Search tips
Search criteria 


Logo of mbioJournal InfoAuthorsReviewersBoard of EditorsJournals ASM.orgmBiomBio Article
mBio. 2013 Jul-Aug; 4(4): e00374-13.
Published online 2013 July 23. doi:  10.1128/mBio.00374-13
PMCID: PMC3735182

Two Independent Pathways for Self-Recognition in Proteus mirabilis Are Linked by Type VI-Dependent Export


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.


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.


The ability to differentiate self from nonself is a behavior observed throughout biology, from animals to single-celled organisms. Self-recognition has been hypothesized to be a cornerstone aspect of territorial behavior, i.e., a preference for kin and aggressiveness toward nonkin (1). Multiple implementations of self-recognition capability have been described in a growing set of bacteria, including Proteus mirabilis, Pseudomonas aeruginosa, Vibrio cholerae, Escherichia coli, Paenibacillus dendritiformis, and Myxococcus xanthus (29). In both P. aeruginosa and V. cholerae, the type VI secretion (T6S) system mediates the exchange of toxins between neighboring cells; kin selection occurs through the strain-specific expression of an antitoxin to the T6S-mediated toxins, which are usually strain- or species-specific effector molecules (35). Similarly, contact-dependent inhibition (CDI) in E. coli and Dickeya dadantii is achieved through the direct exchange of toxin-encoding peptides that are selectively targeted to inhibit growth of nonisogenic strains; these nonisogenic strains do not express the requisite neutralizing antitoxin (7, 1012). It has been proposed that these CDI toxins are linked to rhs (rearrangement hotspot) sequences in bacteria (11). While the molecular mechanisms of these systems are beginning to be described, the native environmental and physiological role for self-recognition in bacteria is poorly understood.

In the model system P. mirabilis, a Gram-negative bacterium and causative agent of urinary tract infections, self-recognition is necessary for territorial behavior. Migrating populations, or swarms, of independent P. mirabilis isolates can recognize each other as foreign and establish a macroscopically visible boundary (of up to 3 mm) where they meet. In contrast, genetically identical swarms merge, forming a single, larger swarm (2). This behavior indicates that P. mirabilis populations are capable of distinguishing self from nonself. P. mirabilis infections have been described as clonal and as a consequence of infection by the host’s endogenous strain (13, 14).

We previously reported that P. mirabilis populations with mutations in the ids operon, consisting of idsABCDEF, do not merge with the wild-type parent, indicating a loss of the ability to correctly recognize self (15, 16). More specifically, we found that idsD and idsE encode strain-specific self-identity determinants in P. mirabilis. Strains in which either idsD or idsE is absent form a territorial boundary with an otherwise genetically identical parent strain, and this behavior is not rescued by expression of idsD and/or idsE alleles from a foreign strain (15). This differs from the Ids proteins IdsA, IdsB, IdsC, and IdsF, which we found do not confer strain-specific self-identity, as replacing their genes with alleles from a foreign strain does not alter boundary formation (15). The Ids proteins, however, are necessary but not sufficient for self-recognition and subsequent boundary formation in P. mirabilis.

To fully understand and model self-recognition behavior in bacteria, we need to know the core components and how they interact with one another. Indeed, the full set of proteins involved in self-recognition in P. mirabilis, as well as their cellular location and the interconnections between them, were previously unknown. Moreover, the role of the Ids proteins, and of self-recognition in general, in social behaviors outside boundary formation has yet to be examined. Here, we have characterized core molecular networks for self-recognition in one strain of P. mirabilis, as well as the interconnections between these proteins.


Self-recognition requires two gene clusters, tss and idr, in addition to the ids genes.

We sought to ascertain the full set of genes necessary for self-recognition by searching for mutants that display a territorial boundary formation phenotype different from that of the wild-type strain and/or an ids-deficient mutant strain. To this end, we generated a library of roughly 13,000 single-insertion transposon mutants in the wild-type strain BB2000, representing an approximately threefold coverage of its genome. Then, we screened each mutant from the library by swarming it against a mutant lacking the ids operon ([increment]ids) and against other mutants from the transposon library, which served as proxies for nonself and self populations, respectively. We isolated mutants that either merged with all strains or formed boundaries with the [increment]ids strain and each other (Fig. 1A). Seven mutants were pursued: five that merged with both the [increment]ids mutant and the wild-type parent (“all-merge”) and two that formed boundaries with both the [increment]ids strain and the wild-type (“no-merge”) (Fig. 1B). We had isolated an additional no-merge mutant in a previous self-recognition screen (15). The isolated mutant strains, like the wild-type parent and the [increment]ids strain, formed boundaries with the independent P. mirabilis wild-type strain HI4320 (Fig. 1B). The eight insertion sites represented by these recovered mutants map to two adjacent, divergently oriented gene clusters.

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 ...

The insertions in the all-merge mutants map to a single 17-gene cluster, tssA-Q, located from base pair 938,609 to 916,585 (Fig. 1C) (NCBI accession number BankIt1590180 BB2000 CP004022). The sequence of tssA-Q reveals similarities to genes encoding core components of the Vibrio cholerae type VI secretion (T6S) system, including the membrane protein genes icmF, dotU, and sciN, as well as the ATPase gene clpV (17). This is the sole locus containing these T6S proteins in the BB2000 genome. To confirm the phenotype associated with tssA-Q, we introduced tssN (icmF) and the three downstream genes, tssOPQ, into a low-copy-number plasmid where gene expression is controlled by the region directly upstream of tssA; we transformed this plasmid, pLW100, into a tssN-deficient mutant (tssN*) in which tssN is disrupted by a transposon insertion. The genes tssNOPQ were included on pLW100, because the tssN mutation likely disrupts expression of the downstream genes. The plasmid pLW100 complements the tssN mutation; the resultant strain forms a boundary with [increment]ids (Fig. 1B). We did not see complementation with a plasmid containing solely tssN using the same promoter region, suggesting that a disruption in tssO, tssP, or tssQ may also be responsible for the all-merge phenotype and that the upstream promoter is not contributing to the complementation phenotype (see the supplemental material). Therefore, we conclude that disruption of T6S function is responsible for the all-merge phenotype.

The no-merge mutants contain transposon insertions in three separate genes of a previously uncharacterized five-gene locus, located from base pair 940,506 to 949,474, that we name idr for “identity recognition” (Fig. 1D). The first gene, idrA, shares high sequence similarity with idsA (95%) and the T6S-related gene hcp, whereas the second gene, idrB, has some sequence similarity to idsB (55%) and the T6S-related gene vgrG (Fig. 1D). The idrB-deficient (idrB*) mutant strain in which idrB is disrupted by a transposon insertion served as the idr-deficient strain throughout our studies. The remaining genes, idrC, idrD, and idrE, are predicted to encode polypeptides of unknown function. The idrD gene contains rhs sequences. Some genes containing rhs sequences have been shown to encode antibacterial toxins (11).

We observe that the ids, idr, and tss gene clusters are all present in the genome of the independent strain HI4320 (18). The Ids proteins share more than 97% sequence identity among strains, except for IdsD and IdsE, which share 96% and 93% sequence identity, respectively (15). The polypeptides encoded by the tss locus are highly similar (over 97% sequence identity) between strains BB2000 and HI4320 (Fig. 1C). However, the idr locus differs in both nucleotide sequence and gene content between strains BB2000 and HI4320, suggesting that the idr locus encodes as-yet uncharacterized strain-specific factors necessary for self-recognition (Fig. 1D).

The ids, tss, and idr loci are each critical for competition on surfaces.

We next examined the role of each gene cluster in self-recognition and territorial behaviors. We predicted that self-recognition capability likely provides an increased ability to survive against other organisms. As self-recognition-dependent boundary formation in P. mirabilis is principally apparent on surfaces, we investigated whether the loss of self-recognition capability decreases a population’s ability to compete on surfaces. In equal initial ratios, we mixed cells of the parent BB2000, which is fully capable of self-recognition, with those of either the [increment]ids, tssN*, or idrB* mutant strains, all of which are deficient in one or more self-recognition protein. We placed each mixed population on a nutrient surface in a single spot from which the population migrated outward as a single swarm. Then, we analyzed for dominance by measuring whether the mixed population merged with a pure swarm of either BB2000 or an isolated swarm of the tested mutant strain. The parent BB2000 prevailed in virtually every mixed population (Fig. 2A).

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 ...

To determine how the parent strain achieves dominance, we sampled for the presence of the parent and mutant strains at discrete locations within the swarm of the mixed population. Notably, parent BB2000 cells migrated to the periphery of the swarm more rapidly than any of the mutant strains (Fig. 2B). None of the mutant strains have a motility defect, compared to BB2000, when migrating alone (Fig. 1). Therefore, loss of self-recognition capability diminishes a population’s relative rate of movement to, and dominance of, the leading edge of a swarm colony when it is growing with an otherwise genetically identical strain fully capable of self-recognition (Fig. 2B).

We next assessed how the BB2000 parent and mutant strains fared in competition with the independent wild-type P. mirabilis strain HI4320. In similar assays for dominance as described above, we mixed an equal ratio of HI4320 and BB2000 cells and then placed the mixed population onto a nutrient surface in a spot from which the cells migrated outward as a single swarm. We measured for dominance by examining whether the swarm of the mixed population formed a boundary with an adjacent pure swarm of either HI4320 or BB2000. Most mixtures of HI4320 and BB2000 yielded boundaries with the neighboring HI4320 swarm but merged with the BB2000 swarm, indicating that BB2000 cells dominated at the leading edges of mixed populations (Fig. 2C). Likewise, mixtures of HI4320 and the [increment]ids mutant strain primarily formed boundaries with a pure HI4320 swarm but merged with a pure [increment]ids swarm, indicating that the [increment]ids strain was dominant in these mixed population and that the ids genes are not needed for competition between strains (Fig. 2C). In contrast, mixtures of HI4320 with either the tssN* or idrB* mutant strain primarily yielded swarms that merged with a pure HI4320 swarm but formed boundaries with pure swarms of the tssN* or idrB* mutant strain, respectively, indicating that HI4320 dominated in these mixed populations (Fig. 2C). The presence of the Idr and T6S proteins, but not the Ids proteins, is therefore advantageous in competitions against the independent strain HI4320. Further, the Idr and Ids proteins have discrete roles in competitions; while Ids and Idr proteins are necessary for competitions with the parent strain, only Idr proteins are involved in competition with foreign strains.

Type VI secretion is required for export of components involved in self-recognition.

The phenotypes observed during the competition assays suggest a dynamic connection between these three gene clusters that together contribute to self-recognition and territorial behaviors. Since T6S is needed for the export of proteins in other bacterial systems, we predicted that self-recognition products in P. mirabilis are likely exported from the cell via this system. As such, we examined the secretion profiles of the wild-type, [increment]ids, idrB*, and tssN* strains for proteins involved in self-recognition using liquid chromatography-tandem mass spectrometry (LC-MS/MS). We detected the self-identity determinant protein IdsD, as well as IdsA and IdsB, in the extracellular fraction of the wild-type P. mirabilis strain BB2000 but not in that of the [increment]ids mutant strain (Fig. 3A). None of the remaining Ids proteins were present in any of the extracellular fractions by LC-MS/MS analysis. The newly identified IdrA and IdrB proteins were also present in the extracellular fractions for both the wild-type and the [increment]ids mutant strains, indicating that export of the Idr proteins is independent of the Ids proteins (Fig. 3A). Conversely, IdsA, IdsB, and IdsD, as well as IdrA, were detected by LC-MS/MS analysis in supernatant isolated from the idrB* mutant strain, providing further support for the idea that the Ids and Idr proteins likely function independently in export from the cell (Fig. 3A).

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 [increment]ids mutant strains. For BB2000 and the idrB* strains, an additional 6 unique (74 ...

We readily observed IdsA and IdrA in the extracellular fraction of the wild-type strain as discrete bands in a Coomassie blue-stained protein gel. We excised these bands and confirmed by LC-MS/MS that they were indeed IdsA and IdrA (Fig. 3B). Only a single polypeptide band corresponding to the molecular weight of IdrA was present in the [increment]ids extracellular fraction, confirming the LC-MS/MS results (Fig. 3B). In contrast, neither IdsA nor IdrA was visible in the extracellular fraction of the tssN* mutant strain (Fig. 3B). Indeed, neither Ids nor Idr proteins were detected above background levels in the supernatant of the tssN* mutant strain analyzed by LC-MS/MS (see the supplemental material).

To further confirm the LC-MS/MS data, we attached a FLAG epitope to the C terminus of IdsA in a low-copy-number plasmid containing the entire ids operon with its native promoter, resulting in plasmid pLW101, and then introduced this plasmid construct into the [increment]ids strain, into wild-type BB2000, and into the tssN* mutant strain. In these plasmid-carrying strains, IdsA-FLAG was absent in the extracellular fraction of the tssN* mutant strain but was present in that of the [increment]ids and wild-type strains, as detected by Western blotting (Fig. 3C). Of note, IdsA-FLAG was detected not only in the supernatant but also on the surface of intact BB2000 carrying pLW101 cells (see the supplemental material). The lack of extracellular IdsA-FLAG in the tssN* mutant strain was not due to reduced production, as IdsA-FLAG was present at equivalent levels in the whole-cell extracts of all plasmid-carrying cells (Fig. 3C).

We performed a similar Western blot analysis using custom-raised antibodies to IdsB. IdsB was detected at equivalent levels in the whole-cell extracts of the wild-type, idrB*, and tssN* strains but was detected only in the supernatant fractions of the wild-type and idrB* strains (Fig. 3D). Unfortunately, we were unable to directly localize epitope-tagged variants of IdsD in vivo, perhaps due to low expression of IdsD and/or to steric hindrance of the epitope by a putative identity complex. However, based on the LC-MS/MS and Western blot analyses, export of the self-identity determinant protein IdsD, as well as the non-identity determinant proteins IdsA, IdsB, IdrA, and IdrB, requires a functional T6S system. Moreover, export of the Ids components is independent of the Idr components and vice versa.


Here we report the discovery of two additional gene clusters that together with the ids operon comprise a network of self-recognition genes (Fig. 4A). One locus, idr, encodes proteins necessary for merging with the parent BB2000 strain, while the other locus, tss, encodes a type VI secretion system that mediates export of Ids and Idr proteins. Significantly, we found that multiple Ids and Idr proteins are exported from the cell, including the self-identity determinant protein, IdsD.

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 ...

The Idr and Ids proteins represent two separate mechanisms for self-recognition. Export of the Idr proteins is independent of the ids gene cluster, and likewise, export of the Ids proteins is not dependent on the idr gene cluster. Moreover, strains with mutations in the ids or idr genes have different phenotypes in intraspecies competitions. The Ids proteins, which are needed only for competition with the parent strain, encode strain-specific self-identity determinants (15). Interestingly, the idrD gene contains rhs sequences, which are commonly found in bacterial species. Recent research has implicated that these rhs-encoded proteins, as well as proteins involved in contact-dependent inhibition, such as CdiA in E. coli, encode toxin elements in the C-terminal domain (6, 11, 19). However, the rhs-containing sequences may also encode adhesion molecules, because they share some sequence similarity to YD-repeat-containing teneurin proteins (11). Either of these proposed idrD functions could explain why the idr genes are required for increased competition (and/or population migration) against foreign strains.

Indeed, we demonstrate that the self-recognition capability of P. mirabilis provides a competitive advantage for the population specifically on surfaces. We did not observe similar advantages when the wild type was grown with self-recognition mutant strains in liquid (see the supplemental material). Growth on surfaces induces a broad developmental change in P. mirabilis where increased cell-cell contact yields increased population-wide coordination that is integral to migration and outward expansion of the swarm (reviewed in references 20 and 21). Perhaps the behavior of self-recognition is most beneficial in environments where social interactions are more frequent and thus potentially more impactful.

Our research supports a model in which P. mirabilis self-recognition involves the display of self-identity proteins that are likely interpreted via a direct physical interaction with other cells; this communication then yields a self-versus-nonself assessment that guides whether boundaries are formed between populations (Fig. 4B). Some self-identity proteins are likely displayed on or near the cell surface, as physical contact between cells is required for boundary formation (22). This extracellular exposure may serve to communicate a cell’s identity represented by the self-identity determinant molecules IdsD and IdsE during interactions with neighboring cells (15, 16). Indeed, Ids and Idr proteins are transported out of the cell via the T6S system and either are transported into the neighboring cell or localize on the cell surface (see the supplemental material). However, we have not yet found evidence for the transfer of self-recognition proteins into a neighboring cell.

We propose that boundary formation can result from the actions of a single population, which has queried on a cell-cell level whether the neighboring cell is self. For each population of P. mirabilis strain BB2000, “self” is defined by the combined actions of the Ids and Idr pathways. Self-recognition occurs when both the expected cognate Ids and Idr proteins are present in (or on) the neighboring cell, ultimately resulting in merging with the neighboring swarm (Fig. 4B). In contrast, we predict that boundary formation results from the absence of the cognate Ids and Idr self determinants in the neighboring cell (Fig. 4B). Both the Ids and Idr proteins likely initiate downstream signaling pathways that are altered depending on whether the interactions are with cognate or noncognate Ids and Idr proteins, respectively.

This two-part network for self-recognition appears analogous to aspects of the innate immune system and indeed has many parallels to the immune surveillance of natural killer (NK) cells. In current models for NK cell activity, the presence of self cells (i.e., of one’s own organism) is conveyed by the combined detection of two surface receptors (an activation receptor ligand and class I major histocompatibility complex [MHC]), resulting in no killing of the self cell. In contrast, the absence of either receptor leads to the NK cell’s determination of an absence of self and the subsequent killing of the nonself (or receptor-deficient) cell, as reviewed in reference 23. Intriguingly, these results in P. mirabilis further support the idea that cellular self-recognition is a behavior shared among organisms at many levels of biological complexity.

While the capability for self-recognition is broadly found, it remains unclear why and how bacteria utilize this ability. In P. mirabilis, self-recognition is necessary for territorial expansion when one population interacts with competing nonisogenic populations. Recently, other research groups have shown that type VI secretion systems confer a fitness advantage in interbacterial and interkingdom competitions, likely through transport of small peptides, but their role in intraspecies interactions is only beginning to emerge (35, 2431). Our demonstration that a T6S system functions directly in self-recognition-dependent territoriality expands the set of known applications for this widely conserved export machinery. We still need to explore the mechanisms of T6S in P. mirabilis and its relative functional capabilities compared to T6S systems described in other bacteria. Importantly, we still need to understand the dynamics of Idr and Ids protein-protein interactions within and between cells. Indeed, the Ids, Idr, and T6S proteins together form a mechanistic foundation for examining the basic biological phenomena of territoriality and self-recognition in a bacterial model system.


Bacterial strains and media.

All strains are listed in Table 1. Escherichia coli strains were maintained on LB agar and Proteus mirabilis strains were maintained on L swarm minus (LSW) agar (32). P. mirabilis was grown on CM55 blood agar base agar (Remel Inc., Lenexa, KS) for swarm colony growth. For broth cultures, all strains were grown in LB broth under aerobic conditions at 37°C. Antibiotics were used at the following concentrations: 15 µg/ml tetracycline (Tet); 100 µg/ml rifampin (Rif); 50 µg/ml kanamycin (Kn); 35 µg/ml chloramphenicol (Cm) for E. coli and 50 µg/ml Cm for P. mirabilis. All media contained antibiotics appropriate for selection or maintenance of plasmids. Strains carrying plasmids are so denoted with a “c.”

Bacterial strains and plasmids

Plasmid construction.

The tssN (icmF) complementation plasmid pLW100 carries the last four genes of the tss gene cluster (tssN through tssQ) under the transcriptional control of the proposed promoter contained in the region immediately upstream of the tss gene cluster. This plasmid was constructed in two steps: the 1,200 bp upstream of tssA was amplified by PCR from the BB2000 genome and inserted into the pBBR1-NheI (15) plasmid using the Infusion HD system (Clontech Laboratories, Mountain View, CA); tssN, tssO, tssP, and tssQ were then PCR amplified and inserted into the above plasmid (with Infusion HD), resulting in pLW100. To construct pLW101, which is the plasmid encoding IdsA-FLAG, a FLAG epitope (N-DYKDDDDK-C) was inserted immediately before the idsA stop codon in the pidsBB plasmid (15) using QuikChange site-directed mutagenesis (Agilent Technologies, Santa Clara, CA). Plasmids were propagated in E. coli XL10-Gold and then conjugated into P. mirabilis via mating with E. coli S17-1λpir carrying the target plasmid as described earlier (15).

Swarm boundary assay.

Cells were inoculated from overnight or mid-logarithmic-phase cultures with an inoculation needle onto the surface of CM55 agar. Swarm plates were incubated for 18 to 24 h at 37°C and screened by eye for the presence or absence of a visible boundary.

Transposon library construction and screen.

A library of P. mirabilis transposon insertion mutants was generated by mating P. mirabilis strain BB2000 with E. coli strain SM10λpir carrying pUTmini-Tn5-Cm as described previously (15). Matings were performed on LSW agar plates in the absence of selection for 8 to 16 h, and the resultant populations spread on 22-cm by 22-cm LSW+Tet+Cm agar trays (Genetics/Molecular Devices, United Kingdom) and incubated at 37°C for 24 to 36 h. Colonies were picked using a robotic colony picker (Qbot/Molecular Devices, Genetix, United Kingdom) and arrayed into 96-well master plates. In total, 12,960 transposon insertion mutants were arrayed from 96-well master plates onto swarm agar trays (Nunc OmniTray; Nalge Nunc International, Rochester, NY) using the gridding head of the robotic colony picker. The swarm agar trays were arrayed in one of two patterns: (i) the mutants alone were arrayed and screened for boundary formation between independent mutants, or (ii) the mutants were arrayed alternating with the [increment]ids mutant. After 24 and 48 h, each mutant was scored for boundary formation or merger with neighboring colonies. From the initial high-throughput, robotic screen of ~26,000 interactions, 192 potential mutants were selected for further retesting. Potential mutants were then examined in individual swarm boundary assays against the BB2000 parent, the [increment]ids mutant strain, and wild-type P. mirabilis strain HI4320 (33) to confirm the phenotype. Of those tested, 21 mutant strains were confirmed. Eight mutants contained disruptions in eight different loci, six strains contained mutations in the ids locus, and the remaining mutants contained disruptions in the tss (five) and idr (two) loci.

Mapping the transposon insertion sites.

Arbitrary PCR was used to map the sites of the mini-Tn5-Cm transposon insertions as described previously (3436). Briefly, genomic DNA was isolated from each transposon mutant of interest by phenol-chloroform extraction, and the transposon insertion sites were amplified using Vent polymerase (New England Biolabs, Ipswich, MA) and primers Tn5Ext and ARB6 for the first round and then oNS054 (5′ TTCACACAGGAAACAGCTATGACCGCATTAAAATCTAGCGAGG 3′) and ARB2 for the second round. Samples were treated with ExoSAP-IT (New England Biolabs, Ipswich, MA) between rounds and prior to sequencing. Sanger sequencing was performed using primer oNS056 (5′ TTCACACAGGAAACAGCTATGACC 3′) by Genewiz, Inc. (South Plainfield, NJ). Results were mapped against the HI4320 (18) and the BB2000 (NCBI accession number BankIt1590180 BB2000 CP004022) genome sequences using ViroBLAST (37).

Sequence alignments.

The predicted polypeptide sequences for the ids, tss, and idr gene clusters were compared between independent P. mirabilis strains BB2000 and HI4320. Percent identities for the entire peptide were calculated in pairwise comparisons using ClustalW2 (38, 39).

Surface competitions.

To observe the spatial distribution of coswarming P. mirabilis strains over time, BB2000 c. pBBR2-GFP (16) was competed against BB2000 or the [increment]ids, tssN*, or idrB* mutant. Overnight cultures were normalized to an optical density at 600 nm (OD600) of 0.1. Competing strains were mixed together in a 1:1 ratio, and 0.5 µl of each coculture was spotted onto the center of a CM55 agar plate. 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). Select swarms (i.e., BB2000 versus BB2000) were also plated nonselectively onto LSW agar. To determine which strain was dominant in surface competitions, overnight cultures, normalized to OD600 of 1.0, of BB2000 were mixed at a 1:1 ratio with cultures of the [increment]ids, tssN*, or idrB* strain. Mixed populations were inoculated onto CM55 agar using an inoculation needle; monocultures of the boundary indicator strains were inoculated approximately 1 cm away. After overnight incubation at 37°C, the presence or absence of boundaries between swarms was used to assess strain dominance at the leading edge. Dominance of BB2000 was assessed as the merger of mixed populations with a monoswarm of BB2000; dominance of the mutant was assessed as a merger with a monoswarm of itself or, in the case of the tssN* strain, the [increment]ids strain.

For competition between independently derived strains, the surface competition described above was repeated using the P. mirabilis wild-type strain HI4320 competed against BB2000 or the [increment]ids, tssN*, or idrB* strain, with the modifications that a monoculture of HI4320 was used as the indicator strain on each plate instead of BB2000 and dominance of HI4320 was assessed as the merger of mixed populations with the HI4320 monoswarm. For all assays, unclear boundaries were recorded as “neither.”

TCA precipitation.

Overnight cultures were diluted to an OD600 of 0.1 in fresh LB+Kn and grown at 37°C with shaking to an OD600 of 3.5 to 4.5. For whole-cell extracts, 1 ml of culture was centrifuged, and the pellet was resuspended in 100 µl SDS-PAGE sample buffer. For supernatant samples, 30 ml of culture was clarified by centrifugation and filter sterilized (0.22-µm filters). The filtered supernatant was treated with trichloroacetic acid (10% final concentration) and incubated on ice for 30 min. Precipitated proteins were collected by centrifugation, washed twice with prechilled 100% acetone, dried, and resuspended in 100 µl 2× SDS-PAGE sample buffer.

Protein sequence analysis by LC-MS/MS.

TCA-precipitated samples were analyzed by electrophoresis using 10% SDS-PAGE gels and then stained with Coomassie blue. Gel regions of interest were excised and analyzed by liquid chromatography-mass spectrometry/mass spectrometry (LC-MS/MS) by the Taplin Biological Mass Spectrometry Facility (Harvard Medical School, Boston, MA). The unique peptide results for the Ids and Idr proteins are presented in Tables S1 to S4.

Antiserum preparation.

Polyclonal antiserum against residues Cys713 to Ala723 of IdsB was raised in rabbits according to the standard protocols (Covance Research Products, Denver, CO).

Gel electrophoresis and Western blot.

Protein samples were separated by gel electrophoresis using 15% Tris-Tricine gels and were either stained with Coomassie blue or transferred to nitrocellulose for Western blot analysis. Western blot membranes were probed with primary antibody (either 1:5,000 mouse anti-FLAG [Sigma-Aldrich, Allentown, PA], 1:1,000 mouse anti-sigma70 [Thermo Scientific, West Palm Beach, FL], or 1:1,000 rabbit anti-IdsB antiserum) for 1 h and with secondary antibody (1:5,000 goat anti-mouse-HRP [KPL, Gaithersburg, MD] or 1:5,000 goat anti-rabbit-HRP [KPL, Gaithersburg, MD]) for 1 h and visualized using Immun-Star HRP luminol/enhancer (Bio-Rad, Hercules, CA) and the ChemiDoc XRS system (Bio-Rad, Hercules, CA).

Nucleotide sequence accession number.

The sequence for tssA-Q was deposited in NCBI under accession number BankIt1590180 BB2000 CP004022.


Figure S1

Boundary assays with tssN* mutant strains. (A) Diagrammatic representations of the boundary behavior patterns exhibited by the indicated strains, matched with the swarm plate to the right. (B) On this swarm agar plate, the tssN* mutant strain carrying plasmid pLW103, which encodes tssN expression alone, merged with both the [increment]ids strain (top) and the parent BB2000 (left), as did the tssN* mutant strain. The tssN-expressing plasmid, pLW103, carries the tssN gene under the transcriptional control of the proposed promoter contained in the region immediately upstream of the tss gene cluster. This plasmid was constructed as follows: pLMW100 was digested at NheI and XmaI sites to obtain the vector backbone; tssN and the 1,200-bp upstream region were then PCR amplified from pLMW100 using primers 5′ ATAGCTAGCTCGAGGCCTCTCATTACAGTAGCAATATTGAGAGAAGATT 3′ and 5′ ATACCCGGGCCCGCGGTTAATAAAGCGTTTCAGGTAAACGGA 3′; this product was then digested with NheI and XmaI and ligated with the vector backbone. The plasmid pLW103 was then transformed into E. coli S17λpir using standard protocols and subsequently conjugated into the tssN* mutant strain. Download

Figure S2 

There is no clear disadvantage to loss of self-recognition in liquid competitions. No significant difference was seen between the growth of BB2000 and that of each self-recognition mutant when the strains were grown together in liquid broth after 3 h (as measured by a two-tailed t test; P = 0.38, 0.39, and 0.14 for the [increment]ids, tssN*, and idrB* strains, respectively) and after 20 h (as measured by a two-tailed t test; P = 0.22, 0.89, and 0.31 for [increment]ids, tssN*, and idrB*, respectively). These results suggest that the self-recognition components do not confer a competitive advantage under liquid-growth conditions. The constitutive lacZ expression plasmid pLW102 carries the lacZ gene under the transcriptional control of the fla and lac promoters. This plasmid was constructed as follows: pKG105 (16) was digested at SacI and AgeI sites to obtain the vector backbone; gene lacZ was then PCR-amplified from pQF50 (M. A. Farinha and A. M. Kropinski, J. Bacteriol. 172:3496-3499, 1990) using primers 5′ CATGAGCTCATGAAAGGGAATTCACTGGCC 3′ and 5′ TAAACCGGTTTATTTTTGACACCAGACCAACTG 3′; after digestion with SacI and AgeI, this product was then ligated with the vector backbone. The ligation reaction was transformed into Stellar competent cells (Clontech Laboratories, Mountain View, CA). The plasmid pLW102 was then transformed into E. coli S17λpir using standard protocols and subsequently conjugated into wild-type BB2000. BB2000 c. pLW102 was competed against BB2000 c. pKG101, the [increment]ids c. pKG101 strain, the tssN* c. pKG101 strain, or the idrB* c. pKG101 strain. Overnight cultures were back-diluted to an OD600 of 0.1 in 3 ml LB+Kn and rotated at 37°C until late-log-phase growth. Cultures were then normalized to an OD600 of 3.5. The competing strains were mixed together in a 1:10 BB2000-to-mutant ratio, back-diluted to a 1:3 ratio in LB+Kn for a total volume of 1.5 ml, and incubated at 37°C while shaking at 225 rpm. After 3 h and 20 h of growth, cells were spotted on nonselective (LSW+Kn and 300 µg/ml X-Gal [5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside]) and, when applicable, selective (LSW+Cm) plates to measure for colony-forming units (CFU). The resultant ratio of CFU of each strain was compared to the initial inoculation ratio to calculate fold change. Statistical analysis was performed using GraphPad Prism 5 (GraphPad Software, Inc., La Jolla, CA). Download

Figure S3 

Dot blot analysis of IdsA-FLAG. IdsA-FLAG, expressed in trans in P. mirabilis cells, was found in the supernatant and on the cell surface of wild-type BB2000 and in severely decreased amounts in the tssN* mutant strain. The blot on the left was probed with an anti-FLAG antibody primary, while an anti-sigma -70 antibody was used to probe the blot on the right as a control for cell lysis. These results confirm that IdsA requires tssN for proper export out of the cell, as indicated by the LC-MS/MS and Western blot analyses. This result suggests that IdsA is normally located and exposed on the cell surface. Samples for whole-cell immunoblots were prepared as described previously (D. Newell, R. D. Monds, and G. A. O’Toole, Proc. Natl. Acad. Sci. U. S. A. 106:3461-3466, 2009). Briefly, cell cultures were grown to late-logarithmic phase. “Cell culture” samples were prepared by spotting cell culture directly onto nitrocellulose membrane in five 5-µl aliquots for a total of 25 µl in each spot. For “cell surface” samples, the loosely adhered portion of a cell pellet from 5 ml of cell culture was gently resuspended in 1 ml LB and spotted on the membrane as described above. For “lysed cell” samples, cells from 10 ml of culture were collected by centrifugation and resuspended in 1 ml lysis buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, and protease inhibitor cocktail [Roche, Indianapolis, IN]). Cells were lysed by vortexing with cell disruptor beads (Electron Microscopy Sciences, Hatfield, PA) and centrifuged to remove cell debris. The soluble fraction was spotted onto the membrane as described above. The dot immunoblot was then developed similarly to the above-described Western blots using one of two primary antibodies: mouse anti-FLAG (Sigma-Aldrich, Allentown, PA) or mouse anti-sigma70 (Pierce Biotechnology, Rockford, IL). Download

Table S1 

For the wild-type parent strain BB2000, listed are the unique peptide results for Ids and Idr proteins, acquired by LC-MS/MS. Peptide fragments that could correspond to either IdsA or IdrA are marked as such. The minimum detection cutoff recommended by the Taplin Biological Mass Spectrometry Facility was three unique peptides.

Table S2

For the [increment]ids mutant strain, listed are the unique peptide results for Ids and Idr proteins, acquired by LC-MS/MS. The minimum detection cutoff recommended by the Taplin Biological Mass Spectrometry Facility was three unique peptides

Table S3

For the idrB* mutant strain, listed are the peptide results for Ids and Idr proteins, acquired by LC-MS/MS. Peptide fragments that could correspond to either IdsA or IdrA are marked as such. The minimum detection cutoff recommended by the Taplin Biological Mass Spectrometry Facility was three unique peptides.

Table S4

For the tssN* mutant strain, listed are the unique peptide results for Ids and Idr proteins, acquired by LC-MS/MS. Peptide fragments that could correspond to either IdsA or IdrA are marked as such. Consistent with the dot blots, the tssN* mutant strain is greatly attenuated for IdsA or IdrA export, though not completely deficient. The tssN* mutant strain does not export any of the remaining Ids and Idr proteins. The minimum detection cutoff recommended by the Taplin Biological Mass Spectrometry Facility was three unique peptides, which IdrA/IdsA did not achieve for the tssN* sample.


We thank Andrew Fields, Clarissa Scholes, Mark Urbanowski, and Pete Greenberg for contributing experimental materials to this project. Richard Losick, Alexander Schier, Feng Chen, Melissa Marks, and Philippe Cluzel provided valued advice on the manuscript.

Harvard University and the David and Lucile Packard Foundation funded our research.

L.M.W., N.L.S., L.C., A.N.S., and K.A.G. designed and performed research, as well as analyzed data. L.M.W., L.C., A.N.S., and K.A.G. wrote the paper.


Citation Wenren LM, Sullivan NL, Cardarelli L, Septer AN, Gibbs KA. 2013. Two independent pathways for self-recognition in Proteus mirabilis are linked by type VI-dependent export. mBio 4(4):e00374-13. doi:10.1128/mBio.00374-13.


1. Wilson EO. 2000. Sociobiology: the new synthesis, 25th anniversary edition. Belknap Press, Cambridge, MA
2. Dienes L. 1946. Reproductive processes in Proteus cultures. Proc. Soc. Exp. Biol. Med. 63:265–270 [PubMed]
3. Hood RD, Singh P, Hsu F, Güvener T, Carl MA, Trinidad RR, Silverman JM, Ohlson BB, Hicks KG, Plemel RL, Li M, Schwarz S, Wang WY, Merz AJ, Goodlett DR, Mougous JD. 2010. A type VI secretion system of Pseudomonas aeruginosa targets a toxin to bacteria. Cell Host Microbe 7:25–37 [PMC free article] [PubMed]
4. LeRoux M, De Leon JA, Kuwada NJ, Russell AB, Pinto-Santini D, Hood RD, Agnello DM, Robertson SM, Wiggins PA, Mougous JD. 2012. Quantitative single-cell characterization of bacterial interactions reveals type VI secretion is a double-edged sword. Proc. Natl. Acad. Sci. U. S. A. 109:19804–19809 [PubMed]
5. Basler M, Ho BT, Mekalanos JJ. 2013. Tit-for-tat: type VI secretion system counterattack during bacterial cell-cell interactions. Cell 152:884–894 [PMC free article] [PubMed]
6. Aoki SK, Diner EJ, de Roodenbeke CT, Burgess BR, Poole SJ, Braaten BA, Jones AM, Webb JS, Hayes CS, Cotter PA, Low DA. 2010. A widespread family of polymorphic contact-dependent toxin delivery systems in bacteria. Nature 468:439–442 [PMC free article] [PubMed]
7. Aoki SK, Pamma R, Hernday AD, Bickham JE, Braaten BA, Low DA. 2005. Contact-dependent inhibition of growth in Escherichia coli. Science 309:1245–1248 [PubMed]
8. Be'er A, Zhang HP, Florin EL, Payne SM, Ben-Jacob E, Swinney HL. 2009. Deadly competition between sibling bacterial colonies. Proc. Natl. Acad. Sci. U. S. A. 106:428–433 [PubMed]
9. Pathak DT, Wei X, Bucuvalas A, Haft DH, Gerloff DL, Wall D. 2012. Cell contact-dependent outer membrane exchange in myxobacteria: genetic determinants and mechanism. PLoS Genet. 8:e1002626.10.1371/journal.pgen.1002626 [PMC free article] [PubMed] [Cross Ref]
10. Aoki SK, Webb JS, Braaten BA, Low DA. 2009. Contact-dependent growth inhibition causes reversible metabolic downregulation in Escherichia coli. J. Bacteriol. 191:1777–1786 [PMC free article] [PubMed]
11. Koskiniemi S, Lamoureux JG, Nikolakakis KC, T’Kint de Roodenbeke C, Kaplan MD, Low DA, Hayes CS. 2013. Rhs proteins from diverse bacteria mediate intercellular competition. Proc. Natl. Acad. Sci. U. S. A. 110:7032–7037 [PubMed]
12. Webb JS, Nikolakakis KC, Willett JL, Aoki SK, Hayes CS, Low DA. 2013. Delivery of CdiA nuclease toxins into target cells during contact-dependent growth inhibition. PLoS One 8:e57609 [PMC free article] [PubMed]
13. Mathur S, Sabbuba NA, Suller MT, Stickler DJ, Feneley RC. 2005. Genotyping of urinary and fecal Proteus mirabilis isolates from individuals with long-term urinary catheters. Eur. J. Clin. Microbiol. Infect. Dis. 24:643–644 [PubMed]
14. Sabbuba NA, Stickler DJ, Mahenthiralingam E, Painter DJ, Parkin J, Feneley RC. 2004. Genotyping demonstrates that the strains of Proteus mirabilis from bladder stones and catheter encrustations of patients undergoing long-term bladder catheterization are identical. J. Urol. 171:1925–1928 [PubMed]
15. Gibbs KA, Urbanowski ML, Greenberg EP. 2008. Genetic determinants of self identity and social recognition in bacteria. Science 321:256–259 [PMC free article] [PubMed]
16. Gibbs KA, Wenren LM, Greenberg EP. 2011. Identity gene expression in Proteus mirabilis. J. Bacteriol. 193:3286–3292 [PMC free article] [PubMed]
17. Heidelberg JF, Eisen JA, Nelson WC, Clayton RA, Gwinn ML, Dodson RJ, Haft DH, Hickey EK, Peterson JD, Umayam L, Gill SR, Nelson KE, Read TD, Tettelin H, Richardson D, Ermolaeva MD, Vamathevan J, Bass S, Qin H, Dragoi I, Sellers P, McDonald L, Utterback T, Fleishmann RD, Nierman WC, White O, Salzberg SL, Smith HO, Colwell RR, Mekalanos JJ, Venter JC, Fraser CM. 2000. DNA sequence of both chromosomes of the cholera pathogen Vibrio cholerae. Nature 406:477–483 [PubMed]
18. Pearson MM, Sebaihia M, Churcher C, Quail MA, Seshasayee AS, Luscombe NM, Abdellah Z, Arrosmith C, Atkin B, Chillingworth T, Hauser H, Jagels K, Moule S, Mungall K, Norbertczak H, Rabbinowitsch E, Walker D, Whithead S, Thomson NR, Rather PN, Parkhill J, Mobley HL. 2008. Complete genome sequence of uropathogenic Proteus mirabilis, a master of both adherence and motility. J. Bacteriol. 190:4027–4037 [PMC free article] [PubMed]
19. Poole SJ, Diner EJ, Aoki SK, Braaten BA, t’Kint de Roodenbeke C, Low DA, Hayes CS. 2011. Identification of functional toxin/immunity genes linked to contact-dependent growth inhibition (CDI) and rearrangement hotspot (Rhs) systems. PLoS Genet. 7:e1002217.10.1371/journal.pgen.1002217 [PMC free article] [PubMed] [Cross Ref]
20. Mobley HL, Belas R. 1995. Swarming and pathogenicity of Proteus mirabilis in the urinary tract. Trends Microbiol. 3:280–284 [PubMed]
21. Manos J, Belas R. 2006. The genera Proteus, Providencia, and Morganella. Prokaryotes 6:245–269
22. Budding AE, Ingham CJ, Bitter W, Vandenbroucke-Grauls CM, Schneeberger PM. 2009. The dienes phenomenon: competition and territoriality in swarming Proteus mirabilis. J. Bacteriol. 191:3892–3900 [PMC free article] [PubMed]
23. Lanier LL. 2005. NK cell recognition. Annu. Rev. Immunol. 23:225–274 [PubMed]
24. Mougous JD, Cuff ME, Raunser S, Shen A, Zhou M, Gifford CA, Goodman AL, Joachimiak G, Ordoñez CL, Lory S, Walz T, Joachimiak A, Mekalanos JJ. 2006. A virulence locus of Pseudomonas aeruginosa encodes a protein secretion apparatus. Science 312:1526–1530 [PMC free article] [PubMed]
25. Pukatzki S, Ma AT, Revel AT, Sturtevant D, Mekalanos JJ. 2007. Type VI secretion system translocates a phage tail spike-like protein into target cells where it cross-links actin. Proc. Natl. Acad. Sci. U. S. A. 104:15508–15513 [PubMed]
26. Russell AB, Hood RD, Bui NK, LeRoux M, Vollmer W, Mougous JD. 2011. Type VI secretion delivers bacteriolytic effectors to target cells. Nature 475:343–347 [PMC free article] [PubMed]
27. Schell MA, Ulrich RL, Ribot WJ, Brueggemann EE, Hines HB, Chen D, Lipscomb L, Kim HS, Mrázek J, Nierman WC, Deshazer D. 2007. Type VI secretion is a major virulence determinant in Burkholderia mallei. Mol. Microbiol. 64:1466–1485 [PubMed]
28. Basler M, Mekalanos JJ. 2012. Type 6 secretion dynamics within and between bacterial cells. Science 337:815 [PMC free article] [PubMed]
29. Unterweger D, Kitaoka M, Miyata ST, Bachmann V, Brooks TM, Moloney J, Sosa O, Silva D, Duran-Gonzalez J, Provenzano D, Pukatzki S. 2012. Constitutive type VI secretion system expression gives Vibrio cholerae intra- and interspecific competitive advantages. PLoS One 7:e48320 [PMC free article] [PubMed]
30. MacIntyre DL, Miyata ST, Kitaoka M, Pukatzki S. 2010. The Vibrio cholerae type VI secretion system displays antimicrobial properties. Proc. Natl. Acad. Sci. U. S. A. 107:19520–19524 [PubMed]
31. Pukatzki S, Ma AT, Sturtevant D, Krastins B, Sarracino D, Nelson WC, Heidelberg JF, Mekalanos JJ. 2006. Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system. Proc. Natl. Acad. Sci. U. S. A. 103:1528–1533 [PubMed]
32. Belas R, Erskine D, Flaherty D. 1991. Transposon mutagenesis in Proteus mirabilis. J. Bacteriol. 173:6289–6293 [PMC free article] [PubMed]
33. Jones BD, Mobley HL. 1988. Proteus mirabilis urease: genetic organization, regulation, and expression of structural genes. J. Bacteriol. 170:3342–3349 [PMC free article] [PubMed]
34. Caetano-Anollés G. 1993. Amplifying DNA with arbitrary oligonucleotide primers. PCR Methods Appl. 3:85–94 [PubMed]
35. O’Toole GA, Kolter R. 1998. Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signalling pathways: a genetic analysis. Mol. Microbiol. 28:449–461 [PubMed]
36. O’Toole GA, Pratt LA, Watnick PI, Newman DK, Weaver VB, Kolter R. 1999. Genetic approaches to study of biofilms. Methods Enzymol. 310:91–109 [PubMed]
37. Deng W, Nickle DC, Learn GH, Maust B, Mullins JI. 2007. ViroBLAST: a stand-alone BLAST web server for flexible queries of multiple databases and user’s datasets. Bioinformatics 23:2334–2336 [PubMed]
38. Goujon M, McWilliam H, Li W, Valentin F, Squizzato S, Paern J, Lopez R. 2010. A new bioinformatics analysis tools framework at EMBL-EBI. Nucleic Acids Res. 38:W695–W699 [PMC free article] [PubMed]
39. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ, Higgins DG. 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23:2947–2948 [PubMed]
40. Warren JW, Tenney JH, Hoopes JM, Muncie HL, Anthony WC. 1982. A prospective microbiologic study of bacteriuria in patients with chronic indwelling urethral catheters. J. Infect. Dis. 146:719–723 [PubMed]
41. de Lorenzo V, Cases I, Herrero M, Timmis KN. 1993. Early and late responses of TOL promoters to pathway inducers: identification of postexponential promoters in Pseudomonas putida with lacZ-tet bicistronic reporters. J. Bacteriol. 175:6902–6907 [PMC free article] [PubMed]

Articles from mBio are provided here courtesy of American Society for Microbiology (ASM)