The indole pulse: a new perspective on indole signalling in Escherichia coli

Abstract

Indole has diverse signalling roles, including modulation of biofilm formation, virulence and stress responses. Changes are induced by indole concentrations of 0.5-1.0 mM, similar to those found in the supernatant of Escherichia coli stationary phase culture. Here we describe an alternative mode of indole signalling that promotes the survival of E. coli cells during long-term stationary phase. A mutant that has lost the ability to produce indole demonstrates reduced survival under these conditions. Significantly, the addition of 1 mM indole to the culture supernatant is insufficient to restore long-term survival to the mutant. We provide evidence that the pertinent signal in this case is not 1 mM indole in the culture supernatant but a transient pulse of intra-cellular indole at the transition from exponential growth to stationary phase. During this pulse the cell-associated indole reaches a maximum of approximately 60 mM. We argue that this is sufficient to inhibit growth and division by an ionophore-based mechanism and causes the cells to enter stationary phase before resources are exhausted. The unused resources are used to repair and maintain cells during the extended period of starvation.

Figure 1. Indole is produced rapidly over a 30 minute period during the onset to stationary phase.

A culture of growing BW25113 cells was sampled regularly. The OD₆₀₀ was measured and the samples were centrifuged to remove cells, and the supernatant was assayed for indole using Kovacs assay. Data shown are the mean values ± standard deviation for three independent repeats.

Figure 2. Non indole producing mutants initially grow to a higher density and are more viable than indole producing counterparts, but are significantly less viable in the long term.

The density (OD₆₀₀: A) and viability (CFU ×10 ⁻⁶/ ml :B) of wild-type, mutant and mutant with 1 mM indole added were assessed over 10 days. Data shown are the mean values ± standard deviation for three independent repeats.

Figure 3. Apparent cell associated indole accumulates rapidly at the onset of stationary phase.

A culture of growing BW25113 cells was sampled regularly The OD₆₀₀ was measured and the samples were centrifuged to pellet cells. The resultant cell pellet was assayed for indole using Kovacs assay. Data shown are the mean values ± standard deviation for three independent repeats.

Figure 4. The relationship between apparent cell associated concentrations of indole and supernatant concentrations of indole.

Known concentrations of indole were added to the supernatant of stationary phase BW25113 Δ tnaA cells, the mixture was vortexed and centrifuged, and the resultant pellet assayed for indole using Kovacs assay. Dotted lines indicates the apparent cell associated indole at the peak of the pulse (60 mM) and the corresponding supernatant concentration and indicates the apparent cell associated indole when cells are in stationary phase and the corresponding supernatant concentration (0.75 mM). Data shown are the mean values ± standard deviation for at least three independent repeats.