Two independent pathways for self-recognition in Proteus mirabilis are linked by type VI-dependent export

Abstract

Swarming colonies of the bacterium Proteus mirabilis are capable of self-recognition and territorial behavior. Swarms of independent P. mirabilis isolates can recognize each other as foreign and establish a visible boundary where they meet; in contrast, genetically identical swarms merge. The ids genes, which encode self-identity proteins, are necessary but not sufficient for this territorial behavior. Here we have identified two new gene clusters: one (idr) encodes rhs-related products, and another (tss) encodes a putative type VI secretion (T6S) apparatus. The Ids and Idr proteins function independently of each other in extracellular transport and in territorial behaviors; however, these self-recognition systems are linked via this type VI secretion system. The T6S system is required for export of select Ids and Idr proteins. Our results provide a mechanistic and physiological basis for the fundamental behaviors of self-recognition and territoriality in a bacterial model system.

Importance: Our results support a model in which self-recognition in P. mirabilis is achieved by the combined action of two independent pathways linked by a shared machinery for export of encoded self-recognition elements. These proteins together form a mechanistic network for self-recognition that can serve as a foundation for examining the prevalent biological phenomena of territorial behaviors and self-recognition in a simple, bacterial model system.

FIG 1

The tss and idr genes are necessary for self-recognition. (A) Diagrammatic representations of the boundary behavior patterns exhibited by mutants isolated in the screen, matched with the swarm plates below. (B) Swarm agar plates inoculated with P. mirabilis strains exhibited the boundary formation behavior of two representative mutants isolated from the self-recognition screen: the tssN* mutant strain, which merged with all other BB2000-derived strains (left), the complemented tssN* mutant strain carrying plasmid pLW100, which formed a boundary with the ∆ids strain (center), and the idrB* mutant, which formed boundaries with all other strains (right). (C) Diagram of the putative type VI secretion (tss) gene locus with sites of the transposon insertions, as depicted by lollipops. (D) Diagram of the idr gene locus with sites of the transposon insertions, as depicted by lollipops. For panels C and D, the dark gray shading indicates 97% or higher identity for the predicted polypeptide sequences of the tss and idr genes between strains BB2000 and HI4320; otherwise, specific identities are provided underneath. The dashed box indicates the region of idrD that shares sequence similarity between strains BB2000 and HI4320. Slanted lines indicate a break in the genomic regions, corresponding to approximately 8 kb.

FIG 2

Competitions between P. mirabilis strains. (A) Competitions between mutant strains and the parent strain BB2000 on surfaces were initiated at a 1:1 ratio, and the mixed populations were permitted to swarm against either BB2000 or the mutant strain (n = 12). Population dominance was measured as the ability of the mixed swarm to merge with either BB2000, indicating BB2000 dominance, or the mutant strain, indicating dominance of the mutant strain. Unclear boundaries were classified as “neither.” (B) To observe the spatial distribution of coswarming P. mirabilis strains over time, BB2000 c. pKG101 (16) was competed against BB2000 or the ∆ids, tssN*, or idrB* mutant. Overnight cultures were normalized to an OD₆₀₀ of 0.1. Competing strains were mixed in a 1:1 ratio, and 0.5 µl of each coculture was spotted onto the center of a CM55 agar plate (n = 3). After incubation first at room temperature for 22 h and then at 37°C for 6 h, each swarm consisted of four swarm rings and was patched using a half-plate 48-prong device onto selective plates that could detect the marked BB2000 strain (LSW⁻ Kn) and, when applicable, the mutant strain (LSW⁻ Cm). Swarms of BB2000 versus BB2000 were also plated nonselectively onto LSW⁻ agar. Representative photographs of the swarm plates, after sampling for migration distance, are depicted in the lower portion. (C) Competitions between BB2000, the BB2000 mutant strains, and an independent strain, HI4320, were initiated at a 1:1 ratio, and the mixed populations were swarmed against either a BB2000 mutant strain or HI4320. The BB2000 mutant strain was defined as dominant when the mixed population merged with the BB2000 mutant strain, while HI4320 was dominant if the mixed population merged with the HI4320 swarm. n = 6 for the tssN* strain, 12 for the ∆ids and idrB* strains, and 18 for BB2000.

FIG 3

Proteins involved in self-recognition are exported outside the cell. (A) LC-MS/MS peptide hits for proteins in the culture supernatants of wild-type BB2000 and the ∆ids mutant strains. For BB2000 and the idrB* strains, an additional 6 unique (74 total) and 4 unique (28 total) peptides, respectively, could be assigned to either IdsA or IdrA, due to high similarity of the two proteins (indicated by a plus sign). (B) The secretion profiles of the wild-type, ∆ids, and tssN* strains were examined by gel electrophoresis followed by Coomassie blue staining. The identity of bands corresponding to IdsA and IdrA were confirmed by LC-MS/MS. (C) Western blots of extracellular secretions (left) and whole-cell extracts (right) isolated from strains expressing IdsA-FLAG. The ∆ids c. pids_BB strain was included as a negative control for the FLAG epitope. For ∆ids expressing IdsA-FLAG in trans, the FLAG epitope was engineered in frame into an expression plasmid that contains the entire ids operon under native control. (D) Western blots of extracellular secretions (left) and whole-cell extracts (right) isolated from the indicated strains using a polyclonal anti-IdsB antibody. The asterisks mark the size of the expected band.

FIG 4

Model for Ids and Idr functional roles in self-recognition. (A) Functional flow chart for the roles of the Ids, Idr, and T6S proteins in self-recognition and territorial behaviors. A subset of Ids and Idr proteins are primarily exported via a shared T6S system (tss) and are necessary for competition on surfaces with the parent strain. Idr proteins are also needed for competition against foreign strains. (B) Our proposed model for self-recognition predicts that the combined actions of interactions between cognate Ids and Idr proteins between two neighboring cells result in the determination that self is present, ultimately resulting in the merging of two swarms. Expression of the self-recognition components within the cells is sufficient, though in wild-type strains, some of these components are exported from the cell by a T6S system. In contrast, absence of one or more of the Ids and Idr self-recognition systems leads to the determination that self is absent and ultimately to boundary formation.