Visco-elastic membrane tethers extracted from Escherichia coli by optical tweezers

Abstract

Tethers were created between a living Escherichia coli bacterium and a bead by unspecifically attaching the bead to the outer membrane and pulling it away using optical tweezers. Upon release, the bead returned to the bacterium, thus showing the existence of an elastic tether between the bead and the bacterium. These tethers can be tens of microns long, several times the bacterial length. Using mutants expressing different parts of the outer membrane structure, we have shown that an intact core lipopolysaccharide is a necessary condition for tether formation, regardless of whether the beads were uncoated polystyrene or beads coated with lectin. A physical characterization of the tethers has been performed yielding visco-elastic tether force-extension relationships: for first pull tethers, a spring constant of 10-12 pN/mum describes the tether visco-elasticity, for subsequent pulls the spring constant decreases to 6-7 pN/mum, and typical relaxation timescales of hundreds of seconds are observed. Studies of tether stability in the presence of proteases, lipases, and amylases lead us to propose that the extracted tether is primarily composed of the asymmetric lipopolysaccharide containing bilayer of the outer membrane. This unspecific tethered attachment mechanism could be important in the initiation of bacterial adhesion.

FIGURE 1

A rough sketch of a lipopolysaccharide (LPS) of a Gram-negative bacteria. From left (outside) to right (inside) a LPS molecule consists of an o-antigen part linked together by glucosidic bonds, a core region divided in an inner and outer region, and the lipid A. Parts of LPS expressed in the different chemotypes are also shown.

FIGURE 2

This figure shows from which of the strains tethers could be created successfully. The fraction of pulls resulting in the creation of tethers is labeled tethers. The number of pulls that did not create tethers is labeled failures. The value n is the total number of experiments of one column. From left, the bars shows the distributions for the smooth CS1861, the rough CS180, the rough S2188, and the deep rough CS2429 strains.

FIGURE 3

Force-extension relations for two cycles of stretch and relax of a tether extracted from S2188 at a constant velocity of 0.2 μm/s. The upper stretch1 and relax1 are the curves for the first time the tether is stretched and relaxed. The labels stretch6 and relax6 are curves resulting from the sixth cycle. Inset shows a sketch of the experiment.

FIGURE 4

Within the first 40 seconds, a tether is extracted from the rough strain S2188 to a distance of 4 μm with a velocity of 0.1 μm/s. At this extreme position the stage is stopped and the figure shows how the force in the tether relaxes as a function of time.

FIGURE 5

Sensitivity of tethers toward various enzymes. The value n is the total number of experiments of a given kind. (A) Stability of tethers from the smooth mutant CS1861 in the presence of protease, lipase, or amylase. The control experiment does not contain any enzymes but BSA in an equivalent concentration. (B) Stability of tethers from the rough mutant S2188 in the presence of protease, lipase, or amylase. Again, the control only contains BSA. The last column shows the sensitivity of the tether when all enzymes are present.

FIGURE 6

Proposed model of the observed visco-elastic bacterial tether stemming from a rough bacterial strain. The tether spanning the distance between the bead and the body of the bacterium is proposed to consist of the asymmetric LPS and phospholipid bilayer but not of the peptid-o-glycan layer or anything beneath it. Possible membrane proteins are not thought to play any crucial role and, hence, are not depicted. The o-antigen part is not drawn on this picture, and for tethers extracted from smooth strains, this should be envisioned too.