Suicidal chemotaxis in bacteria

Abstract

Bacteria commonly live in communities on surfaces where steep gradients of antibiotics and other chemical compounds routinely occur. While many species of bacteria can move on surfaces, we know surprisingly little about how such antibiotic gradients affect cell motility. Here we study the behaviour of the opportunistic pathogen Pseudomonas aeruginosa in stable spatial gradients of a range of antibiotics by tracking thousands of cells in microfluidic devices as they form biofilms. Unexpectedly, these experiments reveal that individual bacteria use pili-based ('twitching') motility to actively navigate towards regions with higher antibiotic concentrations. Our analyses suggest that this biased migration is driven, at least in part, by a direct response to the antibiotics. Migrating cells can reach antibiotic concentrations hundreds of times higher than their minimum inhibitory concentration in a few hours and remain highly motile. However, isolating these cells - using fluid-walled microfluidic devices that can be reconfigured in situ - suggests that these bacteria are terminal and not able to reproduce. In spite of moving towards their death, we show that migrating cells are capable of entering a suicidal program to release bacteriocins that are used to kill other bacteria. Our work suggests that bacteria respond to antibiotics as if they come from a competing colony growing in the neighbourhood, inducing them to invade and attack. As a result, clinical antibiotics have the potential to serve as a bait that lures bacteria to their death.

Reconfigurable microfluidic circuits for isolating and retrieving cells of interest

“…For these experiments, one input contains ciprofloxacin at 100 times the concentration required to inhibit cell growth, known as the minimum inhibitory concentration (MIC, 10 µg/ml) [10]. Consequently, all cells initially exposed to high ciprofloxacin concentrations either fail to grow and/or are lost from the substrate ( Fig.…”

“…After ∼24 h, many cells have undergone chemotaxis towards ciprofloxacin and so we next isolate this migrating sub-population from those establishing a biofilm in antibiotic-free regions of the circuit. As we have seen, this can only be achieved with continuous flow [10]. Therefore, we use the 3-D printed adaptors discussed above to attach a PTFE stylus to the microscope condenser ( Fig.…”

“…The wild-type P. aeruginosa strain used throughout this study was PAO1 (Kolter collection ZK2019) expressing yellow fluorescent protein (YFP) from a constitutive promoter [25]. As an in-experiment control, we used a PAO1 strain harbouring an in-frame deletion of PilG ( ΔpilG , [10]). All bacteria were grown from frozen stocks overnight in LB medium (ThermoFisher; 37°C; 250 rpm) and sub-cultured (1:30 dilution) in tryptone broth (TB, 10 g L -1 Bacto Tryptone; ThermoFisher) for 2.5 h (37°C; 250 rpm).…”

“…Isolated bacteria remained in fluid chambers containing ciprofloxacin at 3 times the MIC for 2 h in order to kill any antibiotic-naïve cells [10]. Cells were monitored by imaging (0.1 frame min -1 ) and whilst most cells remained surface-attached, some detached and were lost from the field of view.…”

“…As proofs of concept, we isolate and extract sub-populations from two very different cell types – murine macrophages and early-stage biofilms of Pseudomonas aeruginosa – as cells migrate up flow-generated gradients of two very different chemoattractants (i.e., macrophages towards complement component 5a (C5a), [8], [9], and the bacteria – counter-intuitively – towards the lethal antibiotic, ciprofloxacin [10]). We first show that fluid-walled circuits can be reconfigured using the printer to isolate a microscopically-targeted population of migrating macrophages.…”