Quorum sensing in Vibrio controls carbon metabolism to optimize growth in changing environmental conditions

Abstract

Bacteria sense population density via the cell-cell communication system called quorum sensing (QS). The evolution of QS and its maintenance or loss in mixed bacterial communities is highly relevant to understanding how cell-cell signaling impacts bacterial fitness and competition, particularly under varying environmental conditions such as nutrient availability. We uncovered a phenomenon in which Vibrio cells grown in minimal medium optimize expression of the methionine and tetrahydrofolate (THF) synthesis genes via QS. Strains that are genetically "locked" at high cell density grow slowly in minimal glucose media and suppressor mutants accumulate via inactivating mutations in metF (methylenetetrahydrofolate reductase) and luxR (the master QS transcriptional regulator). In mixed cultures, QS mutant strains initially coexist with wild-type, but as glucose is depleted, wild-type outcompetes the QS mutants. Thus, QS regulation of methionine/THF synthesis is a fitness benefit that links nutrient availability and cell density, preventing accumulation of QS-defective mutants.

Fig 1. Growth trends of V. campbellii quorum sensing mutants differ in the absence of amino acids.

(A) V. campbellii QS system and mutant alleles. At LCD, LuxO is phosphorylated and along with Sigma-54 activates transcription of the Qrr small RNAs. The Qrrs together with Hfq repress LuxR production. At HCD, unphosphorylated LuxO is unable to activate transcription of the Qrr small RNAs, allowing for translation of LuxR and production of proteases. The luxO D61E allele produces a locked-LCD phenotype. Deletion of luxO or the qrrs results in phenotypically HCD-locked strains. Created with BioRender.com. (B–D) Growth curves for V. campbellii DS40M4 strains in LM (B), M9G (C), or M9GC (D) media. The y-axis represents cell density OD_600. For all panels, the data are from a single experiment that is representative of at least 3 independent biological experiments for each strain under every condition. The data underlying this figure can be found in S1 Data. HCD, high cell density; LCD, low cell density; LM, lysogeny broth marine; QS, quorum sensing; WT, wild-type.

Fig 2. Mutation of metF or luxR restores growth to HCD-locked strains.

(A) Mutations identified in the Δqrr1-5 suppressor mutant strain. (B) Culture growth of V. campbellii DS40M4 strains in M9G medium. (C) Growth rates of strains from panel B are plotted. (D) Growth curves of V. campbellii DS40M4 strains in M9GC medium. The inset in panels B and D are zoomed-in views of the same data to enable visualization of strains at early time points. (E) Growth rates of strains from panel D are plotted. For panels B and D, the data are from a single experiment that is representative of at least 3 independent biological experiments for each strain under every condition and the y-axis represents cell density OD_600. For panels C and E, error bars show the mean and standard deviation of at least 3 biological replicates. The slope of the line was derived from the data in exponential phase with at least 5 data points per growth curve. Statistical analyses were performed using the Kruskal–Wallis test (nonparametric) followed by Dunn’s multiple comparisons test (n = at least 3). Different letters above the bars indicate that the pair-wise comparison is significantly different; the same letters above 2 bars indicate no significant differences (p < 0.05). There were no significant differences in panel D. The data underlying this figure can be found in S2 Data. HCD, high cell density; WT, wild-type.

Fig 3. LuxR G37V has decreased DNA-binding activity.

(A) The protein sequences for LuxR_Vcamp (V. campbellii; AAA27539), SmcR (V. vulnificus; AAF72582), HapR (V. cholerae; ABD24298), OpaR (V. parahaemolyticus; NP_798895), LitR (V. fischeri; YP_205560), and LuxR_Va (V. alginolyticus; PDB: 7AMN) aligned to E. coli TetR (P0ACT4) were aligned using Clustal Omega [19] and the diagram generated using Jalview 2.0 [57]. (B) Structure of LuxR from V. alginolyticus bound to a repressed DNA sequence [14]. (C) Bioluminescence production (light divided by OD₆₀₀) throughout a growth curve for DS40M4 strains. The data shown are from a single experiment that is representative of at least 3 independent biological experiments for each strain under every condition. (D, E) EMSAs with either purified WT LuxR (D) or G37V LuxR (E) with 5′ IR700 Dye (Integrated DNA Technologies) DNA substrate PmetJ region (ZC011 and ZC012). Protein concentrations are 0.0005, 0.005, 0.05, .5, 5, 50, and 500 nM, compared to no protein control (“-”). The data underlying this figure can be found in S2 Data and S1 Raw Images. EMSA, electrophoretic mobility shift assay; WT, wild-type.

Fig 4. Predicted AMC and folate cycle enzymes in V. campbellii DS40M4.

This is based on V. campbellii BB120 KEGG gene predictions. Created with BioRender.com.

Fig 5. LuxR regulation of methionine biosynthesis genes.

(A) Predicted regulation hierarchy of the methionine biosynthesis genes. Created with BioRender.com. (B–G) Relative transcript levels determined by RT-qPCR for cultures grown in M9GC medium (B–D) or M9G medium (E–G). Error bars show the mean and standard deviation of 3 biological replicates. (B–D) Different letters above the bars indicate that the pair-wise comparison is significantly different; the same letters above 2 bars indicate no significant differences (p < 0.05). One-way analysis of variance (ANOVA) was performed on log-normalized data (normally distributed, Shapiro–Wilk test; n = 3 biological replicates, Tukey’s multiple comparisons test). (E–G) Student’s–test was performed data (normally distributed, Shapiro–Wilk test; n = 3 biological replicates; p < 0.05, *; p < 0.001, ***). (H) Quantification of DPDQ (derivatization of DPD, the precursor to AI-2) via mass spectrometry from supernatants from strains grown in M9G medium and collected at HCD. DPDQ levels were normalized to an internal control (Verapamil) spiked in every sample. The mean of 3 independent biological replicates is shown. The data underlying this figure can be found in S3 Data. HCD, high cell density; WT, wild-type.

Fig 6. Altering flux between the AMC and folate cycle affects culture growth.

(A–C) Growth curves for V. campbellii DS40M4 strains grown in M9G (A, B) or M9GC (C) medium. The y-axes represent cell density OD_600. The data shown are from a single experiment that is representative of at least 3 independent biological experiments for each strain under every condition. The data underlying this figure can be found in S4 Data. AMC, activated methyl cycle; WT, wild-type.

Fig 7. The Qrrs regulate growth of V. campbellii in minimal medium.

(A, B) Growth curves for V. campbellii DS40M4 strains in M9G medium. The y-axis represents cell density OD_600. The data shown are from a single experiment that is representative of at least 3 independent biological experiments for each strain under every condition. The data underlying this figure can be found in S4 Data. WT, wild-type.

Fig 8. WT cells outcompete locked-LCD cells in co-culture in minimal medium.

(A) WT (marked with Tm) and luxO D61E (marked with Spec) strains of DS40M4 were inoculated at initial frequencies from 1%–99% WT in M9G medium and allowed to grow for 48 h. (B, C) WT (marked with Tm) and luxO D61E (marked with Spec, panel B) or Δqrr1-5 (marked with Spec, panel C) strains of DS40M4 were inoculated at initial frequencies of 80%, 50%, or 20% WT in (B) M9G medium or (C) M9 Casein medium and grown to near stationary phase at t = 16 h, and then diluted to maintain growth in log-phase. Colony-forming units (cfu/mL) were measured at each time point using selective antibiotics for each strain. The total population number was calculated, and the data are represented as the percent of WT cells in the total population of WT and luxO D61E or WT and Δqrr1-5. The data are representative of 3 biological experiments. The data underlying this figure can be found in S5 Data. cfu, colony-forming unit; LCD, low cell density; WT, wild-type.

Fig 9. Model of QS regulation of methionine flux in M9G.

Wild-type V. campbellii switches between cell density states during growth, and regulatory feedback loops promote or repress expression of genes in the AMC and folate pathways, resulting in a balanced flux of both systems. Strains locked at LCD or HCD cannot oscillate between states, and thus have growth defects under nutrient-limited conditions. Created with BioRender.com. AMC, activated methyl cycle; HCD, high cell density; LCD, low cell density; QS, quorum sensing.