The lethal cargo of Myxococcus xanthus outer membrane vesicles

Abstract

Myxococcus xanthus is a bacterial micro-predator known for hunting other microbes in a wolf pack-like manner. Outer membrane vesicles (OMVs) are produced in large quantities by M. xanthus and have a highly organized structure in the extracellular milieu, sometimes occurring in chains that link neighboring cells within a biofilm. OMVs may be a vehicle for mediating wolf pack activity by delivering hydrolytic enzymes and antibiotics aimed at killing prey microbes. Here, both the protein and small molecule cargo of the OMV and membrane fractions of M. xanthus were characterized and compared. Our analysis indicates a number of proteins that are OMV-specific or OMV-enriched, including several with putative hydrolytic function. Secondary metabolite profiling of OMVs identifies 16 molecules, many associated with antibiotic activities. Several hydrolytic enzyme homologs were identified, including the protein encoded by MXAN_3564 (mepA), an M36 protease homolog. Genetic disruption of mepA leads to a significant reduction in extracellular protease activity suggesting MepA is part of the long-predicted (yet to date undetermined) extracellular protease suite of M. xanthus.

Keywords: fruiting body; predation; predator-prey interactions; predatory rippling; secondary metabolism and enzymes.

Figure 1

OMV structures observed in situ and after purification. (A) Flow chart of experimental procedure with representative analytics (B–E). (B) Negative staining EM of an M. xanthus wild type DZ2 cell showing a mixture of extracellular structures associated with the cell including isolated vesicles and vesicle chains (Scale bar = 200 nm). After purification, these structures maintain their distinctive shape as (C) vesicles and (D) vesicle chains (Scale bars = 40 nm). (E) SDS-PAGE analysis showing consistency of protein profiles in OMV fractions used for protein MS: lane 1 standards, lane 2–4 OMV fractions from 3 independent cultures.

Figure 2

Cell fraction comparison and predicted function. (A) Conservative set of 548 (46 OMV only, 188 shared, 314 OM only) protein IDs from MS analysis of >10,000 peptides. Putative function binning of (B) 46 OMV-specific and (C) 188 OMV-enriched proteins.

Figure 3

Cell fraction vs. protein localization prediction comparison. Distribution of cellular localization predictions of the entire 7331 M. xanthus proteome, the 234 proteins detected in the OMV preparations, and the 502 proteins detected in the membrane preparations, with 188 shared between the two. Protein localization predictions show enrichment for unknown, extracellular and OM proteins in OMVs.

Figure 4

Secondary metabolites detected in OMV fractions. Representative structures from tandem MS analysis of M. xanthus OMV fractions are shown, based on comparison to known M. xanthus products. Six chemical families were identified, most of which have some antibiotic activity. Note that for the dkxanthene chemical family, nine distinct members were detected but only DKxanthene534 is shown. Further detail can be found in Table 3 and Supplement Table 2.

Figure 5

Physiological analysis of a mepA metalloprotease mutant. (A,B) 72 h incubation of a 10 μl aliquot of mepA cells induced to form fruiting bodies on CFL media. Normal fruiting body timing, distribution (A) and sporulation (B) were observed. (C,D) 48 h incubation of a 2 μl aliquot of mepA cells incubated adjacent to a 5 μl aliquot of E. coli on CFL media for observation of predatory behavior. Both clearing (C) and rippling (D) were observed. Electron microscopy of mepA cells shows the presence of vesicles with both (E) Scanning electron microscopy and (F) Transmission electron microscopy. (G) Resorufin-conjugated Casein was incubated with whole cell (solid marks) and vesicle preparations (open marks) of wild type DZ2 (black) and mepA (red) and proteolysis measured by the colorimetric release of soluble resorufin over 90 min, relative to control with no addition. mepA whole cells and OMV fractions show reduced proteolytic activity. (H) Analysis of prey cell lysis of mCherry expressing fluorescent E. coli cells by wild type DZ2 (black) and mepA (red) shows that both strains reduce fluorescent signal with similar kinetics, relative to the E. coli only control (blue). Error bars represent standard deviation from 3 independent assays, scale bars are 0.5 microns.