Robust evidence is somewhat lacking for biofilm susceptibility to bacteriophages in nature, contrasting often substantial laboratory biofilm vulnerability to phages. To help bridge this divide, I review a two-part scenario for ‘heterogeneous’ phage interaction even with phage-permissive single-species biofilms. First, through various mechanisms, those bacteria which are both more newly formed and located at biofilm surfaces may be particularly vulnerable to phage adsorption, rather than biofilm matrix being homogeneously resistant to phage penetration. Second, though phage infection of older, less metabolically active bacteria may still be virion productive, nevertheless the majority of phage population growth in association with biofilm bacteria could involve infection particularly of those bacteria which are more metabolically active and thereby better able to support larger phage bursts, versus clonally related biofilm bacteria equivalently supporting phage production. To the extent that biofilms are physiologically or structurally heterogeneous, with phages exploiting particularly relatively newly divided biofilm-surface bacteria, then even effective phage predation of natural biofilms could result in less than complete overall biofilm clearance. Phage tendencies toward only partial exploitation of even single-species biofilms could be consistent with observations that chronic bacterial infections in the clinic can require more aggressive or extensive phage therapy to eradicate.
INTRODUCTION
Bacteriophages are the world's most abundant semi-autonomous genetic entities and/or the most plentiful of viruses e.g. Diaz-Munoz and Koskella (2014). Bacteria potentially are the world's most abundant cellular organisms (Whitman, Coleman and Wiebe 1998), and biofilms appear to be the predominant (Hall, McGillicuddy and Kaplan 2014) or even ‘most successful’ (Flemming and Wingender 2010) bacteria state. Numerous studies have demonstrated that biofilms can be susceptible to phage-mediated clearance in the laboratory; for references, see e.g. Abedon (2011, 2015b), Sillankorva and Azeredo (2014) or Chan and Abedon (2015). It is possible, therefore, that natural biofilms form particularly when phage predation pressure is low (Abedon 2012). Alternatively, as single-dose phage application to even single-strain laboratory biofilms can result in incomplete elimination of phage-sensitive bacteria—as reviewed in Abedon (2011) and Sillankorva and Azeredo (2014)—it is likely that biofilms can be at least partly resistant to phage attack. Furthermore, it is conceivable that mature, naturally formed biofilms are particularly phage resistant.
Clonal bacterial groupings such as microcolonies may serve as targets within biofilms for phages rather than individual, isolated bacteria. When phages acquire biofilm-associated bacteria, the number of phages produced, overall, therefore might be greater than would be seen with phage infection of isolated bacteria (Azeredo and Sutherland 2008). Poorly appreciated, however, is how phage virions penetrate into along with subsequently exploit such clonally related, contiguously located bacteria (Fig. 1), particularly if biofilm properties are heterogeneous. This commentary addresses aspects of these latter issues, providing a scenario for phage exploitation of bacterial biofilms in nature that could be consistent with observed susceptibilities of chronic bacterial infections to phage therapy.
Factors impacting phage (ϕ) ability to reduce numbers of biofilm bacteria within environments. These can be summarized—clockwise, starting upper-left—as consisting of (1) initial virion penetration to biofilm or microcolony bacteria, with weaker penetration resulting in reduced phage impact (EPS stands for Extracellular Polymeric Substances i.e. biofilm matrix); (2) sequential phage adsorption and infection particularly of clonally related bacteria found immediately local to the initial site of biofilm or microcolony penetration, with virions potentially constrained in their movement by EPS and infections constrained in their productivity by bacteria physiological state; (3) virion diffusion out of biofilms, which ultimately can be hazardous to diffusing virions as well as lengthen phage generation times; and (4) numbers (#s) of phages of specific host ranges undergoing longer distance virion movement, with smaller numbers of surviving virions expected to result in lower eventual phage impact on biofilm bacteria. This commentary focusses on factors (1) and (2); see Abedon (2011; 2015b) for consideration of (3), and note that the specifics of (4) are not well appreciated.
IMPORTANCE OF NEW, SURFACE BACTERIA
Impediments to phage penetration into biofilms, perhaps creating refuges from phage attack (Schrag and Mittler 1996; Heilmann, Sneppen and Krishna 2012; Schmerer et al.2014), likely include biofilm matrix i.e. extracellular polymeric substances or EPS (Sutherland et al.2004; Azeredo and Sutherland 2008). So long as EPS is not produced instantaneously by replicating biofilm bacteria then EPS surrounding more recently divided bacteria might be more readily penetrated by phage virions e.g. such as due to newer, peripheral EPS being purer, less complex (Flemming and Wingender 2010), less dense (Briandet et al.2008; Hu, Miyanaga and Tanji 2010) or thinner (Wilking et al.2011). Certain phages produce or induce EPS depolymerases which could help mitigate this issue (Shen et al.2012; Harper et al.2014), though not necessarily under all conditions (Schmerer et al.2014). Phage tails, to the extent that they display smaller diameters than phage virions as a whole, might contribute to non-enzymatic virion translocation into EPS, perhaps with longer, narrower tails permitting deeper or faster local penetration to biofilm-surface located bacteria. Bacteria found in biofilm peripheries also can display more active metabolisms (Brown, Allison and Gilbert 1988; Azeredo and Sutherland 2008; Stewart and Franklin 2008), thereby potentially supporting larger phage bursts (Hadas et al.1997; Miller and Day 2008).
Virions released from biofilm surfaces could infect other biofilm-surface bacteria (Fig. 2), could dissemination toward more distantly located bacteria (Abedon 2011), could be delayed in their movement by biofilm matrix (Briandet et al.2008) or could diffuse towards more inwardly located biofilm bacteria. Diffusion inward is not necessarily beneficial to phages, however (Sutherland et al.2004), and could even constitute a phage detriment, see ‘dead cells’, p. 263 (Azeredo and Sutherland 2008), unless reasonably large burst sizes (≫0) can on average subsequently occur e.g. due to post-infection improvements in bacteria physiology (Harper et al.2014) or given higher than anticipated biofilm-interior metabolic rates (Liu et al.2015). Lysogen formation, too, might be beneficial given such highly localized phage ‘migration’ (Berngruber, Lion and Gandon 2015).
Clonal bacterial microcolony or biofilm as a heterogeneous phage-exploitable resource. Represented are bacteria as circles, EPS collectively as a blob-like shape, surface substrate as a disk (bottom), a seven-pointed star as the initially phage-infected bacterium, and arrows as post-release movement of phage virions, with the lone gray arrow indicating less effective phage penetration to or exploitation of microcolony-interior bacteria. The darkness of EPS indicates EPS resistance to virion penetration, with darker EPS (bottom, lower right) corresponding to greater resistance; note that presence of adsorbable biofilm-surface bacteria should also interfere at least temporarily with phage penetration into biofilms (Briandet et al. 2008; Hu, Miyanaga and Tanji 2012). The lightness of cells indicates their metabolic activity with lighter coloration, e.g. white (upper, outer), denoting greater metabolic activity and, potentially, ability to support larger and/or sooner phage bursts. As indicated, microcolony or biofilm surface bacteria thus may be more readily reached by phage virions and more effectively support phage population growth once reached (Hu, Miyanaga and Tanji 2012). These properties could be present primarily with bacteria that have more recently replicated, particularly if surrounded with EPS that has not yet fully matured. Phages therefore might be viewed simply as typical ‘predators’ and biofilm bacteria as typical ‘prey’: biofilm bacteria as prey, even if genetically similar, may vary in their vulnerability as well as amount of resources that they can supply to phages as predators, particularly with bacteria that are both biofilm peripheral and younger not only more vulnerable to phages but also more valuable.
Bacteria located on biofilm or microcolony surfaces, that are metabolically active (Heilmann, Sneppen and Krishna 2012), and that are newly formed thus could logically serve as primary targets for phage exploitation of biofilm bacteria. Through lysis and virion release, phages could infect additional biofilm bacteria, but with infection of recently replicated bacteria predominantly contributing to overall phage population growth. Infections of biofilm-interior bacteria, to the extent such penetration is possible, may be less robust, resulting in production of comparatively few phages and/or less overall bacterial clearance. Phage exploitation of clonal groupings of biofilm bacteria – especially of more desirable, less mature, and thereby more vulnerable bacterial prey – in other words could occur even if not all bacteria are easily reached or can sustain lytic or productive phage infections.
PHAGES AS ANTIBIOFILM AGENTS
Phage therapy is the use of bacterial viruses as antibacterial agents; phage use as antibiofilm agents has been considered in a number of recent reviews: Abedon (2011, 2015b), Brüssow (2013), Fan et al. (2013), Harper et al. (2014), Parasion et al. (2014), Sillankorva and Azeredo (2014) and Chan and Abedon (2015). In addition to phage use explicitly as antibiofilm agents, also relevant is the use of phages to treat especially chronic bacterial infections. This is because chronic bacterial infections are thought to typically involve bacterial biofilms (Bjarnsholt 2013; Scali and Kunimoto 2013; Cooper, Bjarnsholt and Alhede 2014; Macia, Rojo-Molinero and Oliver 2014). Though animal models exist—for example of phage treatment of infected wounds (Loc-Carrillo, Wu and Beck 2012), which even as relatively new infections can involve bacterial biofilms (Percival, McCarty and Lipsky 2015)—the majority of experience in using phages to treat chronic bacterial infections has been acquired within clinical settings. English-language-published clinical work particularly has been Polish (Slopek et al.1987; Weber-Dabrowska, Mulczyk and Górski 2000), though such efforts appear to date back to early in the phage era (d'Hérelle and Smith 1930). In particular, phage therapy of chronic bacterial infections can require weeks, typically employing repeated phage application.
This observation that elimination of chronic bacterial infections, potentially consisting of mature bacterial biofilms, can require long treatment periods as well as repeated dosing may be consistent with the scenario for phage-biofilm infection dynamics suggested in the previous section. Specifically, if phages are reaching and successfully lytically infecting only a subset of biofilm bacteria, then any given phage dose may directly impact only a fraction of bacteria targeted. Lytic phage infections, or other means of biofilm disruption such as debridement (Scali and Kunimoto 2013), could result in improvement in bacteria physiology (Harper et al.2014) and/or in phage penetration to bacteria e.g. such as by making biofilm-interior bacteria more surface located. Particularly, it likely is inefficiencies in phage penetration to, and subsequent bactericidal infection of target bacteria that results in requirements for extensive phage treatment to achieve biofilm eradication in the clinic.
Proposed inefficiencies in phage exploitation of biofilm bacteria in nature, in other words, could be equivalent to inefficiencies in phage-mediated eradication of biofilm bacteria in the clinic. Such inefficiencies, however, do not imply either that phages generally will be unable to productively exploit biofilm bacteria in nature or that treatment of chronic bacterial infections using phages is ineffective. Instead, the effectiveness per biofilm-encountering phage or per therapeutic phage dose may simply be limited, that is, to impacting particularly the most vulnerable of biofilm bacteria rather than lytically infecting every bacterium found within a given biofilm, even for clonal, single-species biofilms.
CONCLUSIONS
Elsewhere, I have argued that bacteriophages may serve as inherently more effective antibiofilm agents than antibiotics due to limitations in the ecological utility to producing organisms of antibiotics as biofilm disruptors (Abedon 2015a). Phage-mediated clearance of older biofilms nonetheless can be challenging (Azeredo and Sutherland 2008; Sillankorva and Azeredo 2014), though not necessarily impossible, the latter particularly given extended treatment periods and/or repeated phage dosing. Perhaps phages, both in nature and during therapy of chronic bacterial infections, tend to target only a subset of otherwise clonal biofilm bacteria—a subset that happens, concurrently, to be less mature, more readily reached, and more effectively phage infected.
The author has advised companies with phage therapy interests and maintains the websites phage.org and phage-therapy.org, but received no support in the writing of this manuscript.
Conflict of interest. None declared.
REFERENCES
