Bacterial Biofilms: Development, Dispersal, and Therapeutic Strategies in the Dawn of the Postantibiotic Era

Abstract

Biofilm formation constitutes an alternative lifestyle in which microorganisms adopt a multicellular behavior that facilitates and/or prolongs survival in diverse environmental niches. Biofilms form on biotic and abiotic surfaces both in the environment and in the healthcare setting. In hospital wards, the formation of biofilms on vents and medical equipment enables pathogens to persist as reservoirs that can readily spread to patients. Inside the host, biofilms allow pathogens to subvert innate immune defenses and are thus associated with long-term persistence. Here we provide a general review of the steps leading to biofilm formation on surfaces and within eukaryotic cells, highlighting several medically important pathogens, and discuss recent advances on novel strategies aimed at biofilm prevention and/or dissolution.

Some medically important pathogens (e.g., Pseudomonas aeruginosa, which causes upper respiratory infections) form biofilms. Strategies to disrupt biofilm formation (e.g., using antiadhesion agents) are being explored.

Biofilm formation enables single-cell organisms to assume a temporary multicellular lifestyle, in which “group behavior” facilitates survival in adverse environments. What was once defined as the formation of a community of microorganisms attached to a surface has come to be recognized as a complex developmental process that is multifaceted and dynamic in nature. The transition from planktonic growth to biofilm occurs in response to environmental changes, and involves multiple regulatory networks, which translate signals to concerted gene expression changes thereby mediating the spatial and temporal reorganization of the bacterial cell (Pratt and Kolter 1998; O’Toole et al. 2000; Prigent-Combaret et al. 2001; Parsek and Singh 2003; Lenz et al. 2008; Monds and O’Toole 2009). This cellular reprogramming alters the expression of surface molecules, nutrient utilization, and virulence factors and equips bacteria with an arsenal of properties that enable their survival in unfavorable conditions (Whiteley et al. 2001; Schembri et al. 2003; Stanley et al. 2003; Bagge et al. 2004; Beloin et al. 2004; Vuong et al. 2004; Lenz et al. 2008; Zhang and Mah 2008; Klebensberger et al. 2009).

Within the biofilm, bacteria are cocooned in a self-produced extracellular matrix, which accounts for ∼90% of the biomass (Flemming and Wingender 2010). The matrix is composed of extracellular polymeric substances (EPS) that, along with carbohydrate-binding proteins (Tielker et al. 2005; Branda et al. 2006; Diggle et al. 2006), pili, flagella, other adhesive fibers (Zogaj et al. 2001; Pinkner et al. 2006; Cegelski et al. 2009), and extracellular DNA (eDNA) (Whitchurch et al. 2002; Palchevskiy and Finkel 2006; Qin et al. 2007; Yang et al. 2007; Thomas et al. 2008, 2009; Guiton et al. 2009; Vilain et al. 2009), act as a stabilizing scaffold for the three-dimensional biofilm structure (Fig. 1). In the matrix, nutrients are trapped for metabolic utilizations by the resident bacteria and water is efficiently retained through H-bond interactions with hydrophilic polysaccharides (Conrad et al. 2003; Flemming and Wingender 2010). Enzymes secreted by the bacteria modify EPS composition in response to changes in nutrient availability (Sauer et al. 2004; Gjermansen et al. 2005), thereby tailoring biofilm architecture to the specific environment (Sauer et al. 2004; Ma et al. 2009). Thus, the structural components of the matrix give rise to a highly hydrated, robust structure with high tensile strength that keeps bacteria in close proximity, enabling intimate cell-to-cell interactions and DNA exchange (Flemming and Wingender 2010; Koo et al. 2010), while protecting the biomass from desiccation, predation, oxidizing molecules, radiation, and other damaging agents (Walters et al. 2003; Jefferson et al. 2005; Mai-Prochnow et al. 2008; Flemming and Wingender 2010). The resilient nature of biofilms is also partly attributed to the presence of environmental gradients within the biomass, which give rise to community “division of labor” with subpopulations of bacteria showing differential gene expression in response to local nutrient and oxygen availability (Lewis 2005; Domka et al. 2007). Studies have shown the presence of metabolically inactive nondividing persister cells within biofilms, which are tolerant to a number of antibiotics despite the fact that they are genetically identical to the rest of the bacterial population (Lewis 2005, 2008). These are believed to be responsible for the reseeding of biofilms on cessation of antibiotic treatment in the clinical setting (Lewis 2005, 2008).

Figure 1.

Schematic of the extracellular matrix composition in E. coli. Structural components include the EPS molecules colonic acid, cellulose, and PGA (polyglucosamine), which enable intercellular interactions, keeping bacteria in close proximity with each other. eDNA also serves as a connecting agent, as well as a nutritional source. Extracellular organelles such as flagella and CUP (chaperone usher pathway) pili enable bacterial aggregation strengthening the biofilm lattice. Secreted enzymes modify EPS components in response to environmental changes.

Inside the host, the matrix protects biofilm bacteria from exposure to innate immune defenses (such as opsonization and phagocytosis) and antibiotic treatments (Jesaitis et al. 2003; Walters et al. 2003; Jefferson et al. 2005; Leid et al. 2005; Cerca et al. 2006, 2007). Interbacterial interactions can promote the spread of drug-resistance markers and other virulence factors (Vuong et al. 2004). As a result, biofilm-forming pathogens persist, establishing chronic and recalcitrant infections such as upper respiratory infections (Pseudomonas aeruginosa) (Koch and Hoiby 1993; Govan and Deretic 1996), urinary tract infections (UTIs) (uropathogenic Escherichia coli [UPEC], Klebsiella pneumoniae) (Foxman 2010), periodontitis (mixed biofilms of Streptococcus mutans and other bacteria) (Kuramitsu and Wang 2011), catheter-induced and other device-associated infections (E. coli, Enterococcus faecalis, and others) (Venditti et al. 1993; Ferrieres et al. 2007; Jacobsen et al. 2008; Fey 2010). Especially in immunocompromised patients, the manifestation of infections by opportunistic biofilm-forming pathogens can be devastating, leading to severe symptoms and, in many instances, death.

Here we review the processes leading to the formation of extracellular and intracellular biofilms, highlighting several medically important pathogens. Given the prevalence and recalcitrance of biofilm-related infections, we also provide a synopsis of the most recent advances in the development of novel antibiofilm strategies.

EXTRACELLULAR BIOFILM FORMATION

Bacterial Adherence on Surfaces—What Does It Take to Stick and Stick Around?

Bacterial aggregation and subsequent biofilm maturation consists of reversible and irreversible stages and involves numerous conserved and/or species-specific factors. The first step involves the introduction of bacteria to a surface, a process which is at least in part stochastic, driven by Brownian motion and gravitational forces, and influenced by surrounding hydrodynamic forces (Donlan 2002; Beloin et al. 2008). Within a niche, bacteria encounter attractive or repelling forces that vary depending on nutrient levels, pH, ionic strength, and temperature. Medium properties, along with bacterial cell-surface composition affect velocity and direction toward or away from the contact surface (Donlan 2002). Motile bacteria have a competitive advantage, utilizing flagella to overcome hydrodynamic and repulsive forces. The importance of flagellar motility for initial attachment has been documented for several pathogens, including P. aeruginosa, Vibrio cholerae, Listeria monocytogenes, and E. coli (O’Toole and Kolter 1998; Pratt and Kolter 1998; Watnick and Kolter 1999; Klausen et al. 2003a,b; Lemon et al. 2007; Toutain et al. 2007). In some bacterial species, chemotaxis also plays a role in directing attachment in response to nutrient composition; mutations in the CheR1 methyltransferase have been shown to alter the amino acid response of P. aeruginosa and impair attachment and biofilm maturation (Schmidt et al. 2011). Previous studies showed that chemotaxis is dispensable in E. coli (Pratt and Kolter 1998); however, recent investigations have revealed that disruption of the methyl-accepting chemotaxis protein II (tar), imparts biofilm defects in UPEC (Hadjifrangiskou et al. 2012).

Upon intercepting the surface, adherence is mediated by additional extracellular adhesive appendages and secreted adhesins. However, the decision to “stick” is not absolute; initial attachment is dynamic and reversible, during which bacteria can detach and rejoin the planktonic population if perturbed by hydrodynamic forces (sloughing bacteria off the surface), repulsive forces (Dunne 2002), or in response to nutrient availability (Banin et al. 2005; Anderson et al. 2008; Wu and Outten 2009).

Irreversible attachment is attained by bacteria that can weather shear forces and maintain a steadfast grip on the surface. UPEC and other E. coli pathotypes rely heavily on type 1 pili (Mulvey et al. 1998; Pratt and Kolter 1998; Martinez et al. 2000; Hung et al. 2002; Anderson et al. 2003; Beloin et al. 2008), which are multisubunit adhesive organelles assembled by the chaperone usher pathway (CUP) (Waksman and Hultgren 2009). UPEC harbor numerous CUP pili systems, which are differentially expressed and are presumed to facilitate adherence in a niche-specific manner (Welch et al. 2002; Chen et al. 2006, 2009; Hadjifrangiskou et al. 2011; Spurbeck et al. 2011). Adherence is mediated by the FimH adhesin at the tip of type 1 pili, which recognizes mannosylated moieties (Thankavel et al. 1997; Martinez et al. 2000; Zhou et al. 2001; Hung et al. 2002; Bouckaert et al. 2005; Eto et al. 2007; Nilsson et al. 2007; Wellens et al. 2008; Thumbikat et al. 2009). FimH is thought to play a critical role in UPEC pathogenesis; it mediates binding and invasion to human bladder epithelial cells, binds to human uroplakin, and is critical in a murine preclinical model of cystitis, which mimics human disease (Mulvey et al. 1998; Martinez et al. 2000; Kau et al. 2005; Bishop et al. 2007; Eto et al. 2007; Garofalo et al. 2007; Rosen et al. 2007; Wright et al. 2007; Chen et al. 2009). FimH is found under positive selection in UPEC, consistent with its role as a virulence factor in human disease (Sokurenko et al. 1994, 1995, 1998, 2004; Chen et al. 2006, 2009; Weissman et al. 2007; Wright et al. 2007) and has been said to have fulfilled Koch’s postulates (Connell et al. 1996; Snyder et al. 2006). In addition to type 1 pili, curli fibers and Antigen 43 have been shown to mediate attachment and interbacterial interactions on abiotic surfaces (Henderson et al. 1997; Hasman et al. 1999; Danese et al. 2000a; Kjaergaard et al. 2000; Ulett et al. 2007; Cegelski et al. 2009). Curli also facilitates binding to the eukaryotic extracellular matrix components laminin, fibronectin, and plasminogen (Vidal et al. 1998; Cookson et al. 2002; Uhlich et al. 2006).

P. aeruginosa, an important pathogen and avid biofilm former, also uses several attachment organelles to irreversibly adhere to a surface. Besides flagella, P. aeruginosa uses type IV pili-mediated twitching motility to wade through the liquid interface and contact the surface, maintain adherence, and move across the attachment plane (O’Toole and Kolter 1998; Klausen et al. 2003a,b). Similar to UPEC, P. aeruginosa express numerous CUP fimbriae, of which CupA is involved in surface adherence and autoaggregation (Vallet et al. 2001; Klebensberger et al. 2009).

In contrast to Pseudomonas and UPEC, the Gram-positive Enterococci are nonmotile and, up until recently, were thought to possess no adhesive pili. Over the years, investigations identified a panel of enterococcal adhesins that mediate adherence to eukaryotic extracellular matrix components. Examples include SagA, Acm (E. faecium), and Ace (E. faecalis), which bind collagen (Mohamed et al. 2006), and the surface protein Esp, which has been shown to promote biofilm formation on abiotic surfaces in esp-expressing E. faecalis strains (Toledo-Arana et al. 2001). Recent studies elucidated the presence of Enterococcal biofilm pili (Ebp) in E. faecalis and showed their contribution to biofilm formation, endocarditis, and urinary tract infection (Ton-That and Schneewind 2003; Ton-That et al. 2004; Nallapareddy et al. 2006; Kemp et al. 2007; Guiton et al. 2009; Kline et al. 2010).

Biofilm Maturation—Keeping It Together

Surface contact triggers responses that lead to gene expression changes, up-regulating factors favoring sessility, such as those implicated in the formation of the extracellular matrix (Prigent-Combaret and Lejeune 1999; Otto and Silhavy 2002; Inagaki et al. 2005; Morici et al. 2007; Beloin et al. 2008; Bhomkar et al. 2010). In the case of E. coli, relatively little is known about matrix constituents. Cellulose was first identified as an important component of commensal E. coli pellicle biofilms, and was later shown to be coexpressed with curli in UPEC and gastrointestinal E. coli isolates (Zogaj et al. 2001, 2003; Romling 2002; Bokranz et al. 2005; Kai-Larsen et al. 2010). Curli are amyloid fibers that are critical for the formation of pellicle biofilms, as curli inhibitors (curlicides) inhibit pellicle formation and curli mutants cannot form pellicles (Cegelski et al. 2009). Additional studies showed that polyglucosamine (PGA) and colanic acid contribute to biofilm architecture (Prigent-Combaret and Lejeune 1999; Danese et al. 2000b; Prigent-Combaret et al. 2001; Wang et al. 2004; Agladze et al. 2005), with PGA being prevalent among clinical isolates, including UPEC (Cerca et al. 2007). More detailed analyses are required for a complete characterization of the extracellular matrix in pathogenic E. coli.

Extracellular matrix composition has been more extensively investigated in P. aeruginosa, and has been shown to vary depending on environmental conditions (Harmsen et al. 2010). Two primary EPS components are Pel and Psl (Friedman and Kolter 2004a,b; Jackson et al. 2004; Matsukawa and Greenberg 2004; Vasseur et al. 2005; Ma et al. 2006). Psl augments Pseudomonas attachment to mucin and airway epithelial cells (Ma et al. 2006), whereas increased expression of pel in small colony variants isolated from cystic fibrosis patients has been associated with P. aeruginosa persistence in lung airways (Starkey et al. 2009). Recently, Borlee and colleagues identified CdrA, a large secreted adhesin, which is expressed in the biofilm in response to high levels of the universal signal 3,5-cyclic diguanylic acid (c-di-GMP) and binds Psl, stabilizing biofilm structures (Borlee et al. 2010). Alginate, another P. aeruginosa EPS component, has been associated with increased resistance to antibiotic treatments and host immune defenses during chronic infection (Govan and Deretic 1996; Hatch and Schiller 1998; Hentzer et al. 2001; Leid et al. 2005). As is the case with Pel and Psl, alginate production is subject to regulation by fluctuating levels of c-di-GMP. Recent studies have shown that a surface-bound diguanylate cyclase MucR positively activates alginate synthesis, presumably through high local concentrations of c-di-GMP (Hay et al. 2009). In addition to EPS, several studies have shown that eDNA is critical for cell-to-cell connections and stabilization of Pseudomonas biofilms (Whitchurch et al. 2002; Yang et al. 2007). Young Pseudomonas biofilms are more sensitive to DNase treatment compared with mature biofilms, indicating a stabilizing role for eDNA during the initial biofilm stages when EPS components are not as abundant (Whitchurch et al. 2002). As the biofilm matures, eDNA amounts increase through lysis of a bacterial subpopulation in response to the P. aeruginosa quinolone signal (Pqs) quorum sensing system (Allesen-Holm et al. 2006). Allesen-Holm et al. showed that eDNA is organized in distinct patterns and localizes in the stalk portion of the mushroom-shaped biofilms (Allesen-Holm et al. 2006). This localization may act as a scaffold for the formation of the mushroom structure, as type IV pili show high eDNA binding affinity, inducing the accumulation of migrating bacteria toward areas of high eDNA concentration (Barken et al. 2008).

The contribution of eDNA to biofilm architecture has also been reported for E. faecalis, making it one of the few known E. faecalis matrix components. Thomas et al. first reported that eDNA is critical for E. faecalis biofilms and identified that the secreted enzymes GelE (zinc metalloprotease) and SprE (serine protease) influence biofilm formation by affecting cellular autolysis and DNA release (Thomas et al. 2008, 2009). In a separate study, Mohamed et al. reported that a mutant lacking the Atn autolysin had 30% reduction in biofilm (Mohamed et al. 2004). Guiton and colleagues later established that Atn plays a role in the temporal regulation of DNA release at specific stages during biofilm formation (Guiton et al. 2009).

Escape from the Matrix—Dispersing Mechanisms

Within the mature biofilm there is a bustling community that actively exchanges and shares products that play a pivotal role in maintaining biofilm architecture and providing a favorable living environment for the resident bacteria. However, as biofilms mature, dispersal becomes an option. Besides passive dispersal, brought about by shear stresses, bacteria have evolved ways to perceive environmental changes and gauge whether it is still beneficial to reside within the biofilm or whether it is time to resume a planktonic lifestyle. Biofilm dispersal can be the result of several cues, such as alterations in nutrient availability, oxygen fluctuations and increase of toxic products, or other stress-inducing conditions (Sauer et al. 2004; Karatan and Watnick 2009; Hong et al. 2010; Rowe et al. 2010). In UPEC, increase in extracellular iron induces biofilm dispersal (Rowe et al. 2010), whereas P. aeruginosa biofilms disperse in response to increased amounts of various carbon and nitrogen sources (Sauer et al. 2004; Karatan and Watnick 2009). Several sensory systems monitor the levels of small molecules, as a proxy to environmental changes, and alter gene expression accordingly, promoting dispersal (Hammer and Bassler 2003; Kaplan 2010). Among other signals, the universal c-di-GMP has been extensively implicated in the shift between sessility and motility in bacteria, including P. aeruginosa and E. coli. Typically, an increase in c-di-GMP favors sessility, whereas reduced c-di-GMP leads to up-regulation of motility (Morgan et al. 2006; Pruss et al. 2006; Barraud et al. 2009; Wood et al. 2010). Ma et al. recently reported that a c-di-GMP binding protein, BdcA, is at least partly responsible for the reduction of available c-di-GMP in biofilm communities, down-regulation of EPS, and up-regulation of swimming and swarming motility; a phenomenon that the investigators showed also occurs in Pseudomonas species and Rhizobium mellioti (Ma et al. 2011a,b).

EPS-degrading enzymes, such as alginate lyase in P. aeruginosa, also contribute to bacterial detachment from the matrix (Boyd and Chakrabarty 1994). In E. coli, the CsrA protein was shown to repress PGA synthesis, also aiding in dispersion (Wang et al. 2005). Besides down-regulating EPS, surfactant molecules are produced, reducing surface-bacterial interactions; for example, although controlled rhamnolipid production contributes to channel formation within mature P. aeruginosa biofilms, an increase in rhamnolipid levels aids bacterial dispersal (Boles et al. 2005; Dong et al. 2008; Harmsen et al. 2010). In addition, studies have identified flagellated subpopulations within P. aeruginosa biofilms, which emigrate from the biofilm, creating microcolonies with a central void (Purevdorj-Gage et al. 2005; Harmsen et al. 2010). Voids within the biofilm are also created by cell death, serving as an additional dispersal mechanism that frees resident live bacteria, as shown by studies in P. aeruginosa (Webb et al. 2003). Dispersing bacteria have the capacity to reinitiate the process of biofilm formation, on encountering a suitable environment.

Studies using Bacillus subtilis as a model organism revealed another sophisticated dispersal mechanism that may be widespread among bacteria. B. subtilis forms robust biofilms, which lose their integrity after 5–8 d; Kolodkin-Gal and colleagues found that biofilm disassembly is facilitated by a mixture of d-amino acids (d-leucine, d-methionine, d-tyrosine, and d-tryptophan) that are produced during the stationary phase of growth and get incorporated into the peptide side chains of peptidoglycan in place of the terminal d-alanine (Lam et al. 2009; Kolodkin-Gal et al. 2010). This d-amino acid incorporation interferes with the anchoring of adhesive fibers on the cell surface, leading to fiber dissociation and loss of bacterial adherence, without influencing bacterial growth or expression of matrix components (Kolodkin-Gal et al. 2010). Exogenous addition of the d-amino acid mixture or the individual d-amino acids disrupted preformed biofilms of B. subtilis and other bacterial species (Kolodkin-Gal et al. 2010). Further studies revealed that d-amino acids work together with norspermidine, another factor produced by B. subtilus, to cause biofilm disassembly (Kolodkin-Gal et al. 2012). Thus, d-amino acid/norsperimidine treatment may hold promising potential in preventing or eradicating biofilms.

THE LIFE WITHIN—INTRACELLULAR BIOFILMS

Accumulating evidence indicates that many bacterial pathogens previously considered as strictly extracellular can persist inside the host by adapting an intracellular lifestyle that involves the formation of bacterial communities with biofilm-like properties. These intracellular bacterial communities (IBCs) were first documented for UPEC, using a murine model of infection (Mulvey et al. 1998; Anderson et al. 2003; Justice et al. 2004). UPEC use type 1 pili to bind mannosylated receptors on the superficial umbrella bladder cells (Zhou et al. 2001; Hung et al. 2002; Bouckaert et al. 2005; Eto et al. 2007; Wellens et al. 2008; Thumbikat et al. 2009), triggering events that lead to bacterial internalization. Although internalized UPEC are expelled in a TLR-4-dependent process (Bishop et al. 2007), some bacteria avoid the exocytic process and escape into the host-cell cytoplasm, where they replicate into IBCs (Anderson et al. 2003; Justice et al. 2004).

IBCs progress through several developmental stages that show distinct morphological characteristics (Fig. 2) (Justice et al. 2004). During the first 6 h following bladder inoculation, UPEC divide rapidly (doubling time of ∼30–35 min) resulting in small clusters of loosely associated rods (early IBCs), morphing into coccoid-shaped bacteria, with an average length of 0.7 µm that begin packing into a tight biomass. Then, between 6 and 8 h, the growth rate drops dramatically, resulting in doubling times >60 min. At this stage, bacteria are tightly packed together forming a highly organized sphere inside the cell that comprises the mature middle-stage IBC (Fig. 2). The number of IBCs can range between 3 and 700 IBCs in an infected bladder; each IBC is clonal and composed of ∼10⁴–10⁵ bacteria (Anderson et al. 2003; Schwartz et al. 2011). IBC bacteria are surrounded by numerous fibers that emanate from the bacterial surface, resembling an extracellular matrix and encasing bacteria in individualized compartments (Anderson et al. 2004). Polysaccharides, such as the sialic acid capsule, are also present throughout the IBC and function, in part, to protect the bacteria from neutrophil attack (Anderson et al. 2010). Similar to extracellular biofilms, IBCs are heterogeneous, composed of subpopulations with different gene expression patterns (Anderson et al. 2004).

Figure 2.

Schematic of the IBC developmental cascade in UPEC (uropathogenic Escherichia coli), accompanied by SEM (scanning electron microscopy) images depicting the distinct morphological changes from attachment and invasion to filamentation and dispersal. (SEM images from Anderson et al. 2003, Hultgren Lab.)

As IBCs enlarge, the bacterial mass pushes against the host-cell membrane creating a podlike protrusion on the surface of the infected cell (Anderson et al. 2003). Eventually, UPEC at the IBC periphery detach as single rods or filaments, and flux out of the infected cell into the bladder lumen where they can reinitiate the process by binding and invading naive epithelial cells (Justice et al. 2004). The cell division inhibitor SulA has been shown to be important for filamentation and dispersal of UPEC from the biomass and, thus, establishment of next-generation IBCs (Justice et al. 2006). UPEC filaments have been shown to be a common feature in the urines of patients with UTI, but not in otherwise healthy controls (Rosen et al. 2007). Further, UPEC isolated from the urine of patients with a UTI have been shown to form IBCs when inoculated into the bladders of six different strains of mice, indicating that IBCs are important for human infection (Garofalo et al. 2007). IBC formation is restricted by severe molecular bottlenecks, and higher IBC numbers during acute infection are associated with founding the development of chronic cystitis, indicating the importance of the intracellular pathway in UTI pathogenesis (Hannan et al. 2010; Schwartz et al. 2011).

The IBC cycle is FimH dependent, as interruption of type 1 pili expression after the invasion step, disrupts normal IBC development and leads to UPEC attenuation (Wright et al. 2007). The QseBC two-component system is one of the factors influencing type 1 pili, curli expression, and IBC formation. Recent studies showed that deletion of the QseC sensor results in overactivation of the cognate response regulator QseB, which leads to virulence gene down-regulation by interfering with core metabolic processes (Kostakioti et al. 2009; Hadjifrangiskou et al. 2011). These studies also showed that the UPEC intracellular pathway requires completion of the TCA cycle (Hadjifrangiskou et al. 2011). Microarray and qPCR analyses probing the expression patterns of UPEC within IBCs revealed that iron-acquisition systems are highly up-regulated, indicating the importance of these systems for intracellular biofilm formation (Reigstad et al. 2007). Henderson et al. later showed that the same iron-acquisition systems are prevalent among UPEC isolates (Henderson et al. 2009).

Intracellular communities have also been reported for K. pneumoniae, which accounts for up to 5% of community-acquired UTIs, and is more prevalent in diabetic patients and in the nosocomial setting (Lye et al. 1992; Hansen et al. 1998). Similar to UPEC, type 1 pili mediate K. pneumoniae invasion and IBC formation, albeit with differences in pili expression kinetics and numbers of formed IBCs and filaments (Rosen et al. 2008a,b).

The ability to occupy an intracellular niche and persist inside the host by transitioning from single cell to a multicellular community is not confined to uropathogens. Using cell lines and animal models of acute lung infection, Garcia-Medina and colleagues showed that following infection, P. aeruginosa can form clusters within the airway cells that matured to a podlike structure, similar in morphology to UPEC and Klebsiella IBCs (Garcia-Medina et al. 2005). Bacteria within the Pseudomonas pod structure showed regional variation in their expression patterns similar to what was reported for UPEC, which is a typical characteristic of extracellular biofilms (Garcia-Medina et al. 2005).

The ability to form intracellular biofilms may be an evolutionary adaptation that facilitates bacterial persistence to a level extending even beyond that attained by extracellular biofilms. For example, during UTI, hordes of neutrophils infiltrate the bladder migrating toward the infected superficial umbrella cells, but are unable to effectively penetrate the IBC or engulf dispersing filamentous bacteria (Justice et al. 2008; Horvath et al. 2010). The ability of IBCs to repel neutrophil penetration is lost in K1 capsule mutants (Anderson et al. 2010). Moreover, Blango and Mulvey showed that 17 different antibiotics capable of killing the virulent cystitis isolate UTI89 in vitro or in tissue culture were unable to eliminate UTI89 from bladder tissue during infection (Blango and Mulvey 2010). These findings indicate that IBC formation is a mechanism that enables rapid bacterial expansion within the host and contributes to bacterial persistence.

BIOFILM INHIBITION: TREATMENT STRATEGIES IN THE POSTANTIBIOTIC ERA

Antibiotics are currently the preferred treatment strategy for bacterial infections. Conventional antibiotics work by either preventing bacterial cell division (bacteriostatic) or killing the cell (bactericidal). Although over the years antibiotics have proven critical in eliminating bacterial pathogens, overwhelming evidence indicates that they extensively damage the host microbiota, creating an environment where opportunistic pathogens can prevail, and they increase the selective pressure toward antibiotic resistance (Dethlefsen et al. 2008; Dethlefsen and Relman 2010; Ubeda et al. 2010). Moreover, although prophylactic antibiotic administration preceding surgery is highly successful in reducing infection rate, it has little or no protective effects in surgical procedures involving implants or prostheses (Secinti et al. 2011). In most cases, the best treatment for foreign body-associated biofilm infections is to remove the infected device. However, in cases like implantable prostheses, pacemakers, and cardiac implants, device removal is difficult (Fey 2010).

Biofilm bacteria are particularly recalcitrant to antibiotic treatments not only owing to increased transmission of resistance markers within the biofilm community, but also because of diffusion limitations posed by the extracellular matrix, antibiotic inactivation by high metal ion concentration and low pH, and the presence of metabolically inactive persister cells that survive treatment (Mack et al. 2004; Lewis 2005; Costerton et al. 2007; Lewis 2008). Combined, these attributes make biofilm bacteria up to 1000-fold more tolerant and/or resistant to antibiotics than planktonic cells (Hoiby et al. 2010). Thus, the need for more effective biofilm dissolution treatments becomes imperative. Below we present some of the most recent advances in strategies designed to thwart biofilm formation by killing the bacteria or targeting different biofilm developmental stages (also summarized in Fig. 3).

Figure 3.

Schematic outlining the stages in biofilm development and listing the strategies aimed at inhibiting and/or disrupting biofilm formation at specific stages.

Bactericidal Strategies

Phage Therapy

Phage therapy is a promising alternative to antibiotic treatments (Donlan 2009); phages are abundant and can be easily isolated from a wide range of environments, they are usually specific to narrow host ranges (thus not likely to perturb the host microbiota), and their self-replicating mode permits low dosage (Burrowes et al. 2011). Moreover, the high phage mutation rate facilitates adaptation as the corresponding bacterial hosts accumulate mutations to persist in a given environment. Phage therapy takes advantage of lytic phages that do not enter a prophage state and thus rarely contain or transfer virulence genes, although they result in rapid destruction of the bacterial cell. Many phages have been shown to encode EPS-degrading enzymes (Hughes et al. 1998a,b; Sutherland et al. 2004), or propagate on stationary-phase bacteria, making them more likely to persist within the biofilm (Burrowes et al. 2011).

Silver Nanoparticles

Impregnation of medical devices with antimicrobial agents has been the most commonly used approach for preventing device-associated biofilms (Fey 2010). One of the frequently used agents is silver, which has been used as an anti-infective for hundreds of years and has been extensively used to sterilize wound infections during World War I (Rupp et al. 2005; Chen and Schluesener 2008). The positively charged silver ions facilitate electrostatic attractions between the metal and the negatively charged bacterial membrane, augmenting uptake and antimicrobial activity (Kim et al. 2007). The lethality of silver for bacteria is partly owing to thiol-group reactions that inactivate enzymes (Chen and Schluesener 2008). As a result, silver treatment inhibits DNA replication, expression of ribosomal and other cellular proteins, and interferes with the bacterial electron transport chain (Bragg and Rainnie 1974; Feng et al. 2000; Yamanaka et al. 2005).

The potential toxicity of silver in humans led to its dwindling use for some time. However, the popularity of silver has been revived with the advent of nanotechnology (Chen and Schluesener 2008). Nanoparticles are typically no greater than 100 nm in size and their biocidal effectiveness is suggested to be owing to a combination of their small size and high surface-to-volume ratio, which enable intimate interactions with microbial membranes (Morones et al. 2005; Allaker 2010). Silver nanoparticles have been shown to inhibit P. aeruginosa and Staphylococcus epidermidis biofilms by >95%; studies in rabbits showed that nanoparticle silver ion-coated implants inhibited Staphylococcus aureus biofilm formation without causing silver accumulation in host tissues, even 28 d after impregnation (Kalishwaralal et al. 2010; Secinti et al. 2011).

Antimicrobial Peptides

Antimicrobial peptides are produced by the innate immune response system and have been proposed as attractive candidates for the development of novel types of antibiotics (Yang et al. 2002). However, their activity spectrum and mechanism of action need to be more precisely defined before they can be considered as possible therapeutic strategies (Pompilio et al. 2011). Cathelicidins constitute one of the most important classes of antimicrobial peptides. Recent work indicated that SMAP-29, BMAP-28, and BMAP-27 significantly reduced biofilm formation by multidrug-resistant (MDR) P. aeruginosa strains isolated from patients with CF (cystic fibrosis), and killed bacteria within preformed biofilms (Pompilio et al. 2011). This study compared the bactericidal activity of cathelicidins to that of tobramycin, the frontline antibiotic used to treat P. aeruginosa airway infections in CF patients. Pompilio et al. found that in contrast to tobramycin, the active cathelicidin peptides showed faster kinetics, exerting a rapid bactericidal activity regardless of the species tested. Although the extent of bacterial killing was overall higher with tobramycin, this study shows that cathelicidins may hold potential as antibiofilm agents in the case of MDR strains (Pompilio et al. 2011).

Lytic peptides are another group of antimicrobial peptides assessed for their inhibitory effects on biofilm formation. Lytic peptides bind the LPS (lipopolysaccharide) moieties of the bacterial cell membrane, disrupting membrane stability. Studies in Staphylococcus aureus have shown that the lytic peptide PTP-7 prevented in vitro biofilm formation and was also capable of diffusing into the deep layer of preformed biofilm, killing 99.9% of biofilm bacteria. This peptide retained activity under highly acidic environments and in the presence of excess of metals, conditions that mimic the S. aureus biofilm environment (Kharidia and Liang 2011).

Antiadhesion Agents

Mannosides, Pilicides, and Curlicides

Attachment constitutes the first step in virtually all types of biofilm formation, thus numerous studies have focused on ablating bacterial adherence. In UPEC, efforts have concentrated on the development of compounds that interfere with the adhesive properties or assembly of type 1 pili, because they are the prevalent means for UPEC adherence during biofilm formation in vitro and within the host. The X-ray crystal structures of the FimH adhesin bound to mannose have been used to rationally design molecules, termed mannosides, which fit the FimH mannose binding pocket and competitively inhibit FimH binding to its host receptor (Fader and Davis 1980; Schaeffer et al. 1980; Hung et al. 2002; Bouckaert et al. 2005; Sperling et al. 2006; Wellens et al. 2008; Klein et al. 2010). Recent advances in this field led to the development of monomeric biphenyl mannosides with largely enhanced potency, relative to previously reported FimH inhibitors (Han et al. 2010). These optimized mannosides prevented UPEC biofilm formation in vitro and were shown to disrupt preformed biofilms (Cusumano et al. 2011). In vivo studies indicated that mannoside administration as a prophylactic measure for UTIs interfered with UPEC adherence and invasion, reducing IBC formation and attenuating UPEC during the acute infection stages (Cusumano et al. 2011). Moreover, mannosides enhanced the antimicrobial effects of trimethoprim-sulfamethoxazole (TMP-SMZ) preventing infection by PBC-1, a UPEC isolate that was resistant to TMP-SMZ treatment in the clinical setting (Cusumano et al. 2011). Given that the intracellular niche protects UPEC from antibiotic treatment (Blango and Mulvey 2010), the conferred PBC-1 susceptibility on dual mannoside/TMP-SMZ treatment is likely owing to bacterial sequestration in the extracellular environment, where antibiotic concentrations are higher relative to the intracellular compartment (Patel and Welling 1980; Cusumano et al. 2011). Mannosides were also efficient as a therapeutic strategy for chronic UTIs, significantly reducing the bladder bacterial load of orally treated mice within 6 h (Cusumano et al. 2011). Thus, if translated to the clinic mannosides, they hold great potential for the elimination of complicated UTIs associated with antibiotic-resistant UPEC strains.

In parallel, efforts have been made to inhibit assembly of type 1 pili and other CUP pili, through the use of pilicides, which are compounds rationally designed to interfere with export of the corresponding pilin subunits. Pilicides were shown to inhibit UPEC biofilm formation in vitro by 50%, at concentrations as low as 3 µM (Pinkner et al. 2006; Berg et al. 2008; Chorell et al. 2010, 2011). Similar compounds have been shown to be effective against curli (“curlicides”), inhibiting in vitro curli biogenesis, biofilm formation, and potentiating UPEC clearance from the urinary tract (Cegelski et al. 2009).

Polysaccharides

Exopolysaccharides mediate cell-to-surface and cell-to-cell interactions that are critical for biofilm formation and stabilization. Mutants unable to synthesize or export such polysaccharides are typically deficient in adherence and biofilm formation and thus are highly sensitive to killing by antibiotics and host immune defenses (Rendueles et al. 2012). However, recent evidence indicates that some bacterial exopolysaccharides inhibit or destabilize biofilm formation by other species. For example, in the case of P. aeruginosa, Qin and colleagues showed that Pel and Psl-containing culture supernatants disrupted preformed S. epidermidis and S. aureus biofilms without inhibiting bacterial growth (Qin et al. 2009). Moreover, the presence of P. aeruginosa inhibited S. epidermidis biofilm formation in dual-species in vitro biofilm experiments (Pihl et al. 2010). Polysaccharides with nonbiocidal antibiofilm properties have also been isolated from cell-free biofilm extracts of several species (Rendueles et al. 2012). Their antibiofilm properties are believed to lie on their ability to (a) alter the physical characteristics of bacterial cells or abiotic surfaces; (b) act as signaling molecules that impact the gene expression patterns of susceptible bacteria; or (c) competitively inhibit multivalent carbohydrate–protein interactions, thereby interfering with adhesion. Most antibiofilm polysaccharides show a broad spectrum of biofilm inhibition, whereas some are capable of dispersing preformed biofilms. Given their nonbiocidal mode of action, as well as their biocompatibility and biodegradability, antibiofilm polysaccharides could be a promising strategy suitable for the treatment and prevention of biofilm-related infections. Furthermore, several studies suggest that antibiofilm polysaccharides could be valuable as an adjuvant, because they enhance antibiotic functions when administered together. Other potential applications could be coating surfaces of indwelling medical devices or even using antibiofilm polysaccharide-producing bacteria in probiotics to outcompete pathogens (Rendueles et al. 2012).

Signal Transduction Interference

Many studies have focused on inhibiting biofilm initiation by interfering with bacterial signaling cascades, given that two-component systems constitute a central means of intercepting and translating environmental changes (Lyon et al. 2000; Okada et al. 2007; Cegelski et al. 2008; Rasko et al. 2008; Watanabe et al. 2008; Njoroge and Sperandio 2009). Inhibition of signal transduction systems poses an attractive means of antivirulence therapy, because interference with signaling does not kill the bacteria but rather deprograms optimal gene expression and ablates virulence without applying pressure for selection of resistance. Among the systems that appear to be attractive drug target candidates are the QseBC two-component system that is common among biofilm-forming Gram-negative pathogens (Clarke et al. 2006; Bearson and Bearson 2008; Rasko et al. 2008; Kostakioti et al. 2009; Khajanchi et al. 2011; Wang et al. 2011). Previous studies have investigated the potential of inhibiting QseC kinase activity and showed efficacy in reducing enterohemorrhagic Escherichia coli (EHEC) virulence (Rasko et al. 2008). Studies in UPEC and EHEC have shown that deletion of QseC results in the overactivation of the QseB response regulator owing to the specific phosphatase activity of QseC required for QseB deactivation. Thus, targeting QseC phosphatase activity would be an optimized strategy to decouple normal gene expression in QseC-bearing pathogens (Kostakioti et al. 2009; Hadjifrangiskou et al. 2011).

In E. faecalis, studies have targeted the FsrC/FsrA TCS. FsrC/FsrA controls the expression of fsrBDC and gelE-sprE, leading to increased production of gelatinase and serine protease, both of which are required for appropriate production of eDNA (Qin et al. 2000; Thomas et al. 2008, 2009). High-throughput screening of compounds that inhibited gelatinase and the gelatinase biosynthesis-activating pheromone (GBAP), identified a peptide antibiotic, Siamycin I, as an inhibitor of GBAP signaling by FsrC/FsrA (Nakayama et al. 2007; Gotoh et al. 2010).

“Antimatrix” Agents

Besides ablating adherence, bacterial aggregation can be targeted by disrupting components of the extracellular matrix. Several investigations exploited the potential of inhibiting enzymes involved in the synthesis or modification of cell wall-associated or secreted EPS components and other matrix constituents. These studies use the direct use of naturally occurring or engineered enzymes, use bacteriophages (phage therapy) as a vehicle of enzyme delivery and expression, or take advantage of metal chelators as a means to disrupt matrix integrity.

Enzymes

N-acetyl-d-glucosamine-1-phosphate acetyltransferase (GlmU), which is involved in the biosynthesis of activated UDP-GlcNAc, an essential peptidoglycan and lipopolysaccharide (LPS) precursor in Gram-positive and Gram-negative pathogens, respectively, is among the enzymes targeted for matrix disruption (Burton et al. 2006). Burton et al. tested the effects of GlmU inhibitors, including N-ethyl maleimide (NEM), and the NEM analogs N-phenyl maleimide, N,N′-(1,2-phenylene)dimaleimide (oPDM), and N-(1-pyrenyl)maleimide (PyrM), on catheter-associated uropathogen biofilms. All NEM analogs showed antibiofilm activity against clinical isolates of E. coli, P. aeruginosa, K. pneumoniae, S. epidermidis, and E. faecalis (Burton et al. 2006). The same study showed that coating silicone catheters with a mixture of oPDM and the cationic polypeptide protamine sulfate (PS) enhanced the antibiofilm activity of PS against P. aeruginosa and S. epidermidis, demonstrating that this dual administration could be used as a broad-spectrum anti-infective coating for medical devices (Burton et al. 2006).

The enzymes DNase I and Dispersin B have also recently gained attention as potential antibiofilm agents, particularly against Gram-positive pathogens. The effects of DNase I lie on its ability to digest the eDNA found within the biofilm structure (Qin et al. 2007; Zhu et al. 2007; Guiton et al. 2009). DNase treatment prevented Staphylococcus and Enterococcus biofilm formation (Guiton et al. 2009; Mann et al. 2009) and dispersed preformed biofilms in vitro (Guiton et al. 2009). A recombinant form of DNase I, pulmozyme, is used in certain cases to treat patients with CF (Shak et al. 1990; Fey 2010). Dispersin B is a glycoside hydrolase produced by Actinobacillus actinomycetemcomitans that cleaves β 1–6 N-acetylglucosamine polymers (PNAG) in the bacterial peptidoglycan layer (Fey 2010). Dispersin-B treatment has been shown to be effective against S. aureus and S. epidermidis biofilms and other PNAG-containing bacteria (Izano et al. 2008; Kaplan 2010).

Engineering Dispersin B into a phage that replicates in stationary-phase cells led to complete disruption of preformed E. coli biofilms in vitro, an effect that was more dramatic than administration of either nonengineered phage or Dispersin B alone, likely because Dispersin B-mediated degradation of EPS allowed the phage to access the deeper layers of the biofilm structure (Lu and Collins 2007).

Chelating Agents

Metal cations, such as calcium, magnesium, and iron have been implicated in maintaining matrix integrity (Patrauchan et al. 2005; Raad et al. 2008). Consistent with this observation, chelating agents have been shown to destabilize biofilm architecture besides interfering with bacterial membrane stability (Donlan 2011). For example, sodium citrate inhibited biofilm formation by several Staphylococci species in vitro (Shanks et al. 2006). In addition, tetrasodium-EDTA eradicated biofilms in an in vitro biofilm model and on explanted hemodialysis catheters (Kite et al. 2004; Percival et al. 2005), whereas disodium-EDTA, in combination with tigecyclin or gentamicin, reduced biofilm formation by Staphylococcus species and P. aeruginosa on Hickman catheter segments in vitro (Bookstaver et al. 2009). Raad et al. showed efficacy of a combination of minocycline and disodium-EDTA against biofilms in vitro or on explanted catheter tips, as well as in the treatment of catheter-related bloodstream infections in three different patient studies (Raad et al. 1997, 2003). A minocycline-EDTA solution was also successfully used to prevent indwelling implantable-port infections in children with cancer; no port infections or other adverse effects were observed in patients whose ports were flushed with the monocycline-EDTA solution, whereas 21% of the patients in the untreated control group developed infection (Chatzinikolaou et al. 2003). Moreover, reduction of catheter-related bloodstream infections was observed in hemodialysis patients after treating catheters with minocycline-EDTA (Bleyer et al. 2005; Feely et al. 2007).

Manipulating Dispersal Signals to Disassemble Biofilms

Given that planktonic cells are more susceptible to treatments, a novel treatment strategy in which a signal for biofilm dispersion is combined with administration of an antimicrobial agent for killing the dispersed organisms could be successful. As we discussed earlier, one of the signals most associated with bacterial dispersal is c-di-GMP. In a recent study, Ma et al. exploited the potential of engineering a protein that causes biofilms to disperse (Ma et al. 2011b). Using a knockout library of previously uncharacterized genes known to be influenced by impaired autoinducer-2 secretion, the investigators identified BdcA as a protein that enhances biofilm dispersal, by sequestering c-di-GMP and reducing its local concentration (Ma et al. 2011b). Subsequent analyses showed that mutating the BdcA E50 residue to V or Q, resulted in higher induction of motility by increasing the c-di-GMP binding affinity of BdcA (Ma et al. 2011b). Given that BdcA is conserved in several pathogens, this analysis provides new tools, which in combination with novel delivery strategies, such as phage therapy, could facilitate active biofilm dispersal as a therapeutic approach.

d-Amino Acids and Norspermidine

The idea of manipulating natural dispersion factors to combat biofilms has also been exploited for Gram-positive organisms. Losick and colleagues showed that exogenous addition of the d-amino acids produced by dispersing B. subtilis disrupted preformed biofilms and were also effective in preventing biofilm formation by S. aureus and P. aeruginosa (Kolodkin-Gal et al. 2010). It is likely that d-amino acids promote biofilm disassembly by disrupting adhesive fiber interactions (Cava et al. 2010). Indeed, studies by Hochbaum et al. showed that d-amino acids inhibit S. aureus biofilm formation by preventing protein localization to the cell surface (Hochbaum et al. 2011). Given that d-amino acids are produced by many bacterial species, they may provide a general strategy for biofilm disassembly (Kolodkin-Gal et al. 2010) and thus might be useful in medical and industrial antibiofilm applications. A second biofilm-disassembly molecule was recently discovered in B. subtilus; norspermidine works in a manner complimentary to d-amino acids by targeting the exopolysaccharide (Kolodkin-Gal et al. 2012). As is the case for d-amino acids, the biofilm-inhibiting properties of norspermidine were not limited to B. subtilus, but biofilm inhibition was also observed in the case of S. aureus and E. coli pellicle biofilm. Thus, norspermidine and other polyamines synthesized to bind to specific exopolysaccharides could be exploited in conjunction with d-amino acids as a novel antibiofilm approach (Kolodkin-Gal et al. 2012).

CONCLUDING REMARKS

Biofilm formation enables bacterial pathogens to colonize a wide variety of host niches and persist in harsh environments, making their eradication particularly difficult. Biofilm characteristics determine whether, to what extent, and which antimicrobial treatments may be effective. The age and composition of the biofilm are the major factors influencing the susceptibility of the resident microorganisms. As the biofilm matures, increased EPS accumulation, combined with the nutrient and oxygen gradients that affect cell metabolism and growth rates, result in reduced entry and activity of antimicrobial agents making biofilm-forming pathogens progressively more resistant to antibiotic regimens. Thus, novel strategies, designed to block a specific biofilm step without killing the bacteria, such as the use of antiadhesion agents, or using natural, bacterially produced signals to promote bacterial dispersal, are exciting avenues for exploration and ultimately the development of fast-acting, potent, and bioavailable treatment strategies.