A mechanical trigger for biofilm formation

Biofilm related infections kill tens of thousands of people each year. And yet, we still poorly understand why and how biofilms form. In our recent paper (1), we uncovered a surprising mechanism robustly triggering biofilm development.

Like Comment

An upgrade in the social status

What makes bacteria switch from the solitary planktonic existence to the collective life in a complex community of biofilm? The triggers of this tremendous change in social behavior of bacterial ensembles that can cause thousands of deaths are still far from being clear. Social, collective behavior presupposes close contact and coordination of action. And yet, if you are a cell, there are disadvantages of being too close to your peers. For one thing, your neighbors (who are frequently your siblings) may eat your lunch. For another, being surrounded by neighbors in a tight room may leave little physical space for you and your progeny to grow. And yet, these disadvantages may not be too high of a price to pay for the advantage of symbiotic defense against external challenges that is provided by the communal life in biofilm, which can enable collective cell behavior and sharing of the load and dividing the labor in the harsher environments.

Forces and stresses galore

 It has long been known that bacterial cells tend to gather together into either suspended clouds or find the smallest room in the mazes built for them by enterprising human scientists. As a result, cells may find themselves packed in small cavities or other confined spaces, which may both enhance their collective existence and create boundaries to the colony growth. What is the effect of boundaries, in native environments or in man-made microfluidic devices, on bacterial colonies? In our experiments with the first microfluidic chemostats (2) performed more than a decade ago, bacterial colonies were locked in small chambers, which were connected to continuously replenished reservoirs of fresh medium through tiny slits so small that cells would not be able to escape. We found that the colonies were apparently well-supplied through the slits, because they kept growing and eventually filled the chambers completely, becoming tightly packed. What happened next was a surprise. In spite of the fact that the microfluidic device was built to withstand the pressure, the dense bacterial colonies often managed to exert enough force on the chamber roof to lift the roof and separate the chip from the coverglass. After that, in an act of daring jailbreak, the colony rapidly spread into the newly formed gap, in a kind of an avalanche. The force lifting the roof and pushing cells out comes from the turgor pressure that drives the cell growth itself and can be substantial. We became curious about how large this force  may actually be, and so, we built a new kind of device, in which the growth chambers have flexible roofs that undergo measurable deformations as colonies push on them. By observing the roof deformations, we were able to evaluate the pressures mounted by the growing colonies. These forces turned out to be indeed very large, enough for uropathogenic E. coli to break eukaryotic cells that bacteria can initially invade. However, surprisingly, the forces were smaller than those expected from the turgor pressure, hinting at something more complex. 

Generate, sense and respond to mechanical stress; a recipe for biofilm formation

 As we are all taught by good old Newton, anything pressed upon is going to press back. As a colony presses on thewalls of the chamber, so do the chamber walls press on the colony, and as a result, cells in the colony become physically stressed. 

Biofilm forming in a chamber of the microfluidic device. Note the characteristic 'butterfly' shape. The biofilm forming in a chamber of the microfluidic device. Note the 'characteristic butterfly' shape at the bottom of the chamber.

We surmised that this physical stress leads to physiological stress, eventually altering what the cells do and triggering a stress signal and response. Why did we feel that way? For one thing, we saw that the bulging of the chamber roofs on top of the growing colonies stopped way sooner than one could have expected from the previously reported turgor pressure values for E. coli. Apparently, there was something else, in addition to the pressure from the roof, that held cells back, not letting them grow, and that something might have been of cells’ own doing. Secondly, as soon as the cells started deforming the roof, we observed an increase in the expression of a sigma factor that is usually associated with various environmentally triggered stresses. Hence, apparently, even the relatively small physical pressure from the roof could trigger various biological stress responses associated with the sigma factor, and one of these responses could be biofilm formation. These observations raised the possibility that biofilm formation could be triggered by the biological stress, which is itself caused by a self-imposed physical stress originating from the colony growth. The biofilm comes with secretion of gooey exopolysaccharides and other compounds that can make the environment very viscous or maybe even jelly-like. The altered mechanical properties of the cellular environment can explain why the cells in the emerging biofilm would have trouble expanding. It is hard to stir, deform, or penetrate something that is very viscous. Just try stirring honey or jelly with a spoon. Pieces of the puzzle appeared to fit together, but we still had to show that what formed in response to the self-imposed physical stress had the essential features of biofilm.

 Biofilm formation by uropathogenic E. coli, our chosen species, has telltale signs, biofilm markers. Those include proteins and sugars secreted by cells, which can glue cells into a cohesive ensemble and protect them by creating biochemical barriers to the outside world hazards, such as antibiotics. Both structural (proteins and sugars) and functional (increase in antibiotic resistance) markers of biofilms were clearly on display after the onset of physical stress (bulging of the chamber roofs). Remarkably, inside the chambers, the expression of the sigma factor and of the biofilm markers were correlated in space, both forming consistent spatial patterns, remotely resembling butterflies. To explain these expression patterns, we came up with two models, which both postulated that, within the viscous biofilm, cells growing more rapidly experience greater stresses. The growth rates are higher for cell in those areas of the chambers, where the access to nutrients is better. Higher growth rate leads to higher physical stress and to greater secretion of the gooey compounds specific for biofilms. These compounds increase the viscosity, further increasing the physical stress on the growing cells and providing a positive feedback mechanism, which is modulated by nutrient supply. Strikingly, the models incorporating this positive feedback mechanism closely reproduced the experimental observations, including the ‘butterfly’ patterns observed in our experiments, thus providing support for the proposed feedback mechanism. Predictions of the models also matched our experimental findings on E. coli inside a hydrogel, an environment physically resembling the interior of a eukaryotic cell, such as a human urinary tract cell hosting uropathogenic E. coli during a poorly treatable chronic infection. 

Biofilm in the gel
A biofilm forming inside a 3D gel resembling the cellular cytosol.

Lessons learned

What should we take away from this analysis? First, our study provides yet another indication that cells can sense and respond to various facets of their environment, in this case – mechanical cues. Second, our study suggests a general mechanism of how biofilm formation may be triggered, which is still a matter of some debate. Harnessing this mechanism for the generation of precisely controlled arrays of biofilm could potentially enable high throughput drug screening. Third, our analysis suggests mechanisms for sculpting complex biofilm architectures that may help explain diverse shapes observed in natural biofilm formation. Finally, and perhaps most importantly, any new facet of biofilm formation arising from research can help us better understand this key mechanism of antibiotic tolerance and take us closer to successfully dealing with biofilm related clinical and industrial problems. 

Andre Levchenko with Eric Chu, Onur Kilic and Alex Groisman

Yale University and UC San Diego 


1. Chu EK, Kilic O, Cho H, Groisman A, Levchenko A. Self-induced mechanical stress can trigger biofilm formation in uropathogenic Escherichia coli Nature Communications 9: 4087 (2018)


2. Groisman A, Lobo C, Cho H, Campbell JK, Dufour YS, Stevens AM,   Levchenko A.   A microfluidic chemostat for experiments with bacterial and yeast cells.                 Nat Methods. 2(9):685-9 (2005).

Andre Levchenko

John C. Malone Professor and Director of Yale Systems Biology Institute , Yale University