Activated sludge foaming: can phage therapy provide a control strategy?

Foaming in activated sludge systems is a global problem leading to environmental, cosmetic and operational problems. Proliferation of filamentous hydrophobic bacteria (including the Mycolata) are responsible for the stabilisation of foams. Currently no reliable methods exist to control these. Reducing the levels of the filamentous bacteria with bacteriophages below the threshold supporting foaming is an attractive approach to control their impact. We have isolated 88 bacteriophages that target members of the foaming Mycolata. These double stranded DNA phages have been characterised and are currently being assessed for their performance as antifoam agents.

The activated sludge process is a robust and proven system for treating domestic and industrial wastewater and is used globally¹. It relies on a specialised community of microbes organised into structures called flocs, which metabolise organic nutrients and remove inorganic nitrogen and phosphorus compounds so that the treated effluent can be discharged safely into a receiving body of water without leading to eutrophication from growth of toxic Cyanobacteria¹.

These systems are no longer considered as wastewater water disposal systems, but as valuable sources of purified water for reuse and useful chemicals. Despite their popularity most suffer from the problem of foaming where a brown foam layer develops on the reactor surface and leaves in the treated effluent².

Foaming, which increases plant operating costs, reduces effluent quality and acts as a source of opportunistic human pathogens, is a flotation event, requiring three components; air bubbles, surfactants and hydrophobic particles (bacterial cells), which act to stabilise it (Figure 1). With only air bubbles and surfactants, an unstable foam forms, and is often seen in start-up, where abundances of hydrophobic bacteria are below the threshold supporting foam formation³. With insufficient levels of surfactants, air bubbles collapse and a greasy surface layer, a scum, forms, consisting of hydrophobic bacteria. There are no reliable control measures to deal with an already established foam, but any proposed strategy should target the hydrophobic bacterial cells, since control of the other two is impractical.

Figure 1. Foam covering the surface of an aeration basin and walkways.

It is now clear that a diverse range of bacteria are responsible for foaming episodes^4–8. Theoretically, any sufficiently hydrophobic cell can stabilise this foam, but surveys suggest that the unbranched actinobacterial filamentous organism ‘Microthrix parvicella’ and the right angled branching mycolic acid producing filaments placed in the Mycolata (include members of the genus Gordonia, Nocardia, Rhodococcus, Tsukamurella, Skermania and related members)^5,7 are the main culprits (Figure 2). Being strongly hydrophobic, these organisms escape the plant bulk liquid to the air liquid interface, often carrying biomass or sludge with them, and there attach to the liquid air films of the bubbles, preventing liquid drainage from them and hence stabilise them.

Figure 2. Gram stain of (a) ‘Microthrix parvicella’, (b) Gordonia amarae like organism (GALO) and (c) Skermania piniformis. Gordonia amarae like organism (GALO) and Skermania piniformis are right angled branching mycolic acid producing filaments and belong to the Mycolata.

The conventional way to deal with existing foams is commonly non-selective using bactericidal chemicals, where organisms other than those causing foaming are likely to be harmed. Others include changing aeration rates, or reducing sludge ages hoping that foaming organisms, assumed to be slow growing, are washed out. Unfortunately other desirable bacteria are also lost. All reflect our inadequate understanding of the microbial ecology of foams.

What is needed is a specific control strategy, which is environmentally safe and importantly only removes the nuisance organisms^9,10. Bacteriophages (or phages), viruses that specifically target only their bacterial hosts, and are naturally occurring and self-dosing, seem especially attractive. They are used clinically to treat infectious bacterial diseases, where the causative organism is antibiotic resistant¹¹. As they infect their hosts, they replicate and upon lysis, release often hundreds of new phages that then infect other host cells.

The general experience has been that wherever bacteria are present, phages able to lyse them will also be present. Consequently, phages lytic for members of the foaming Mycolata should be plentiful in activated sludge. Thomas and colleagues⁹, demonstrated that phages, some polyvalent, are isolated readily from activated sludge plants, capable of killing their Mycolata hosts under laboratory conditions. What we know of phage/host population dynamics suggest that their presence would not lead to the total loss of their bacterial hosts. Such outcomes would be disastrous, since Mycolata play important roles in metabolising recalcitrant xenobiotics there. Strategically the aim is to reduce Mycolata numbers below individual threshold levels needed for stable foam formation. This requires identifying which are the causative organisms. FISH probing provides the tools to screen foam samples, and while their true level of biodiversity, is not known, FISH data suggest a limited number of foaming bacteria are common in plants.

The advent of Next Generation DNA Sequencing (NGS) has revolutionised our understanding of phage genomics and allowed us to screen those attractive for phage therapy, avoiding those possessing virulence or toxin genes. We have isolated 88 double stranded DNA phages that seem suitable for further study. These include phages against foaming Gordonia^12,13, Rhodococcus^14–16, Nocardia¹⁷, Skermania¹⁸ and Tsukamurella^19,20. While most are monovalent, polyvalent phages are clearly more attractive, since most foams contain more than one Mycolata member. Not surprisingly, sequencing reveals that all are highly novel at the DNA level, but share the same genomic arrangements. They have all been screened against foaming Mycolata hosts using a simple foaming apparatus²¹. Almost always their foaming abilities were reduced to the point where no foam was detected as previously described^13,22.

The next step is to scale up the system. Before this is warranted, it is important to determine their host specificities and burst sizes in situ, their persistence times in full scale plants, the host cell threshold values for foam production and how much inoculum is required. Equally, the location for introducing the phages into the system is likely to be important. These parameters are plant specific, and so will need to be determined on an individual basis. Whether these phages are involved in gene transfer between host cells (transduction), and whether they acquire, as a consequence, antibiotic resistant genes and hence pose a possible threat upon release into the environment, will need investigation. In addition, the possibility of the bacterial strains developing phage resistance will need to be investigated and one possible solution would be to add multiple phages for the same host as multiple mutations is less likely.