Factors affecting daughter cells' arrangement during the early bacterial divisions

Abstract

On agar plates, daughter cells of Escherichia coli mutually slide and align side-by-side in parallel during the first round of binary fission. This phenomenon has been previously attributed to an elastic material that restricts apparently separated bacteria from being in string. We hypothesize that the interaction between bacteria and the underneath substratum may affect the arrangement of the daughter bacteria. To test this hypothesis, bacterial division on hyaluronic acid (HA) gel, as an alternative substratum, was examined. Consistent with our proposition, the HA gel differs from agar by suppressing the typical side-by-side alignments to a rare population. Examination of bacterial surface molecules that may contribute to the daughter cells' arrangement yielded an observation that, with disrupted lpp, the E. coli daughter cells increasingly formed non-typical patterns, i.e. neither sliding side-by-side in parallel nor forming elongated strings. Therefore, our results suggest strongly that the early cell patterning is affected by multiple interaction factors. With oscillatory optical tweezers, we further demonstrated that the interaction force decreased in bacteria without Lpp, a result substantiating our notion that the side-by-side sliding phenomenon directly reflects the strength of in-situ interaction between bacteria and substratum.

Figure 1. Appearance of daughter bacteria growing on LB-HA gel.

JW1923, a fliG-deleted K-12 strain, was grown on a LB-HA (0.5%)-coated slide and the first two rounds of bacterial division were imaged by time-lapsed microscopy. (A) Formation of 4-cell arrays; (B) Patterning in a string; (C) Representatives of non-typical arrangements. Scale bar: 5 µm.

Figure 2. Patterning of daughter bacteria in liquid without underneath substratum.

Bacterial divisions in LB medium were followed similar to that described in Figure 1. (A) Parental strain BW25113; (B) flagellum-null strain JW1923. In (A), arrowhead indicates a cell separating and disappearing from the bacterial string in the subsequent image. In (B), the arrow indicates bacteria that appeared as a string of four cells while the arrowhead marks a string of 16 offspring bacteria. Note: the balloon-type bacteria were not exactly on the focal plane of the microscope. No bacteria were observed to have a parallel or 4-cell array arrangement in all fields. Scale bar: 5 µm.

Figure 3. Measurement of the interaction of the bacterium with the surrounding LB agar in terms of the interaction force constant (K_bio).

The K_bio values, representing the interacting forces between bacterium and 0.2% LB-agar, were measured by oscillatory optical tweezers. In each group, about 20 measurements were performed and every single spot represents a measured K_bio value; the horizontal line marks the average of the group. Asterisk indicates that a statistically significant difference is observed between the paired strains (p<0.001 by t-test).

Figure 4. Comparison of flagellar rotation frequency.

The flagellar rotation frequencies measured for strains BW25113 and JW1923 in LB medium were compared. A distinct peak was shown at around 185 Hz in the power spectrum of the quadrant photodiode signal for the case of the wild type strain BW25113 (A) but not in the flagella mutant strain JW1923 (B). Arrow indicates the position of the peak.

Figure 5. Measurement of bacterial motility.

BW25113 and JW1923 in motility medium were tracked by time-lapsed microscopy for 10 min and the recorded images were analyzed to determine the average motility of each bacterium by MetaMorph software. The motilities of the two bacteria differed significantly at P<0.0001 (by t-test).

Figure 6. Bacterial motility detected on agar plates.

Appropriately diluted bacteria were grown overnight at 37°C on LB plates containing 0.6% or 0.27% agar. No differences of bacterial colonies are seen in panel (A) (0.6% agar) whereas panel B (0.27% agar) shows that the parental BW25113 gave strong swarming activity that is not seen with the ΔfliG mutant.

Figure 7. Interaction force constants measured for JW1667 and its related plasmid-transformed strains.

JW1667 is an lpp-disrupted strain derived from BW25113 and has a phenotype with a low rate of daughter cells' patterning in parallel. K_bio measurement and illustrations of the measured values are as in Figure 3. Transformation of JW1667 with plasmid pACYC184-lpp was to complement the lost lpp whereas that with pACYC184 transformation was to give a plasmid control. The arrowhead marks the mean K_bio value of BW25113 shown in Figure 3. Asterisks mark the pairs with statistically significant differences (p<0.0001 by t-test).