ReViews

The biofilm life cycle: expanding the

conceptual model of biofilm formation
Karin Sauer   1,2, Paul Stoodley   3,4,5,6, Darla M. Goeres   7, Luanne Hall-​Stoodley   3,
Mette Burmølle   8, Philip S. Stewart   7 and Thomas Bjarnsholt   9,10 ✉
Abstract | Bacterial biofilms are often defined as communities of surface-​attached bacteria and
are typically depicted with a classic mushroom-​shaped structure characteristic of Pseudomonas
aeruginosa. However, it has become evident that this is not how all biofilms develop, especially
in vivo, in clinical and industrial settings, and in the environment, where biofilms often are
observed as non-​surface-​attached aggregates. In this Review, we describe the origin of the cur­
rent five-​step biofilm development model and why it fails to capture many aspects of bacterial
biofilm physiology. We aim to present a simplistic developmental model for biofilm formation that
is flexible enough to include all the diverse scenarios and microenvironments where biofilms are
formed. With this new expanded, inclusive model, we hereby introduce a common platform for
developing an understanding of biofilms and anti-​biofilm strategies that can be tailored to the
microenvironment under investigation.

Aggregates
Cohesive groups of microbial
cells surrounded by extracellu-
lar polymeric substances and
other entrapped abiotic or
biotic materials. Microbial
aggregates can be surface
attached, matrix associated or
free floating in the liquid phase
and display a biofilm-​like
phenotype.
Biofilms
Microbial aggregates attached
or associated with a surface
and embedded in a matrix.
These can include single or
multiple discrete aggregates or
more continuous films.
Aggregation
Any biological, chemical or
physical process that enables
microbial cells to form an
aggregate.
✉e-​mail: tbjarnsholt@
sund.ku.dk
https://doi.org/10.1038/
s41579-022-00767-0
Over the past 40 years, microbiologists have catego-
rized bacteria as displaying two life forms in nature. In
one, bacteria appear as single, independent, free-​floating
cells (planktonic); in the other, bacteria are organized
in microbial aggregates (biofilms). In addition, the word
‘biofilm’ originally referred to biomaterial on a sur-
face; however, more recently, non-​surface-​attached
aggregated bacteria have also been recognized as
biofilms1–3. Likewise, non-​surface-​associated aggre-
gates are now recognized in clinical settings, and
chronic biofilm infections are further divided into
surface-​associated or non-​surface-​associated infec-
tions. Surface-​associated infections are commonly
observed in patients with implants or medical devices.
Non-​surface-​associated infections include respiratory
tract infections with impaired host mucociliary clear-
ance (in viscous airway mucus in individuals with
cystic fibrosis) or persistent soft tissue infections that
are associated with comorbidities such as diabetes and
impaired vascularization of the lower limbs predispos-
ing to non-​healing wounds4. Until recently, bacteria
growing planktonically have been associated with acute
infections that are generally treatable with antibiot-
ics, although successful treatment mainly depends on
accurate and fast diagnosis. When bacteria succeed in
forming biofilms within the host, the infection is often
untreatable and, sustained by low-​grade inflammation,
develops into a chronic state5. However, this dogma is
challenged by recent findings that suggest that the differ-
ence between bacteria in acute and chronic infections is
due to metabolic activity rather than aggregation6.
In the environment, the functional consequences
of bacterial life in biofilms have been associated with
enhanced protection from shear stress, desiccation,
toxic compounds and protozoan grazing7. Moreover,
retention of enzymes in the biofilm matrix was pro-
posed to improve the efficiency and diversity of organic
matter decomposition, and biofilm formation on plant
roots and fungal cells may respectively promote bac-
terial nutrient acquisition and transport8. Pathogenic
biofilms that form on plants may also have serious dis-
ease consequences9–11. Although (motile) planktonic
cells are primarily found in water columns and soil
pores, the predominant forms of microbial life in nat-
ural environments are linked to highly diverse biofilm
communities in aquatic environments (including sedi-
ments, submerged surfaces, as free-​floating flocs and on
higher organisms) and soil (for example, on litter, plant
roots and soil particles)12. Likewise, biofilms dominate
in industrial microbial applications such as cleaning of
wastewater and bioremediation of soil and water13,14.
Biofilms are associated with microbially induced
corrosion in oil-​f ield pipelines, plugging of pipes,
fouling of ship hulls creating drag and increased fuel
costs, reductions in heat transfer in cooling towers and
fouling of manufacturing lines resulting in product
contamination14. In all these instances, the industrial sys-
tem is not sterile, and so it is not necessarily an issue that
bacteria are present but more that the biofilm compro-
mises a product or system performance. In these cases,
biocide manufacturers develop clean-​in-​place proce-
dures to control biofilm growth, but, in reality, the biofilm

Nature Reviews | Microbiology

0123456789();:
Reviews

Growth
is never completely eliminated, and, thus, as with dental of Pseudomonas aeruginosa (a nosocomial pathogen),
Expansion of aggregates by biofilms, routine cleaning and maintenance is required presenting the current five-​step biofilm model27,30 (Fig. 1).
microbial growth and concomi- to keep biofouling in check, for example, in the souring The developmental stages of biofilms are referred to as
tant production of extracellular of oil reservoirs by sulfate-​reducing bacteria15. reversible and irreversible attachment, biofilm matura-
polymeric substance, whether
in suspension or attached to a
A common denominator of bacterial biofilms is tion I and II (involving cluster and microcolony forma-
surface. the distinction between surface-​attached and non- tion, respectively), and dispersion27,30. Variations of this
​surface-​attached bacterial aggregates despite new evi- model have also been developed for other species such
Attachment dence showing that these share similar phenotypes16,17. as Staphylococcus aureus31 and the soil bacterium Bacillus
Suspended single cells or
For both of these phenotypes, the bacteria create subtilis32 as well as for algal biofilms33.
aggregates adhere to a host
cellular surface or an abiotic
microenvironments that in turn influence bacterial Although the schematic conceptual biofilm develop-
surface, either directly to the community and behaviour in an interdependent and mental model based on P. aeruginosa in vitro biofilm
substratum or to previously dynamic manner18,19. In this Review, we define bacte- formation is easy to understand and has been widely gen-
attached microbial cells or rial aggregates as biofilms irrespective of attachment to eralized to describe all biofilms, this model does not nec-
clusters.
a biotic or abiotic surface and define the aggregation of essarily describe the complexity of biofilms in real-​world
Detachment bacteria as the central hallmark of bacterial biofilms. industrial, natural and clinical settings. Importantly, this
An overarching term Although bacteria have been studied in the labora- model does not reflect the relevant microenvironments
encompassing all phase tory for well over 100 years, biofilms were first studied that develop within these biofilms. It is therefore impor-
transfer processes in which
after surface-​attached bacteria were observed attached to tant to consider the substantial differences between the
microbial cells and extracellular
polymeric substances move
the pacemaker lead in a patient suffering from recurrent processes occurring in a laboratory flow cell and those
from the surface-​attached bacteraemia20 and growing on glass slides inoculated leading to biofilm formation in real-​world scenarios,
phase to a fluid-​borne phase. with sea water21. The bacteria attached to the pacemaker including in the rhizosphere, in oil pipelines, in chronic
This term is specific to lead mark one of the first references to ‘biofilm-​growing wounds, in the respiratory tract at an air–water inter-
surface-​attached biofilms.
bacteria’ in medicine, with a subsequent increased inter- face (that is, a pellicle), around a prosthetic joint, in a
Dispersal est in biofilm infections. Numerous in vitro systems have wastewater granule, in microcosm ecosystems34–36 and
Specifically connotes an active been devised to study biofilm formation22–24 and how on ex vivo organoid-​associated biofilms. In such diverse
and biologically regulated biofilm bacteria differ from planktonic cells, including systems, the processes of attachment, aggregation, inter-
release of microbial cells from
their hallmark property of increased antibiotic tolerance action with biotic or abiotic materials and interfaces (for
a suspended or attached
biofilm aggregate.
or the presence of a matrix composed of extracellular example, roots, tissue, a gas phase, environmental pol-
polymeric substances (EPS), which is a hydrogel-​like ymers or corrosion deposits), growth and maturation,
Disaggregation matrix encasing biofilm cells25,26. These initial findings and detachment and dispersal are potentially quite differ-
Aggregated cells, whether in supported the notion that microorganisms undergo ent and do not necessarily occur sequentially. Given the
suspension or associated with
a surface, that shed smaller
substantial changes in their phenotypic repertoire dur- variety of systems and conditions, we propose to expand
microbial aggregates or ing the transition from planktonic to biofilm growth the existing model to include a wider range of real-​world
individual cells into the fluid (Box 1) and revealed the potential for new ways to con- scenarios.
phase. trol or manipulate biofilms. In vitro systems commonly In this Review, we describe the origin of the cur-
used shaken, well-​mixed cultures and led to most bio- rent biofilm model and its shortcomings. Additionally,
film experiments being initiated by using single-​cell we discuss differences in biofilm formation by diverse
planktonic cultures with one controlled seeding event. types of bacteria in varied experimental systems, both
Likewise, the transformation of single cells into ses- in vitro and in vivo, and focus on new findings, such as
sile biofilm communities has been thoroughly studied the common lack of an attachment surface, differences
in closed, surface-​based in vitro systems without the in matrix properties and transcriptional profiles, among
influx of new cells during the biofilm formation and others, that warrant amending the current model. We
maturation process27–29. Such studies led to a key pub- suggest a revised conceptual model that encompasses the
lication in the field describing the developmental stages three major steps of biofilm life, aggregation, growth and
disaggregation independently of surfaces, and initiation
from single-​cell planktonic bacteria, to present a revised
Author addresses and simple model that we believe represents a broader
1
Department of Biological sciences, Binghamton university, Binghamton, NY, usa. range of biofilm systems.
2
Binghamton Biofilm research Center, Binghamton university, Binghamton, NY, usa.
3
Department of Microbial infection and immunity, College of Medicine, the Ohio state The origin of the five-​step biofilm model
university, Columbus, OH, usa. Numerous studies support the notion that biofilm for-
4
Department of Orthopaedics, the Ohio state university, Columbus, OH, usa. mation starts with the initial surface attachment of sin-
5
Department of Microbiology, the Ohio state university, Columbus, OH, usa. gle, free-​floating, planktonic cells and that biofilm cells
6
National Biofilm innovation Centre (NBiC) and National Centre for advanced tribology differ from their planktonic counterparts in the genes
at southampton (nCats), Mechanical engineering, university of southampton, and proteins that they express (Box 1). Given the pro-
southampton, uK. found changes that microorganisms undergo during
7
Department of Chemical and Biological engineering and Center for Biofilm engineering,
their transition from planktonic organisms to cells that
Montana state university, Bozeman, Mt, usa.
8
Department of Biology, Faculty of science, university of Copenhagen, Copenhagen, are part of a complex, surface-​attached community, it is
Denmark. not surprising that the transition from the planktonic
9
Department of immunology and Microbiology, Faculty of Health and Medical science, to the biofilm mode of growth is a complex and highly
university of Copenhagen, Copenhagen, Denmark. regulated process that is often regarded to be develop-
10
Department of Clinical Microbiology, rigshospitalet, Copenhagen, Denmark. mental. However, although it was widely accepted that

www.nature.com/nrmicro

0123456789();:

Box 1 | Studying biofilms
Biofilm research in the early years primarily focused on engineering applications and
observational descriptions of biofilms. However, biofilm research changed with the
observation of surface attachment-​specific gene regulation in vitro and the introduc-
tion of in vitro systems to study biofilm formation and phenotypes in the laboratory. This
facilitated the study of specific and differential gene expression upon surface attach-
ment in vitro42,43,49, including the role of cell signalling in genetic regulation from a
population49 as well as the use of genetic tools to identify genes required for in vitro
surface and subsequent biofilm formation39,161,162.
the idea that biofilms are amenable to molecular genetic studies39,161 also opened the
door to the exploration of factors beyond early surface attachment, including those con-
tributing to biofilm architecture, metabolic interactions, phylogenetic groupings, com-
petition and cooperation. Molecular genetic applications furthermore led to exciting
progress in the development of new technologies for the study of biofilm communities,
advanced our understanding of the ecological significance of surface-​attached bacteria
and provided new insights into the molecular genetic basis of biofilm development40.
what followed was extensive research on genes that are required for bacteria to asso-
ciate with surfaces and investigations of differences in the transcriptional abundance of
bacterial genes when growing planktonically and as biofilms. although some studies
failed to detect differences in the transcriptomes of planktonic and surface-​associated
cells163, the majority of studies confirmed that planktonic and sessile biofilm cells dis-
play distinct transcriptomic profiles, with the number of genes changing in transcript
abundance upon surface-​associated growth ranging from less than twenty to several
hundred164,165. Moreover, numerous reports indicate that biofilms are heterogeneous,
with bacteria residing at different locations within the biofilm structure experiencing
steep gradients in the concentration of nutrient resources, oxygen and waste products
(such as acids produced by fermentation in oxygen-​depleted zones) as well as of
metabolites and extracellular signalling molecules, resulting in the modulation of meta-
bolic rates, dormancy, stress responses and mutation rates155,166–168. transcriptome anal-
yses of in vitro-​grown biofilms confirmed that biofilm cells experience various stresses,
including hypoxia or oxygen deprivation, nutrient stress, and slow growth, which
increase as the biofilm grows in size59,169,170. imaging and transcriptome imaging (parallel
sequential fluorescence in situ hybridization (par-​seq FISH)) have provided visual evi-
dence of the presence of subpopulations and associated gradients (that is, chemical
and signalling molecules) within the biofilm structure114,134,171,172. additionally, changes
in cell–cell signalling, virulence gene expression and the biosynthesis of matrix compo-
nents have been reported173–177. Notably, many of these findings have been confirmed
using in vivo biofilms (animal models), although not in human infections178,179.
the transition to a surface lifestyle was a highly regu-
lated process, it remained unknown whether subse-
quent surface-​associated growth progressed simply as
an accumulation of cells due to cell division or instead
coincided with distinct events indicative of progressive
or transitional changes over the course of biofilm for-
mation. In 2002, in an effort to better understand the
progression of biofilm formation, researchers27 used a
combination of direct observation by microscopy, eval-
uation of biofilm morphology, matrix polymer produc-
tion, activation of quorum sensing-​regulated genes and
the quantitative analysis of protein abundance. The anal-
ysis led to the realization that, over the course of biofilm
formation, P. aeruginosa displays multiple phenotypes
with distinct physiological characteristics (structural and
metabolic changes) that can be correlated to distinct epi-
sodes or stages of biofilm development (Fig. 1). These
stages were referred to as reversible and irreversible
Accumulation attachment, maturation (maturation I and II stages), and
The net result of attachment, dispersion, with each biofilm developmental stage corre-
aggregation, growth, sponding to unique patterns of protein production and
disaggregation and gene expression30,37–41. The distinction between reversible
detachment processes that
leads to expansion or
and irreversible attachment was based on the timescale
shrinkage of a biofilm or of the fate (whether it remains attached or detaches) of a
aggregate. cell over a few minutes once it contacts a surface.
Nature Reviews | Microbiology
0123456789();:
Reviews
In P. aeruginosa, the reversible attachment stage is
characterized by cells attaching to a surface by a sin-
gle pole (Fig. 1). Most surface contact is unstable and
cells are often seen returning to the bulk phase. Once
rod-​shaped cells commit to a more stable surface
existence, cells attach via their longitudinal axis; this
phenomenon is referred to as ‘irreversible attachment’
(Fig. 1). Furthermore, reports suggest that irreversible
attachment initiates a cascade of changes in bacterial
cells. Apparent changes following bacterial attachment
include cessation of flagella-​mediated motility while at
the molecular level, changes include surface-​induced
gene activation of P. aeruginosa algC (a gene involved
in lipopolysaccharide core and exopolysaccharide
alginate biosynthesis 42,43) and induction of genes
involved in the biosynthesis of the Psl matrix polymer44
and of genes linked to antibiotic resistance, including
β-​lactamase45, phenazine46, SagS and BrlR47. The find-
ings suggest that committing to the surface-​associated
mode of growth not only coincides with the produc-
tion of biofilm matrix components that enable cells to
cement themselves more firmly to the surface but also
with biofilm antimicrobial tolerance (a hallmark charac­
teristic of biofilms) as an early adaptive response to the
sessile lifestyle. Once attached, cells will grow into a
more complex multicellular mature form, which in some
bacterial species, including P. aeruginosa, is character-
ized by the presence of differentiated, mushroom-​like
or pillar-​like structures or microcolonies interspersed
with fluid-​filled channels48 (Fig. 1). The structuring of
biofilms in microcolonies with water channels has been
shown to be dependent on intercellular small messenger
molecules (acylated homoserine lactones) that are used
for bacterial communication49, rhamnolipids50,51 and
regulatory proteins, mostly two-​component regulatory
systems (TCSs)52–54. However, in P. aeruginosa, even cell
signalling knockout mutants have been shown to form
such channels, which suggests that their structure is
determined by the interplay between intrinsic bacterial
regulation and environmental conditions55. As biofilms
develop a three-​dimensional (3D) structure, resident
bacteria near the base increasingly undergo physical sep-
aration from the bulk–liquid interface and from essential
sources of energy or nutrients, with biofilm cells expe-
riencing an ever-​changing microenvironment. Changes
are driven by cellular crowding, chemical gradients and
nutrient competition, which lead to stratification within
the biofilm and the creation of subpopulations56,57. Thus,
bacteria residing at different locations within the biofilm
structure experience concentration gradients of nutri-
ent resources, oxygen, waste products (such as acids
produced by fermentation in oxygen-​depleted zones)
and extracellular signalling molecules56–58. Supporting
evidence is provided by the observation that resident
biofilm cells express genes linked to oxygen depriva-
tion, general stress and stationary phase conditions,
nutrient stress and slow growth 56–60. Importantly,
cells can leave the biofilm structure and return to the
planktonic mode of growth by a process referred to as
dispersion61. Dispersion is an active event whereby ses-
sile, matrix-​encased biofilm cells escape from the bio-
film, leaving behind eroded biofilms and biofilms with
Reviews
Planktonic cell
Microcolonies
Cell cluster
Irreversible
EPS
attachment
Surface
Reversible Maturation I Maturation II Dispersion
attachment
Fig. 1 | The original five-step model of biofilm development. The original model of bio-
film formation is based on key publications investigating Pseudomonas aeruginosa. The
model proposed that the formation of biofilms is a cyclic process that occurs in a five-​stage-
specific and progressive manner. The process is initiated following surface contact by
single planktonic cells. Several developmental steps are discernible, including reversible
attachment, irreversible attachment, biofilm maturation (maturation I and maturation II)
and, finally, dispersion27,37. During reversible attachment, bacteria attach to the substra-
tum via the cell pole or via the flagellum, followed by longitudinal attachment. Transition
to irreversible attachment coincides with a reduction in flagella reversal rates, a reduction
in flagella gene expression and the production of biofilm matrix components. This stage is
also characterized by attached cells demonstrating drug tolerance47. Biofilm maturation
stages are characterized by the appearance of cell clusters that are several cells thick and
are embedded in the biofilm matrix (maturation I stage), which subsequently fully mature
into microcolonies (maturation II stage)27,37. Dispersion has been reported to coincide with
the decrease in and degradation of matrix components, with dispersed cells being motile
and demonstrating increased drug susceptibility relative to biofilm cells. A shortcoming of
the current biofilm model is that it does not account for non-​surface-​attached aggregates
that are often observed in clinical or environmental settings (see Fig. 5). EPS, extracellular
polymeric substances. Adapted with permission from ref.30, Annual Reviews.
central voids27,30,62–64. Not surprisingly, dispersion (also
referred to as seeding dispersal) is also considered to be
the next stage of biofilm formation, an active event that
leads to bacterial dissemination and the colonization
of new locations65. An additional key regulator of the
biofilm developmental life cycle is the ubiquitous bac-
terial second messenger cyclic diguanylate (c-​di-​GMP),
with high c-​di-​GMP levels favouring the biofilm mode
of growth, whereas low levels have been associated with
planktonic and dispersed cells66–68. c-​di-​GMP is required
to mediate surface sensing, repress motility upon surface
attachment, and increase biosynthesis of biofilm matrix
components, with reduced c-​di-​GMP levels contributing
to dispersion66–69.
The findings above suggested an expanded model
of biofilm development by P. aeruginosa that detailed
the progression of biofilm formation as well as the
stage-​specific formation of biofilms30. Although the
model represented the developmental stages specifically
for P. aeruginosa biofilms, it has become widely used to
represent biofilm formation by diverse biofilm-​forming
microorganisms in various settings, for example, bio-
films growing in extreme environments70 and microalgal
biofilms71.
Owing to its elegant simplicity, the developmen-
tal model of microbial biofilm formation was adopted
quickly by the scientific community as the major con-
ceptual framework for biofilm research on which to base
empirical research and scientific inference. As discussed
below, the ability to extrapolate this model to biofilms
0123456789();:
outside the laboratory — in nature, engineered systems
and medicine — has been limited because the diver-
sity and complexity of biofilm structures and processes
in real-​world, non-​laboratory systems were not consid-
ered. Also importantly, the logistical difficulty in stud-
ying the initiation and development of biofilms in vivo
in real time means that the bulk of our understanding
is extrapolated from snapshots in time. The study of
suspended aggregates in the bulk liquid is even more
challenging as these do not stay in the same place over
timescales relevant to observe developmental processes.
Limitations of the five-​step model
We have identified at least four limitations of the con-
ceptual model described above. First, it has not yet been
determined whether biofilm formation can be described
as a true developmental process when we consider bio-
films formed outside of the flow cell and by species other
than P. aeruginosa and S. aureus as model biofilm species.
Moreover, the model does not capture the wide variety
of biofilm architectures observed in real-​world systems
such as microbial mats, which can be highly stratified
along horizontal layers72. The model also does not
incorporate the diversity of aggregation (see below) and
detachment mechanisms now recognized in the field by
both motile and non-​motile organisms such as S. aureus;
however, the model has been adapted to accommodate
this organism31. Finally, the model does not consider the
succession of events in biofilms formed in open systems
with a continuous influx of new colonizers. Indeed, even
for dental biofilms, for which it is recognized that biofilm
progression proceeds as an ecological succession with
new species proliferating in different parts of the bio-
film, the single-​species model is commonly depicted73.
Likewise, applying the five-​step biofilm model to indus-
trial systems is limited. These systems are so large and
complex that it is likely that all stages of attachment,
growth and detachment occur simultaneously at vari-
ous points in the system. The tidy description of how
biofilms form in a simple, rich media laboratory sys-
tem does not necessarily capture the complexity of
biofilms in most industrial or environmental systems,
where surface characteristics, such as scale or corro-
sion, the chemical properties of the bulk fluid and the
fluid dynamics, will influence how the biofilm attaches,
grows and detaches. This also applies to infection sites,
where it remains unknown if the site is seeded with sin-
gle cells or aggregates, or if bacteria are trapped within
host material in a complex environment rather than just
forming aggregates by clonal expansion. Furthermore,
there are no in situ sensors that can be incorporated
into these complex systems to directly monitor bio-
films on surfaces, in fluid suspensions or associated
with host materials. Sections of the system can be sam-
pled during upgrades or replacement but these only
give a snapshot in time at specific locations. Although
sampling fluids can give clues that biofilms might be
present through the capture and release of shed cells
or aggregates, all that is known is that these originated
upstream in the system.
Most host-​associated biofilms are subject to strong
host selection factors. In the gut of humans and higher
www.nature.com/nrmicro

Removal
animals, biofilm diversity is regulated through com-
Implies the response to a pounds excreted from gut epithelial cells74. This is also
mechanical, chemical or the case for other host-​associated biofilms, including
enzymatic intervention that those formed on plant roots that are shaped by plant exu-
causes attached aggregates or
cells to be released from the
dates to attract growth-​promoting or nutrient-​capturing
surface. bacteria75, and biofilms formed on algae specifically select
for a stable core set of functional genes76. Selective host–
microorganism interactions may lead to very complex
and less sequential events of bacterial attachment and
detachment that are not considered in the five-​step model.
The paradigmatic value of this model was first chal-
lenged by the research community as early as 2009
(ref.77), questioning the validity of the model and the
concept of biofilm formation as a developmental process.
Based on definitions by several researchers78,79, reviewed
elsewhere77, development coincides with changes in
form and function that are part of the normal life cycle
of the cell. This is regulated by a dedicated, hierarchically
ordered genetic pathway and stage-​specific transitions
in response to environmental cues. If biofilm formation
is indeed a regulated developmental process, the for-
mation of biofilms would require genetic pathways that
evolved to facilitate cooperation among members of the
biofilm. Although it is undeniable that a community of
cells form a biofilm, that biofilm formation coincides
with surface structure and temporal changes, and that
several regulators affecting biofilm formation have
been identified80–82, no such genetic pathway regulating
these morphological changes and stage-​specific transi-
tions in a hierarchically ordered manner has yet been
identified77. However, in the same year, another group37
reported a previously uncharacterized signal transduc-
tion network regulating committed biofilm develop-
mental steps by P. aeruginosa following attachment, in
which phospho-​relays and response regulators seemed
to be key components of the regulatory machinery that
coordinates gene expression during biofilm develop-
ment in response to environmental cues. More specifi-
cally, the signalling network is composed of several TCSs
named SagS, BfiSR, BfmRS and MifRS37. TCS activation
occurred sequentially (SagS–BfiSR–BfmRS–MifRS)
over the course of biofilm formation, whereas inactiva-
tion of these systems arrested biofilm formation at dis-
tinct developmental stages. ΔsagS and ΔbfiSR biofilms
a Glucose minimal medium b Citrate minimal medium c Benzoate minimal medium
20 μm
20 μm 20 μm
Fig. 2 | Variation in Pseudomonas aeruginosa biofilm architecture. Pseudomonas
aeruginosa PAO1 grown in flow cells under flow conditions but with different carbon
sources shows remarkably different three-​dimensional architectures. Biofilms grown on
glucose minimal medium demonstrated the typical mushroom-​shaped multicellular
biofilm structures (part a); growth in minimal medium containing citrate, casamino acids
(part b) or benzoate (part c) as the carbon source led to the formation of flat unstructured
biofilms29. Reprinted with permission from ref.29, Wiley.
Nature Reviews | Microbiology
0123456789();:
Reviews
arrested at the irreversible attachment stage whereas bio-
films formed by ΔbfmRS and ΔmifRS mutants arrested
at the maturation I and maturation II stages of biofilm
development, respectively37,53,83–85.
Although the discovery of the signal transduction
network strongly supported the idea that the formation
of biofilms was a biologically regulated developmental
process, at least for P. aeruginosa grown under labora-
tory conditions, other concerns remained, including the
validity of the representation of the biofilm structure or
architecture being composed of mushroom-​like micro-
colonies. In fact, several reports demonstrated that, even
for P. aeruginosa, the biofilm architecture varied with
growth conditions and the growth medium. For instance,
a study29 demonstrated that, although P. aeruginosa PAO1
biofilms grown on glucose minimal medium demon-
strated the typical mushroom-​shaped multicellular bio-
film structures, growth in minimal medium containing
citrate, casamino acids or benzoate as the carbon source
led to the formation of flat unstructured biofilms (Fig. 2).
In multispecies biofilms, different medium composi-
tion affects not only biofilm morphology but also spe-
cies composition86. In addition to growth medium and
nutrient sources, other variations in growth conditions
have been reported to influence the biofilm architecture.
Although P. aeruginosa forming mushroom-​shaped bio-
films has been associated with growth under relatively
low flow rate conditions87, static growth conditions favour
the formation of pellicles that form at the air–liquid
interface88. However, at higher flow rates, structures
such as streamers and ripples can form, demonstrating
the remarkable ability of biofilms to adapt to the physical
conditions under which they grow55,89.
Additionally, the organisms composing the biofilm
also have a marked effect on the biofilm structure. For
example, in comparison to pure cultures of laboratory-
​g rown biofilms of either Klebsiella pneumoniae or
P. aeruginosa, biofilms containing both species were
thicker90,91. Moreover, in a mixed species biofilm com-
posed of four bacterial soil isolates, removal of one biofilm
member completely changed biofilm morphology, spe-
cies structural organization and relative abundance, even
when the species removed was initially of low abundance
and intrinsically weak in biofilm formation capability92.
Biofilms formed by the Gram-​positive bacteria S. aureus93
and Streptococcus pneumoniae94, while having a hetero-
geneous appearance indicative of the presence of water
channels, lack the distinct microcolonies that had become
a distinct feature of the biofilm architecture. By contrast,
studies with pneumococcal biofilms formed under static
conditions were used to investigate chronic otitis media
with effusion as fluid and flow is severely disrupted in
the middle ear during infection. Biofilm structures were
also dependent on bacterial strains but were smaller
(5–15 μm), resembling the appearance of pneumococ-
cal biofilms from ex vivo middle ear mucosa samples
from children with chronic otitis media95,96. Similarly,
non-​typeable Haemophilus influenzae, a Gram-​negative
bacterium, also formed biofilm aggregates in these otitis
media samples as recapitulated in a chinchilla model of
otitis media97. Non-​typeable H. influenzae biofilms also
formed on differentiated airway epithelial cultures from
Reviews

a Mushroom structure of P. aeruginosa Fig. 3 | Diverse surface-attached and non-surface-

biofilm in vitro in a flow cell
b Mucus-embedded aggregates of P. aeruginosa surrounded b | Mucus-​embedded aggregates of P. aeruginosa sur-
by PMLs in a cystic fibrosis lung
~20 µm
c Wound-embedded aggregates of P. aeruginosa
surrounded by PMLs
attached biofilm structures. Variety of biofilm structures
underscoring differences between in vitro and in vivo or
environmental biofilms. Original images are shown in the
left column and a schematic drawing of the structures and
their organization are provided in the right column with
shading denoting water (blue), aggregated microbial cells
(purple) and their extracellular polymeric substances
(yellow), host cells and other material, including mucus
or tissue (red), and the attachment surface (hatched grey).
a | Mushroom structure of Pseudomonas aeruginosa biofilm
in vitro in a flow cell (GFP-​tagged bacterial cells (green)).
rounded by polymorphonuclear leukocytes (PMLs) in a
cystic fibrosis lung (bacterial cells (red), host DNA (blue))105.
c | Wound-​embedded aggregates of P. aeruginosa sur-
rounded by PMLs (bacterial cells (red), host cells/DNA
(blue), autofluorescence (green))157. d | Aerobic granules
from a full-​scale AquaNereda wastewater treatment pro-
cess (lipids (red), proteins (green), extracellular polysaccha-
rides (pink, white)). e | Striated microbial mat from a
Brazilian lake (natural coloration of the microbial mat)158.
Part a courtesy of T.B. Part b reprinted with permission from
ref.105, Wiley. Part c reprinted with permission from ref.157,
American Society for Microbiology. Part d courtesy of
K. Bodle and C. Kirkland, Center for Biofilm Engineering,
Montana State University, USA. Part e reprinted with
permission from ref.158, Elsevier.

antibiotic therapy and persistence in the host despite

d Aerobic granules from a full-scale wastewater
treatment process
600 µm
e Striated microbial mat from a Brazilian lake
0 which select for a microbial community capable of iron
1
2
3
patients with primary ciliary dyskinesia98. These smaller
aggregated structures suggest that biofilms with highly
complex 3D architectures may be less likely to form in
the host microenvironments99. Importantly, even discrete
biofilm aggregates may still be able to induce inflamma-
tion and tissue destruction that leads to sustained chronic
infection because they are associated with tolerance to
robust innate immune responses.
Mixed-​species biofilms taken from the environment
are structurally very diverse. As an example, microbial
mats are thick and layered, whereas bacterial aggregates
on sand grains are thin and small72. In addition, environ-
mental biofilms often form on biodegradable materials,
and thus the nutrients are not provided only from ‘above’,
potentially affecting the growth zones and structure of
the biofilm. Biofilm structure and community members
can also be strongly influenced by the underlying sub-
stratum as observed in marine biofilms examined ex
vivo100. Generally, surface topography dictates specific
bacterial adhesion, directly affecting the biofilm com-
position, and thus biofilm structure, owing to intrinsic
biofilm formation capabilities (including matrix com-
ponent production) of the individual species present.
More specifically, substrata may select for specific meta­
bolic community profiles, exemplified by surfaces con-
taining elemental iron and its various oxidative states,
oxidation and reduction100.
Absence of an attachment surface
As reviewed by several groups4,101, many of the chronic
bacterial infections linked to biofilms that involve aggre-
gated bacteria and antimicrobial recalcitrance may not
involve hard surface attachment, even if a surface is
present (Fig. 3). Likewise, biofilms in the environment
are often free floating, including diverse bacterial
aggregates (granules) formed in wastewater treatment
plants102 or those in marine, lake and river habitats, com-
monly referred to as ‘marine snow’12. A study103 used a
sophisticated experimental system to hold a sinking
diatom aggregate in place while measuring the internal

www.nature.com/nrmicro

0123456789();:

Polymer bridging
The aggregation of microbial
cells in suspension caused by
polymers that adhere to cell
wall components forming
bridging bonds between
multiple cells.
Sloughing
The release of coherent layers
of surface-​attached biofilm by
adhesive failure (that is, at the
biofilm–substratum interface),
generally by fluid shear. This
mechanism is specific to
surface-​attached biofilms.
Co-​aggregation
The formation of aggregates
(also known as clumps) in
suspension by bacteria of
different species.
Depletion aggregation
The formation of aggregates in
suspension through a colloidal
physics phenomenon that
occurs when polymers in
solution are of high enough
concentration and molecular
weight to initiate phase
separation, ‘forcing’ microbial
cells together.
Nature Reviews | Microbiology
oxygen profiles of the aggregates as well as the flow field
by particle imaging velocimetry, showing the presence
of strong oxygen gradients, similar to those seen in
attached biofilms. In a similar manner to attached bio-
films, the researchers found that, although the irregular
surface of the aggregate influenced flow and mass trans-
fer locally, there was no flow within the EPS, which in
this field is referred to as ‘transparent exopolymer par-
ticles’. Similarly, granular-​activated sludge aggregates of
microorganisms create strong gradients of oxygen, pH,
substrates and metabolites, which leads to microenvi-
ronments that enable simultaneous aerobic and anaer-
obic digestion in the same aggregate, with concomitant
redox gradient-​based physiological stratifications104.
Two types of chronic infection linked to aggregated
rather than surface-​associated bacteria are soft tissue
infections, such as chronic lung infection in individuals
with cystic fibrosis105, and chronic dermal wounds106.
Similarly, in other biofilm-​associated respiratory infec-
tions, such as chronic otitis media, rhinosinusitis or
biofilms on differentiated ciliated cells from individuals
with primary ciliary dyskinesia, aggregates (~10–20 μm)
may adhere to mucosal epithelia or grow as aggregates in
effusion, mucus and airway surface liquid. The bacterial
aggregates seen in these infections are not necessarily
modelled well by flow biofilm experimental systems,
although shear is present in the bronchial airways107.
Osteomyelitis with and without an implant also
belongs to this category. In the case of osteomyelitis
with implants, it is generally assumed that the bac-
teria are attached to the implant surface and so can
be described by the current biofilm model. This has
led to much research into designing antibacterial and
anti-​adhesive surfaces. However, current studies show
that, even though the bacteria can be associated with
the surface, the implant does not have to be colo-
nized to cause persisting infection108,109. Samples from
implant-​associated infections show that bacteria can be
present both in peri-​prosthetic tissue and on the implant
but not necessarily both110. Importantly, detached aggre-
gates recapitulated the antibiotic tolerance observed in
surface-​attached biofilms111.
Aggregates have also been reported for non-​infectious
biofilms. Consider microbiota on the skin, where the
majority of bacteria are organized as small aggregates112,113.
Bacteria clearly attach and grow on the enamel surface
of teeth; however, these structures tend not to be the 3D
mushroom structures seen in flow cells114. Observation
of polymicrobial aggregates in human saliva demon-
strates a progression in structural and community
complexity and in the flow cells that these aggregates
attach to when forming surface-​adhered biofilms114,115.
Similarly, on the skin, bacteria are distributed in small
heterogeneously distributed aggregates and as single
cells116.
In addition to infectious and other host-​associated
biofilms, bacteria in the environment are present as both
surface-​attached colonies and free-​floating or embed-
ded aggregates. In biological wastewater treatment pro-
cesses, dense multispecies aggregates of microorganisms
self-​assemble in both aerobic and anaerobic processes.
The overarching observation is that the environmental
0123456789();:
Reviews
microbiota are dominated by heterogeneous patterns
of aggregated bacteria12 rather than continuous films of
bacteria over large (centimetre) areas; however, to a cer-
tain extent, this is an issue of scale. Algal biofilms on ship
hulls may appear macroscopically continuous and, in
localized areas, as a uniform flat layer but appear patchy
under microscopic examination117.
Aggregate formation
As outlined above, a shortcoming of the current bio-
film model is that it does not account for non-​surface-​
attached aggregates that are often observed in clinical
or environmental settings. Other than their descriptive
observation when the model was first published in 2002,
little was known mechanistically about suspended bac-
terial aggregates. Since then, several publications have
reported on surface-​independent bacterial aggregation,
with bacteria in aggregates displaying simi­lar pheno-
types to bacteria present in surface-​attached commu-
nities, including increased tolerance to antibiotic and
host defence as well as matrix production and slow
growth16,118–121.
Examples of different types of aggregates include
bacteria embedded in host material such as mucus
from the lungs of patients with cystic fibrosis, slough
in the chronic wound bed, or external material flocs in
wastewater treatment plants and soil. Host fluids, includ-
ing synovial fluid and human serum, can induce rapid
(within minutes) aggregation in both Gram-​positive
and Gram-​negative bacteria in vivo and in vitro118,122,123,
which suggests that host components, such as fibronec-
tin, form bridging connections (polymer bridging). Such
suspended biofilm-​like aggregates have also been seen
ex vivo124 and in shaken in vitro cultures125–127. Bacteria in
a shaken liquid culture have, until recently, been assumed
to be entirely planktonic, independent single cells (or
short chains or clusters). However, recent publications
challenge the conceptual separation between plank-
tonic and biofilm bacteria by showing that both S. aureus
and P. aeruginosa can grow as a mixture of planktonic and
aggregate cells in liquid batch cultures126–128.
Based on several laboratory studies, literature cur-
rently points to five mechanisms for the formation of
free-​floating aggregates (Fig. 4) that are discussed here in
the order they were recognized. The first is the detach-
ment of pieces of attached biofilm due to changes in
hydrodynamic shear, nutrient reduction, physical abra-
sion, or exogenously added or endogenously produced
dispersal agents129; loss of biofilm bacterial cells due to
this process has often been referred to as sloughing. The
second is through growth in the planktonic phase128.
As cells divide, the daughter cells remain with the
mother cells rather than dispersing, presumably through
interactions of self-​recognizing surface adhesion mole­
cules or simultaneous production of EPS. The pres-
ence of surface adhesins may also contribute to the
co-​aggregation of cells in the planktonic state, which leads
to the formation of aggregates in the absence of growth.
More recently, it has been proposed that aggregation can
occur in the liquid phase mediated by host polymers
such as mucin and DNA130. One potential mechanism
is depletion aggregation, which occurs as entropic forces
Reviews

a Surface-associated biofilm formation Fig. 4 | Mechanisms of microbial aggregate formation.

Attachment followed by clonal growth and EPS production a | The current model for biofilm formation: attachment of
single planktonic cells to a smooth surface followed by cell
division and production of extracellular polymeric sub-
stances (EPS) to form three-​dimensional surface-​attached
aggregate structures. b | Different mechanisms for gener-
ating free-​floating biofilm-​like aggregates. (1) Sloughing
detachment of aggregates from attached biofilms.
b Biolfilm-like aggregate formation in the liquid phase (2) Clonal growth (division) in the liquid, which may occur
with or without facilitation by bacterially produced EPS.
1 Aggregates formed due to sloughing from surface-attached biofilm
(3) Auto-​aggregation or co-​aggregation (for a single species
or multiple species, respectively), whereby bacteria attach
to each other through mutual attraction of surface mole-
cules, such as adhesins or EPS bridging interactions, followed
by clonal growth and EPS production. (4) Polymer depletion-
​driven aggregation, whereby, under high concentrations of
polymers in the liquid, phase separation occurs and the
polymers and cells separate out such that the polymers
2 Aggregates grown in planktonic phase surround groups of aggregated cells. In the case of bacteria
entering a wound or surgical site, the polymers can be pro-
duced by the host as is the case for hyaluronic acid, a com-
ponent of synovial fluid159. Polymer depletion-​driven
EPS polymer aggregation can also be facilitated through bacterially
produced EPS130. (5) Polymer bridging aggregation occurs
when polymers in the liquid environment160 form bridges
Aggregate without EPS
that connect individual cells, such as fibronectin in synovial
3 Aggregate formation initiated through cell surface components of single cells fluid and fibronectin-​binding proteins and clumping factor
expressed on the cell wall of staphylococci161.
?
factors, such as adhesin proteins, and host factors such
as fibrinogen, fibronectin and hyaluronic acid132,133.
Notably, aggregate formation in liquid or in response
to host polymers includes both co-​aggregation without
bacterial growth as well as clonal growth of trapped bac-
4 Polymer depletion teria, coinciding with a continually increasing aggregate
Host skin size. Moreover, very little is known about how cells leave
barrier
? aggregates, including whether aggregates disassemble by
dispersion or sloughing, or simply turn into single cells.

Host Expanding the five-​step biofilm model

polymer
5 Polymer bridging
Auto-​aggregation
The formation of aggregates
(also known as clumps) in
suspension by bacteria of the
same species.
?
between uncharged or like-​charged polymers cause par-
ticles (single bacterial cells in the case of our discussion)
in the suspension to ‘push out’ polymers between the
cells as they come close together, thus forcing the for-
mation of aggregates131. Another possible aggregation
mechanism is that bacteria bind to molecules in host
fluids through surface adhesion interactions. For exam-
ple, staphylococci have been shown to aggregate in syn-
ovial fluid, which has been a proposed mechanism for
the initiation of peri-​prosthetic joint infection118. This
aggregation is a binding interaction between bacterial
Visualization of biofilms and bacterial aggregates in
other in vitro experimental systems, in the environment
and in infections reveals major disparities with the orig-
inal model (Fig. 1). As evaluation of biofilms from diverse
environments has become more sophisticated, major
differences in the microenvironment of the individual
biofilm or aggregate in regards to substrates, oxygen and
exposure to secreted products have been observed56,134–136.
This varies depending on whether bacteria are directly
adjacent to the growth medium or entrapped in some
sort of biological material (for example, mucus, tissue
or infection wound bed) or non-​biological material (for
example, within corrosion or hard water deposits) as
opposed to a self-​produced biofilm matrix. The microen-
vironments that develop due to an interplay between
microbial physiology, substrates and physicochemical
conditions (redox and pH), and mass transfer within bio-
films and aggregates have a dominant role in determin-
ing the metabolism and behaviour of bacteria, affecting
characteristics such as antibiotic tolerance, growth rate
and expression of virulence factors19,137–139.
Therefore, we propose an updated, more encompass-
ing model describing the three major events in biofilm
formation: aggregation, growth and disaggregation (Fig. 5).

www.nature.com/nrmicro

0123456789();:

This new model represents three basic events that we
believe can be used to depict most of the different scenar-
ios for biofilm formation independently of whether the
condition is in vitro, in situ or in vivo. The model bridges
and combines the different possibilities and pathways of
biofilm aggregate development in an inclusive manner.
We acknowledge that this is a work in progress based on
what we know to date. Thus, we expect that the model is
not final but will undergo future revision.
In contrast to the five-​step model, the present model
considers open systems that may be encountered in
the environment or the human airways or gut, where
a continuous influx of new biofilm members is likely.
Most importantly, there are no known correlations
suggesting that a particular biofilm structure is either
‘better’ or ‘worse’ in any given situation. Ex vivo and
ex situ observations suggest that mushroom structures
and surface-​attached 3D structures are just as likely to
occur as the aggregates observed in chronic infections
and the natural environment with and without surface
association. A key concept is that biofilm aggregates are
heterogeneous, diverse microbial communities, shaped
and influenced by different environmental cues, that
represent multiple discrete microenvironments.
The original model was largely derived based on data
from in vitro flow cell experiments; however, snapshots
Aggregation Disaggregation
and attachment and detachment
Growth and
accumulation
Fig. 5 | Expanded conceptual model of biofilm formation. The schematic presents the
proposed updated model of biofilm formation, including the three main events in biofilm
formation independently of surfaces and initiation from single-​cell planktonic bacteria,
thus encompassing in vitro, in situ and in vivo systems. In contrast to the five-​step model,
the present model considers different habitats, conditions and microenvironments as
well as the possible influx of new cells. In addition, the new model also encompasses both
surface-​attached (outer ring) and non-​surface-​attached (inner ring) biofilms with possi-
ble exchange between the two. Bacteria can enter the model at any given point; thus,
we have removed the restrictive developmental scheme the five-​step model represented.
This enables a more dynamic overview that can be used to explain most biofilm scenarios
in clinical, environmental and industrial habitats. Aggregation and attachment: during this
event, bacteria aggregate to each other or attach to biotic and abiotic surfaces. Growth
and accumulation: during this event, aggregated and attached bacterial colonies expand
by growth and recruitment of surrounding cells. Disaggregation and detachment: during
this event, bacteria can leave the biofilm as aggregates and as single cells, depending
on the mechanism. These three events characterize and represent most, if not all, biofilm
scenarios independently of time and maturity.
Nature Reviews | Microbiology
0123456789();:
Reviews
of biofilms from environmental systems and from in vivo
and ex vivo studies suggest that this type of development
is not always supported in vivo or in open systems. This
led to questioning the general validity of the original
model. From in vitro investigations, we know that flow
and nutrients are important in the experimental sys-
tems shaping the 3D architecture of the surface-​attached
biofilms27,140. The question then becomes, how much do
we really know about the microenvironment and bio-
film development in environmental, in vivo and ex vivo
examples? Photosynthetic mats are well-​described flat
biofilms where the penetration of sunlight and metabolic
activity of the organisms leads to stratified species dis-
tribution and microenvironments141. Suspended biofilm
aggregates used for wastewater treatment, such as aerobic
granules, are another example of a stratified biofilm. In
this case, the aggregates are generally spherical. Although
direct measurements of the microenvironment are dif-
ficult because they are free floating, stratification show-
ing aerobes on the outside and anaerobes on the inside
provides evidence of oxic and anoxic zones142. These
microenvironments enable simultaneous aerobic diges-
tion and anaerobic denitrification of wastewater. Patchy
aggregates of bacteria in industrial systems allow the
formation of strong oxygen gradients accelerating pit-
ting and the deposition of corrosion products producing
large tubercules protruding from the metal surface143. In
iron and steel industrial pipes, biofilms can cause micro-
bially induced corrosion owing to the development of
microenvironments144. These biofilms tend to be present
as mound-​shaped aggregates on metal surfaces and con-
sist of bacteria and corrosion products. The stratification
of organisms such as iron-​oxidizing and sulfur-​reducing
bacteria creates anoxic zones within the tubercule, which
become anodic relative to the surrounding metal, caus-
ing pitting corrosion below the tubercule and rust dep-
osition at the surface. These examples illustrate how the
interplay between the original external environmental
conditions and the physiology of biofilm microorgan-
isms leads to the creation of different biofilm structures
and microenvironments in situ. Mechanical forces can
also shape biofilm architecture, microbial community
and microenvironment development. Samples from river
biofilms growing under higher turbulence were thinner,
more compact and formed more homogenous layers
than those growing under lower hydrodynamic shear145.
In a medical context in the lung of an individual with
cystic fibrosis, bacteria can be present and form aggre-
gates independently of the epithelial surface121. Thus, the
new proposed model includes a variety of conditions and
biofilm developmental pathways to embrace multiple
diverse habitats and microenvironments in the environ-
ment, industry and medicine (Fig. 5). What we do know
is that the microenvironment depends on the immediate
milieu surrounding a single cell, the aggregate itself and
the close proximity of the aggregate146.
As for the hallmark mushroom-​shaped structures
adopted for the original developmental model, these
seem to be formed primarily by P. aeruginosa, and are
highly dependent on the flow conditions, surface mate-
rial and carbon source29. For most other species, even
under flow conditions and in the presence of glucose,
Reviews
mushroom structures do not form. In the environment
outside of stromatolites and some hot spring structures,
mushroom structures seem to be uncommon147.
Conclusions
The most cited and used model (Fig. 1) for biofilm devel-
opment is extremely intuitive, which explains in part why
it has become the preferred model to describe all types of
biofilm formation. However, as discussed in this Review,
the five-​step developmental model of P. aeruginosa
biofilms grown in flow cells is limited in its scope. We
argue that the proposed conceptual model presented in
this Review is broader and more inclusive of different
types of biofilms by including attached and suspended
aggregates and that attached and suspended pathways
are not mutually exclusive of one another. Biofilms do
not necessarily form a mushroom-​shaped structure as
the final culminating structure, nor is there an absolute
dependence on a surface. Currently, no developmental
model accurately depicts biofilm formation of all micro-
organisms, habitats and microenvironments. With the
inclusive model, we depict three major steps of bio-
film growth irrespective of the presence of a surface:
aggregation, growth and disaggregation.
Growing evidence indicates that biofilms do not nec-
essarily require surface attachment to form. Aggregates
formed in fluids, due to clonal growth, co-​aggregation,
or aggregates induced by bacterial EPS or host fluids,
demonstrate many of the characteristics previously
attributed only to surface-​associated biofilms. These
aggregates are not limited to laboratory conditions but
may be found as part of the human microbiota in several
chronic infection sites and in the environment99,148–151.
Two decades of biofilm research indicate that the model
depicted in Fig. 1 was incomplete because it did not
capture the multiple biofilm structures and phenotypes
that can form with different bacteria and in different
microenvironments. This has important implications for
how we study biofilms specifically and bacteria in gen-
eral, as different biofilm experimental systems in vitro
or experimental animals in vivo cannot encompass all
the factors important for different microenvironments22.
Given the strong influence of laboratory experimental
conditions on biofilm structure, we caution researchers
to consider this in the design of experiments to address
specific research questions, as well as the possibility of the
overgeneralization of conclusions. This is also important
for how we extrapolate from the experimental situation
to the native scenario. We need to understand biofilms in
the context of the relevant microenvironment.
Given that free-​f loating and surface-​associated
aggregates are now accepted as sharing similarities to
surface-​associated biofilms, several questions remain
to be addressed. For example, it is not known what
drives aggregate formation in the absence of a surface
— that is, does bacteria–bacteria adhesion involve the
same mechanisms as the attachment of single cells to
surfaces? Additionally, do aggregates interact with sur-
faces and can aggregates attach to surface biofilms and,
if so, how? Are the same surface properties commonly
associated with initial cell-​surface adhesion (stiff-
ness and surface energy, which in turn is a function
0123456789();:
of electrostatic charge, wettability, surface tension and
roughness) as important for attachment as the macro­
scale topographical features (such as edges, screw
holes, expansions and contractions and threads, among
others) that may physically entrap aggregates? It is well
established that in vitro biofilms actively disperse, but
do aggregates actively disassemble and/or disperse cells
to the surroundings? These questions could be investi-
gated by analysing gene expression profiles during the
different stages of biofilm development in the absence
and presence of a surface. How do the transcriptional
profiles of bacteria in aggregates that have developed
through chemical and/or physical interaction or growth
differ from each other and from biofilms formed on
surfaces? Furthermore, are successional dynamics and
community assembly processes similar for aggregates
and surface-​associated biofilms? A fundamental open
question is whether aggregation protects bacteria from
antimicrobials. Recent work suggests that it is not the
aggregation alone that promotes tolerance towards anti-
microbial agents and host defences, but rather gradients
of oxygen and nutrients that may become pronounced
in aggregates as they increase in size152. The aggregate
size may also determine how easily phagocytes engulf
the aggregates153. In flow cells and as depicted in the
original five-​step model, this results in stratified growth
with a fast-​growing exterior and a dormant inner sub-
population154. In infections, host material can surround
microbial aggregates causing strong chemical gradi-
ents. Thus, the original five-​step model does not accu-
rately represent the microenvironment around these
aggregates and likely also fails to capture the reality of
biofilms in complex environmental and industrial sys-
tems. Concentration gradients influence and regulate
bacterial physiology and metabolism and are recipro-
cally controlled by the microenvironment as well as by
matrix components57,155. However, an important ques-
tion that arises is what is the threshold aggregate size
for tolerance manifestation and how do the microenvi-
ronment and access to nutrients and electron acceptors
influence aggregate size? The size of biofilms has been
shown to vary considerably between in vitro and in vivo
biofilms99. The questions of tolerance and matrix pro-
duction and physiology in general might be addressed
by controlling the microenvironment, possibly in 3D
experimental models, to move beyond the attachment
surface as the main constraint controlling immediate
access to nutrients and electron acceptors.
The original five-​step model has provided a unifying
yet possibly unintentionally biased understanding of
biofilm morphology. This may have caused an unfor-
tunate division of the research area as some researchers
studying biofilms that deviated from this model (flocs,
granules, particles, aggregates or mats) may have been
excluded from interacting with and interpreting work
from researchers studying in vitro model biofilms. We
hope that a simpler, more inclusive model can help to
unite the biofilm research community and enable more
cross-​disciplinary collaboration and knowledge sharing.
Specifically, the five-​step model may become challenging
when used to describe clinical manifestations and devise
new in vitro test methods to evaluate medical implants,
www.nature.com/nrmicro
 Reviews

drugs and treatments as these may fail owing to lack of
extrapolation. Crucially, differences in the microenvi-
ronment between in vitro and in vivo may underpin why
direct extrapolation is not possible. Additionally, relying
on the original biofilm model (Fig. 1), health-​care profes-
sionals may have a conceptual framework that markedly
differs from clinical findings and observations, leading
to the erroneous conclusion that a biofilm is not pres-
ent in a given clinical sample and to treatment regimens
that will not effectively treat infections156. It is our hope
also to broaden this framework, ultimately leading to
improved infection diagnostics and selection of efficient,
targeted treatment regimens.
In summary, we suggest a new overall model for bio-
film formation that considers the most inclusive recent
insights. It demonstrates the three major events: aggrega-
tion, growth and disaggregation (Fig. 5). Our intent is that
this simpler model will alleviate some of the miscon-
ceptions of how biofilms form in diverse environments,
ranging from industrial systems to environmental hab-
itats and medical settings. We hope that, as a scientific
community, we can expand on this model to facilitate an
inclusive, less controversial interdisciplinary discussion
on biofilms and biofilm formation.
Published online xx xx xxxx

1. Costerton, J. W., Geesey, G. G. & Cheng, K. J. How
bacteria stick. Sci. Am. 238, 86–95 (1978).
2. McCoy, W. F., Bryers, J. D., Robbins, J. &
Costerton, J. W. Observations of fouling biofilm
formation. Can. J. Microbiol. 27, 910–917 (1981).
3. Flemming, H.-C. et al. Who put the film in biofilm?
The migration of a term from wastewater engineering
to medicine and beyond. NPJ Biofilms Microbiomes 7,
10 (2021).
4. Lebeaux, D., Chauhan, A., Rendueles, O. & Beloin, C.
From in vitro to in vivo models of bacterial biofilm-​related
infections. Pathogens 2013, 288–356 (2013).
5. Hoiby, N. et al. ESCMID guideline for the diagnosis
and treatment of biofilm infections 2014. Clin.
Microbiol. Infect. 21 (Suppl. 1), S1–S25 (2015).
6. Kolpen, M. et al. Bacterial biofilms predominate in both
acute and chronic human lung infections. Thorax https://
doi.org/10.1136/thoraxjnl-2021-217576 (2022).
7. Raghupathi, P. K. et al. Synergistic interactions within
a multispecies biofilm enhance individual species
protection against grazing by a pelagic protozoan.
Front. Microbiol. https://doi.org/10.3389/
fmicb.2017.02649 (2018).
8. Jass, J., Roberts, S. K. & Lappin-​Scott, H. M. in
Enzymes in the Environment. Activity, Ecology and
Applications 307–326 (Marcel Dekker Inc., 2002).
9. Tkacz, A. & Poole, P. The plant microbiome: the dark
and dirty secrets of plant growth. Plants People Planet
3, 124–129 (2021).
10. Annous, B. A., Solomon, E. B., Cooke, P. H. & Burke, A.
Biofilm formation by Salmonella spp. on Cantaloupe
melons. J. Food Saf. 25, 276–287 (2005).
11. Rudrappa, T., Biedrzycki, M. L. & Bais, H. P. Causes
and consequences of plant-​associated biofilms. FEMS
Microbiol. Ecol. 64, 153–166 (2008).
12. Flemming, H. C. & Wuertz, S. Bacteria and archaea
on earth and their abundance in biofilms. Nat. Rev.
Microbiol. 17, 247–260 (2019).
Excellent review that quantitatively explores and
proves the long-​standing notion that biofilm is the
predominant form of prokaryotic life.
13. Costerton, J. W. et al. Bacterial biofilms in nature and
disease. Annu. Rev. Microbiol. 41, 435–464 (1987).
14. Vishwakarma, V. Impact of environmental biofilms:
industrial components and its remediation. J. Basic
Microbiol. 60, 198–206 (2020).
15. Jurelevicius, D. et al. Long-​term souring treatment
using nitrate and biocides in high-​temperature oil
reservoirs. Fuel 288, 119731 (2021).
16. Alhede, M. et al. Phenotypes of non-​attached
Pseudomonas aeruginosa aggregates resemble
surface attached biofilm. PLoS ONE 6, e27943
(2011).
17. Vitzilaiou, E., Kuria, A. M., Siegumfeldt, H.,
Rasmussen, M. A. & Knøchel, S. The impact of
bacterial cell aggregation on UV inactivation kinetics.
Water Res. 204, 117593 (2021).
18. Bjarnsholt, T. et al. The importance of understanding
the infectious microenvironment. Lancet Infect. Dis.
22, e88–e92 (2022).
19. Cornforth, D. M., Diggle, F. L., Melvin, J. A.,
Bomberger, J. M. & Whiteley, M. Quantitative
framework for model evaluation in microbiology
research using Pseudomonas aeruginosa and cystic
fibrosis infection as a test case. mBio https://doi.
org/10.1128/mBio.03042-19 (2020).
20. Marrie, T. J., Nelligan, J. & Costerton, J. W. A scanning
and transmission electron microscopic study of an
infected endocardial pacemaker lead. Circulation 66,
1339–1341 (1982).
21. Zobell, C. E. The effect of solid surfaces upon bacterial
activity. J. Bacteriol. 46, 39–56 (1943).
Simple observation results in a paradigm-​changing
idea: bacteria like to live in communities.
22. Thaarup, I. C. & Bjarnsholt, T. Current in vitro biofilm-​
infected chronic wound models for developing new
treatment possibilities. Adv. Wound Care 10, 91–102
(2021).
23. Sternberg, C., Bjarnsholt, T. & Shirtliff, M. Methods
for dynamic investigations of surface-​attached in vitro
bacterial and fungal biofilms. Methods Mol. Biol.
1147, 3–22 (2014).
24. Azeredo, J. et al. Critical review on biofilm methods.
Crit. Rev. Microbiol. 43, 313–351 (2017).
25. Hall-​Stoodley, L., Costerton, J. W. & Stoodley, P.
Bacterial biofilms: from the natural environment to
infectious diseases. Nat. Rev. Microbiol. 2, 95–108
(2004).
26. Irie, Y. et al. Self-​produced exopolysaccharide is a
signal that stimulates biofilm formation in
Pseudomonas aeruginosa. Proc. Natl Acad. Sci. USA
109, 20632–20636 (2012).
27. Sauer, K., Camper, A. K., Ehrlich, G. D., Costerton, J. W.
& Davies, D. G. Pseudomonas aeruginosa displays
multiple phenotypes during development as a biofilm.
J. Bacteriol. 184, 1140–1154 (2002).
The original key paper that started the concept of
a biofilm life cycle and is revisited in this Review.
28. Pamp, S. J., Sternberg, C. & Tolker-​Nielsen, T. Insight
into the microbial multicellular lifestyle via flow-​cell
technology and confocal microscopy. Cytom. A 75,
90–103 (2009).
29. Klausen, M. et al. Biofilm formation by Pseudomonas
aeruginosa wild type, flagella and type IV pili mutants.
Mol. Microbiol. 48, 1511–1524 (2003).
30. Stoodley, P., Sauer, K., Davies, D. G. & Costerton, J. W.
Biofilms as complex differentiated communities. Annu.
Rev. Microbiol. 56, 187–209 (2002).
31. Moormeier, D. E. & Bayles, K. W. Staphylococcus
aureus biofilm: a complex developmental organism.
Mol. Microbiol. 104, 365–376 (2017).
32. Vlamakis, H., Chai, Y., Beauregard, P., Losick, R.
& Kolter, R. Sticking together: building a biofilm
the Bacillus subtilis way. Nat. Rev. Microbiol. 11,
157–168 (2013).
33. Hu, Y., Xiao, Y., Liao, K., Leng, Y. & Lu, Q. Development
of microalgal biofilm for wastewater remediation:
from mechanism to practical application. J. Chem.
Technol. Biotechnol. 96, 2993–3008 (2021).
34. Liu, J., Lu, H., Wu, L., Kerr, P. G. & Wu, Y. Interactions
between periphytic biofilms and dissolved organic
matter at soil-​water interface and the consequent
effects on soil phosphorus fraction changes. Sci. Total
Environ. 801, 149708 (2021).
35. Wu, B. C. et al. Human organoid biofilm model for
assessing antibiofilm activity of novel agents. NPJ
Biofilms Microbiomes 7, 8 (2021).
36. Zhao, Y., Liu, H., Wang, R. & Wu, C. Interactions
between dicyandiamide and periphytic biofilms in
paddy soils and subsequent effects on nitrogen
cycling. Sci. Total Environ. 718, 137417 (2020).
37. Petrova, O. E. & Sauer, K. A novel signaling network
essential for regulating Pseudomonas aeruginosa
biofilm development. PLoS Pathog. 5, e1000668
(2009).
Findings reported here demonstrated, for the first
time, that the formation of biofilms is coordinated
by a genetic pathway that regulates morphological
changes of biofilms and stage-​specific transitions
in a hierarchically ordered manner. Components
of the genetic pathways only seemed to have a role
under biofilm growth conditions.
38. Petrova, O. E., Gupta, K., Liao, J., Goodwine, J. S.
& Sauer, K. Divide and conquer: the Pseudomonas
aeruginosa two-​component hybrid SagS enables
biofilm formation and recalcitrance of biofilm cells
to antimicrobial agents via distinct regulatory circuits.
Environ. Microbiol. 19, 2005–2024 (2017).
39. O’Toole, G. A. & Kolter, R. Initiation of biofilm
formation in Pseudomonas fluorescens WCS365
proceeds via multiple, convergent signalling pathways:
a genetic analysis. Mol. Microbiol. 28, 449–461
(1998).
40. Davey, M. E. & O’Toole, G. A. Microbial biofilms:
from ecology to molecular genetics. Microbiol. Mol.
Biol. Rev. 64, 847–867 (2000).
41. Characklis, W. G. Attached microbial growths-​II.
Frictional resistance due to microbial slimes. Water
Res. 7, 1249–1258 (1973).
42. Davies, D. G., Charabarty, A. M. & Geesey, G. G.
Exopolysaccharide production in biofilms: substratum
activation of alginate gene expression by
Pseudomonas aeruginosa. Appl. Environ. Microbiol.
59, 1181–1186 (1993).
43. Davies, D. G. & Geesey, G. G. Regulation of the
alginate biosynthesis gene algC in Pseudomonas
aeruginosa during biofilm development in continuous
culture. Appl. Environ. Microbiol. 61, 860–867
(1995).
44. Colvin, K. M. et al. The Pel and Psl polysaccharides
provide Pseudomonas aeruginosa structural
redundancy within the biofilm matrix. Environ.
Microbiol. 14, 1913–1928 (2012).
45. Bagge, N. et al. Dynamics and spatial distribution of
β-​lactamase expression in Pseudomonas aeruginosa
biofilms. Antimicrob. Agents Chemother. 48,
1168–1174 (2004).
46. Wood, D. W., Gong, F., Daykin, M. M., Williams, P.
& Pierson, L. S. N-​acyl-homoserine lactone-​mediated
regulation of phenazine gene expression by
Pseudomonas aureofaciens 30-84 in the wheat
rhizosphere. J. Bacteriol. 179, 7663–7670
(1997).
47. Gupta, K., Marques, C. N. H., Petrova, O. E. & Sauer, K.
Antimicrobial tolerance of Pseudomonas aeruginosa
biofilms is activated during an early developmental
stage and requires the two-​component hybrid SagS.
J. Bacteriol. 195, 4975–4987 (2013).
Findings reported here indicated that biofilm cells
gain heightened tolerance to antimicrobial agents
in a manner independent of biofilm biomass
accumulation (demonstrating mature biofilm
architecture).
48. Wood, S. R. et al. Architecture of intact natural human
plaque biofilms studied by confocal laser scanning
microscopy. J. Dent. Res. 79, 21–27 (2000).
49. Davies, D. G. et al. The involvement of cell-​to-cell
signals in the development of a bacterial biofilm.
Science 280, 295–298 (1998).
50. Lequette, Y. & Greenberg, E. P. Timing and localization
of rhamnolipid synthesis gene expression in
Pseudomonas aeruginosa biofilms. J. Bacteriol. 187,
37–44 (2005).
51. Espinosa-​Urgel, M. Resident parking only:
rhamnolipids maintain fluid channels in biofilms.
J. Bacteriol. 185, 699–700 (2003).
52. Kuchma, S. L., Connolly, J. P. & O’Toole, G. A. A
three-​component regulatory system regulates biofilm
maturation and type III secretion in Pseudomonas
aeruginosa. J. Bacteriol. 187, 1441–1454 (2005).

Nature Reviews | Microbiology

0123456789();:
Reviews
53. Petrova, O. E., Schurr, J. R., Schurr, M. J. & Sauer, K.
Microcolony formation by the opportunistic pathogen
Pseudomonas aeruginosa requires pyruvate and
pyruvate fermentation. Mol. Microbiol. 86, 819–835
(2012).
54. Sriramulu, D. D., Lünsdorf, H., Lam, J. S. & Römling, U.
Microcolony formation: a novel biofilm model of
Pseudomonas aeruginosa for the cystic fibrosis lung.
J. Med. Microbiol. 54, 667–676 (2005).
55. Purevdorj, B., Costerton, J. W. & Stoodley, P. Influence
of hydrodynamics and cell signaling on the structure
and behavior of Pseudomonas aeruginosa biofilms.
Appl. Environ. Microbiol. 68, 4457–4464 (2002).
Understanding the relationship between biofilm
and fluid dynamics is crucial and completely
overlooked in traditional microbiological systems.
It impacts not only mass transfer rates but also
the biofilm architecture, spatial organization and
detachment.
56. Stewart, P. S. & Franklin, M. J. Physiological
heterogeneity in biofilms. Nat. Rev. Microbiol. 6,
199–210 (2008).
57. Serra, D. O. & Hengge, R. Stress responses go three
dimensional–the spatial order of physiological
differentiation in bacterial macrocolony biofilms.
Environ. Microbiol. 16, 1455–1471 (2014).
58. Williamson, K. S. et al. Heterogeneity in Pseudomonas
aeruginosa biofilms includes expression of ribosome
hibernation factors in the antibiotic-​tolerant
subpopulation and hypoxia-​induced stress response
in the metabolically active population. J. Bacteriol.
194, 2062–2073 (2012).
59. Heacock-​Kang, Y. et al. Spatial transcriptomes within
the Pseudomonas aeruginosa biofilm architecture.
Mol. Microbiol. 106, 976–985 (2017).
60. Haussler, S. & Fuqua, C. Biofilms 2012: new
discoveries and significant wrinkles in a dynamic field.
J. Bacteriol. 195, 2947–2958 (2013).
61. Rumbaugh, K. P. & Sauer, K. Biofilm dispersion.
Nat. Rev. Microbiol. 18, 571–586 (2020).
62. Petrova, O. E. & Sauer, K. Escaping the biofilm in more
than one way: desorption, detachment or dispersion.
Curr. Opin. Microbiol. 30, 67–78 (2016).
63. Davies, D. G. in Biofilm Highlights (eds Flemming,
H.-C., Wingender, J. & Szewzyk, U.) 1–28 (Springer,
2011).
64. Steinberg, N. et al. The extracellular matrix protein
TasA is a developmental cue that maintains a motile
subpopulation within Bacillus subtilis biofilms.
Sci. Signal. 13, eaaw8905 (2020).
65. Purevdorj-​Gage, B., Costerton, W. J. & Stoodley, P.
Phenotypic differentiation and seeding dispersal in
non-​mucoid and mucoid Pseudomonas aeruginosa
biofilms. Microbiology 151, 1569–1576 (2005).
66. Valentini, M. & Filloux, A. Biofilms and cyclic di-​GMP
(c-​di-GMP) signaling: lessons from Pseudomonas
aeruginosa and other bacteria. J. Biol. Chem. 291,
12547–12555 (2016).
67. Römling, U., Galperin, M. Y. & Gomelsky, M. Cyclic
di-​GMP: the first 25 years of a universal bacterial
second messenger. Microbiol. Mol. Biol. Rev. 77,
1–52 (2013).
68. Jenal, U., Reinders, A. & Lori, C. Cyclic di-​GMP:
second messenger extraordinaire. Nat. Rev. Microbiol.
15, 271–284 (2017).
69. Purcell, E. B. & Tamayo, R. Cyclic diguanylate signaling
in Gram-​positive bacteria. FEMS Microbiol. Rev. 40,
753–773 (2016).
70. Yin, W., Wang, Y., Liu, L. & He, J. Biofilms: the
microbial “protective clothing” in extreme
environments. Int. J. Mol. Sci. 20, 3423 (2019).
71. Mantzorou, A. & Ververidis, F. Microalgal biofilms:
a further step over current microalgal cultivation
techniques. Sci. Total Environ. 651, 3187–3201
(2019).
72. Prieto-​Barajas, C. M., Valencia-​Cantero, E. &
Santoyo, G. Microbial mat ecosystems: structure types,
functional diversity, and biotechnological application.
Electron. J. Biotechnol. 31, 48–56 (2018).
73. Hao, Y. et al. Influence of dental prosthesis and
restorative materials interface on oral biofilms.
Int. J. Mol. Sci. 19, 3157 (2018).
74. Chang, C.-S. & Kao, C.-Y. Current understanding of the
gut microbiota shaping mechanisms. J. Biomed. Sci.
26, 59 (2019).
75. Pii, Y. et al. Microbial interactions in the rhizosphere:
beneficial influences of plant growth-​promoting
rhizobacteria on nutrient acquisition process. A
review. Biol. Fertil. Soils https://doi.org/10.1007/
s00374-015-0996-1 (2015).
76. Roth-​Schulze, A. J. et al. Functional biogeography and
host specificity of bacterial communities associated
with the Marine Green Alga Ulva spp. Mol. Ecol. 27,
1952–1965 (2018).
77. Monds, R. D. & O’Toole, G. A. The developmental
model of microbial biofilms: ten years of a paradigm
up for review. Trends Microbiol. 17, 73–87 (2009).
78. Dworkin, M. Developmental Biology of the Bacteria
(Benjamin/Cummings Pub. Co., 1985).
79. Brun, Y. V. & Shimkets, L. J. Prokaryotic Development
114 (ASM Press, 2000).
80. Goodman, A. L. et al. A signaling network reciprocally
regulates genes associated with acute infection
and chronic persistence in Pseudomonas aeruginosa.
Dev. Cell 7, 745–754 (2004).
81. Merritt, J. H., Brothers, K. M., Kuchma, S. L.
& O’Toole, G. A. SadC reciprocally influences biofilm
formation and swarming motility via modulation of
exopolysaccharide production and flagellar function.
J. Bacteriol. 189, 8154–8164 (2007).
82. Chambers, J. R. & Sauer, K. Small RNAs and their role
in biofilm formation. Trends Microbiol. 21, 39–49
(2013).
83. Petrova, O. E., Schurr, J. R., Schurr, M. J. & Sauer, K.
The novel Pseudomonas aeruginosa two-​component
regulator BfmR controls bacteriophage-​mediated lysis
and DNA release during biofilm development through
PhdA. Mol. Microbiol 81, 767–783 (2011).
84. Petrova, O. E. & Sauer, K. The novel two-​component
regulatory system BfiSR regulates biofilm
development by controlling the small RNA rsmZ
through CafA. J. Bacteriol. 192, 5275–5288 (2010).
85. Petrova, O. E. & Sauer, K. SagS contributes to the
motile-​sessile switch and acts in concert with BfiSR
to enable Pseudomonas aeruginosa biofilm formation.
J. Bacteriol. 193, 6614–6628 (2011).
86. Lee, K. W. K. et al. Biofilm development and enhanced
stress resistance of a model, mixed-​species
community biofilm. ISME J. 8, 894–907 (2014).
87. Barken, K. B. et al. Roles of type IV pili, flagellum-​
mediated motility and extracellular DNA in the
formation of mature multicellular structures in
Pseudomonas aeruginosa biofilms. Environ.
Microbiol. 10, 2331–2343 (2008).
88. Friedman, L. & Kolter, R. Genes involved in matrix
formation in Pseudomonas aeruginosa PA14 biofilms.
Mol. Microbiol. 51, 675–690 (2004).
89. Fabbri, S. et al. Fluid-​driven interfacial instabilities
and turbulence in bacterial biofilms. Env. Microbiol.
19, 4417–4431 (2017).
90. James, G. A., Beaudette, L. & Costerton, J. W.
Interspecies bacterial interactions in biofilms. J. Ind.
Microbiol. 15, 257–262 (1995).
91. Murga, R., Stewart, P. S. & Daly, D. Quantitative
analysis of biofilm thickness variability. Biotechnol.
Bioeng. 45, 503–510 (1995).
92. Liu, W. et al. Low-​abundant species facilitates specific
spatial organization that promotes multispecies biofilm
formation. Env. Microbiol. 19, 2893–2905 (2017).
93. Sauer, K., Steczko, J. & Ash, S. R. Effect of a solution
containing citrate/methylene blue/parabens on
Staphylococcus aureus bacteria and biofilm, and
comparison with various heparin solutions. J.
Antimicrob. Chemother. 63, 937–945 (2009).
94. Allegrucci, M. et al. Phenotypic characterization of
Streptococcus pneumoniae biofilm development.
J. Bacteriol. 188, 2325–2335 (2006).
95. Hall-​Stoodley, L. et al. Direct detection of bacterial
biofilms on the middle-​ear mucosa of children with
chronic otitis media. JAMA 296, 202–211 (2006).
First paper to directly link biofilms with otitis
media using molecular probes specific for three
pathogens, which cause otitis media, on human
middle ear mucosal epithelial biopsy samples
from children with chronic otitis media and not
on uninfected biopsy samples.
96. Hall-​Stoodley, L. et al. Characterization of biofilm
matrix, degradation by DNase treatment and
evidence of capsule downregulation in Streptococcus
pneumoniae clinical isolates. BMC Microbiol. 8,
173 (2008).
97. Bakaletz, L. O. Bacterial biofilms in the upper airway -
evidence for role in pathology and implications for
treatment of otitis media. Paediatr. Respir. Rev. 13,
154–159 (2012).
98. Walker, W. T. et al. Primary ciliary dyskinesia ciliated
airway cells show increased susceptibility to Haemophilus
influenzae biofilm formation. Eur. Respir. J. https://
doi.org/10.1183/13993003.00612-2017 (2017).
99. Bjarnsholt, T. et al. The in vivo biofilm. Trends
Microbiol. 21, 466–474 (2013).
100. Tuck, B., Watkin, E., Somers, A. & Machuca, L. L. A
critical review of marine biofilms on metallic materials.
NPJ Mater. Degrad. 6, 25 (2022).
0123456789();:
101. Hall-​Stoodley, L. et al. Towards diagnostic guidelines
for biofilm-​associated infections. FEMS Immunol. Med.
Microbiol. 65, 127–145 (2012).
102. Trego, A. C., Mills, S. & Collins, G. Granular biofilms:
function, application, and new trends as model
microbial communities. Crit. Rev. Environ. Sci. Technol.
51, 1702–1725 (2021).
Focused and informative review on types, functions
and applications of microbial granules.
103. Zetsche, E.-M., Larsson, A. I., Iversen, M. H.
& Ploug, H. Flow and diffusion around and within
diatom aggregates: effects of aggregate composition
and shape. Limnol. Oceanogr. 65, 1818–1833 (2020).
104. Wilén, B.-M., Liébana, R., Persson, F., Modin, O.
& Hermansson, M. The mechanisms of granulation
of activated sludge in wastewater treatment, its
optimization, and impact on effluent quality. Appl.
Microbiol. Biotechnol. 102, 5005–5020 (2018).
105. Bjarnsholt, T. et al. Pseudomonas aeruginosa biofilms
in the respiratory tract of cystic fibrosis patients.
Pediatr. Pulmonol. 44, 547–558 (2009).
First paper to show, using specific 16S molecular
FISH probes, P. aeruginosa in biofilm aggregates
in situ in the respiratory tract in individuals with
cystic fibrosis.
106. Bjarnsholt, T. et al. Why chronic wounds will not heal:
a novel hypothesis. Wound Repair Regen. 16, 2–10
(2008).
107. Gloag, E. S., Wozniak, D. J., Stoodley, P. & Hall-​
Stoodley, L. Mycobacterium abscessus biofilms have
viscoelastic properties which may contribute to their
recalcitrance in chronic pulmonary infections. Sci. Rep.
11, 5020 (2021).
108. Jensen, L. K. et al. Novel porcine model of implant-​
associated osteomyelitis: a comprehensive analysis of
local, regional, and systemic response. J. Orthop. Res.
35, 2211–2221 (2017).
109. Li, C., Renz, N. & Trampuz, A. Management of
periprosthetic joint infection. Hip Pelvis 30, 138–146
(2018).
110. Dudareva, M. et al. Sonication versus tissue sampling
for diagnosis of prosthetic joint and other orthopedic
device-​related infections. J. Clin. Microbiol. https://
doi.org/10.1128/JCM.00688-18 (2018).
111. Fux, C. A., Wilson, S. & Stoodley, P. Detachment
characteristics and oxacillin resistance of Staphyloccocus
aureus biofilm emboli in an in vitro catheter infection
model. J. Bacteriol. 186, 4486–4491 (2004).
First paper demonstrating that aggregates are
continually shed from biofilms and demonstrate
biofilm tolerance to antibiotics.
112. Bay, L. et al. Universal dermal microbiome in human
skin. mBio https://doi.org/10.1128/mBio.02945-19
(2020).
113. Burmolle, M. et al. Biofilms in chronic infections -
a matter of opportunity - monospecies biofilms in
multispecies infections. FEMS Immunol. Med.
Microbiol. 59, 324–336 (2010).
114. Kim, D. et al. Spatial mapping of polymicrobial
communities reveals a precise biogeography
associated with human dental caries. Proc. Natl Acad.
Sci. USA 117, 12375–12386 (2020).
115. Bowen, W. H., Burne, R. A., Wu, H. & Koo, H. Oral
biofilms: pathogens, matrix, and polymicrobial
interactions in microenvironments. Trends Microbiol.
26, 229–242 (2018).
116. Ashrafi, M. et al. Validation of biofilm formation
on human skin wound models and demonstration
of clinically translatable bacteria-​specific volatile
signatures. Sci. Rep. 8, 9431 (2018).
117. Salta, M., Wharton, J. A., Blache, Y., Stokes, K. R.
& Briand, J.-F. Marine biofilms on artificial surfaces:
structure and dynamics. Environ. Microbiol. 15,
2879–2893 (2013).
118. Dastgheyb, S. S. et al. Staphylococcal persistence due
to biofilm formation in synovial fluid containing
prophylactic cefazolin. Antimicrob. Agents Chemother.
59, 2122–2128 (2015).
119. Landry, R. M., An, D., Hupp, J. T., Singh, P. K.
& Parsek, M. R. Mucin-​Pseudomonas aeruginosa
interactions promote biofilm formation and antibiotic
resistance. Mol. Microbiol. 59, 142–151 (2006).
120. Singh, P. K. et al. Quorum-​sensing signals indicate
that cystic fibrosis lungs are infected with bacterial
biofilms. Nature 407, 762–764 (2000).
121. Worlitzsch, D. et al. Effects of reduced mucus oxygen
concentration in airway Pseudomonas infections of
cystic fibrosis patients. J. Clin. Invest. 109, 317–325
(2002).
122. Pestrak, M. J. et al. Investigation of synovial fluid
induced Staphylococcus aureus aggregate
development and its impact on surface attachment
www.nature.com/nrmicro

and biofilm formation. PLoS ONE 15, e0231791
(2020).
123. Macias-​Valcayo, A. et al. Synovial fluid mediated
aggregation of clinical strains of four enterobacterial
species. Adv. Exp. Med. Biol. https://doi.org/10.1007/
5584_2020_573 (2020).
124. Bidossi, A., Bottagisio, M., Savadori, P. & De Vecchi, E.
Identification and characterization of planktonic
biofilm-​like aggregates in infected synovial fluids
from joint infections. Front. Microbiol. 11, 1368
(2020).
125. Kragh, K. N. et al. The inoculation method could
impact the outcome of microbiological experiments.
Appl. Environ. Microbiol. https://doi.org/10.1128/
AEM.02264-17 (2018).
126. Schleheck, D. et al. Pseudomonas aeruginosa PAO1
preferentially grows as aggregates in liquid batch
cultures and disperses upon starvation. PLoS ONE 4,
e5513 (2009).
127. Haaber, J., Cohn, M. T., Frees, D., Andersen, T. J. &
Ingmer, H. Planktonic aggregates of Staphylococcus
aureus protect against common antibiotics. PLoS ONE
7, e41075 (2012).
128. Kragh, K. N. et al. Inoculation method could impact
the outcome of microbiological experiments. Appl.
Environ. Microbiol. https://doi.org/10.1128/
AEM.02264-17 (2017).
129. Hall-​Stoodley, L. & Stoodley, P. Biofilm formation and
dispersal and the transmission of human pathogens.
Trends Microbiol. 13, 7–10 (2005).
130. Secor, P. R., Michaels, L. A., Ratjen, A., Jennings, L. K.
& Singh, P. K. Entropically driven aggregation of
bacteria by host polymers promotes antibiotic
tolerance in Pseudomonas aeruginosa. Proc. Natl
Acad. Sci. USA 115, 10780 (2018).
131. Secor, P. R., Michaels, L. A., Ratjen, A., Jennings, L. K.
& Singh, P. K. Entropically driven aggregation of
bacteria by host polymers promotes antibiotic
tolerance in Pseudomonas aeruginosa. Proc. Natl
Acad. Sci. USA 115, 10780–10785 (2018).
132. Dastgheyb, S., Parvizi, J., Shapiro, I. M., Hickok, N. J.
& Otto, M. Effect of biofilms on recalcitrance of
staphylococcal joint infection to antibiotic treatment.
J. Infect. Dis. 211, 641–650 (2015).
First paper demonstrating the importance
of host factors in rapid aggregation of S. aureus
into biofilm-​like aggregates.
133. Knott, S. et al. Staphylococcus aureus floating biofilm
formation and phenotype in synovial fluid depends
on albumin, fibrinogen, and hyaluronic acid. Front.
Microbiol. 12, 655873 (2021).
134. Dar, D., Dar, N., Cai, L. & Newman, D. K. Spatial
transcriptomics of planktonic and sessile bacterial
populations at single-​cell resolution. Science https://
doi.org/10.1126/science.abi4882 (2021).
Using an innovative transcriptome-​imaging
approach, this article provides visual evidence
of differential gene expression and heterogeneity
within biofilms at the single-​cell level.
135. Dal Co, A., van Vliet, S. & Ackermann, M.
Emergent microscale gradients give rise to metabolic
cross-feeding and antibiotic tolerance in clonal
bacterial populations. Philos. Trans. R. Soc. Lond. B
Biol. Sci. 374, 20190080 (2019).
136. Kowalski, C. H., Morelli, K. A., Schultz, D., Nadell, C. D.
& Cramer, R. A. Fungal biofilm architecture produces
hypoxic microenvironments that drive antifungal
resistance. Proc. Natl Acad. Sci. USA 117,
22473–22483 (2020).
137. Cornforth, D. M. et al. Pseudomonas aeruginosa
transcriptome during human infection. Proc. Natl
Acad. Sci. USA 115, E5125–E5134 (2018).
138. Rossi, E., Falcone, M., Molin, S. & Johansen, H. K.
High-​resolution in situ transcriptomics of
Pseudomonas aeruginosa unveils genotype
independent patho-​phenotypes in cystic fibrosis lungs.
Nat. Commun. 9, 3459 (2018).
139. Koo, H., Allan, R. N., Howlin, R. P., Stoodley, P.
& Hall-​Stoodley, L. Targeting microbial biofilms:
current and prospective therapeutic strategies.
Nat. Rev. Microbiol. 15, 740–755 (2017).
140. Kragh, K. N. et al. Role of multicellular aggregates
in biofilm formation. mBio 7, e00237 (2016).
141. Hawes, I., Sumner, D. & Jungblut, A. D. in The
Structure and Function of Aquatic Microbial
Communities (ed. Hurst, C. J.) 91–120 (Springer
International Publishing, 2019).
Nature Reviews | Microbiology
142. Franca, R. D. G., Pinheiro, H. M.,
van Loosdrecht, M. C. M. & Lourenço, N. D.
Stability of aerobic granules during long-​term
bioreactor operation. Biotechnol. Adv. 36,
228–246 (2018).
143. Li, Y. et al. Anaerobic microbiologically influenced
corrosion mechanisms interpreted using bioenergetics
and bioelectrochemistry: a review. J. Mater. Sci.
Technol. 34, 1713–1718 (2018).
144. Bahrami, A., Khouzani, M. K. & Harchegani, B. B.
Establishing the root cause of a failure in a firewater
pipeline. Eng. Fail. Anal. 127, 105474 (2021).
145. Risse-​Buhl, U. et al. The role of hydrodynamics in
shaping the composition and architecture of epilithic
biofilms in fluvial ecosystems. Water Res. 127,
211–222 (2017).
146. Kirketerp-​Moller, K., Stewart, P. S. & Bjarnsholt, T.
The zone model: a conceptual model for understanding
the microenvironment of chronic wound infection.
Wound Repair Regen. 28, 593–599 (2020).
147. Chan, C. S. et al. The architecture of iron microbial
mats reflects the adaptation of chemolithotrophic iron
oxidation in freshwater and marine environments.
Front. Microbiol. https://doi.org/10.3389/
fmicb.2016.00796 (2016).
148. Bay, L. et al. Bacterial aggregates establish at the
edges of acute epidermal wounds. Adv. Wound Care 7,
105–113 (2018).
149. Ring, H. C. et al. Bacterial biofilm in chronic lesions
of hidradenitis suppurativa. Br. J. Dermatol. 176,
993–1000 (2017).
150. Ring, H. C. et al. The follicular skin microbiome
in patients with hidradenitis suppurativa and
healthy controls. JAMA Dermatol. 153, 897–905
(2017).
151. Qvist, T. et al. Chronic pulmonary disease with
Mycobacterium abscessus complex is a biofilm
infection. Eur. Respir. J. 46, 1823–1826 (2015).
152. Folsom, J. P. et al. Physiology of Pseudomonas
aeruginosa in biofilms as revealed by transcriptome
analysis. BMC Microbiol. 10, 294 (2010).
153. Alhede, M. et al. Bacterial aggregate size determines
phagocytosis efficiency of polymorphonuclear leukocytes.
Med. Microbiol. Immunol. 209, 669–680 (2020).
154. Pamp, S. J., Gjermansen, M., Johansen, H. K.
& Tolker-​Nielsen, T. Tolerance to the antimicrobial
peptide colistin in Pseudomonas aeruginosa biofilms
is linked to metabolically active cells, and depends on
the pmr and mexAB-​oprM genes. Mol. Microbiol. 68,
223–240 (2008).
155. Díaz-​Pascual, F. et al. Spatial alanine metabolism
determines local growth dynamics of Escherichia coli
colonies. eLife 10, e70794 (2021).
156. Bjarnsholt, T. et al. The impact of mental models
on the treatment and research of chronic infections
due to biofilms. APMIS https://doi.org/10.1111/
apm.13163 (2021).
157. Kirketerp-​Moller, K. et al. Distribution, organization,
and ecology of bacteria in chronic wounds. J. Clin.
Microbiol. 46, 2717–2722 (2008).
158. Vasconcelos, C. et al. Lithifying microbial mats in
Lagoa Vermelha, Brazil: modern Precambrian relics?
Sediment. Geol. 185, 175–183 (2006).
159. Gilbertie, J. M. et al. Equine or porcine synovial fluid
as a novel ex vivo model for the study of bacterial
free-​floating biofilms that form in human joint
infections. PLoS ONE 14, e0221012 (2019).
160. Dorken, G., Ferguson, G. P., French, C. E.
& Poon, W. C. K. Aggregation by depletion attraction
in cultures of bacteria producing exopolysaccharide.
J. R. Soc. Interface 9, 3490–3502 (2012).
First paper linking polymer depletion with bacterial
aggregates.
161. O’Toole, G. A. & Kolter, R. Flagellar and twitching
motility are necessary for Pseudomonas aeruginosa
biofilm development. Mol. Microbiol. 30, 295–304
(1998).
162. Schembri, M. A., Kjaergaard, K. & Klemm, P. Global
gene expression in Escherichia coli biofilms. Mol.
Microbiol. 48, 253–267 (2003).
163. Whiteley, M. et al. Gene expression in Pseudomonas
aeruginosa biofilms. Nature 413, 860–864 (2001).
164. Wagner, V. E., Bushnell, D., Passador, L., Brooks, A. I.
& Iglewski, B. H. Microarray analysis of Pseudomonas
aeruginosa quorum-​sensing regulons: effects of
growth phase and environment. J. Bacteriol. 185,
2080–2095 (2003).
0123456789();:
Reviews
165. Bagge, N. et al. Pseudomonas aeruginosa biofilms
exposed to imipenem exhibit changes in global gene
expression and beta-​lactamase and alginate
production. Antimicrob. Agents Chemother. 48,
1175–1187 (2004).
166. Ren, Y. et al. Emergent heterogeneous
microenvironments in biofilms: substratum surface
heterogeneity and bacterial adhesion force-​sensing.
FEMS Microbiol. Rev. 42, 259–272 (2018).
167. Serra, D. O. & Hengge, R. Stress responses go three
dimensional – the spatial order of physiological
differentiation in bacterial macrocolony biofilms.
Environ. Microbiol. 16, 1455–1471 (2014).
168. Povolotsky, T. L., Keren-​Paz, A. & Kolodkin-​Gal, I.
Metabolic microenvironments drive microbial
differentiation and antibiotic resistance. Trends Genet.
37, 4–8 (2021).
169. Liao, J., Schurr, M. J. & Sauer, K. The MerR-​like
regulator BrlR confers biofilm tolerance by activating
multidrug-​efflux pumps in Pseudomonas aeruginosa
biofilms. J. Bacteriol. 195, 3352–3363 (2013).
170. Kowalski, C. H., Morelli, K. A., Stajich, J. E., Nadell, C. D.
& Cramer, R. A. A heterogeneously expressed gene
family modulates the biofilm architecture and hypoxic
growth of Aspergillus fumigatus. mBio https://doi.org/
10.1128/mBio.03579-20 (2021).
171. Nair, H. A., Periasamy, S., Yang, L., Kjelleberg, S.
& Rice, S. A. Real time, spatial, and temporal mapping
of the distribution of c-​di-GMP during biofilm
development. J. Biol. Chem. 292, 477–487 (2017).
172. Klauck, G., Serra, D. O., Possling, A. & Hengge, R.
Spatial organization of different sigma factor activities
and c-​di-GMP signalling within the three-​dimensional
landscape of a bacterial biofilm. Open Biol. 8,
180066 (2018).
173. Lenz, A. P., Williamson, K. S., Pitts, B., Stewart, P. S.
& Franklin, M. J. Localized gene expression in
Pseudomonas aeruginosa biofilms. Appl. Environ.
Microbiol. 74, 4463–4471 (2008).
174. Shrout, J. D. et al. The impact of quorum sensing
and swarming motility on Pseudomonas aeruginosa
biofilm formation is nutritionally conditional.
Mol. Microbiol. 62, 1264–1277 (2006).
175. Krasteva, P. V. et al. Vibrio cholerae VpsT regulates
matrix production and motility by directly sensing
cyclic di-​GMP. Science 327, 866–868 (2010).
176. Boles, B. R. & Horswill, A. R. Agr-​mediated dispersal
of Staphylococcus aureus biofilms. PLoS Pathog. 4,
e1000052 (2008).
177. Rumbo-​Feal, S. et al. Whole transcriptome analysis of
Acinetobacter baumannii assessed by RNA-​sequencing
reveals different mRNA expression profiles in biofilm
compared to planktonic cells. PLoS ONE 8, e72968
(2013).
178. Bielecki, P. et al. In-​vivo expression profiling
of Pseudomonas aeruginosa infections reveals
niche-specific and strain-independent transcriptional
programs. PLoS ONE 6, e24235 (2011).
179. Potvin, E. et al. In vivo functional genomics of
Pseudomonas aeruginosa for high-​throughput
screening of new virulence factors and antibacterial
targets. Environ. Microbiol. 5, 1294–1308 (2003).
Acknowledgements
The authors thank J. Story for help with preparing Fig. 3.
Author contributions
The authors contributed equally to all aspects of the article.
Competing interests
The authors declare no competing interests.
Peer review information
Nature Reviews Microbiology thanks Ilana Kolodkin-​Gal and
the other, anonymous, reviewer(s) for their contribution to the
peer review of this work.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
Springer Nature or its licensor holds exclusive rights to this
article under a publishing agreement with the author(s) or
other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© Springer Nature Limited 2022