Background

Many organisms have evolved adaptations to community living. One common mechanism present throughout nature is kin discrimination: preferential treatment of close kin and hindrance of non-kin (Smith, 1964). Well studied examples of bacterial kin discrimination include contact-dependent inhibition (CDI) (Aoki et al., 2009; Garcia et al., 2016) and toxin transfer through type IV and type VI secretion systems (Brunet et al., 2013; Souza et al., 2015). Typically, kin discrimination in bacteria is mediated by exchange of toxic effector proteins, such as nucleases or peptidoglycan degrading enzymes (Alteri and Mobley, 2016). Each strain expresses characteristic effector-immunity pairs, and cells must either express the cognate immunity protein or die. Proteus mirabilis, however, makes use of a non-lethal self recognition system, the Ids system, to perform kin discrimination during swarming (Gibbs et al., 2008). P. mirabilis displays a sophisticated swarming phenotype on solid surfaces. After surface contact, a subset of cells differentiate into elongated “swarmer” cells that collectively travel outwards. Periods of swarming are regularly followed by non-expansion “consolidation” periods where swarmer cells differentiate into shorter stationary cells. Successive rounds of swarming and consolidation lead to a characteristic bullseye pattern on surfaces (Rauprich et al., 1996). While swarming, different P. mirabilis isolates express different variants of identity factors IdsD and IdsE. IdsD is transferred between cells in a contact-dependent fashion. If cells do not express a cognate IdsE they do not die but are instead forced out of cooperative swarm behavior in a phenomenon termed territorial exclusion. This behavior is due to forced induction into a persister-like state mediated by the alarmone (p)ppGpp (Tipping and Gibbs, 2019).

In lethal kin discrimination systems, progression can be measured through several methods: dilution assays, microscopy to visualize killing (e.g., Basler et al., 2013), or live-dead staining (Virta et al., 1998). Quantifying the efficacy of a non-lethal system is more difficult. The relative swarm extent in situ can be measured through microscopy or replica plating (Wenren et al., 2013), but these methods provide limited information about the swarm state, and none at all about the states of individual cells that have taken part in non-lethal kin discrimination. Isolation of cells from the greater swarm mass is therefore critical to enable their physiological state to be examined in more detail.

We have developed the first method for separating two populations of cells from a mixed P. mirabilis swarm using fluorescence-activated cell sorting (FACS), so that the state of cells that have taken part in self recognition behavior can be measured. This efficient and reliable protocol allows downstream examination of cells using several different methods. Here, we describe how to measure growth rate of cells that have experienced territorial exclusion. We have also used this method to generate samples for RNA-Seq library preparation, and describe our methods for doing so briefly. Furthermore, sorted cells can potentially be used to generate samples for other metabolic or biochemical assays. The protocol as described is optimized for separating P. mirabilis cells and could be adapted to isolate cells from swarms of other species such as Escherichia coli, Vibrio cholerae, Pseudomonas aeruginosa, and Providencia stuartii.