(a) Biotin-labeled DNA corresponding to the rpoS promoter (PrpoS) was incubated with either a 10-fold (lanes 2–3), a 25-fold (lanes 4–5), or a 50-fold (lanes 6–7) excess of MqsA. At all protein concentrations, MqsA is able to bind and shift the labeled PrpoS DNA (lanes 2, 4, 6). Addition of 100-fold excess of unlabeled PrpoS DNA demonstrates specificity of binding (lanes 3, 5, 7). Biotin-labeled PrpoS in the absence of MqsA was used as a control (lane 1). (b) Biotin-labeled DNA corresponding to a mutated rpoS promoter (PrpoS-M) was incubated with MqsA as indicated in (a). MqsA binding to the mutated PrpoS is dramatically reduced at all concentrations tested. (c) Reporter strain Δ6 R1 PrpoS was constructed by conjugating pKNOCK-PrpoS with the wild-type PrpoS into Δ6 ΔlacZ, while reporter strain Δ6 R2 PrpoS-M and Δ6 R3 PrpoS were constructed by conjugating pKNOCK-PrpoS-M with mutated PrpoS into Δ6 ΔlacZ. Transcription start sites of rpoS are indicated by black arrows. Primers used for strain verification by DNA sequencing are indicated by blue arrows. (d) Relative repression of β-galactosidase activity in the three Δ6 reporter strains with MqsA (produced via pCA24N-mqsA) vs. without MqsA (pCA24N). β-galactosidase activity was measured 2 h and 4 h after adding 0.1 mM IPTG when the turbidity (600 nm) was ~1. Error bars indicate standard error of mean (n = 3). Significant changes are marked with an asterisk for P < 0.05.

(a) c-di-GMP concentrations in stationary-phase cultures (starving cells) after 15 h at 37°C. (b) Percentage of cells which survive the oxidative stress induced by 20 mM H₂O₂ for 10 min. (c) Percentage of cells which survive the acid stress induced by pH 2.5 for 10 min. Error bars indicate standard error of mean (n = 3). Significant changes are marked with an asterisk for P < 0.05. (d) Images of MG1655 Δ6 and BW25113 mqsRA cultures (turbidity of 1) 10 min after adding 20 mM H₂O₂. Bubbles are oxygen produced by the decomposition of hydrogen peroxide by catalase: 2 H₂O₂ → 2 H₂O + O₂. mqsA was induced from pCA24N-mqsA via 0.5 mM IPTG for the c-di-GMP assay and via 1 mM IPTG for the stress assays.

(a) Swimming motility after 9 h of growth at 37°C for Δ6 cells overexpressing mqsA or mqsR. (b) Swimming motility after 12 h of growth at 37°C for BW25113 cells overexpressing mqsA. (c) Swimming motility after 12 h of growth at 37°C for the three Δ6 reporter strains with MqsA (right side) and without MqsA (left side). For (a–c), chloramphenicol (30 µg/mL) was used for plasmid maintenance, and kanamycin (50 µg/mL) and tetracycline (5 µg/mL) were used to select the host strains, where appropriate. (d) Swimming motility after 12 h of growth at 37°C for Δ6 ΔKm^R cells overexpressing mqsA was from pBS(Kan)-mqsA (pBS-mqsA) and rpoS from pCA24N-rpoS via 0.1 mM IPTG. Control where rpoS was induced in the absence of MqsA was included to show the direct effect of RpoS on motility. Chloramphenicol (30 µg/mL) and kanamycin (50 µg/mL) were used to maintain pCA24N-based and pBS(Kan)-based plasmids, respectively. Representative images are shown. Petri dishes (8.3 cm diameter) were used for all motility tests. MqsA, MqsR, and RpoS were produced using 0.1 mM IPTG.

(a) Cellulose/curli formation after 180 min at 30°C with oxidative stress (2 mM H₂O₂). (b) Biofilm formation after 24 h in M9 glucose (0.2%). mqsA and mqsR were induced from pCA24N-based plasmids via 0.1 mM IPTG for the cellulose/curli assays and via 1 mM IPTG for the biofilm assays. Error bars indicate standard error of mean (n = 3). Significant changes are marked with an asterisk for P < 0.05.

(a) Lanes 1–5 (Western, short time) show the degradation of His-MqsA detected by a His-tag antibody with 20 mM H₂O₂ (Δ6/pCA24N-mqsA). (b) Lanes 1–6 (Western, longer time) show the degradation of His-MqsA detected by a His-tag antibody with and without 20 mM H₂O₂ (Δ6/pCA24N-mqsA). Lane 1 is the negative control (NC) where His-MqsA is absent (Δ6/pCA24N). mqsA was induced from pCA24N-mqsA via 0.5 mM IPTG. (c) β-galactosidase activity for Δ5 R1 PrpoS (contains MqsA, closed circles) vs. Δ6 R1 PrpoS (lacks MqsA, open circles) with 20 mM H₂O₂ at a turbidity of ~3. Error bars indicate standard error of mean (n = 3). Significant changes are marked with an asterisk for P < 0.05. (d) Lanes 1–6 (Western) show RpoS levels as detected by an anti-RpoS antibody with 20 mM H₂O₂ (upper blot shows BW25113 (active MqsA) vs. mqsRA (no MqsA), lower blot shows MG1655 (active MqsA) vs. Δ6 (no MqsA). Lane 7 is the negative control (NC) where RpoS is absent (BW25113 rpoS). shows full SDS-PAGE and Western blots. (e) Swimming motility after 12 h of growth at 37°C for Δ6 ΔKm^R cells overexpressing mqsA from pBS(Kan)-mqsA (pBS-mqsA) and lon from pCA24N-lon via 0.1 mM IPTG. Chloramphenicol (30 µg/mL) and kanamycin (50 µg/mL) were used for maintaining the pCA24N-based and pBS(Kan)-based plasmids, respectively. Control where lon was induced in the absence of MqsA shows the effect of Lon on motility. Representative image is shown.

Transcription of rpoS is repressed by MqsA through direct binding. Under oxidative stress, MqsA is degraded by protease Lon, and rpoS transcription is derepressed. The increase in rpoS transcription leads to an increase in σ^s activity, which induces genes encoding catalase, c-di-GMP, and curli/cellulose production and represses the genes encoding the master regulator of motility, FlhDC. Upon oxidative stress, OxyR induces katG and the regulatory RNA OxyS which inhibits translation of the rpoS message. In addition, the RNase activity of MqsR may serve to rapidly direct the cell toward the translation of newly-transcribed, stress-related transcripts and leads to the formation of persister cells. The lightning bolt indicates oxidative stress, → indicates induction, and ⊥ indicates repression. Moreover, MqsA represses csgD probably via the mqsRA-like palindrome in the promoter region as indicated by a dotted line (direct binding studies were not performed here).