Review for "Theory for the spatiotemporal interaction between lytic phages and biofilm-dwelling bacteria"

Completed on 11 Nov 2016 by Laura Weber. Sourced from

Login to endorse this review.

Comments to author

Author response is in blue.

Hello all,

As a microbial ecologist interested in aquatic biofilms associated with eukaryotic organisms, I was fascinated to read this article. You have provided me with a number of considerations for understanding the complex interactions and abiotic variables that structure these micro-environments.

I have several questions regarding some of the methods used in this study and I would also like to offer some thoughts that you may find interesting or relevant to explore using the existing model that you have built.

For computing bacterial biomass dynamics, you mention a ‘shoving’ condition. Could you expound upon the need for this condition within the model? Is this condition used to control bacterial density within each grid to account for limited physical space for cells within the biofilm?

I understand that your model uses a fixed incubation time of 28.8 minutes. If this incubation time were to decrease throughout the incubation (say if the phages were to increase the growth rate of their hosts by hijacking the host’s cellular machinery), how might that change the stability of the system?

Would we expect to see differences in host-phage dynamics if there was a concentration gradient in the nutrient (i.e. diffusing from the substratum), causing different bacterial growth rates as a function of depth or position of a cell within the biofilm?

Based on your results that coexistence of phages and bacteria becomes increasingly more likely as phage diffusivity increases, should we expect host-phage coevolution to occur at a faster rate than in the water column? Can biofilms be sources for ‘super’ phages?

Really great work- I am excited to read more!



Hi Laura,

Thanks for taking the time to read and comment!

With respect to the shoving condition, it is exactly what you expected, a condition used to control the density due to the physical limitation of each point in the grid. A conceptual example would just be the overflow of one grid element to another, although we did not implement it mathematically that way.

I don't have any data for the incubation time decreasing throughout the infection, so speculatively I would imagine that it would behave similarly to increasing the probability of infection and the burst size together. So if we were to run the same set of simulations again, with the varying incubation time, we might see a shift in the different regions of figure 3, but they would probably resemble the same shapes in the figure we see now. However, the system can be quite unpredictable and we can only speculate until a feature like that was implemented.

The system does implement a gradient in the nutrient following the diffusion with a monod-type kinetic reaction. Like many biofilm simulations, there's a constant bulk concentration at the top of the system. The different rates of growth as a function of height were explored a bit in the supplementary figure S1. In that figure we initially infected the biofilm with phage at different times, and consequently heights. If you take a look at the supplementary video S4, you see the biofilm growing up, and because the biomass near the substratum has access to fewer nutrients (the biomass at the top eats them), the phage are able to pinch off the growing mass. Note of course we have "strong" detachment, so if anything isn't directly connected to the substratum (phage or biomass) it is removed immediately. The phage are able to diffuse through the "empty" space, but they only remain if they stop on the biomass.

Here Zp is the impedance of the phage (not the diffusivity), so it's essentially the factor by which the diffusivity is reduced while the phage are in the biofilm. Another way to put it is as the phage "stick" more to the biomass, the rate of coexistence increases. I would expect coevolution to occur in this coexistence regime faster, but not resulting in any super phage. As the coevolution occurs, I would expect specifically the phage to be better at infecting biofilms, which may or may not result in phage that are superior at infecting in the water column. So we could probably say that a biofilm CAN be a source for a super phage, but its likelihood of evolving would be pretty unknown. Answering a question like that might require a more in depth simulation, looking at tradeoffs during the evolution.