Evading while invading: induction of CRISPR-Cas by mucin as a mean to escape phages adhered to mucus

Mucosal surfaces are the point of entry used by many pathogenic bacteria to infect their hosts. Some phages are able to adhere to mucus and wait for these bacteria there. Activation of CRISPR by mucin allow the bacteria to evade the mucosal-associated phages while invading the mucosa.
Evading while invading: induction of CRISPR-Cas by mucin as a mean to escape phages adhered to mucus

The idea that originated this paper appeared even before our first main phage-mucus-bacteria publication [1,2] came out. Until then the mucosal interactions seemed too biased towards the phage-metazoan side: increased host encounters for the phage and protection for the animal, with the bacteria being a passive target. But bacteria are remarkably efficient in adapting, so Red Queen Evolution dynamics were to be expected. But how could a bacterium invade a mucosal surface while staying virulent and escape the mucosal-associated phages at the same time?

Bacterial resistance to phages exists and we still know very little about all the strategies. But with such a big arsenal, how bacteria choose which mechanisms to use? Trade-offs in virulence and energy costs are important, so we can expect that the context matter. We investigated this using our bacterial model: Flavobacterium columnare, an opportunistic mucosal pathogen. In this species there are two main strategies to escape phages: change the morphotype via surface modifications, trading virulence for phage resistance or activate CRISPR immunity (something sometimes hard to see happening in vitro). How the bacteria “decide” which to use was still a mystery.

We started by mixing the bacteria and phage in different nutritional conditions: culture media or autoclaved lake water, both supplemented or not with purified mucin as a mean to simulated mucosal surfaces. The lake water conditions, without or with mucin, were particularly interesting, since these are simulations of the environments in which F. columnare lives in nature: lake water alone while in the environment and lake water with mucins while switching to the opportunistic pathogenic state.

Then we followed with 16 weeks of samplings plus months of follow up experiments and data analysis. It is important to note that we did not add more bacteria or phages after the initial time 0 inoculum. The first surprise I got was to see that, in all replicates, phages and bacteria co-existed until the last sampling. As someone with a background in classical animal virology (poxviruses), it is strange to not see one side quickly extinguishing the other. Impossible not to stop and wonder on how deeply “ecological” phages are, since long-term co-existence with their hosts is likely a favoured trait (as opposed to full termination of the host cells by more “conventional” viruses).

Our main finding was that, in the presence of mucin, CRISPR activity (measured by spacer acquisition) increases. It was clear from our previous work that mucin serve as a signal for F. columnare to upregulate virulence factors, in an attempt to be ready to infect its host mucosa. But we also knew that, by being changed by mucins, the bacteria got more susceptible to mucosa-associated phages. Thus, after sensing mucin, an invading bacterium would be in good shape to cause disease but at extreme risk of being infected itself. If phage protection came from morphology changes, the bacterium would be saved from phage infection but lose the chance to cause disease. So, it is clever that the mucin signal also activates the CRISPR system in the cell: by doing so the bacterium can sense the mucosa, upregulate its virulence, and invade while being protected from phages via CRISPR.

Then we followed by analysis of isolates obtained over the experiment. Bacteria developed resistance, while phage genomes stayed remarkably stable over the four months. After the immunity experiments were done, we ended up with additional questions and a lot of data to analyse. That is when our co-authors came and helped to shape the manuscript you read today. In between this and publication, the pandemic came. So in a way this is a pandemic paper. Not because it focuses on coronavirus (far from it, actually), but because it was deeply affected by it. A reader may wonder, for example, why samplings on the competition experiment did not match the timings of the main experiment. The answer? The competition experiment was running when one of the many lockdowns came, and access to the University (with the cultures inside) became harder.

It is satisfying to see that a pilot experiment started in early 2019 led to such nice results, increasing our knowledge on the tripartite interactions between phages, bacteria and animals. This time the bacteria “won”, showing us that CRISPR is the key to be protected while causing disease. But phages are also good players in the host-pathogen co-evolution game, and are likely holding counter-measure surprises to be found in the future.



1) Almeida et al, mBIO, 2019, https://doi.org/10.1128/mBio.01984-19

2) Almeida et al, 2019, https://go.nature.com/3OBbtTo