Toxin-antitoxin systems in Escherichia coli influence biofilm formation through YjgK (TabA) and fimbriae

Abstract

The roles of toxin-antitoxin (TA) systems in bacteria have been debated. Here, the role of five TA systems in regard to biofilm development was investigated (listed as toxin/antitoxin: MazF/MazE, RelE/RelB, ChpB, YoeB/YefM, and YafQ/DinJ). Although these multiple TA systems were reported previously to not impact bacterial fitness, we found that deletion of the five TA systems decreased biofilm formation initially (8 h) on three different surfaces and then increased biofilm formation (24 h) by decreasing biofilm dispersal. Whole-transcriptome profiling revealed that the deletion of the five TA systems induced expression of a single gene, yjgK, which encodes an uncharacterized protein; quantitative real-time PCR (qRT-PCR) confirmed consistent induction of this gene (at 8, 15, and 24 h). Corroborating the complex phenotype seen upon deleting the TA systems, overexpression of YjgK decreased biofilm formation at 8 h and increased biofilm formation at 24 h; deletion of yjgK also affected biofilm formation in the expected manner by increasing biofilm formation after 8 h and decreasing biofilm formation after 24 h. In addition, YjgK significantly reduced biofilm dispersal. Whole-transcriptome profiling revealed YjgK represses fimbria genes at 8 h (corroborated by qRT-PCR and a yeast agglutination assay), which agrees with the decrease in biofilm formation upon deleting the five TA systems at 8 h, as well as that seen upon overexpressing YjgK. Sand column assays confirmed that deleting the five TA systems reduced cell attachment. Furthermore, deletion of each of the five toxins increased biofilm formation at 8 h, and overexpression of the five toxins repressed biofilm formation at 8 h, a result that is opposite that of deleting all five TA systems; this suggests that complex regulation occurs involving the antitoxins. Also, the ability of the global regulator Hha to reduce biofilm formation was dependent on the presence of these TA systems. Hence, we suggest that one role of TA systems is to influence biofilm formation.

FIG. 1.

(A) Normalized biofilm formation (total biofilm/growth) in LB medium at 37°C for MG1655 and Δ5 with 96-well plates of polystyrene (PS), polyvinyl chloride (PVC), and polypropylene (PP) after 8 and 24 h of incubation. (B) Biofilm images under the same culture conditions in 14-ml polystyrene tubes. The data are the average of 10 replicate wells from two independent cultures, and one standard deviation is shown.

FIG. 2.

Time course of biofilm dispersal for MG1655 and Δ5 (A) and BW25113 and BW25113 yjgK (B) in 96-well polystyrene plates with LB medium at 37°C. Dispersal indicates the percentage of normalized biofilm (total biofilm/growth) that was removed at each time point compared to the 15-h normalized biofilm. The data are the averages of 10 replicate wells from two independent cultures, and one standard deviation is shown.

FIG. 3.

(A) Attachment to sand columns for MG1655 and Δ5 in LB medium at 37°C. (B) BW25113 versus BW25113 fimA was used as fimbria minus control. The data in panel A are the averages of two independent cultures, and one standard deviation is shown.

FIG. 4.

Normalized biofilm formation (total biofilm/growth) in LB medium at 37°C for MG1655 and MG1655 fimA with 96-well polystyrene plates. The data are the averages of 10 replicate wells from two independent cultures, and one standard deviation is shown.

FIG. 5.

Normalized biofilm formation (total biofilm/growth) after 8 h and 24 h of incubation in LB medium at 37°C for BW25113, BW25113 yjgK, and BW25113 yjgK/pCA24N-yjgK (complementation study). YjgK was induced by the addition of 0.5, 1, and 2 mM IPTG. The data are the average of 10 replicate wells from two independent cultures, and one standard deviation is shown.

FIG. 6.

(A) Normalized biofilm formation (total biofilm/growth) upon induction of Hha in MG1655 and Δ5 via 1 mM IPTG for 8 h and 24 h in LB at 37°C using plasmid pCA24N-hha. (B) Normalized biofilm formation (total biofilm/growth) upon deleting toxin genes mazF, relE, chpB, yoeB, and yafQ after 8 h in LB at 37°C with BW25113. (C) Normalized biofilm formation (total biofilm/growth) upon overexpressing toxin genes mazF, relE, chpB, yoeB, yafQ after 8 h in LB at 37°C with 1 mM IPTG with BW25113. The data are the average of 10 replicate wells from two independent cultures, and one standard deviation is shown.

FIG. 7.

Normalized biofilm formation (total biofilm/growth) upon overexpressing antitoxin genes relB, dinJ, and yefM with 1 mM IPTG after 24 h in LB at 37°C in MG1655 (A) and in Δ5 (B). The data are the average of 10 replicate wells from two independent cultures, and one standard deviation is shown.

FIG. 8.

Schematic of the mechanism for the impact of the five TA systems and biofilm formation in E. coli: early biofilm formation is increased via fimbriae and repression of YjgK, and late biofilm formation is decreased due to dispersal and repression of YjgK. Also, the ratio of toxins (T) and antitoxins (A) may influence biofilm formation. The “→” indicates induction, and “⊥” indicates repression.