Biofilm thickness matters

Does biofilm thickness matters?

Biofilms are agglomerations of microbial cells attached to each other or to surfaces (like the slime on river rocks or the slimy, sticky goo in bathrooms). Their thickness can vary, and that thickness can be influenced by many factors. For example, older biofilms would be thicker than newer ones. Nutrient availability can also result in bigger biofilms. Other factors like microbial predation, water flow and turbulence may also limit biofilm thickness.

Intuitively, we would think that the tickness of a biofilm matters. A thicker biofilm would have more cells. It could also have a more heterogenous structure. In thick biofilm the cells in the interior may have limited access to nutrients, and even lack access to oxygen, which would influence their metabolism. But the importance of biofilm thickness itself had never been studied, as it is often linked to other parameters, like biofilm age.

We answered this question in two scientific publications: One at Water Research, and another at Scientific Reports. In the two studies we controlled the thickness of a biofilm, to 50 and 400 micrometres, and we kept all the other conditions the same, by keeping both thin and thick biofilms in the same nitrifying bioreactor.

Picture of thick (above) and thin biofilms (below). Scale bar: 100 micrometres.

We discovered that thick and thin biofilms not only look different, but they also behave different, and differ in their microbial community composition. For example, we observed anammox bacteria in thick biofilm, but they were absent in thin biofilms. This was probably due to thin biofilms being fully oxygenated, which would inhibit the growth of anammox bacteria, as these bacteria are sensitive to oxygen.

Biofilms are ubiquitous in nature and engineered environments, and thus these results are not only important for biofilm researchers, but also for people working in wastewater treament, drinking water treatment, medical microbiology environmental microbiology and biofouling.

Read more about biofilm thickness:

Suarez, C., Piculell, M., Modin, O. et al. Thickness determines microbial community structure and function in nitrifying biofilms via deterministic assembly. Sci Rep 9, 5110 (2019). https://doi.org/10.1038/s41598-019-41542-1

Piculell, M. et al. The inhibitory effects of reject water on nitrifying populations grown at different biofilm thickness. Water Res. 104, 292–302 (2016). https://doi.org/10.1016%2Fj.watres.2016.08.027

Fluorescence in situ hybridization (FISH)

Microscopy picture of biofilm

Microorganisms are very small! In the microscopy picture above each tiny circle is a microbial cell, with a size of around one micrometre (0.001 milometers). It is a picture of a microbial biofilm with many species living together, where each colour represents a different population. The big cyan agglomerate in the lower left corner is one of my favourite bacteria: Anammox bacteria.

So, how was this picture made? To prepare the samples, I used a method known as fluorescence in situ hybridization (FISH), which can be used to detect specific microbial populations. All microorganisms have large RNA molecules known as ribosomes, and the sequences of these ribosomes differ between microbial species. In FISH these ribosomes are targeted using specific DNA probes, which have attached a dye (a fluorophore). The FISH probe will only bind to the population we want to target (for example anammox bacteria). Then using a confocal laser microscope or an epifluorescence microscope, we illuminate the sample, and the cells with bound FISH probes will become fluorescent.

Another interesting thing that can be done with FISH is to study the physical location of microbial populations, and to identify the populations they co-exist with. One way to achieve this is with cryosections, where the biofilm sample is frozen using liquid nitrogen, and then using a cryotome, ultra-thin slices are made.

Microscopy picture of biofilm
FISH cryosectioning. From Suarez et al 2015 (https://doi.org/10.1093/femsec/fiv124)

Years ago, I wondered if anammox bacteria living in biofilms could be eaten by other microscopic predators. This was thought to be unlikely as anammox bacteria live in the deeper layers of biofilms, which would likely protect them against predation. Using FISH we targeted both anammox bacteria and protozoa living in the biofilm. We could show that protozoa could go deep in the biofilm, even in the layers where anammox bacteria live. We also observed anammox bacteria inside the food vacuoles of protozoa, showing that these protozoa were eating anammox bacteria.

Microscopy picture of biofilm
FISH picture of amoeba grazing anammox bacteria in a biofilm. Green: Anammox bacteria. Red: Protozoa

You can read more predation of anammox bacteria here:

Carolina Suarez, Frank Persson, Malte Hermansson. Predation of nitritation–anammox biofilms used for nitrogen removal from wastewater. FEMS Microbiology Ecology, Volume 91, Issue 11, November 2015, fiv124, https://doi.org/10.1093/femsec/fiv124.

rRNA/rDNA and looking for active taxa in microbial communities

During my scientific research, many of experiments did not work, some of my hypotheses were wrong, and sometimes I got negative results.

One of these studies where I had high expectations, but I did not get interesting results was years ago during my PhD. There I attempted to identify the metabolically active taxa in a biofilm community, by sequencing ribosomes, i.e. ribosomal rRNA (rRNA). This is because cells need ribosomes to make proteins, and for some Proteobacteria it has been observed that their ribosomal content increases when they are growing. A theoretical advantage of this approach is that is has the potential to distinguish metabolically active, dormant and dead populations, something that ribosomal rRNA gene (rDNA) cannot achieve.

rDNArRNA
Active 😀yesyes
Dormant 😴yeslow
Dead 💀yesno

If 5% of your community in your rRNA dataset species belongs to species X, then we could think that species X is metabolically active. We are detecting their ribosomes. However, the rRNA values you get will depend in (1) the average number of ribosomes your population have, and (2) the abundance of your population. So maybe species X could be very abundant, but it could have a low ribosomal content.

A workaround around this problem is to normalize your rRNA values by the relative abundance of a taxa in your community. These are the so-called rRNA:rDNA ratios. This method was widely used during the 2010s to identify active taxa in all kinds of microbial communities like soil, water, atmosphere, and wastewater bioreactors. However, that approach is far from being perfect. (For a discussion on the topic the read the paper by the Steven Blazewicz et al: https://www.nature.com/articles/ismej2013102)

Sequencing rDNA and rRNA

In my PhD, we used rRNA sequencing to identify active taxa in biofilms from a wastewater bioreactor. In the plot below rDNA and rRNA are compared, with the diagonal line showing an rDNA:rRNA of one. We observed that some taxa like Firmicutes (the pink dots) had high rDNA:rRNA, while others like Acidobacteria (the light blue dots) had low rDNA:rRNA ratios.

So, what do rDNA:rRNA ratios mean? Traditionally it has been assumed that rRNA:rDNA values higher than one are an indication of activity. In other words, taxa overrepresented in the rRNA library are metabolically active. Values lower than one would indicate dormant taxa. But the rDNA:rRNA ratio of one is an arbitrary threshold and it might lead to underestimation of active taxa, as reported by Blaire Steven et al (https://doi.org/10.1128/AEM.00696-17)

Maybe sequencing rRNA is not the best approach

Katerina Papp et al (https://doi.org/10.1128/AEM.02441-17) showed that rRNA:rDNA ratios are a poor proxy to ribosomal production. By using stable isotope probing they quantified ribosomal production and found that sequencing and rRNA:rDNA ratios did not correlate with ribosomal production. Most taxa were producing ribosomes, and thus this supports that rRNA:rDNA ratios underestimate active taxa.

Sequencing of rRNA might not be a valid approach to estimate the activity of a population. Still, it can be useful to sequence ribosomes. For example, Søren M Karst et al (https://doi.org/10.1038/nbt.4045) used a primer-free approach to sequence rRNA, which allow them to identify novel taxa.