New Locus Important for Myxococcus Social Motility and Development

Phenotypic characteristics of the mutant HL-1.

The mutant HL-1 (Table 1) was assessed for motility phenotypes by standard methods (21). As shown in Fig. 1A to D, the swarming colony sizes of HL-1 were 90.4% and 84.2% smaller than those of HW-1 on hard and soft agar, respectively. On soft agar, the mutant produced small colonies with a rough, dentate swarm edge (Fig. 1C), in contrast to the large colonies with the translucent smooth lacy swarm edge of the wild-type strain (Fig. 1D). At their swarming edges on hard agar, HL-1 cells moved mainly as individuals, with a few in small groups (Fig. 1E), whereas HW-1 cells translocated over the agar surfaces either as individuals or in groups (Fig. 1F). The phenotypes of the mutants mimic those of some Myxococcus xanthus A⁺ S⁻ mutants deficient in extracellular polysaccharides (EPS), such as DK3468 (dsp) (22) and YZ603 (ΔdifE) (2), suggesting that the mutant HL-1 is defective in social motility.

• Open in new tab
• Download powerpoint

FIG. 1.

Gliding behavior and fruiting-body development of the M. fulvus wild-type strain HW-1 (B, D, F, and H) and its mutant HL-1 (A, C, E, and G). Colony expansions were done on CTT medium with 1.5% (A and B) or 0.3% (C and D) agar for 3 days. Colony edge morphologies were done on CTT hard (1.5%) agar (E and F). Development of fruiting bodies was carried out on TPM plates (G and H) for 5 days, with inoculation of 5 × 10⁹ cells/ml. The plates were incubated at 30°C. Bars, 0.6 cm for panels A to D, 30 μm for panels E and F, and 1.5 mm for panels G and H.

The mutant HL-1 was assessed for developmental ability on TPM starvation medium by methods described previously (15). The mutant cells formed a weak and rudimentary fruiting-body structure (Fig. 1G), compared to the mature fruiting bodies of wild-type cells (Fig. 1H). The sporulation frequency of HL-1 was only 0.67% that of HW-1, which was able to develop about 3.0 × 10⁶ spores from an initial input of 5 × 10⁷ cells after a 5-day incubation. These results indicate that the mutant HL-1 is also significantly defective in developmental aggregation and sporulation.

Cells of the mutant HL-1 dispersed easily in liquid culture, indicating possible defects in cell cohesion. The amount of EPS was assessed by scanning electron microscopy, which revealed less extracellular matrix on the surfaces of HL-1 cells than on those of HW-1 cells. The dyes Congo red and trypan blue, which bind to EPS (3, 26), were employed for quantitative analysis of the extracellular matrices of the wild-type and mutant strains by the method described previously (2). The wild type strain HW-1 bound 69.8% and 48.4% of Congo red and trypan blue, respectively, compared to 49.8% and 22.9% for HL-1, indicating less cohesion ability of the mutant. A clumping assay, using the method described previously by Shimkets (22), also confirmed that the mutant cells exhibited less cohesion than the wild-type cells, and the relative absorbance readings at the 100-min end point for the mutant and wild-type cells in morpholinepropanesulfonic acid (MOPS) buffer were 0.563 ± 0.086 and 0.075 ± 0.055 optical density units at 600 nm, respectively.

We reported previously that the salt-tolerant Myxococcus strains exhibited enhanced S motility in the presence of seawater on either soft or hard CYE agar (24). Interestingly, the effect of seawater on swarming ability was significantly decreased by the mutation (Fig. 2). These results suggest that the mutated gene(s) is involved in or responsible for the enhancement of surface translocation in response to the presence of seawater.

• Open in new tab
• Download powerpoint

FIG. 2.

Swarming colony sizes of the mutant HL-1 and the wild-type strain HW-1 on the nutrient medium CTT without (dashed lines) or with (solid lines) 20% seawater and with different concentrations of agar.

The mutagenized gene in HL-1 and the related genes in this locus.

The MiniHimar1-lacZ transposon contains the Escherichia coli replication origin R6K. To identify the gene mutated in HL-1, its genomic DNA was digested with SphI and BamHI, self-ligated for E. coli transformation, and then sequenced. Two thermal asymmetric interlaced PCR amplifications (18) were then performed. The nested specific primers and arbitrary degenerate primers (AD primers) used in this study are listed in Table 2. An upstream 6.3-kb segment and a downstream 6.7-kb segment flanking the insertion were obtained. After sequencing, the junction sequence between the two segments was further PCR amplified from the wild-type strain HW-1 and sequenced. By using the FramePlot program, version 2.3.2 (http://www.nih.go.jp/%7Ejun/cgi-bin/frameplot.pl ), the 13-kb segment (deposited in GenBank with the accession number EF371498) was predicted to contain six open reading frames (ORFs), which likely form a gene cluster (Fig. 3). Blastx against the GenBank database revealed that the sequences are significantly homologous to the corresponding ORFs from M. xanthus DK1622 and Stigmatella aurantiaca DW4/3-1. Most of these ORFs are predicted to be putative type 3 thrombospondin genes. Type 3 thrombospondin, reported for eukaryotic cells, has affinity for cell surfaces, calcium ions, and many matrix macromolecules (1, 4, 16). Thus, the gene cluster was designated mts, for myxobacterial thrombospondin-like proteins (MtsA to MtsF). The MiniHimar1-lacZ insertion interrupted the codon for Tyr359, located 83 residues from the C terminus of the predicted MtsC protein. MtsA and MtsE possess transmembrane regions at their N termini, as assessed by SMART (http://smart.embl-heidelberg.de/ ). The SignalP-HMM program (http://www.cbs.dtu.dk/services/SignalP/ ) predicted that MtsA, MtsC, MtsD, MtsE, and MtsF contain signal peptides with probabilities of 0.998, 1.000, 0.997, 0.984, and 0.997 and that the potential cleavage sites are at residues 24, 25, 21, 20, and 22, respectively. Bioinformatics analysis of the predicted mts gene, together with the phenotypes of the mutant, suggested that the Mts proteins are probably involved in the construction of the cell surface matrix for S motility and development.

• Open in new tab
• Download powerpoint

FIG. 3.

Physical organization of the mts region (13.0 kb) on the M. fulvus chromosome and the MiniHimar1-lacZ insertion site (indicated by an open arrowhead). The ORFs and predicted transcription directions are represented by arrows. The locations of the restriction enzyme sites used for the cloning sequence are shown. S, SphI; B, BamHI.

Characteristics of M. xanthus mts mutants.

Genetic manipulation of HW-1 is difficult. Attempts to make targeted mts mutations in HW-1 using established protocols for M. xanthus (12) were unsuccessful (data not shown). Instead, an in-frame deletion of mtsC (429 amino acid residues were deleted from 496 amino acids of MXAN1334) was performed for M. xanthus strains DK1622 (A⁺ S⁺), DK1217 (A⁻ S⁺), and DK10410 (A⁺ S⁻) to determine the function of mtsC. The mutants in the A⁺ S⁺ and A⁻ S⁺ backgrounds formed smaller colonies on soft agar than their parent strains, whereas the mutation in the A⁺ S⁻ background did not lead to changes in the colonies on either hard or soft agar (Fig. 4). The results indicated that mtsC is likely also involved in S motility in M. xanthus.

• Open in new tab
• Download powerpoint

FIG. 4.

Morphological characteristics of the wild-type strain M. xanthus DK1622 and its mutants ZC16-4 (ΔMXAN1334) and ZC16-23 (deletion of MXAN1332 to MXAN1337); the A⁻ S⁺ strain M. xanthus DK1217 and its mutant ZC12-3 (ΔMXAN1334); and the A⁺ S⁻ strain M. xanthus DK10410 and its mutant ZC10-1 (ΔMXAN1334). Colony expansions were done on CTT medium with 1.5% (A to E, N, and O) (bar, 1.0 cm) and 0.3% (F to J, P, and Q) (bar, 0.7 cm) agar for 5 days. Development of fruiting bodies was done on TPM plates (K to M) (bar, 1.5 cm) for 5 days. The plates were incubated at 30°C. (R to T) Cell movement at the swarming edges of DK1622, ZC16-4, and ZC16-23 on 1.5% agar. Bar, 50 μm.

Interestingly, the effect of the in-frame deletion on motility in M. xanthus seems to be less marked than that of the insertion in mtsC in M. fulvus HW-1. To determine whether the Mts proteins are important for S motility and development in M. xanthus, we completely deleted the sequence of MXAN1332 to MXAN1337 from DK1622. Plasmid pZCY9, which contains a deletion of all the mts genes, was transformed into DK1622 to give rise to the mutant ZC16-23. ZC16-23 produced smaller colonies (about 75% the size of colonies of the wild-type strain DK1622) on hard CTT (1% (wt/vol) Casitone, 8 mM MgSO₄, 1 mM K₂HPO₄-KH₂PO₄ [pH 7.6], and 10 mM Tris-HCl [pH 7.6]) agar (Fig. 4). Compared to the change of the colony size in M. fulvus HL-1 by the insertion in mtsC, the effect of the complete deletion of mts in M. xanthus is not as prominent. The MXAN1334 insertion mutant exhibited motility phenotypes similar to those of the deletion mutant ZC16-4 (data not shown). These results indicate that the motility differences between the M. fulvus HW-1 insertion mutant and the M. xanthus DK1622 deletion mutant were not likely due to polar effects of transposon insertion. The mts products obviously play a more important role in S motility in M. fulvus HW-1 than they do in M. xanthus DK1622.

In contrast to their role in S motility, Mts proteins are essential for fruiting-body formation and sporulation in M. xanthus. When inoculated onto TPM (10 mM Tris HCl [pH 7.6], 8 mM MgSO₄, 1 mM K₂HPO₄-KH₂PO₄ [pH 7.6], and 1.5% agar) plates, DK1622 cells formed visible fruiting-body structures from the second day of incubation; whereas the mutants ZC16-4 and ZC16-23 did not form fruiting bodies even after 5 days of incubation (Fig. 4). Under our assay conditions, DK1622 cells produced 1.3 × 10⁶ spores from an initial input of 5 × 10⁷ cells after 5 days of incubation. The sporulation frequency of ZC16-4 cells was only 0.01% that of DK1622 cells, and no spores were detected for ZC16-23. The effects of mts genes on development are rather similar in M. fulvus HW-1 and M. xanthus DK1622.

Concluding remarks.

S motility is a cell-cell contact-dependent mode of movement that is essential for myxobacterial predation and development. Youderian and Hartzell recently suggested that at least 25% of the nonessential genes involved in S motility had not yet been identified, probably due to preferential mutation hot spots (27). Different Myxococcus species or strains may possess genotypes with subtle differences in motility. The diversification of the motility genotypes thus provides a useful source and also an efficient approach to discovering the hypomorphic motility genes in the model strain M. xanthus DK1622. This paper describes a new genetic locus (mts) that is required for S motility and development in Myxococcus. The mts locus is predicted to contain six ORFs. Four components—MtsB, MtsC, MtsD, and MtsE—were predicted to be homologous to the type 3 thrombospondins (Goldman et al. [6] predicted two, MtsC and MtsE), which are multifunctional proteins with affinity for cell surfaces, calcium ions, and many matrix macromolecules (16). The thrombospondins have been reported previously only in eukaryotes. In this paper we determined that the mts locus is also involved in Myxococcus cellular cohesion. Although the mts products are required in development, they affect S motility to different extents in different Myxococcus strains. The Mts proteins probably function cooperatively, serving as components for intercellular cohesion, like the thrombospondins in eukaryotic cells (16). However, the predicted myxobacterial thrombospondin-like proteins are similar to thrombospondin 3 only in the variable region, not in the highly conserved region or the calcium binding motifs. The structure and function of Mts are under further investigation.