Figure 1.
The natural agn43 expression state (ON or OFF) strongly influences E. coli community behavior.
A. Schematic representation of the agn43 phase variation mechanism (not to scale): if OxyR binds to the agn43 promoter, it impedes agn43 transcription (PHASE OFF); however, if Dam methylates GATC sites at the OxyR binding site, agn43 can be expressed (PHASE ON). This mechanism is heritable but reversible upon replication. B. Aggregating and non-aggregating clones from an isogenic wild-type TG strain. Pictures of stationary phase cultures were taken after 6 h settling on the bench. WT (A+): wild-type aggregating culture, WT (A−): wild-type non-aggregating culture, ΔoxyR: agn43 locked-ON strain, Δdam: agn43 locked-OFF strain. C. Immunodetection of Ag43 in aggregating and non-aggregating clones. WT (A+): wild-type aggregating culture, WT (A−): wild-type non-aggregating culture, ΔoxyR: agn43 locked-ON strain, Δdam: agn43 locked-OFF strain. D. Biofilm forming ability of an aggregating clone and a non-aggregating clone. Biofilms were formed in microfermentors for 24 h; quantitative analysis involved measuring the optical density of the resuspended biofilm. WT (A+): wild-type aggregating culture, WT (A−): wild-type non-aggregating culture, ΔoxyR agn43 locked-ON strain, Δdam: agn43 locked-OFF strain, Δagn43: deletion mutant of agn43. ***: p<0.0001. NS: not significant.
Table 1.
Strains and plasmids used in this study.
Table 2.
Primers used in this study.
Figure 2.
Construction and characterization of agn43-lacZ transcriptional fusions.
A. Schematic representation of TG agn43-lacZ and TG Δagn43::lacZ fusions (not to scale). Blue or white colony plated on LB agar+X-gal plates: a blue colony gives rise to blues and whites and vice versa. Switching frequencies (ON or OFF cells/generation) of the transcriptional fusions were calculated as described in the materials and methods. S.D.: standard deviation. B. DAPI and immunofluorescence microscopy of a TG agn43-lacZ blue (ON) colony and a white (OFF) colony. Anti-Ag43 polyclonal antibody was used to detect surface-exposed Ag43. WT(B): wild-type ON colony with OFF cells pointed out by white arrow heads, WT(W): wild-type OFF colony, ΔoxyR: agn43 locked-ON strain, Δdam: agn43 locked-OFF strain. C. Kinetics of aggregation of overnight cultures inoculated with TG agn43-lacZ (WT) and Δagn43::lacZ strains (Δagn43) blue or white colony. D. Immunodetection of Ag43 in TG agn43-lacZ and Δagn43::lacZ cultures started with either an ON or OFF colony, using an anti-Ag43 polyclonal antibody. In a TG agn43-lacZ background, WT(W): OFF colony, WT(B): ON colony, ΔoxyR: locked-ON strain, Δdam: locked-OFF strain. In TG Δagn43::lacZ background, Δagn43 (W): OFF colony, Δagn43 (B): ON colony, ΔoxyRΔagn43: locked-ON strain.
Figure 3.
In vitro biofilm produced in continuous flow culture bioreactors selects for agn43 ON cells.
A. Biofilms of TG agn43-lacZ from an ON or an OFF colony in microfermentors. The biofilms were grown for 24 and 48 h; the biomass growing on the spatula was resuspended in 10 mL M63B1Gluc and the optical density at 600 nm was measured. (B): ON colony, (W): OFF colony; B. Pictures of microfermentors after 48 h growth. ***: p<0.0001. C. Immunodetection of Ag43 in TG agn43-lacZ biofilm (F) or planktonic cultures (P) from ON (B) or OFF (W) colonies. I: inoculum, F: microfermentor biofilm, P: planktonic culture, the α-subunit of RNA polymerase (RNAP) was used as an internal control.
Table 3.
Percentages of ON and OFF cells in 24/48 h-old biofilms or planktonic cultures of strain TG agn43-lacZ.
Table 4.
Percentages of ON and OFF cells in 24/48 h-old biofilms or planktonic cultures of strain TG Δagn43::lacZ.
Figure 4.
Ag43 interferes with YfaL-mediated biofilm formation.
Biofilms were formed in 96-well micro-titer plates for 24 h; biofilm production was quantified by crystal violet staining as described in materials and methods. A TG agn43-lacZ background was used to monitor the ON or OFF state of colonies, ON: wild-type ON colony, OFF: wild-type OFF colony. PcL-yfaL strains constitutively expressing yfaL: agn43 ON/OFF colonies, ΔoxyR: locked-ON strain, ΔoxyR Δagn43: Δagn43::lacZ locked-ON strain. Unless specified statistical analyses were performed using the WT (ON) strain as a reference: NS: not significant, **: p<0.001, ***: p<0.0001.
Figure 5.
Lrp and MqsR do not regulate agn43 expression.
A. Switching frequencies of lrp and mqsR mutants assessed with an agn43-lacZ fusion. S.D.: standard deviation. B. Semi-quantitative RT-PCR analyses of agn43 expression in wild-type (WT), Δlrp, ΔmqsR, ΔoxyR, and Δdam agn43-lacZ strains. Experiments were performed using RNA preparations that were not diluted (1), or diluted 1/10 (2) or 1/100 (3). (B): ON colony, (W): OFF colony, +/− RT: with or without reverse transcriptase polymerase. The 16S (rrsh gene) was used as an internal control. Relative ratio: average of agn43/16S band intensity ratio quantified using ImageJ, WT(B) used as reference; +/−: standard deviation. C. Immunodetection of Ag43 with anti-αAg43 polyclonal antibody in WT, Δlrp and ΔmqsR agn43-lacZ strains. (B): ON colony, (W): OFF colony, the α-subunit of RNA polymerase (RNAP) was used as an internal control. Relative ratios (Ag43/RNAP): band intensity ratio quantified using ImageJ; +/−: standard deviation. D. Kinetics of aggregation of strains TG agn43-lacZ, TG agn43-lacZ Δlrp and TG agn43-lacZ ΔmqsR. (B): ON colony, (W): OFF colony, ΔoxyR: locked-ON strain, Δdam: locked-OFF strain.