A Highly Polymorphic Receptor Governs Many Distinct Self-Recognition Types within the Myxococcales Order

Abstract

Self-recognition underlies sociality in many group-living organisms. In bacteria, cells use various strategies to recognize kin to form social groups and, in some cases, to transition into multicellular life. One strategy relies on a single genetic locus that encodes a variable phenotypic tag ("greenbeard") for recognizing other tag bearers. Previously, we discovered a polymorphic cell surface receptor called TraA that directs self-identification through homotypic interactions in the social bacterium Myxococcus xanthus Recognition by TraA leads to cellular resource sharing in a process called outer membrane exchange (OME). A second gene in the traA operon, traB, is also required for OME but is not involved in recognition. Our prior studies of TraA identified only six recognition groups among closely related M. xanthus isolates. Here we hypothesize that the number of traA polymorphisms and, consequently, the diversity of recognition in wild isolates are much greater. To test this hypothesis, we expand the scope of TraA characterization to the order Myxococcales From genomic sequences within the three suborders of Myxococcales, we identified 90 traA orthologs. Sequence analyses and functional characterization of traAB loci suggest that OME is well maintained among diverse myxobacterial taxonomic groups. Importantly, TraA orthologs are highly polymorphic within their variable domain, the region that confers selectivity in self-recognition. We experimentally defined 10 distinct recognition groups and, based on phylogenetic and experimental analyses, predicted >60 recognition groups among the 90 traA alleles. Taken together, our findings revealed a widespread greenbeard locus that mediates the diversity of self-recognition across the order MyxococcalesIMPORTANCE Many biological species distinguish self from nonself by using different mechanisms. Higher animals recognize close kin via complex processes that often involve the five senses, cognition, and learning, whereas some microbes achieve self-recognition simply through the activity of a single genetic locus. Here we describe a single locus, traA, in myxobacteria that governs cell-cell recognition within natural populations. We found that traA is widespread across the order Myxococcales TraA is highly polymorphic among diverse myxobacterial isolates, and such polymorphisms determine selectivity in self-recognition. Through bioinformatic and experimental analyses, we showed that traA governs many distinct recognition groups within Myxococcales This report provides an example in which a single locus influences social recognition across a wide phylogenetic range of natural populations.

Keywords: cell surface; cellular transfer; kin recognition; myxobacteria; polymorphism.

FIG 1

Identification of traAB orthologs across the order Myxococcales. (A) The monophyletic order Myxococcales currently contains three well-defined suborders, 10 families, and 31 genera. Genera that harbor or lack traAB orthologs are indicated, whereas genera with unavailable genomic sequences are shaded in gray. *, family affiliation not designated. (B) Schematics of multiple-sequence alignments of 78 TraA/B orthologs across the order Myxococcales. Amino acid residues are displayed using the Clustal X default scheme. The traAB operon is shown at the top. Domain architectures of TraA/B are labeled as follows: SS, type I signal sequence; CA, Cys-A region; Mx-CT, MYXO-CTERM (TIGR03901); TSP-3, thrombospondin type 3 (Pfam02412); CTD, carboxyl-terminal domain. Suborders of the aligned TraA/B sequences are indicated on the left.

FIG 2

Functional characterization of diverse traAB orthologs in Myxococcales. (A) Schematic of the stimulation assay that restores motility to certain mutants (recipients) by the transfer of wild-type motility proteins from nonmotile donors. A positive control (donor and recipient, with both harboring traAB^DK1622) and a negative control (donor lacking traAB) are shown. (B) Taxonomic origins of the traAB alleles analyzed. (C) Stimulation assays testing for functional complementation by heterologously expressing different traA alleles in isogenic ΔtraA M. xanthus strains. Each micrograph shows a mixture of donors and recipients bearing identical traA alleles. Allele names and their percentages of identical amino acids relative to traA^DK1622 (full length) are shown. The arrowhead highlights a small emergent flare at the edge of the colony. The inset shows an enlarged view of stimulated cells. (D) Stimulation assays testing for complementation of heterologous traB alleles in isogenic ΔtraB strains. Allele names and allele identities to traB^DK1622 are indicated. Stimulation efficacy was calculated by measuring the distance (d) of the movement of emergent flares from colony edges. Data representing the relative swarming distances determined in traA and traB complementation experiments are shown in panels E and F, respectively. Four experimental replicates were done. Error bars represent standard deviations from the means. Significant differences between the DK1622 group and other groups (i.e., functionally distant alleles) are indicated by asterisks (P < 0.05 [t test]). Strain details are given in Table 2 (see also Table S1). Scale bar, 200 µm.

FIG 3

Deletion analysis of the Cys-rich region within TraA^DK1622. (A) Domain architectures of TraA from a S. cellulosum strain (MSr7282) and a M. xanthus strain (DK1622). Cys-rich repeats are numerically labeled; note that S. cellulosum is missing C2 and C3. (B) Schematic depicting markerless in-frame deletion mutants (deleted residues listed) within the Cys-rich region of TraA^DK1622. The ability of these mutants to complement a ΔtraA mutant was assessed by a stimulation assay. Stimulation efficacy was judged as swarming distance of emergent flares from colony edges compared with a positive control (as described in the Fig. 2D and E legends). Wild-type stimulation, >75% efficacy; weak stimulation, ∼20% efficacy; very weak stimulation, <5% efficacy. See Table S1 for strain details.

FIG 4

Bioinformatic analyses of TraA orthologs from the suborder Cystobacterineae. (A) Maximum likelihood tree showing the relationships among diverse Cystobacterineae TraA orthologs (full length). Families and genera are indicated by outlines and colors, respectively. Each allele prefix indicates its taxonomic origin (see Table 1 for details). TraA orthologs that were functionally characterized are marked with black dots. The scale bar represents the number of amino acid substitutions per residue. Bootstrap values (%) are color coded. (B) Heat maps showing sequence conservation of TraA orthologs from the suborder Cystobacterineae and the genus Myxococcus. TraA domain organization is indicated.

FIG 5

TraA orthologs harboring related VDs display allele-specific recognition. (A) Pairwise amino acid sequence identity among three related VDs. (B) Maximum likelihood tree of the VDs of 59 Cystobacterineae TraA orthologs. Red stars highlight the three related VDs shown in panel A. Each allele prefix shows the taxonomic origin (see Table 1 for details). Different genera are indicated by colors. The dashed border indicates members of recognition group A. Bootstrap values are indicated with color. The scale bar represents the number of substitutions per amino acid site. (C) Stimulation assays showing specific recognition among these traA alleles. Strain details are given in Table 2 (see also Table S1). Scale bar, 200 µm.

FIG 6

Self-recognition among a wide range of myxobacteria is governed by the traA locus. (A) Stimulation assays showing specific recognition among 10 traA alleles (names are shown in bold on the left) in an isogenic set of strains. Black borders highlight 10 distinct recognition groups (groups A to J). Additional group members that have been functionally characterized (5, 10) are also listed on the left. The asterisk indicates a chimeric allele harboring VD^MCy8337 (see Fig. S3 for details). Scale bar, 200 µm. (B) Pairwise plot of (%) identity among VDs of the TraA orthologs tested in panel A (asterisks indicate self-recognition). (C) Same tree as that shown in Fig. 5B, where allele names are given. Shaded areas highlight distinct recognition groups, solid lines indicate characterized recognition groups (letters indicate group names), and dashed lines show predicted recognition groups. Groups are color coded according to the specificity-determining residue at position 205. The scale bar indicates the number of substitutions per amino acid residue.