, aggregated
state, known as ultramicrobacteria, cells can become detached, or roll or
which wait in water, soil, rock or ripple along a surface in sheets and
tissue until conditions are remain in their protected biofilm state.
suitable for active growth.
6 Chemical gradients create
microenvironments for
different microbial species
or levels of activity.
2 Active bacteria can attach to almost

any surface. Changes in gene
‘Persisters’
expression transform ‘swimmers’ to
‘stickers’ within minutes.
5 The close proximity of cells

4 Nutrients diffuse into in the matrix facilitates the
the matrix. exchange of molecular signals
that regulate behaviour.
‘Dispersers’
3 Attached bacteria multiply and
encase the colonies with a
slimy matrix. ‘Wall formers’
Figure 1 | Conceptualization of biofilm development and dynamic behaviours. The figure was compiled from laboratory and natural observations of pure
culture (both Gram-positive and Gram-negative organisms), and mixed-culture biofilms. (For an interactive web-based version of Figure 1 and biofilm movies showing
dynamic processes of growth and detachment, rolling and rippling, see the Online links). Image courtesy of P. Dirckx, Center for Biofilm Engineering, USA.
the Korarchaeota and Aquificales respectively 8,9. Adaptation of biofilm structure for survival in varying
Taken together, the data indicate that the ability to environments. Intriguingly, the visual characteristics
form biofilms is an ancient and integral characteristic of biofilms growing in diverse environments are
of prokaryotes. In the context of evolution and adap- strikingly similar, indicating there are important con-
tation it is likely that biofilms provided homeostasis vergent survival strategies that are conferred in part by
in the face of the fluctuating and harsh conditions of structural specialization (FIG. 2).
the primitive earth (extreme temperatures, pH and Biofilms growing in fast-moving water tend to form
exposure to ultraviolet (UV) light), thereby facilitating filamentous STREAMERS regardless of whether they occur
the development of complex interactions between in the drainage run-off from acid mines12, in hydro-
BIOFILM individual cells and providing an environment which thermal photosynthetic mats (algal or bacterial)6 or as
Microbial biofilms are was sufficient for the development of signalling PERIPHYTON in rivers (FIG. 2). In quiescent waters, biofilms
populations of microorganisms pathways and chemotactic motility10. In addition to tend to form mushroom or mound-like structures that
that are concentrated at an facilitating cell–cell interactions that require close are similar to those of STROMATOLITES. The overall patterns
interface (usually solid–liquid)
and typically surrounded by an
proximity, surfaces can also concentrate nutrients11. are isotropic with no obvious indication of flow direction.
extracellular polymeric substance It is generally assumed that PLANKTONIC CELLS occurred The structure of biofilms also changes with nutrient
(EPS) matrix. Aggregates of cells before the development of more complex biofilm com- conditions13,14. The ability of prokaryotes to adopt dif-
not attached to a surface are munities. However, we hypothesize that the catalytic ferent biofilm structures in response to environmental
sometimes termed ‘flocs’ and
and protective conditions offered by life on surfaces conditions — owing to genetic regulation15, selection13,
have many of the same
characteristics as biofilms. might have led to the concurrent development of or both16, or to localized growth patterns determined by
17
both SESSILE and planktonic forms in biofilm cellular MASS TRANSFER — gives them the flexibility to rapidly
PLANKTONIC CELLS communities 10. This concept of biofilms not only adapt to an extent that is not possible in multicellular
Planktonic (or suspended) cell enhances our existing understanding of prokaryotic eukaryotic organisms. The proclivity of bacteria to
cultures are those grown
primarily as single cells in
behaviour, but also our control strategies against the adhere to surfaces and form biofilms in so many envi-
suspension, either in a renowned tenacity of biofilms that has been acquired ronments is undoubtedly related to the selective
chemostat or a shake flask. through billions of years of evolutionary adaptation. advantage that surface association offers.
96 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

REVIEWS
a b i
c d
e f
g h
Figure 2 | Structural similarity of biofilms growing in hydrothermal hot springs, freshwater rivers and laboratory flow cells. Similar structures are seen in
biofilms growing in hydrothermal hot springs (a–c), biofilms growing in freshwater rivers (d,e) and laboratory flow cells (f–h). Biofilms growing in quiescent or low-shear
environments tend to form circular structures, such as ‘mushrooms’ or mounds (a,d,g). Biofilm streamers (b,e,h,i) and ripple structures (c,f) form in faster, high-shear
flows. Biofilms are from the Biscuit Basin thermal area, Yellowstone National Park, USA (a–c), Gardener River, Yellowstone National Park, USA (d), Hyalite Creek,
Bozeman, Montana, USA (e), mixed species biofilm grown at a flow velocity of 1 m s–1 (REF. 14) and a Pseudomonas aeruginosa PAN067 biofilm grown in a flow cell
with a flow of 0.03 m s–1 (g) or 1 m s–1 (h, i) (REF. 43). Panel c is modified with permission from REF. 53 © (2004) ASM Press.
Biofilm development development was not simple and uniform, but rather
The developmental sequence. The structural and devel- more complex and differentiated. The ability of chan-
opmental complexity of biofilms, and its significance in nels to facilitate efficient nutrient uptake by infusing
SESSILE both natural and man-made environments, has been fluid from the bulk phase into the biofilm20, thereby
In an ecological context, a increasingly appreciated over the past two decades optimizing nutrient and waste-product exchange,
stationary organism such as a
owing to the concomitant development of sophisticated provided the first link between form and function19.
plant or a barnacle. In biofilm
microbiology, it is used to imaging and molecular techniques that have identified More recently, proteomic studies have indicated
distinguish planktonic (free- the mechanisms that are involved in biofilm develop- that biofilm formation in Pseudomonas aeruginosa
floating) prokaryotic cells from ment. In situ observations of biofilm structure using proceeds as a regulated developmental sequence, and
those attached to surfaces. confocal laser microscopy showed sessile bacteria grow- five stages have been proposed10,15. Stages one and two
However, new evidence shows
that these ‘sessile’ cells are often
ing in heterogeneous matrix-enclosed microcolonies are generally identified by a loose or transient associa-
dynamic, at least on the interspersed with open water channels18,19. This complex tion with the surface, followed by robust ADHESION.
microscopic scale. architecture was one of the first indications that biofilm Stages three and four involve the aggregation of cells into

REVIEWS
microcolonies and subsequent growth and maturation. used extensively to identify ‘biofilm-specific’ genes. In
Biofilm structures can be flat or mushroom-shaped these studies, biofilm formation by a mutant strain is
depending on the nutrient source, which seems to compared with that of wild-type bacteria to assess the
influence the interactions between localized clonal influence of a particular gene. To adapt these studies
growth and the subsequent rearrangement of cells for high-throughput screening, biofilms are often
through TYPE IV PILUS-mediated gliding motility in grown in microtitre wells. However, this limits the
response to the nutritional cues13. Stage five is charac- growth conditions to those of a poorly mixed batch
terized by a return to transient motility where biofilm culture, with little shear and no nutrient exchange.
cells are sloughed or shed. Also, comparative biofilm growth is usually assessed
Although there is understandably intense interest in after short periods, which limits interpretation to the
the investigation of the initial stages of biofilm forma- early stages of biofilm development. These studies have
tion, more detailed investigation into biofilm DETACHMENT identified numerous genes or factors as being ‘essential’
STREAMERS as a discrete process that is important to structural or ‘required’ for biofilm formation32–39. Such genes
Filamentous biofilm
microcolonies that form in
development and dispersal is also warranted. Dispersal include those that regulate or express surface-adhesion
flowing water. The streamers are mechanisms are discussed below. There is also evidence proteins, appendages such as pili and flagella, and
atached to the surface by an for developmental sequences in Escherichia coli 21–23 extracellular polymeric substance (EPS) matrix materi-
upstream ‘head’, while the and Vibrio cholerae 24 biofilms. Such behaviour paral- als. However, in many cases, on closer inspection
downstream ‘tail’ can oscillate in
lels the more social behaviour that is observed in biofilm formation is not prevented by ‘knocking out’ a
the current.
myxobacteria25. Until recently, it was thought that specific gene, but is only retarded or reduced. The
PERIPHYTON highly regulated social behaviour in prokaryotes was redundancy of the pathways that are involved in
An assemblage of organisms an unusual feature of the myxobacteria. Now, the biofilm differentiation could reflect the fundamental
attached to and living on more subtle structures and behaviours of biofilms importance of biofilm formation.
submerged solid surfaces in
natural environments such as
indicate that it might be more common within the Also, cell signalling has been shown to control
rivers. proteobacteria. Conceptual ‘motility-based’ models biofilm differentiation40; however, this influence can be
that have been derived from observations of the negated or mitigated by growth under different nutri-
STROMATOLITE Gram-negative proteobacteria are often generalized as ent41 or fluid shear conditions42,43. Taken together,
Mushroom- and tower-shaped
universal biofilm development cycles. However, many observations from laboratory experiments and
structures formed by layers of
cyanobacteria and entrapped of the same dynamic processes — moving over sur- biofilms growing in nature (FIG. 2) show that both the
sediments that grow in quiescent faces, growth cycles and detachment and reattachment environment and the genome significantly influence
(calm) saline and hydrothermal — can also be observed in non-motile species, such as biofilm formation.
waters. the human pathogen Staphylococcus aureus 26 or
MASS TRANSFER
Mycobacterium spp.27. It is interesting that the stages of Biofilm dispersal strategies
In the context of biofilms, mass biofilm development seem to be conserved among a Traditionally, laboratory experiments focus on the
transfer refers to the process by remarkable range of prokaryotes. Similar convergent attachment of planktonic batch-cultured or chemostat-
which dissolved and particulate behaviour in eukaryotes might be seen in the cellular cultured cells to surfaces and the subsequent biofilm
substances (such as nutrients)
slime moulds, which can live either as single motile growth. The detachment and dispersal of cells from
are moved into and out of the
biofilm by the surrounding fluid. cells or as multicellular colonies that are held together biofilms has received less attention. Detachment can be
in an extracellular matrix28. caused by external perturbations, such as increased fluid
ADHESION shear44, by internal biofilm processes, such as endoge-
A stable interaction of a cell with Determinants of biofilm structure nous enzymatic degradation, or by the release of EPS or
respect to a surface. Living cells
actively excrete chemicals from
There is a continuing debate among biofilm surface-binding proteins45–47. Detachment is normally
their surface to anchor researchers concerning the relative contributions of viewed from the perspective of control (biofilm removal
themselves to a substratum. This genetics (active response) and environmental condi- strategies), or the contamination of food and water pro-
is referred to as adhesion or tions (passive response) to the development of biofilm duction facilities48,49, or medical and dental devices50–52.
attachment.
structure and development29,30. These are not mutually In some species, dispersal from biofilms seems to be an
TYPE IV PILUS
exclusive and the relative contributions of each genetic active process, presumably adapted to allow coloniza-
An elongated structure response to a specified set of growth conditions tion of new niches15. Three distinct biofilm dispersal
extending from the surface of requires more multifactorial experiments to determine strategies can be identified:‘swarming/seeding dispersal’,
Gram-negative cells that is how genetic and environmental responses interact. in which individual cells are released from a micro-
independent of flagella and
which can retract and pull the
colony into the bulk fluid or the surrounding substra-
cell forward. Prediction of complex biofilm structure using models tum;‘clumping dispersal’, in which aggregates of cells are
and genetic determination of biofilm structure and shed as clumps or emboli; and ‘surface dispersal’, in
DETACHMENT redundancy. Structural and temporal complexity can which biofilm structures move across surfaces. The sur-
The loss of single cells or
also be successfully modelled using simple rules that are vival advantages and disadvantages of active (motility-
aggregates of cells from the
biofilm, usually into an overlying based on localized growth patterns determined by the driven) and passive (shear-mediated) dispersal strategies
flow of fluid. Detachment can be distribution of nutrients and FLUID SHEAR17. Models also are shown in TABLE 1.
an active process (dispersal), predict that biofilm heterogeneity can be maintained
a passively induced mechanical through the production of diffusible ‘detachment Swarming dispersal. Swarming dispersal has been best
process (for example, through
fluid shear) or a chemical
factors’, which cause localized detachment31. described in non-mucoid P. aeruginosa biofilms. After
process (by adding agents that Molecular techniques, such as random transposon initial biofilm growth, the microcolonies differentiate to
‘dissolve’ the EPS matrix). mutagenesis and knockout mutant studies, have been form an outer ‘wall’ of stationary bacteria (of biofilm

REVIEWS
Table 1 | Hypothetical dispersal mechanisms for biofilms compared with those for single cells
Cellular state Medium Dispersal mechanism Advantages and disadvantages
Single cells Fluid Swimming motility Active directional chemotactic motility, but vulnerable
Sporulation Passive fluid-directed dispersal, but protected
Surface Gliding or twitching motility Active directional chemotactic motility, but vulnerable
Biofilm Fluid Clumping fluid-borne dispersal Passive fluid-directed dispersal, but protected
Surface Shear-mediated rippling or rolling Passive fluid-directed dispersal, but protected
phenotype), while the inner region of the microcolony in some species single cells can actively move across
‘liquefies’, which allows motile cells (of planktonic phe- surfaces through GLIDING OR TWITCHING MOTILITY61, there is
notype) to ‘swim’ out of the microcolony, leaving a evidence that whole biofilms can also move across sur-
hollow mound15,53,54. Liquification has been attributed faces through shear-mediated transport. Migratory
to lysis of a subpopulation due to prophage-mediated ripple structures travelling at velocities of up to 1 mm
cell death55. The lysing population can be regarded as a hour–1 have been reported in laboratory studies on
third phenotype, whereas the remaining swarming cells P. aeruginosa and mixed-species biofilms42,62, and simi-
might be a surviving, apoptosis-negative, ‘persister’ lar structures have also been seen in natural biofilms
phenotype56. Hollow microcolonies have been seen in (FIG. 2). Rippling transport might have consequences in
Staphylococcus epidermidis growing on agar plates medicine. Ripple structures have been reported in
(P. Stewart, personal communication), and transmis- biofilms in endotracheal tubes and it has been hypothe-
sion electron micrographs (TEM) indicate that the sized that the flow of biofilms down the tubes is related
FLUID SHEAR
hollowing occurs through localized lysis. to dissemination into the lungs and subsequent cases of
The mechanical force that is Bacteriophages have also been shown to reduce the ventilator-associated pneumonia63. Similar mechanisms
exerted by a fluid as it moves viscosity of purified P. aeruginosa ALGINATE57. The might be important for dissemination from infected
past a surface. Although shear authors of this study concluded that this increased the catheters. This type of dispersal might be more clinically
exists throughout the fluid in
transport of bacteriophage through the biofilm to challenging than planktonic dispersal because it can be
biofilms,‘shear stress’ is usually
used in the context of the shear enhance infection. However, it is possible that this more difficult to detect and can be more resistant to
exerted at the solid surface — for phenomenon is also important in swarming/seeding antibiotics64. In addition to rippling migration, we have
example, where the biofilm is dispersal. A similar phenomenon has been reported in recently found that S. aureus biofilm microcolonies can
growing. The shear stress will other species, including the non-motile dental ‘roll’ along the lumen of a glass tube in an in vitro cen-
tend to ‘wash away’ the attached
biofilm from the surface on
pathogen Actinobacillus actinomycetemcomitans 45,58. tral venous catheter model (C.J. Rupp, S. Wilson and
which it is growing, and it Instead of active swimming dispersal, it was hypothesized P.S., unpublished observations). VISCOELASTIC tethers
increases as the flow rate is that the released cells took advantage of convective flow were observed between the rolling S. aureus micro-
increased. currents in the overlying fluid for a fluid-borne dispersal colony and the glass surface. The tethers stretched until
strategy 59. breaking at the upstream edge, which allowed the
EPS
Extracellular polymeric microcolony to roll forward and create a new tether at
substance. Polymers of varying Clumping dispersal. In this dispersal strategy, whole the new downstream point of contact.
chemical composition that are aggregates are continually shed from the biofilm60.
excreted by the cells in the The aggregates consist of biofilm cells that are sur- Biofilm viscoelasticity and adaptation to survival on
biofilm. The EPS is the slime
matrix that gives the biofilm
rounded by EPS and which might be more similar surfaces. So far, we have made the case that biofilm
stability and helps it to adhere to physiologically to the attached biofilm than to plank- formation is an important factor in the survival of
a surface. Although generally tonic cells. The tendency to shed clumps containing prokaryotes. We have also discussed aspects of multicel-
assumed to be primarily hundreds of cells by S. aureus, a non-motile human lularity in biofilms and how dynamic biofilm behaviour
composed of polysaccharides,
pathogen, contrasts with the detachment pattern for could be related to dispersal. The mechanical proper-
the EPS can also contain
proteins and nucleic acids. P. aeruginosa biofilms, in which the loss of single cells ties of biofilms might also explain the tenacity of
and small clumps predominates (P.S. and S. Wilson, biofilms that are associated with solid surfaces while, at
ALGINATE unpublished observations). Moreover, the antibiotic the same time, allowing enough flexibility to flow or
An exopolysaccharide produced resistance of detached S. aureus clumps is similar to move over those surfaces. Sessile plants and animals
by P. aeruginosa, which is
believed to contribute to the
the resistance that is associated with attached biofilms. have developed several different strategies to remain
antibiotic resistance of The reattachment of detached, protected emboli attached in moving fluids. Generally, very little is
P. aeruginosa. from S. aureus (FIG. 3) might explain the high fre- known about the mechanical properties of intact
quency of infectious metastasis that is associated with biofilms, but biochemical analysis of the EPS slime
GLIDING OR TWITCHING
this organism50. Although the dispersal direction is matrix indicates that biofilms are HYDROGELS. Recently,
MOTILITY
Movement, predominantly by not actively controlled, it is possible that the biofilm it has been reported that a wide range of biofilms that
Gram-negative cells, that is uses the fluid flow in a similar manner to the wind- are grown under flow, either in vitro or in hot spring
dependent on type IV pili. borne seed-dispersal strategies used by dandelions environments, show classic viscoelastic behav-
and other plants. iour44,65–68. These properties allow the biofilms to with-
VISCOELASTIC
A material that has both elastic
stand the transient periods of rapidly changing shear
(solid-like) and viscous (liquid- Surface dispersal. Another strategy for biofilm dispersal stresses that are expected in many marine and river
like) properties. is movement across surfaces. Although it is known that environments due to seasonal variation, run-off due to

REVIEWS
a b shear environments through EPS modification. A better

characterization of the mechanical properties of
biofilms will elucidate the role of biofilm mechanics for
sessile survival in flowing environments, explain
observed shear-mediated dynamic behaviour and allow
the development of more effective mechanical removal
strategies for biofilm control.
Why do prokaryotes form biofilms?

What are the advantages of biofilm formation and surface
association that make it such a widespread phenomenon?
At present, there are several hypotheses. First, surfaces
provide a space to be occupied and, as discussed previ-
20 µm 20 µm ously, they provide a degree of stability in the growth
c d environment and might have catalytic functions
through localizing cells in close proximity. Second,
biofilm formation affords protection from a wide
range of environmental challenges, such as UV expo-
sure69, metal toxicity70, acid exposure71, dehydration
and salinity72, phagocytosis73 and several antibiotics
and antimicrobial agents64,74,75.
Three mechanisms have been proposed to explain
the general resistance of biofilms to biocidal agents.
The first is the barrier properties of the slime matrix.
This mechanism might be more relevant for reactive
(bleach or superoxides), charged (metals) or large
(immunoglobulin) antimicrobial agents that are neu-
Figure 3 | Staphylococcus aureus biofilms. a | Confocal laser microscopy image of a biofilm
tralized or bound by the EPS and are effectively
growing in a flow cell with a flow rate of 1 ml min–1. The square panel is a plan view and the
side panels are vertical and horizontal cross sections, respectively. b | Detached cells from ‘diluted’ to sublethal concentrations before they can
the biofilm shown in part (a) captured on a polycarbonate filter. The detached fraction reach all of the individual bacterial cells within the
consisted of single cells and large clumps containing thousands of cells. Biofilms in parts a biofilm. The barrier properties of the EPS hydrogel
and b were stained with Molecular Probes Live/DeadTM kit so that live cells appeared green might also protect against UV light and dehydration,
and dead cells appeared red. Images courtesy of Suzanne Wilson, Center for Biofilm and might localize enzymatic activity. For example,
Engineering, USA. c | Biofilm grown in the skin cut of a neutropoenic mouse. The
extracellular β-lactamase enzymatic activity against
extracellular polymeric substance (EPS) of the biofilm was stained with FITC-ConA (green)
and safranine (red). Reproduced with permission from REF. 87 © (2002) Blackwell Publishing.
P. aeruginosa occurs within the matrix76.
d | Scanning electron micrograph of S. aureus biofilm that developed on the distal tip of a The second protective mechanism could involve the
cardiac pacemaker lead in a patient. Reproduced with permission from REF. 85 © (1982) physiological state of biofilm organisms. Although
American Heart Association. many antibiotics can freely penetrate the EPS, cells
within the biofilm are often still protected. The creation
of starved, stationary phase dormant zones in biofilms
rapidly changing weather conditions and changes in seems to be a significant factor in the resistance of
local currents due to variations in bed morphology or biofilm populations to antimicrobials56,77,78, particularly
flow channel. Over short periods (seconds), biofilms against antibiotics such as β−lactams, which are effec-
can absorb elevated shear by behaving elastically, but tive against rapidly dividing Gram-positive bacteria by
over longer periods loading stresses in the biofilm can interruption of cell-wall synthesis. However, arguably
be dissipated through viscous flow, so that instead of all antibiotics require at least some degree of cellular
detaching, the biofilm can either flow over the surface62 activity to be effective, because the mechanism of action
or become streamlined to reduce drag43 (FIG. 2G). of most antibiotics involves disruption of a microbial
Although the viscoelastic response of biofilms is process. Therefore, pockets of cells in a biofilm in sta-
similar, the magnitudes of the viscoelastic parameters tionary phase dormancy might represent a general
are highly variable. For biofilms that are grown in vitro, mechanism of antibiotic resistance.
the shear modulus (G, which is a measure of rigidity) is A third mechanism of protection could be the exis-
10–1–103 Pa, and the viscosity (η) is 105–108 Pa, whereas tence of subpopulations of resistant phenotypes in the
for biofilms from natural hot springs, G is 103–105 Pa biofilm79, which have been referred to as ‘persisters’56.
and η is 107–108 Pa (P.S., T. Shaw, M. Winston and I. Persisters comprise a small fraction of the entire biomass,
Klapper, unpublished observations). The variability in whether in planktonic or biofilm culture, but as distinct
the absolute magnitude of the viscoelastic parameters phenotypes have yet to be cultured, it remains unclear if
HYDROGELS might reflect the diversity of biofilm organisms and the these organisms do indeed represent a distinct phenotype
An extremely hydrated polymer
gel. The polymer chain holds
growth environment. Viscoelasticity might be not only or are simply the most resistant cells within a population
many times its weight in trapped an adaptation for survival on surfaces in flowing water, distribution.Although the relative contribution of each of
water. but might also allow an adaptive response to different these mechanisms (and possibly others) varies with the

REVIEWS
type of biofilm and the nature of the environmental to a substratum; direct examination reveals bacteria in
stress, the result is one of general protection. clusters, encased in a matrix of bacterial or host con-
We have hypothesized that biofilm formation is stituents; the infection is localized; and the infection is
likely to be an ancient adaptation of prokaryotic life. resistant to antibiotic therapy despite the antibiotic
Paradoxically, many of the impacts of the proclivity sensitivity of the constituent planktonic organisms.
of bacteria to stick to surfaces have become more The infections discussed in this review were chosen
obvious with ever-increasing technological develop- because they illustrate consistencies between biofilm
ments. Industrial pipelines, nuclear power stations, growth in the environment and published literature
space stations, air conditioning systems, water distri- investigating clinical infections. Owing to a great
bution systems and one of the fastest technologically increase in the number of medical biofilm papers, how-
advancing settings, the hospital, are all susceptible to ever, space does not allow a comprehensive review of
colonization by microorganisms growing in biofilms. the medically relevant biofilms and the readers are
referred to several reviews80–82. Device-related infections
Biofilm infections were the first clinical infections to be identified as
The first part of this review concentrated on the having a biofilm aetiology and show that biofilm forma-
importance of biofilm formation for the survival of tion can be facilitated by the host inflammatory
prokaryotes in the natural environment. Some of the response because host inflammatory molecules facilitate
developmental processes and observed dynamic phe- adhesion to the surface of the device. Bacterial endo-
nomena of biofilm formation were outlined and inter- carditis shows how microorganisms on the skin or in
preted in the context of convergent survival strategies. the oral cavity that transiently enter the bloodstream
In the second part, we discuss biofilm formation as a can colonize abnormal or implanted valves, or altered
fundamental consequence of bacterial adhesion and endothelial surfaces in the heart (FIG. 4). Surface attach-
biofilm growth in a host. Many of the physiological ment within vegetations occurs as a result of interac-
characteristics of biofilm formation — for example, tions between microbial cells and host products. Cystic
localized clusters of bacteria adhering to a substratum fibrosis (CF) illustrates how the opportunistic pathogen
that are more resistant to antibiotic therapy — are simi- P. aeruginosa exploits the unique environment of the CF
lar whether in the natural environment or in an animal lung and responds to environmental cues by altering its
host. Fundamental knowledge of biofilm formation phenotype.
from environmental studies helped to characterize
biofilms growing on medical devices in the earliest stud- Device-related infections. Intravenous catheters, pros-
ies, and continues to provide insights into biofilm infec- thetic heart valves, joint prostheses, peritoneal dialysis
tions. However, we argue that the complex interaction catheters, cardiac pacemakers, cerebrospinal fluid
between the biofilm pathogen and the host inflamma- shunts and endotracheal tubes save millions of lives,
tory response modifies the host environment, and that but all have an intrinsic risk of surface-associated
successful biofilm parasites respond accordingly by infections. Biofilms associated with medical devices
altering their phenotype to the biofilm mode of growth. were first noted in the early 1980s when electron
microscopy revealed bacteria deposited on the surface
Matrix-enclosed microbial communities adherent to of indwelling devices, such as intravenous catheters
non-biological and biological surfaces. A distinguishing and cardiac pacemakers83–85.
feature of biofilms from that of other colonizing infec- The microorganisms that are most frequently asso-
tions is the presence of aggregated microcolonies of cells ciated with medical devices are the staphylococci
that are attached to a surface. Importantly, biofilm for- (particularly S. epidermidis and S. aureus), followed by
mation as a protective mechanism could have profound P. aeruginosa and a plethora of other environmental
implications for the host, because the microorganisms bacteria that opportunistically infect a host who is
that are growing in these matrix-enclosed aggregates are compromised by invasive medical intervention,
more resistant to antibiotics and host defences. The chemotherapy or a pre-existing disease state. Biofilm
biofilm model proposes that microbial cells growing in formation on medical implants has even led to the char-
biofilms are clustered. It fundamentally challenges the acterization of a new infectious disease called chronic
assumption that infectious agents are evenly distributed polymer-associated infection82,86. Staphylococci com-
and therefore equally susceptible to the host immune monly colonize the skin and are frequently found in
response or antibiotic therapy. It might further account wounds and implants87. Interestingly, S. epidermidis was
for several problematic clinical challenges, such as not considered an opportunistic pathogen until the
symptomatic, but unculturable, inflammation, antibi- widespread use of medical devices. Biofilm forma-
otic resistance, recurrence or persistence, and metastasis tion, then, can be thought of as a virulence factor —
or the spread of infectious emboli. a bacterial strategy that contributes to its ability to
However, a problem with assessing the contribu- cause an infection.
tion of biofilms in human disease is the lack of defined The most notable characteristic of the adherent
criteria with which to characterize biofilm-induced staphylococci colonizing medical implants is the
pathogenesis. Parsek and Singh80 propose four criteria copious amount of EPS (also known as glycocalyx or
for defining a biofilm aetiology of an infection: the ‘slime’) that encases and protects cells from host
pathogenic bacteria are surface associated or adherent defences and antibiotic treatment (FIG. 3c). Biofilm
NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 1 0 1

REVIEWS
Catheter contamination
Gum disease
Implant contamination
Figure 4 | Schematic showing three examples of possible points of entry into the body for infectious biofilms; catheter,
hip replacement, and periodontal disease. Arrows show how the biofilm (green) might be disseminated around the body, either
by single cells or clumps of protected emboli, using the example of native or artificial heart valve infective endocarditis as a common
central location for embolization. Sporadic detachment could lead to cycles of bacteraemia. Image courtesy of P. Dirckx, Center for
Biofilm Engineering.
formation is characterized by two principal stages in specifically to cell-to-cell adhesion mechanisms that are
staphylococci: adhesion of bacteria to a solid surface, fol- associated with a β-1,6-linked glycosaminoglycan poly-
lowed by growth-dependent accumulation of cells which saccharide known as polysaccharide intercellular
generates multiple layers of cell clusters. In S. epidermidis, adhesin (PIA)88. Proteins that are involved in the synthe-
the formation of multiple cell layers has been attributed sis of matrix polysaccharides are regulated by the ica

REVIEWS
gene locus in S. epidermidis 89, a locus that is conserved binding of S. aureus in adhesion to both polyethylene
in S. epidermidis, S. aureus and other phylogenetically- and polyvinyl surfaces96. So, certain bacteria seem to
related staphylococci34. Mutations in this locus disrupt have the ability to exploit host proteins that are pro-
biofilm formation primarily by disrupting cell aggre- duced in wound healing or inflammation, which indi-
gation and accumulation90,91. However, strains with cates that bacterial adhesins provide a mechanism by
PIA and ica might still fail to form biofilms if they are which colonization of the host can occur on viable, but
defective in initial adherence. The extensive literature damaged tissues, and on devices in conditions where
on S. epidermidis and S. aureus shows that the adhesion these host inflammatory molecules are present and
stage alone is multifactorial, and depends on both the might be characterized by biofilm-like infections.
physiochemical properties of the biomedical polymer
material and the nature of the bacterial cell surface. In Infective endocarditis. The biofilm matrix is most fre-
particular, the hydrophobicity and the electrostatic quently referred to as being of bacterial origin. Although
charge of the material will influence interactions this applies to biofilms grown on an abiotic surface in
between the polymer and the surface of the bacterial cell. the laboratory, biofilm infections within the body,
Bacterial surface proteins contribute significantly to which are characterized by adherent bacteria on tissue,
adhesion, and several key proteins have been identified might also include host cells and molecules as part of
as being important in staphylococcal biofilm formation. the surface-associated infection. An interesting example
S. epidermidis adherence to polystyrene is mediated by is bacterial endocarditis.
AtlE, the major autolysin. An AtlE mutant that is defec- Streptococci are the aetiological agents in more than
tive in forming biofilms on polystyrene, but not on glass half of infective endocarditis cases, with staphylococci
surfaces, is also less hydrophobic and forms large cell accounting for another quarter of the cases81. Many
clusters compared with the wild type89. This protein strains are common commensals of the skin and the
also mediates binding to vitronectin, a component of oral cavity. Clinically, bacterial endocarditis lesions are
the host extracellular matrix. In S. aureus, mutants referred to as vegetations and they comprise aggregates
lacking D-alanine in teichoic acid (dltA) exhibited a of bacterial cells, platelets and fibrin which are adherent
change in surface charge that compromised their ability to the damaged epithelium of cardiac valves.
to adhere to polystyrene or glass, although the produc- Endocarditis is associated with congenital heart defects,
tion of PIA was intact92. The adhesion step of this prosthetic heart valves and vascular grafts, and is most
defect in biofilm formation was re-established by the likely caused by clots of platelet and fibrin, which amass
addition of Mg2+. Additionally, S. epidermidis biofilm where turbulent flow is aggravated by abnormal tissue,
formation was enhanced by Mg2+ and inhibited by pre-existing heart disease or an indwelling vascular
EDTA93, illustrating the role of environmental factors catheter. Damaged endothelium exposes the underlying
on biofilm development. Other surface proteins, basement membrane, which consists of collagens,
including the biofilm-associated protein (Bap) and the laminin, vitronectin and fibronectin, thereby providing
accumulation-associated protein (AAP), are important a substratum for bacterial adherence (the initial stage in
in biofilm formation. Mutations in Bap affect biofilm the pathogenesis of endocarditis). In addition, after
formation and pathogenesis in a mouse foreign body endothelial damage and turbulent blood flow, inflam-
infection model33. matory processes stimulate the clotting system, lead-
Finally, the host can contribute significantly to ing to the deposition of fibrin and the creation of an
adhesion in device-related infections, particularly with insoluble clot of fibrin and platelets.
staphylococci94. Multiple specific receptors on the cell Durack showed that Streptococcus sanguis adhered
surface, called adhesins, bind to host molecules (for to the surface of sterile vegetations within 30 minutes
example, protein/glycoprotein components in plasma of injection into catheterized rabbits and began to
or platelets or components of the host extracellular replicate97. Streptococcal microcolonies developed in
matrix). Many of these proteins belong to a family of the thrombus and were surrounded by strands of
microbial surface components that recognize adhesive fibrin ‘capsules’ that seemed to retard leukocyte inter-
matrix molecules (MSCRAMMs), which mediate actions. When the vegetations were examined four
adhesion to various host cell types as well as to polymer weeks post-infection, viable streptococci were present
surfaces coated with host plasma proteins82,95. Several within a calcified lesion surrounded by fibroblasts
bacteria have adhesins for fibronectin, which is a host underlying the endothelium. The metabolic activ-
protein that is frequently associated with bacterial ity of organisms in the vegetation seemed reduced
attachment to surfaces94, followed by fibrinogen/fibrin, compared with bacteria at the periphery — consistent
collagen, laminin and vitronectin. Fibronectin also par- with biofilm formation in the vegetation. Over time,
ticipates in adhesion by bridging associations with fib- the vegetations grew by the addition of layers of fibrin
rin, collagen, heparin and other host cell surface gly- and platelets, with bacterial colonies ‘sandwiched’
cosaminoglycans. Two fibronectin adhesins have been between them, indicating cycles of thrombosis and
identified in S. aureus — FnBPA and FnBPB. further bacterial colonization of the layers97.
A collagen-binding protein (Cna) and two fib- However, using the rabbit model, Höök and Sand
rinogen-binding proteins, known as clumping factors A showed that bacterial colonization occurred even
and B (ClfA and ClfB) also belong to the MSCRAMM when vegetations were prevented with anticoagulant
family. ClfA has also been shown to be important in the treatment, which indicated that the presence of a clot

REVIEWS
is sufficient, but not necessary, for endocarditis. species106. ClfA has also been shown to bind S. aureus to
Interestingly, the course of the disease was remarkably human platelets directly by a previously uncharacterized
different. In rabbits that were treated with anticoagu- platelet-membrane receptor107. These specific interac-
lants, disease was more fulminant and animals exhib- tions show that, similar to the example of biofilms in
ited higher bacteraemia and a lower survival rate. By device-related infections, biofilm development is a multi-
contrast, animals without anticoagulant treatment step process of initial adherence between specific bacterial
exhibited a subacute, chronic infection that was more adhesins and host molecules, followed by intercellular
recalcitrant to antibiotic therapy. The ultrastructure of the accumulation of bacterial cells and host components that
resulting vegetations was comprised of large bacterial generate multiple cell layers of the biofilm vegetation.
colonies, densely packed with fibrin and platelets and Bacterial endocarditis might also illustrate how
surrounded by a fibrin mesh. The authors speculated turbulent flow contributes to the formation of vegeta-
that the bacteria within the vegetation were metaboli- tions. Although turbulence has been traditionally
cally less active98. Similar structures were observed in thought to induce clot formation and tissue damage, it
human vegetations99. Therefore, the appropriate local is conceivable that clumps of biofilm cells respond to the
conditions exist for colonization of underlying host turbulent flow of the cardiac environment by producing
tissues by bacteria growing as microcolonies, and the more EPS. Large vegetations are particularly friable, and
pathogenesis of endocarditis as a biofilm disease is amplify the risk of embolization (detachment through
consistent with bacteria growing in biofilms. However, clumping dispersal) that might cause infarcts and septic
these reports also indicate that the presence of platelets abscesses in other tissues.
and fibrin owing to the host inflammatory response Importantly, antibiotic therapy in the treatment of
confers biofilm characteristics that are observed in endocarditis is also consistent with a role for biofilms.
endocarditis — a persistent, recurring infection that is In vivo studies in a rabbit model with E. coli as the infec-
more resistant to antibiotic treatment. tious agent required sustained antibiotic concentrations
Several studies have examined the ability of bacteria that were 220 times the serum minimum bacteriocidal
to adhere to tissue surfaces and establish a localized concentrations. Even when the vegetations were treated
site of infection by specific interactions between bac- with antibiotic ex vivo, antibacterial effects within the
terial adhesins and host tissue. Ramirez-Rhonda vegetation required 150 times the minimal bacteriocidal
examined streptococcal species for their ability to concentration108.
adhere to cardiac valves and found that strains that
produce EPS consisting of glucans and dextrans Cystic fibrosis pneumonia. CF is an autosomal recessive
adhered better to damaged heart valves100. More disease that is caused by mutations in the cystic fibrosis
recently, in Streptococcus parasanguis, a colonizer of transmembrane conductance regulator (CFTR) gene
the human tooth surface and an opportunist that is that results in dysfunctional electrolyte secretion and
found in both native and prosthetic heart valve endo- absorption. Although multiple, complex physiological
carditis, a gene encoding PERITRICHOUS FIMBRIAE, fap1, dysfunctions are present, the primary site of morbidity
was shown to be associated with biofilm formation is the respiratory system. Reduced hydration of the air-
on plastic. A fap1 mutant showed limited adherence, way surface fluid that renders the respiratory mucous
but primarily failed to aggregate and form micro- more viscous and impairs mucociliary clearance, leads
colonies36. Fey et al. showed a correlation between to the main clinical feature of CF — chronic endo-
haemagglutination and biofilm formation in S. epider- bronchial bacterial infection and airway inflammation
midis, which indicated a link between PIA, inter- — which leads to airway obstruction, progressive
cellular adhesion and adhesion to erythrocytes. destruction of the airway epithelium and, ultimately,
Defective PIA strains were found to be less virulent in respiratory failure.
a rabbit endocarditis model101,102. Similarly, S. aureus Pulmonary colonization of the lower respiratory
mutants that are defective in adherence to platelets tract of CF patients begins in infancy or early child-
correlated with reduced virulence in a rabbit model of hood, most commonly by S. aureus and Haemophilus
endocarditis, which is characterized by fewer bacteria influenzae. However, by adolescence and early adult-
within vegetations and reduced embolization103. hood most CF patients have become colonized with
An S. aureus mutant defective in fibronectin binding P. aeruginosa109,110. The shift from colonization with other
also showed decreased binding to damaged heart bacteria to chronic infection with P. aeruginosa seems
valves104. The clumping factor ClfA, a fibrinogen-binding to be the result of the peculiar environment of the CF
protein in S. aureus which also mediates adhesion of lung, which includes asialylated receptors on epithelial
S. aureus to plastic, seems to have a s pecific role in the cells that facilitate pseudomonal attachment in addition
rat model of endocarditis. Mutant and complementa- to impaired mucociliary clearance109–111.
tion studies showed that clfA was important in both There are two salient features regarding coloniza-
adherence to surfaces and virulence; however, endo- tion of the CF lung with P. aeruginosa. First, P. aerugi-
PERITRICHOUS FIMBRIAE carditis still occurred with a larger innoculum of clfA nosa grows in biofilms within the CF lung 110,112–115.
The many short extracellular pili mutants105. However, when ClfA and FnbA (another Microscopic analysis of sputum from CF patients
appendages that protrude from
the surface of some prokaryotic
adhesin) were co-expressed in a non-virulent organism showed that P. aeruginosa forms biofilm-like struc-
cells. Fimbriae are used for — Lactococcus lactis — endocarditis in the rat model tures consisting of clusters of bacteria surrounded by
attachment to surfaces. was comparable to that observed for pathogenic a dense matrix113,114. A similar structural morphology

REVIEWS
has been observed in CF lung specimens 116. These results indicate that the presence of P. aeruginosa
Moreover, homoserine lactone (HSL) QUORUM SENSING induces the release of inflammatory mediators and
(QS) signals measured in CF sputum were consistent leukocyte infiltration into lung tissue that is not
with the QS profile of P. aeruginosa grown in associated with clearance of the pathogen, but rather
biofilms, not the profile from planktonically grown with continuing inflammation and lung pathology.
P. aeruginosa cultures115. Another hypothesized mechanism of bacterial
The second salient feature of P. aeruginosa coloniza- colonization in CF implicates the mucous layer. In this
tion of the CF lung is the selection of mucoid variants of model, the increased viscosity of CF airway mucous
P. aeruginosa, which are characterized by overproduc- acts as a matrix scaffold and is important in decreased
tion of the exopolysaccharide alginate, and a resistance clearance. Recently, Worlitzsch et al.124 studied CF
to antibiotic therapy110,111. Initially, P. aeruginosa isolated patients with chronic lung disease directly using elec-
from the lungs of CF patients is non-mucoid. However, tron microscopy and tissue explants, and found that
mucoid isolates typically coincide with persistent P. aeruginosa was present in mucopurulent hypoxic
chronic infection. Interestingly, mucoid variants are ‘macrocolonies’ of 100-µm diameter in the airway
absent among environmental isolates of P. aeruginosa, lumen, rather than attached to the epithelium. Both
although non-mucoid strains seem to have the geno- in vivo and in vitro experiments using microelectrodes
type for mucoidy. Research indicates that the host showed that oxygen was depleted in these mucoid
inflammatory response contributes to mucoid conver- macrocolonies. Furthermore, motile P. aeruginosa pen-
sion. Mathee et al. grew P. aeruginosa in biofilms and etrated the hypoxic mucous layers and responded to
subjected these biofilms to either exogenous hydrogen anaerobic conditions by producing more alginate.
peroxide (H2O2) or activated human polymorphoneu- These results argue that the local environment in CF
trophils (PMNs) in vitro 117. They observed that mucoid lungs, which is characterized by thick mucous plaques
conversion was consistent with a deletion in the mucA and depleted O2 in the respiratory epithelium, leads to
open reading frame; the same deletion was also observed colonization by P. aeruginosa that further exacerbates
in 25% of mucoid isolates from CF patients. Therefore, the pathology of CF pneumonia.
P. aeruginosa responds to the microenvironment of Yoon et al.125 have also shown that anaerobic
the CF lung by modifying its phenotype. growth might be an important feature of P. aeruginosa
The specific mechanism of P. aeruginosa coloniza- growing in biofilms in CF patients. P. aeruginosa
tion of the CF lung is not known. One hypothesis pro- formed vigorous biofilms under anaerobic conditions,
poses that airway inflammation leads to the attachment leading to the build-up of toxic nitrogen metabolites.
of P. aeruginosa to denuded airway epithelium. One Proteomic analysis identified an outer-membrane
study found that CF bronchial secretions possess prote- porin, OprF, the concentration of which increased
olytic activity against fibronectin that is associated with 40-fold in anaerobic culture. OprF was also detected in
the respiratory mucosa118, which provides a mechanism secretions from CF lungs and circulating antibodies
that might favour P. aeruginosa colonization over against OprF were found in chronically infected CF
S. aureus colonization. Another study indicated that patients. This study indicates a mechanism that
P. aeruginosa binding to nasal polyp primary cultures explains several clinical aspects of CF P. aeruginosa
was due to modification of epithelial cells by bacterial pneumonia: the ineffectiveness of phagocytes against
exoproducts that exposed asialoganglioside-binding P. aeruginosa, the mucoid P. aeruginosa phenotype and
sites which facilitate pseudomonal adherence 119. the resistance of P. aeruginosa biofilms to tobramycin78,126.
Using a similar ex vivo model, P. aeruginosa was Interestingly, P. aeruginosa antibiotic resistance and
found to adhere to undifferentiated epithelial cells biofilm formation seem to be induced at the same
undergoing repair, specifically between αvβ1 integrins time127. When antibiotic resistance was studied in a
and the fibronectin RGD (Arg-Gly-Asp) receptor on clinical isolate of P. aeruginosa, a different phenotype
epithelial cells and a P. aeruginosa outer-membrane that is associated with both an enhanced ability to
protein 120. P. aeruginosa has several adhesins and form biofilms and increased antibiotic resistance was
binds to a broad range of receptors and cell types in observed. The phenotype was observed both in vitro
the respiratory tract121. and in CF patients undergoing antibiotic therapy, but
The QS molecule 3-oxo-C12-HSL that is produced not in untreated patients. Experimentally, resistance
by P. aeruginosa also seems to have modulatory effects variants also arose more frequently in response to
on the respiratory epithelium. Cultured bronchial environmental cues, such as alterations in salt con-
epithelial cells produced interleukin (IL)-8, an centration. It was speculated that antibiotic-resistant
inflammatory cytokine, in response to P. aeruginosa, phenotypic variants of P. aeruginosa observed in CF
and it was later shown that this effect was due to infections were either selected within biofilms by
QUORUM SENSING
A system by which bacteria 3-oxo-C12-HSL, not to other signalling molecules122. sub-lethal antibiotic treatment, or by the specific
communicate. Signalling Additionally 3-oxo-C12-HSL induced the expression environment of the CF lung (characterized by osmotic
molecules — chemicals similar of other inflammatory cytokines and chemokines, and oxidative stress). This study showed that both are
to pheromones that are such as IL-1, IL-6 and interferon (INF)-γ, and several responsible for the resistant phenotype127.
produced by an individual
bacterium — can affect the
macrophage inflammatory proteins. 3-oxo-C12-HSL In the complex environment of the CF lung, it is
behaviour of surrounding also induced cyclooxygenase 2 (COX-2) and unlikely that there is an exclusive mechanism of pathol-
bacteria. prostaglandin E2 (PGE2) in human lung fibroblasts123. ogy, and therefore the host inflammatory response

REVIEWS
undoubtedly induces changes in the local microenvi- together and to surfaces, simple volume-independent
ronment to which P. aeruginosa responds. What is clear concentration-based manipulations, such as dilutions
is that in CF pneumonia, the complex interactions or concentration series that are used to calculate
between bacteria and the local environmental cues in inhibitory and bacteriocidal concentrations (which are
the host owing to the inflammatory response contribute routinely used in batch cultures) become difficult.
to the complex pathology of this disease. Calculating the concentration of antibiotic per cell in
planktonic cultures is trivial when each cell is assumed
From planktonic to biofilm microbiology to encounter the same level of antibiotic. In biofilms
Although it is accepted that biofilms are found ubiqui- this is difficult given local heterogeneities and unchar-
tously in natural environments, the significance of acterized growth parameters, such as the surface
biofilms in infectious disease is often not recognized or area:volume ratio, residence time (the time a volume of
still debated. For those with experience of the problematic fluid is in the system) or the nutrient loading rate (the
nature of biofilms in the clinic, this might be seen as concentration of nutrients per unit area per time).
academic. However, to those who have been studying The lack of standard methods for growing, quantify-
planktonic bacterial cultures to investigate fundamental ing and testing biofilms in continuous culture results in
aspects of microbial physiology, pathogenesis and incalculable variability between laboratory systems. So
control — such as minimum inhibitory concentrations far there is only one ASTM (American Society for Testing
(MIC) of antibiotics — this is a matter of intense and Materials) standard method for growing biofilms
importance. The lag in studying surface-attached (E-2196-02), which uses a rotating disc reactor
bacteria rather than the planktonic complement is (BioSurface Technologies, Bozeman, Montana), although
understandable given the difficulties of working with the Centers for Disease Control (CDC) have also recently
surface-attached populations compared with homoge- developed a biofilm growth reactor130. Quantification is
nous batch-culture planktonic populations. Organisms being addressed with the development of image analysis
are more laborious to culture as biofilms; the inherent packages, such as COMSTAT131 and ISA132.
heterogeneity of spatial distribution18 leads to the creation Biofilm microbiology is complex and not well repre-
of localized zones that vary widely in both physiological sented by flask cultures. Although homogeneity allows
conditions19 and cellular physiologies128 over distances of statistical enumeration, the extent to which it reflects
only tens of micrometres. Another complexity when the real, less orderly world is questionable. Arguably, it
dealing with culturing on surfaces is that mass transfer is the complexity of biofilms that helps make them so
(diffusion and flow through the biofilm), which con- resilient. Biofilms present the next challenge in micro-
trols nutrient exchange, becomes an important consid- biology — to confront this complexity and devise more
eration129, as do the fluid forces (shear and drag) that relevant testing protocols to deal with demanding
act on the biofilm14. Also, because biofilm cells stick microbial problems in industry and medicine.
1. Zobel, C. E. The effect of solid surfaces upon bacterial Discovered that there could be a ‘division of labour’ in 22. Reisner, A., Haagensen, J. A., Schembri, M. A.,
activity. J. Bacteriol. 46, 39–56 (1943). bacterial biofilm populations such that a subset that Zechner, E. L. & Molin, S. Development and maturation
2. Doyle, R. (ed.). Microbial growth in biofilms, part A: remained attached could degrade the chitin of Escherichia coli K-12 biofilms. Mol. Microbiol. 48,
developmental and molecular biological aspects. Methods substratum and provide nutrients for a detached, 933–946 (2003).
Enzymol. 336 (2001). planktonic subset. 23. Tremoulet, F., Duche, O., Namane, A., Martinie, B. &
3. Doyle, R. (ed.) Microbial growth in biofilms, part B: special 12. Edwards, K. J., Bond, P. L., Gihring, T. M. & Banfield, J. F. Labadie, J. C. A proteomic study of Escherichia coli
environments and physiochemical aspects. Methods An archaeal iron-oxidizing extreme acidophile important in O157:H7 NCTC 12900 cultivated in biofilm or in planktonic
Enzymol. 337 (2001). acid mine drainage. Science. 287, 1731–1732 (2000). growth mode. FEMS Microbiol. Lett. 215, 7–14 (2002).
4. Westall, F. et al. Early Archean fossil bacteria and biofilms in 13. Klausen, M. et al. Biofilm formation by Pseudomonas 24. Watnick, P. I. & Kolter, R. Steps in the development of a Vibrio
hydrothermally-influenced sediments from the Barberton aeruginosa wild type, flagella and type IV pili mutants. Mol cholerae El Tor biofilm. Mol. Microbiol. 34, 586–595 (1999).
greenstone belt, South Africa. Precambrian Res. 106, Microbiol. 48, 1511–1524 (2003). 25. Kaiser, D. Coupling cell movement to multicellular
93–116 (2001). 14. Stoodley, P., Dodds, I., Boyle, J. D. & Lappin-Scott, H. M. development in myxobacteria. Nature Rev. Microbiol. 1,
5. Rasmussen, B. Filamentous microfossils in a 3,235-million- Influence of hydrodynamics and nutrients on biofilm 45–54 (2003)
year-old volcanogenic massive sulphide deposit. Nature structure. J. Appl. Microbiol. 85, 19S–28S (1999). 26. Fux, C. A., Stoodley, P., Hall-Stoodley, L. & Costerton, W. J.
405, 676–679 (2000). 15. Sauer, K., Camper, A. K., Ehrlich, G. D., Costerton, J. W. & Bacterial biofilms — a diagnostic and therapeutic challenge.
6. Reysenbach, A. L. & Cady, S. L. Microbiology of ancient and Davies, D. G. Pseudomonas aeruginosa displays multiple Expert Rev. Anti-Infective Ther. 1, 667–683 (2003).
modern hydrothermal systems Trends Microbiol. 9, 79–86 phenotypes during development as a biofilm. J. Bacteriol. 27. Hall-Stoodley, L., Keevil, C. W. & Lappin-Scott, H. M.
(2001). 184, 1140–1154 (2002). Mycobacterium fortuitum and Mycobacterium chelonae
7. Taylor, C. D., Wirsen, C. O. & Gaill, F. Rapid microbial Linked structural development with protein expression form biofilms under high and low nutrient conditions.
production of filamentous sulfur mats at hydrothermal vents. in Pseudomonas biofilms thereby demonstrating that J. Appl. Microbiol. 85, S60–S69 (1999).
Appl. Environ. Microbiol. 65, 2253–2255 (1999). biofilms can have regulated ‘life-cycles’. 28. Bracco, E. et al. Cell signaling and adhesion in phagocytosis
8. Jahnke, L. L. et al. Signature lipids and stable carbon isotope 16. Ghigo, J.-M. Are there biofilm-specific physiological pathways and early development of Dictyostelium. Int. J. Dev. Biol. 4,
analyses of octopus spring hyperthermophilic communities beyond a reasonable doubt? Res. Microbiol. 154, 1–8 (2003). 733–742 (2000).
compared with those of aquificales representatives. Appl. 17. van Loosdrecht, M. C., Heijnen, J. J., Eberl, H., Kreft, J. & 29. Hall-Stoodley, L. & Stoodley, P. Development regulation of
Environ. Microbiol. 67, 5179–5189 (2001). Picioreanu, C. Mathematical modelling of biofilm structures. microbial biofilms. Curr. Opin. Biotech. 13, 228–233 (2002).
9. Reysenbach, A. L., Ehringer, M. & Hershberger, K. Antonie Van Leeuwenhoek. 81, 245–256 (2002). 30. Kjelleberg, S. & Molin, S. Is there a role for quorum sensing
Microbial diversity at 83 degrees C in Calcite Springs, 18. Lawrence, J. R., Korber, D. R., Hoyle, B. D., Costerton, J. signals in bacterial biofilms? Curr. Opin. Microbiol. 5,
Yellowstone National Park: another environment where the W. & Caldwell, D. E. Optical sectioning of microbial biofilms. 254–258 (2002).
Aquificales and ‘Korarchaeota’ coexist. Extremophiles 4, J. Bacteriol. 173, 6558–6567 (1991). Discusses the relative contribution of various
61–67 (2000). 19. deBeer D., Stoodley, P., Roe, F. & Lewandowski, Z. Effects environmental factors (such as flow and nutrients) and
10. Stoodley, P., Sauer, K., Davies, D. G. & Costerton, J. W. of biofilm structures on oxygen distribution and mass genetic factors (cell signalling) on biofilm structure.
Biofilms as complex differentiated communities. Annu. Rev. transport. Biotechnol. Bioeng. 43, 1131–1138 (1994). 31. Hunt, S. M., Hamilton, M. A., Sears, J. T., Harkin, G. &
Microbiol. 56, 187–209 (2002). 20. Stoodley, P., de Beer, D., & Lewandowski, Z. Liquid flow in Reno, J. A computer investigation of chemically mediated
11. Baty, A. M., Eastburn, C. C., Techkarnjanaruk, S., biofilm systems. Appl. Environ. Microbiol. 60, 2711–2716 detachment in bacterial biofilms. Microbiology 149,
Goodman, A. E. & Geesey, G. G. Spatial and temporal (1994). 1155–1163 (2003).
variations in chitinolytic gene expression and bacterial 21. Pratt, L. A. & Kolter, R. Genetic analysis of Escherichia coli 32. Caiazza, N. C. & O’Toole, G. A. Alpha-toxin is required for
biomass production during chitin degradation. Appl. biofilm formation: roles of flagella, motility, chemotaxis and biofilm formation by Staphylococcus aureus. J. Bacteriol.
Environ. Microbiol. 66, 3574–3585 (2000). type I pili. Mol. Microbiol. 30, 285–293 (1998). 185, 3214–3217 (2003).

REVIEWS
33. Cucarella, C. et al. Bap, a Staphylococcus aureus surface Links the antibiotic resistance of biofilms to the 81. Donlan, R. M. & Costerton, J. W. Biofilms: survival
protein involved in biofilm formation. J. Bacteriol. 183, stationary phase physiology of cells within the mechanisms of clinically relevant microorganisms. Clin.
2888–2928 (2001). biofilms and the presence of a small phenotypically Microbiol. Rev. 15, 167–193 (2002).
34. Cramton, S. E., Gerke, C., Schnell, N. F., Nichols, W. W. & distinct ‘persister’ population. An excellent, comprehensive review of medically
Cotz, F. The intercellular adhesion (ica) locas is present in 57. Hanlon, G. W., Denyer, S. P., Olliff, C. J. & Ibrahim, L. J. relevant biofilms.
Staphylococcus aureus and is required for biofilm formation. Reduction in exopolysaccharide viscosity as an aid to 82. Gotz, F. Staphylococcus and biofilms. Mol. Microbiol. 43,
Infect. Immun. 67, 5427–5433 (1999). bacteriophage penetration through Pseudomonas aeruginosa 1367–1378 (2002).
35. Hamon, M. A. & Lazazzera, B. A. The sporulation biofilms. Appl. Environ. Microbiol. 67, 2746–2753 (2001). An excellent review of staphylococcal biofilms and
transcription factor Spo0A is required for biofilm 58. Kaplan, J. B., Meyenhofer, M. F. & Fine, D. H. Biofilm growth the molecular mechanisms of adhesion and biofilm
development in Bacillus subtilis. Mol. Microbiol. 42, and detachment of Actinobacillus actinomycetemcomitans. development in staphylococci.
1199–1209 (2001). J. Bacteriol. 185, 1399–1404 (2003). 83. Peters, G., Locci, R. & Pulverer, G. Microbial colonization of
36. Froeliger, E. H. & Fives-Taylor, P. Streptococcus parasanguis 59. Kaplan, J. B. & Fine, D. H. Biofilm dispersal of Neisseria prosthetic devices. II. Scanning electron microscopy of
fimbria-associated adhesion Fap1 is required for biofilm subflava and other phylogenetically diverse oral bacteria. naturally infected intravenous catheters. Zentralb. Bacteriol.
formation. Infect. Immun. 69, 2512–2519 (2001). Appl. Environ. Microbiol. 68, 4943–4950 (2002). Mikrobiol. Hyg. 173, 293–299 (1981).
37. Gavín, R. et al. Lateral flagella of Aeromonas species are 60. Stoodley, P. et al. Growth and detachment of cell clusters 84. Christensen, G. D. Simpson, W. A., Bisno, A. L. & Beachey,
essential for epithelial cell adherence and biofilm formation. from mature mixed species biofilms. Appl. Environ. E. H. Phenotypic variation of Staphylococcus epidermidis
Mol. Microbiol. 43, 383–397 (2002). Microbiol. 67, 5608–5613 (2001). slime production in vitro and in vivo. Infect. Immun. 55,
38. Whitchurch, C. B., Tolker-Nielsen, T., Ragas, P. C. & Mattick, 61. Mattick, J. S. Type IV pili and twitching motility. Annu. Rev. 622–628 (1982).
J. S. Extracellular DNA required for bacterial biofilm Microbiol. 56, 289–314 (2002). 85. Marrie, T. J., Nelligan, J. & Costerton, J. W. A scanning and
formation. Science 295, 1487 (2002). 62. Stoodley, P., Lewandowski, Z., Boyle, J. D. & Lappin-Scott, transmission electron microscopic study of and infected
39. Valle, J. et al. SarA and not sigmaB is essential for biofilm H. M. The formation of migratory ripples in a mixed species endocardial pacemaker lead. Circulation 66, 1339–1341
development by Staphylococcus aureus. Mol. Microbiol. 48, bacterial biofilm growing in turbulent flow. Environ. Microbiol. (1982).
1075–1087 (2003). 1, 447–457 (1999). A seminal paper showing biofilm formation on a
40. Davies, D. G. et al. The involvement of cell-to-cell signals in 63. Inglis, T. J. J. Evidence for dynamic phenomena in residual medical device.
the development of a bacterial biofilm. Science 280, trachael tube biofilm. Br. J. Anaesth. 70, 22–24 (1993). 86. von Eiff, C., Heilmann, C., Hermann, M. & Peters, G. Basic
295–298 (1998). 64. Stewart, P. S. & Costerton, J. W. Antibiotic resistance of aspects of the pathogenesis of staphylococcal polymer-
Demonstrated that cell–cell communication bacteria in biofilms. Lancet. 358, 135–138 (2001). associated infections. Infection 27, S7–S10 (1999).
molecules, associated with the production of 65. Klapper, I., Rupp, C. J., Cargo, R., Purevdorj, B. & Stoodley, 87. Akiyama, H., Huh, W. K., Yamasaki, O., Oono, T. & Iwatsuki, K.
virulence factors, have a role in the structure of P. A viscoelastic fluid description of bacterial biofilm material Confocal laser scanning microscopic observation of
Pseudomonas biofilms, opening the concept that properties. Biotech. Bioeng. 80, 289–296 (2002). glycocalyx production by Staphylococcus aureus in mouse
biofilm structure was genetically regulated. 66. Korstgens, V., Flemming, H. C., Wingender, J. & Borchard, W. skin: does S. aureus generally produce a biofilm on
41. Heydorn, A. et al. Statistical analysis of Pseudomonas Uniaxial compression measurement device for investigation damaged skin? Br. J. Dermatol. 147, 879–885 (2002).
aeruginosa biofilm development: impact of mutations in of the mechanical stability of biofilms. J. Microbiol. Meth. 46, 88. Mack, D. et al. The intercellular adhesion involved in biofilm
genes involved in twitching motility, cell-to-cell signaling, and 9–17 (2001). accumulation of Staphylococcus epidermidis is a linear
stationary-phase sigma factor expression. Appl. Environ. 67. Towler, B. W., Rupp, C. J., Cunningham, A. B. & Stoodley, P. β-1,6-linked glycosaminoglycan: purification and structural
Microbiol. 68, 2008–2017 (2002). Viscoelastic properties of a mixed culture biofilm from analysis. J. Bacteriol. 178, 175–183 (1996).
42. Purevdorj, B., Costerton, J. W. & Stoodley, P. Influence of rheometer creep analysis. Biofouling 19, 279–285 (2003). 89. Heilmann, C., Hussain, M., Peters, G. & Gotz, F. Evidence
hydrodynamics and cell signaling on the structure and 68. Vinogradov, A. M., Winston., M., Rupp., C. J. & Stoodley, P. for autolysin-mediated primary attachment of
behavior of Pseudomonas aeruginosa biofilms. Appl. Rheology of biofilms formed from the dental plaque Staphylococcus epidermidis to a polystyrene surface.
Environ. Microbiol. 68, 4457–4464 (2002). pathogen Streptococcus mutans. Biofilms (in the press). Mol. Microbiol. 24, 1013–1024 (1997).
Demonstrated that environmental factors (in this 69. Espeland, E. M. & Wetzel, R. G. Complexation, stabilization, 90. Heilmann, C., Gerke, C., Perdreau-Remington, F. & Gotz, F.
case, flow) can ‘override’ cell–cell communications as and UV photolysis of extracellular and surface-bound Characterization of Tn917 insertion mutants of
a principal determinant of biofilm structure, glucosidase and alkaline phosphatase: implications for Staphylococcus epidermidis affected in biofilm formation.
illustrating that biofilm development is a multifactorial biofilm microbiota. Microb. Ecol. 42, 572–585 (2001). Infect. Immun. 64, 277–282 (1996).
process influenced by both environmental and 70. Teitzel, G. M. & Parsek, M. R. Heavy metal resistance of 91. Heilmann, C. et al. Molecular basis of intercellular adhesion
genetic factors. biofilm and planktonic Pseudomonas aeruginosa. Appl. in the biofilm-forming Staphylococcus epidermidis.
43. Stoodley, P., Jørgensen, F., Williams, P. & Lappin-Scott, H. M. Environ. Microbiol. 69, 2313–2320 (2003). Mol. Microbiol. 20, 1083–1091 (1996).
in Biofilms: The Good, the Bad, and the Ugly (eds Bayston, 71. McNeill, K. & Hamilton, I. R. Acid tolerance response of 92. Gross, M., Cramton, S. E., Gotz, F. & Peschel, A. Key role of
R., Brading, M., Gilbert, P., Walker, J. & Wimpenny, J. W. T.) biofilm cells of Streptococcus mutans. FEMS Microbiol. Lett. teichoic acid net charge in Staphylococcus aureus colonization
323–330 (BioLine, Cardiff, UK, 1999) 221, 25–30 (2003). of artificial surfaces. Infect. Immun. 69, 3423–3426 (2001).
44. Stoodley, P., Cargo, R., Rupp, C. J., Wilson, S., & Klapper, I. 72. Le Magrex-Debar, E., Lemoine, J., Gelle, M. P., Jacquelin, L. F. 93. Dunne, W. M. Jr & Burd, E. M. The effects of magnesium,
Biofilm mechanics and shear-induced deformation and & Choisy, C. Evaluation of biohazards in dehydrated biofilms calcium, EDTA and pH on the in vitro adhesion of
detachment. J. Industrial Microbiol. Biotech. 29, 361–368 on foodstuff packaging. Int. J. Food Microbiol. 55, 239–234 Staphylococcus epidermidis to plastic. Microbiol. Immunol.
(2002). (2000). 36, 1019–1027 (1992).
45. Boyd, A. & Chakrabarty, A. M. Role of alginate lyase in cell 73. Leid, J. G., Shirtliff, M. E., Costerton, J. W. & Stoodley, P. 94. Vaudaux, P. E, Lew, D. P. & Waldvogel, F. in Infections
detachment of Pseudomonas aeruginosa. Appl. Environ. Human leukocytes adhere, penetrate, and respond to Associated with Indwelling Medical Devices. 2nd Edition
Microbiol. 60, 2355–2359 (1994). Staphylococcus aureus biofilms. Infect. Immun. 70, (eds Bisno, A. L. & Waldvogel, F. A.) (ASM Press,
46. Kaplan, J. B., Ragunath, C., Ramasubbu, N. & Fine, D. H. 6339–6345 (2002). Washington DC, USA, 1994).
Detachment of Actinobacillus actinomycetemcomitans 74. Gilbert, P., Allison, D. G. & McBain, A. J. Biofilms in vitro and 95. Foster, T. J. & Höök, M. Surface protein adhesins of
biofilm cells by an endogenous β-hexosaminidase activity. in vivo: do singular mechanisms imply cross-resistance? Staphylococcus aureus. Trends Microbiol. 6, 484–488 (1998).
J. Bacteriol. 185, 4693–4698 (2003). J. Appl. Microbiol. 92, S98–S110 (2002). 96. Vaudaux, P. E. et al. Use of adhesion defective mutants of
47. Lee, S. F., Li, Y. H. & Bowden, G. H. Detachment of 75. Mah, T. F. & O’Toole, G. A. Mechanisms of biofilm Staphylococcus aureus to define the role of specific plasma
Streptococcus mutans biofilm cells by an endogenous resistance to antimicrobial agents. Trends Microbiol. 9, proteins in promoting adhesion to arteriovenous shunts.
enzymatic activity. Infect. Immun. 64, 1035–1038 (1996). 34–39 (2001). Infect. Immun. 63, 585–590 (1995).
48. Piriou, P., Dukan, S., Levi, Y. & Jarrige, P. A. Prevention of 76. Dibdin, G. H., Assinder, S. J., Nichols, W. W. & Lambert, P. A. 97. Durack, D. T. Experimental bacterial endocarditis. IV.
bacterial growth in drinking water distribution systems. Mathematical model of β-lactam penetration into a biofilm of Structure and evolution of very early lesions. J. Pathol. 115,
Water Sci. Technol. 35, 283–287 (1997). Pseudomonas aeruginosa while undergoing simultaneous 81–89 (1975).
49. Zottola, E. A. & Sasahara, K. C. Microbial biofilms in the food inactivation by released β-lactamases. 98. Höök, E. W. & Sande, M. A. Role of the vegetation in
industry — should they be a concern? Int. J. Food J. Antimicrob. Chemother. 38, 757–769 (1996). experimental Streptococcus viridans endocarditis. Infect.
Microbiol. 23, 125–148 (1994). 77. Anderl, J. N., Zahller, J., Roe, F. & Stewart, P. S. Role of Immun. 10, 1433–1438 (1974).
50. Lowy, F. D. Staphylococcus aureus infections. N. Engl. J. nutrient limitation and stationary-phase existence in 99. Marrie, T. J, Cooper, J. H. & Costerton, J. W. Ultrastructure
Med. 339, 520–532 (1998). Klebsiella pneumoniae biofilm resistance to ampicillin and of cardiac bacterial vegetations on native valves with
51. Pankhurst, C. L., Johnson, N. W. & Woods, R. G. Microbial ciprofloxacin. Antimicrob. Agents Chemother. 47, emphasis on alterations in bacterial morphology following
contamination of dental unit waterlines: the scientific 1251–1256 (2003). antibiotic treatment. Can. J. Cardiol. 3, 275–280 (1987).
argument. Int. Dent. J. 48, 359–368 (1998). 78. Walters, M. C., Roe, F., Bugnicourt, A., Franklin, M. J. & 100. Ramirez-Rhonda, C. H. Adherence of glucan-positive and
52. Raad, I. I. Catheter-related septicemia: risk reduction. Infect. Stewart, P. S. Contributions of antibiotic penetration, oxygen glucan-negative streptococcal strains to normal and
Med. 13, 807–812, 815–816, 823 (1996). limitation, and low metabolic activity to tolerance of damaged heart valves. J. Clin. Invest. 62, 805–814 (1978).
53. Purevdorj, B. & Stoodley, P. in Microbial Biofilms. (eds Pseudomonas aeruginosa biofilms to ciprofloxacin and 101. Fey, P. D. et al. Characterization of the relationship between
Ghannoum, M. A. and O’Toole, G.) (ASM Press, Washington tobramycin. Antimicrob. Agents Chemother. 47, 317–323 polysaccharide intercellular adhesin and hemagglutination in
DC, USA, in the press). (2003). Staphylococcus epidermidis. J. Infect. Dis. 179, 1561–1564
54. Tolker-Nielsen, T. et al. Development and dynamics of 79. Suci, P. A. & Tyler, B. J. A method for discrimination of (1999).
Pseudomonas sp. biofilms. J. Bacteriol. 182, 6482–6489 subpopulations of Candida albicans biofilm cells that exhibit 102. Shiro, H. et al. Transposon mutants of Staphylococcus
(2000). relative levels of phenotypic resistance to chlorhexidine. epidermidis deficient in elaboration of capsular
55. Webb, J. S. et al. Cell death in Pseudomonas aeruginosa J. Microbiol. Methods 53, 313–325 (2003). polysaccharide/adhesin and slime are avirulent in a rabbit
biofilm development. J. Bacteriol. 185, 4585–4592 80. Parsek, M. R. & Singh, P. K. Bacterial biofilms: an emerging model of endocarditis. J. Infect. Dis. 169, 1042–1049 (1994).
(2003). link to disease pathogenesis. Annu. Rev. Microbiol. 57, 103. Sullam, P. M., Bayer, A. S., Foss, W. M. & Cheung, A. L.
56. Spoering, A. L. & Lewis, K. Biofilms and planktonic cells 677–701 (2003). Diminished platelet binding in vitro by Staphylococcus
of Pseudomonas aeruginosa have similar resistance to Discusses biofilm pathogenesis and defines some aureus is associated with reduced virulence in a rabbit
killing by antimicrobials. J. Bacteriol. 183, 6746–6751 clinical criteria for classifying infections with a biofilm model of infective endocarditis. Infect. Immun. 64,
(2001). aetiology. 4915–4921 (1996).

REVIEWS
104. Kuypers, J. M. & Proctor, R. A. Reduced adherence to bronchial tree from patients with cystic fibrosis. Pediatr. active phase over very small distances (micrometres),
traumatized rat heart valves by a low-fibronectin-binding Path. Lab. Med. 15, 707–721 (1995). due to the heterogeneity in nutrient distribution (in
mutant of Staphylococcus aureus. Infect. Immun. 57, 117. Mathee, K. et al. Mucoid conversion of Pseudomonas this case, oxygen) resulting from the mass transfer
2306–2312 (1989). aeruginosa by hydrogen peroxide: a mechanism for characteristics and the shape of the biofilm
105. Moreillon, P. et al. Role of Staphylococcus aureus coagulase virulence activation in the cystic fibrosis lung. Microbiology microcolonies.
and clumping factor in pathogenesis of experimental 145, 1349–1357 (1999). 127. Drenkard, E. & Ausubel, F. M. Pseudomonas biofilm
endocarditis. Infect. Immun. 63, 4738–4743 (1995). Links inflammatory responses of the host to the formation and antibiotic resistance are linked to phenotypic
106. Que, Y. A. et al. Reassessing the role of Staphylococcus emergence of a virulent P. aeruginosa mucoid variation. Nature 416, 740–743 (2002).
aureus clumping factor and fibronectin-binding protein by phenotype. 128. Huang, C. T., Xu, K. D., McFeters, G. A. & Stewart, P. S.
expression in Lactococcus lactis. Infect. Immun. 69, 118. Suter, S., Schaad, U. B., Morgenthaler, J. J., Chevallier, I. & Spatial patterns of alkaline phosphatase expression
6296–6302 (2001). Schnebli, H. P. Fibronectin-cleaving activity in bronchial within bacterial colonies and biofilms in response to
107. Siboo, I. R., Cheung, A. L., Bayer, A. S. & Sullam, P. M. secretions of patients with cystic fibrosis. J. Infect. Dis. 158, phosphate starvation. Appl. Environ. Microbiol. 64,
Clumping factor A mediates binding of Staphylococcus aureus 89–100 (1988). 1526–1531 (1998).
to human platelets. Infect. Immun. 69, 3120–3127 (2001). 119. Saiman, L. & Prince, A. Pseudomonas aeruginosa pili bind 129. deBeer, D. & Stoodley, P. Relation between the structure of
108. Joly, V. et al. Value of antibiotic levels in serum and cardiac to asialoGM1 which is increased on the surface of cystic an aerobic biofilm and mass transport phenomena. Water
vegetations for predicting antibacterial effect of ceftriaxone in fibrosis epithelial cells. J. Clin. Invest. 92, 1875–1880 Sci. Tech. 32, 11–18 (1995).
experimental Escherichia coli endocarditis. Antimicrob. (1993). 130. Donlan, R. M. et al. in Legionella (eds Marre, R. et al.)
Agents Chemother. 31, 1632–1639 (1987). 120. Roger, P. et al. Fibronectin and α5β1 integrin mediate 406–410 (ASM Press, Washington DC, USA, 2002).
109. Lyczak, J. B., Cannon, C. L. & Pier, G. B. Lung infections binding of Pseudomonas aeruginosa to repairing airway 131. Heydorn, A. et al. Quantification of biofilm structures by the
associated with cystic fibrosis. Clin. Microbiol. Rev. 15, epithelium. Eur. Respir. J. 13, 1301–1309 (1999). novel computer program COMSTAT. Microbiology 146,
194–222 (2002). 121. Ofek, I., Hasty, D. L. & Doyle, R. J. (eds). Bacterial Adhesion 2395–2407 (2000).
110. Koch, C. & Hoiby, N. Pathogenesis of cystic fibrosis. Lancet to Animal Cells and Tissues (ASM Press, Washington DC, 132. Yang, X., Beyenal, H., Harkin, G. & Lewandowski, Z.
341, 1065–1069 (1993). USA, 2003). Quantifying biofilm structure using image analysis.
111. Govan, J. R. & Deretic, V. Microbial pathogenesis in cystic 122. Smith, R. S., Harris, S. G., Phipps, R. & Iglewski, B. The J. Microbiol. Methods 39, 109–119 (2000).
fibrosis: mucoid Pseudomonas aeruginosa and Pseudomonas aeruginosa quorum-sensing molecule N-(3-
Burkholderia cepacia. Microbiol. Rev. 60, 539–574 (1996). oxododecanoyl) homoserine lactone contributes to virulence Acknowledgements
112. Costerton, J. W., Irvin, R. T. & Cheng, K. J. The role of and induces inflammation in vivo. J. Bacteriol. 184, Financial support was provided by the National Institutes
bacterial surface structures in pathogenesis. Crit. Rev. 1132–1139 (2002). of Health.
Microbiol. 8, 303–338 (1981). 123. Smith, R. S., Kelly, R., Iglewski, B. H. & Phipps, R. P.
113. Costerton, J. W., Lam, J., Lam, K. & Chan, R. The role of the The Pseudomonas autoinducer N-(3-oxododecanoyl) Competing interests statement
microcolony mode of growth in the pathogenesis of homoserine lactone induces cyclooxygenase-2 and The authors declare that they have no competing financial interests.
Pseudomonas aeruginosa infections. Rev. Infect. Dis. 5, prostaglandin E2 production in human lung fibroblasts:
S867–S873 (1983). implications for inflammation. J. Immunol. 169, 2636–2642
114. Lam, J., Chan, R., Lam, K., & Costerton, J. W. Production of (2002). Online links
mucoid microcolonies by Pseudomonas aeruginosa within 124. Worlitzsch, D. et al. Effects of reduced mucus oxygen
infected lungs in cystic fibrosis. Infect Immun. 28, 546–556 concentration in airway Pseudomonas infections of cystic DATABASES
(1980). fibrosis patients. J. Clin. Invest. 109, 317–325 (2002). The following terms in this article are linked online to:
A seminal paper suggesting Pseudomonas 125. Yoon, S. S. et al. Pseudomonas aeruginosa anaerobic Entrez: http://www.ncbi.nlm.nih.gov/Entrez/
pneumonia in cystic fibrosis is a biofilm infection. respiration in biofilms: relationships to cystic fibrosis Escherichia coli | Pseudomonas aeruginosa
115. Singh, P. K. et al. Quorum-sensing signals indicate that pathogenesis. Dev. Cell. 3, 593–603 (2002).
cystic fibrosis lungs are infected with bacterial biofilms. 126. Xu, K. D., Stewart, P. S., Xia, F., Huang, C. T. & McFeters, G. A. FURTHER INFORMATION
Nature 407, 762–764 (2000). Spatial physiological heterogeneity in Pseudomonas biofilm movies: http://www.erc.montana.edu/Res-Lib99-
Presents several criteria to show that biofilm aeruginosa biofilm is determined by oxygen availability. Appl. SW/Movies/Database/MD-DisplayScript.asp
infections are present in cystic fibrosis. Environ. Microbiol. 64, 4035–4039 (1998). web-based version of Figure 1:
116. Potts, S. B., Roegli, V. L. & Spock, A. Immunohistologic Showed that biofilm cells could exhibit a wide range http://www.erc.montana.edu/MultiCellStrat/default.html
quantification of Pseudomonas aeruginosa in the tracheo- of physiologies from stationary phase to a highly Access to this interactive links box is free online.
View publication stats

Biofilm Review2004

Uploaded by

Copyright:

Available Formats

Biofilm Review2004

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Biofilm Review2004

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Bacterial Bioﬁlms: From the Natural Environment to Infectious Diseases

Article in Nature Reviews Microbiology · March 2004

Luanne Hall-Stoodley Paul Stoodley

SEE PROFILE SEE PROFILE

Biofilms in upper respiratory infections View project

Biofilm Conceptual Models View project

The user has requested enhancement of the downloaded file.

BACTERIAL BIOFILMS: FROM

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 9 5

7 Although antimicrobials damage

1 Starvation can induce bacteria

2 Active bacteria can attach to almost

5 The close proximity of cells

96 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 9 7

98 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 9 9

a b shear environments through EPS modification. A better

Why do prokaryotes form biofilms?

100 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 1 0 1

102 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 1 0 3

104 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 1 0 5

106 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

NATURE REVIEWS | MICROBIOLOGY VOLUME 2 | FEBRUARY 2004 | 1 0 7

108 | FEBRUARY 2004 | VOLUME 2 www.nature.com/reviews/micro

View publication stats

You might also like