Bacterial-induced pH shifts link individual cell physiology to macroscale collective behavior

Abstract

Bacteria have evolved a diverse array of signaling pathways that enable them to quickly respond to environmental changes. Understanding how these pathways reflect environmental conditions and produce an orchestrated response is an ongoing challenge. Herein, we present a role for collective modifications of environmental pH carried out by microbial colonies living on a surface. We show that by collectively adjusting the local pH value, Paenibacillus spp., specifically, regulate their swarming motility. Moreover, we show that such pH-dependent regulation can converge with the carbon repression pathway to down-regulate flagellin expression and inhibit swarming in the presence of glucose. Interestingly, our results demonstrate that the observed glucose-dependent swarming repression is not mediated by the glucose molecule per se, as commonly thought to occur in carbon repression pathways, but rather is governed by a decrease in pH due to glucose metabolism. In fact, modification of the environmental pH by neighboring bacterial species could override this glucose-dependent repression and induce swarming of Paenibacillus spp. away from a glucose-rich area. Our results suggest that bacteria can use local pH modulations to reflect nutrient availability and link individual bacterial physiology to macroscale collective behavior.

Keywords: Paenibacillus spp.; carbon catabolite repression; pH modulation; swarming.

Fig. 1.

Effect of glucose on Paenibacillus sp. motility. (A) Representative images of Paenibacillus sp. colonies inoculated on 1.5% LBA plates, with or without 1% glucose (LBAG and LBA, respectively). Images were taken at 72 h postinoculation. (B) Swarming area of Paenibacillus sp. at 48 h postinoculation. (C) Growth of Paenibacillus sp. in liquid LB medium with or without 1% glucose (LBG and LB, respectively), y-axis presented in log₂. (D) qRT-PCR of Paenibacillus sp. flagellin (hag) gene expression in LBA and LBAG. Mean ± SEM; n = 12; Student’s t test was performed. ****P < 0.0001 in B; ***P < 0.001 in D. The plate diameter was 9 cm.

Fig. 2.

Effect of X. perforans volatiles on swarming of Paenibacillus sp. in the presence of glucose. (A) Experiments were carried out using bipartite Petri dishes. Representative images of Paenibacillus sp. colonies inoculated on LBA with 1% glucose media (LBAG) (upper compartment); X. perforans was inoculated on LBA (bottom compartment, right plate). X. perforans was inoculated 24 h prior to Paenibacillus sp. Image was taken 36 h after Paenibacillus sp. inoculation. (B) Swarming area of Paenibacillus sp. on LBAG in the absence (-Xp) and presence (+Xp) of X. perforans. Swarming area was measured at 36 h of post-Paenibacillus sp. inoculation. (C) qRT-PCR of Paenibacillus sp. flagellin (hag) gene expression in the absence (-Xp) and presence (+Xp) of X. perforans. Mean ± SEM; n = 12; Student’s t test was performed. ***P < 0.001 in B and C. The plate diameter was 9 cm.

Fig. 3.

Involvement of ammonia in swarming induction. (A–D) Detection of ammonia in the headspace of X. perforans cultures. (A) Isobutyl chloroformate reacts with ammonia to produce isobutyl carbamate. (B) Total ion chromatogram of a SPME derivatization experiment with X. perforans culture, showing the occurrence of an isobutyl carbamate peak. (C) Control experiment with an agar plate without X. perforans culture. (D) Electron ionization mass spectrum of isobutyl carbamate and major fragments, verifying the structure. (E–G) Effect of ammonia vapors on swarming of Paenibacillus sp. (E) Experiments were carried out using bipartite Petri dishes. Representative images of Paenibacillus sp. colonies inoculated on LBAG (upper compartment); X. perforans was inoculated on LBA (bottom compartment). Ten microliters of 32% ammonium hydroxide (i.e., 100 mM ammonia) were placed on the bottom compartment 24 h prior to Paenibacillus sp. Image was taken 36 h after Paenibacillus sp. inoculation. (F) Swarming area of Paenibacillus sp. on LBAG in the absence (control) and presence (ammonia) of ammonia. X. perforans serves as positive control. Swarming area was measured at 36 h after Paenibacillus sp. inoculation. (G) qRT-PCR of Paenibacillus sp. flagellin (hag) gene expression in the absence (Control) and presence (Ammonia) of ammonia. Mean ± SEM; n = 10. In F and G, Student’s t test was performed. ***P < 0.001; **P < 0.01. The plate diameter was 9 cm.

Fig. 4.

Effect of pH on glucose-dependent swarming inhibition of Paenibacillus sp. (A) Representative images of Paenibacillus sp. colonies inoculated on 1.5% LBA with 1% glucose (LBAG), LBAG + 100 mM Ammonium chloride (LBAG NH₄Cl), LBAG + 10 mM NaOH to adjust the pH to 8 (LBAG pH8), LBAG + 100 mM MOPS pH8 (LBAG MOPS pH8), and LBA + 100 mM MOPS pH6 (LBA MOPS pH6). Image was taken at 48 h after Paenibacillus sp. inoculation. (B) Swarming area of Paenibacillus sp. colonies at 36 h postinoculation. (C) qRT-PCR of Paenibacillus sp. flagellin (hag) gene expression in LBAG, LBAG NH₄Cl, LBAG pH8, LBAG 100 mM MOPS pH8, and LBA 100 mM MOPS pH6. Mean ± SEM; n = 5; the lowercase letter is used to denote significant differences in treatments using Tukey’s test at 0.05% significance level. Plates were supplemented with the pH indicator phenol-red (0.018 g/L); yellow color indicates pH ≤7; pink color indicates pH ≥8. The plate diameter was 9 cm.

Fig. 5.

Swarming of Paenibacillus sp. and P. mirabilis colonies inoculated on LBA with phenol red. (A) Representative images of Paenibacillus sp. colonies inoculated on 1.5% LBA plates with phenol red as pH indicator. The left image was taken at ∼18 h before swarming commenced colony; the right image was taken at ∼48 h. (B) P. mirabilis was inoculated on rich 1.5% LBA plates supplemented with 0.018 g/L phenol red. Images were taken at 4 h (Left) and 8 h (Right) after P. mirabilis inoculation. Plate diameter 9 cm. Yellow color indicates pH ≤7, pink color indicates pH ≥8.

Fig. 6.

Coinoculation of X. perforans and Paenibacillus sp. on custom-made LBA–LBAG swarm plates. (A) Schematic representation of the interaction between X. perforans (orange) and Paenibacillus sp. (light blue) demonstrating how X. perforans cells inoculated on LBA gain accesses to a glucose-rich area (LBAG) by exploiting the swarming of Paenibacillus sp. (B) Representative images of Paenibacillus sp. (PYH6) and X. perforans (XP) inoculated on custom-made LBA–LBAG. Left plate X. perforans inoculated on LBA (Bottom) and Paenibacillus sp. inoculated on LBAG (Top); right plate X. perforans inoculated on LBA (Bottom), and the top part contains LBAG without bacteria. Black horizontal lines indicate place of inoculation. Images taken at plates at 60 h after bacterial inoculation. The plate diameter was 9 cm.