BAM’ing with Bacteriophage

ResearchBlogging.orgYou know, I sometimes think that animals are just walking colonies of multiple kinds of cells that evolved to live in a symbiotic relationship with one another. Seriously, it is highly unlikely that we would be able to survive without all the viruses and bacteria that live in and on us. We need these little microorganisms for all sorts of things from immune system development to nutrient acquisition. Moreover, increasing amount of evidence shows that our microbiome and virome are an essential part of who we are, and perhaps humans would be a very different species today if some of these animal-microbe interactions would have not occurred during the course of evolution.
web-M0900084-Bacteriophages_attacking_bacterium-SPL

An important part of our virome are bacteriophages (or phages for short), i.e. viruses that infect bacteria (in contrast, to the general word ‘virus’, which is usually reserved to viruses infecting animal or plant cells). I always thought that phages have been somewhat underrated in terms of their importance to human health. Although, I suppose it is perhaps not all that surprising because at first instance one would not necessary think that there is a direct interaction between a phage and a eukaryotic cell. But then again, how likely is it that something, which is so abundant in our environment, and interacts with microorganisms that we know are essential to us, would be just a neutral passer-by? (I’d like to give you an actual number for how many phages live in/on a given person but I cannot find any good source for that. Nevertheless, if say there are 10 times more bacterial cells in a person’s body than human cells and some estimates suggest that there are 10 phages per single bacterium then we are talking about hundred phages per one human cell). And actually increasing evidence suggests that our own immune system does produce a response to phages and in turn phages themselves can influence an outcome of infection. Antibodies have been found to be produced against phage capsid proteins, macrophages have been observed to “eat” phage particles and the presence of phages can reduce host’s inflammatory response. All in all it’s clear that bacteriophages are not just neutral passers-by. So, when I recently came across a study, which suggested a mutualistic relationship between a eukaryotic host and a bacteriophage and presented some interesting evolutionary dynamics between the two, I though finally phages are getting some credit!

The story goes this way. In any given animal there are many tissues, which are covered with a layer of mucus (think the surface of lungs, eyes, stomach etc.). The mucus is large part water but also contains, various cell debri, proteins, DNA and salts. The largest macromolecules in mucus are glycoproteins (proteins with sugar chains attached to them), also called mucins, which provide rigidity to the mucus layer. Among many functions, mucus provides a nutrient-rich place to live in for the mutualistic bacteria and is at the same time a first point of contact for an incoming pathogen with a host. In addition, mucus also contains a large population of phages. In fact, sampling of mucosal surfaces from various animals has showed that mucus layers have significantly larger numbers of phage particles as compared to surrounding environments (figure 1). In addition, laboratory tests have confirmed that cultured cells, which produce mucus, are better at retaining phages on their surfaces compared cells that are mucus deficient and it is the mucins, rather then DNA or other proteins, in mucus that are important for phage binding (figure 2).

Figure 1. Abundance of phages and bacteria in mucosal surfaces and surrounding environments from different animals.
Figure 1. Abundance of phages and bacteria in mucosal surfaces and surrounding environments from different animals.
Fig. 2. B) Adherence of Phages to mucus producing  cells (TC cells), cells where mucus have been chemically removed (mucolytic agent), and cells that have genetic knock down of mucus (knockdown). C) Phages adhere to surfaces with covered with mucins but not proteins or DNA
Fig. 2. B) Adherence of Phages to mucus producing cells (TC cells), cells where mucus have been chemically removed (mucolytic agent), and cells that have genetic knock down of mucus (knockdown mutant). C) Phages adhere to surfaces which are covered with mucins but not proteins or DNA

A natural question perhaps arises: does the presence of phages in mucus can reduce microbe colonisation? To answer the question the following system was used: mucus producing and non-producing cells were plated in a dish and incubated with a T4 phage suspension. After 30min suspension was washed off to remove any residual phage and a new suspension, this time containing E. coli bacteria, was added. After 4h cells were washed again and the numbers of remaining E. coli on cells were counted. As expected, cells, which were producing mucus, were able to retain T4 phage after washing and in turn had reduced bacterial colonisation (figure 3 A).

To show that reduced colonisation was a consequence of T4 phage infecting and lysing bacteria a so-called amber system was used. The amber system involves a T4 phage, which has a mutation that prevents it from lysing bacteria. However, if an infected bacterium is a strain, called SupD, which expresses a protein that can reverse the effect of mutation in T4 phage, then a mutant phage can lyse the bacterium. So, amber phage cannot lyse “normal” E. coli but can lyse E. coli which expresses SupD. When this system was used in exactly the same experiment as described above, the numbers of SupD E. coli on the mucus producing cell surface were significantly reduced in the presence of amber T4 as compared to in the absence of T4. Numbers of T4 phage itself were increased in SupD system but as expected not in wild-type E.coli system as T4 cannot replicate in it (figure 3 B).

Figure 3. A) T4 phage reduces bacterial colonisation of cells which secrete mucus. B) Amber T4 phage causes lysis of SupD expressing  E. coli cells on mucus producing cells (control are cells that have not been pretreated with T4, CFU are bacterial and PFU are phage counts). C) T4 pretreatment prevents bacteria-induced cell death in mucus producing cells.
Figure 3. A) T4 phage reduces bacterial colonisation of cells which secrete mucus. B) Amber T4 phage causes lysis of SupD expressing E. coli cells on mucus producing cells (control are cells that have not been pretreated with T4, CFU are bacterial and PFU are phage counts). C) T4 pre-treatment prevents bacteria-induced cell death in mucus producing cells.

So, if the phage can protect mucus from being colonised by bacteria can it also protect the eukaryotic cells from bacteria-induced cell death? Indeed, it turns out that eukaryotic cells that can make mucus and in turn retain the phage show significantly reduced cell death in the presence of E. coli Figure 3 C).

Figure 4. Head of T4 phage Hoc protein is in yellow (source).
Figure 4. Head of T4 phage Hoc protein is in yellow (source).
The capsid of T4 phage contains multiple copies of Hoc protein (figure 4). Hoc has an Ig-like domain, which in various other proteins is known to be involved in protein-protein and protein-ligand interactions. Therefore, the authors hypothesised that Hoc may mediate phage-mucin binding. Surprisingly, deletion of Hoc itself is not required for phage replication, which allowed to compare phages with (hoc+) and without (hoc-) Hoc in their ability to bind mucus. Indeed, hoc+ phages bound mucins significantly more then hoc- strains and the presence of hoc reduced the speed at which a phage was able to move through mucin-enriched buffer because the Hoc was binding to the glycoproteins (figure 5).

Figure 5. A) hoc+ phages can bind to mucins.  C) hoc+ phages move through mucin containing buffer slower then hoc- mutants.
Figure 5. A) hoc+ phages can bind to mucins. C) hoc+ phages move through mucin containing buffer slower then hoc- mutants.

Finally, the authors propose a model, which they call ‘Bacteriophage Adherence to Mucus (BAM) Model’, which postulates the presence of non-host derived immunity in animals. According to the model mucins in the mucus bind to phage capsid proteins and prevent quick clearance of phages from the surface. The longer the phage resides in mucus the more likely it is to encounter an invading bacterium, thus not only increasing its own reproductive success but also providing protection for the host (see summary figure 6).

Figure 6. Bacteriophage Adherence to Mucus Model
Figure 6. Bacteriophage Adherence to Mucus Model

Obviously, the study described here is only based on in vitro model and in vivo systems are so much more complicated. Not only there are phages with lytic and lysogenic life cycles and multiple species of bacteria in the mucus but the mucus itself is continually produced and removed from the surfaces; i.e. as ever, the actual in vivo system is much more dynamic then our simplified models can depict. Nevertheless, the study does provide a renewed way of thinking about phages as an integral part of our adaptive immune system. Moreover, it also suggest that phage Ig-like capsid proteins and mucins may be coevolving to retain the binding to each other and maintain the mutualistic relationship. Interestingly, phage Ig-like protein sequences do appear to be under significant selection, which is partially caused by the fact that these proteins are also required for binding to glycans on the bacterial surface during invasion. It appears now then that the selection may be mediated by balancing act between ability to bind the mucus of eukaryotic hosts and invade the target bacterial species.

Reference for paper and figures:
Barr, J., Auro, R., Furlan, M., Whiteson, K., Erb, M., Pogliano, J., Stotland, A., Wolkowicz, R., Cutting, A., Doran, K., Salamon, P., Youle, M., & Rohwer, F. (2013). Bacteriophage adhering to mucus provide a non-host-derived immunity Proceedings of the National Academy of Sciences, 110 (26), 10771-10776 DOI: 10.1073/pnas.1305923110

Leave a Reply

Your email address will not be published. Required fields are marked *