Short-range C-signaling restricts cheating behavior during Myxococcus xanthus development

Abstract

Myxococcus xanthus uses short-range C-signaling to coordinate multicellular mound formation with sporulation during fruiting body development. A csgA mutant deficient in C-signaling can cheat on wild type (WT) in mixtures and form spores disproportionately, but our understanding of cheating behavior is incomplete. We subjected mixtures of WT and csgA cells at different ratios to co-development and used confocal microscopy and image analysis to quantify the arrangement and morphology of cells. At a ratio of one WT to four csgA cells (1:4), mounds failed to form. At 1:2, only a few mounds and spores formed. At 1:1, mounds formed with a similar number and arrangement of WT and csgA rods early in development, but later the number of csgA spores near the bottom of these nascent fruiting bodies (NFBs) exceeded that of WT. This cheating after mound formation involved csgA forming spores at a greater rate, while WT disappeared at a greater rate, either lysing or exiting NFBs. At 2:1 and 4:1, csgA rods were more abundant than expected throughout the biofilm both before and during mound formation, and cheating continued after mound formation. We conclude that C-signaling restricts cheating behavior by requiring sufficient WT cells in mixtures. Excess cheaters may interfere with positive feedback loops that depend on the cellular arrangement to enhance C-signaling during mound building. Since long-range signaling could not likewise communicate the cellular arrangement, we propose that C-signaling was favored evolutionarily and that other short-range signaling mechanisms provided selective advantages in bacterial biofilm and multicellular animal development.

Importance: Bacteria communicate using both long- and short-range signals. Signaling affects community composition, structure, and function. Adherent communities called biofilms impact medicine, agriculture, industry, and the environment. To facilitate the manipulation of biofilms for societal benefits, a better understanding of short-range signaling is necessary. We investigated the susceptibility of short-range C-signaling to cheating during Myxococcus xanthus biofilm development. A mutant deficient in C-signaling fails to form mounds containing spores (i.e., fruiting bodies) but cheats on C-signaling by wild type in starved cell mixtures and forms spores disproportionately. We found that cheating requires sufficient wild-type cells in the initial mix and can occur both before mound formation and later during the sporulation stage of development. By restricting cheating behavior, short-range C-signaling may have been favored evolutionarily rather than long-range diffusible signaling. Cheating restrictions imposed by short-range signaling may have likewise driven the evolution of multicellularity broadly.

Keywords: Myxococcus xanthus; bacterial development; biofilms; cheating; evolution; extracellular signaling; fruiting body; multicellular development; short-range signaling; spores.

Fig 1

Co-development of wild type (WT) and csgA mutant cells at different ratios. WT and csgA cells were mixed at ratios indicated on the left. Half of the WT cells were the labeled strain YH7 (green fluorescence), and half were the unlabeled strain DK1622. Likewise, half of the csgA cells were the labeled strain YH11 (red fluorescence), and half were the unlabeled strain MRR33. Vanillate (0.5 mM) was added, and the mixtures were starved under submerged culture conditions. Confocal images of the same field of view were acquired near the bottom (i.e., the first optical section above the bottom of the well in which cells could be clearly visualized, so ~0.25 to 0.5 µm above the bottom of the well) of the biofilm or a mound at the indicated times poststarvation (PS). Images show the green and red channels merged and are representative of five biological replicates. Bar, 20 µm.

Fig 2

Radial patterns of cell density, neighbor alignment, and tangential orientation in co-developed mixtures at different times poststarvation. In the experiment described in the Fig. 1 legend, z-stacks of optical sections were collected from near the bottom of the same nascent fruiting body (NFB) to 5 µm up for each of the five biological replicates, and segmented cells were classified as rods, transitioning cells, or spores. The combined results for both WT green- and csgA red-labeled cells are shown. Line, median. Shaded region, 90% credible interval. (A) Combined density of all cell classes from the center (0 µm) to the edge (60 µm) of NFBs. The cartoon at the right depicts an early NFB (gray circle) with decreasing rod density ~30–60 µm from the radial center (arrow, radius). (B) Neighbor alignment of rods radially in NFBs over time. Alignment is the weighted average of the cosine of the angle of a rod to each of its neighbors (using a Gaussian kernel with sigma = 2.5 µm) (1, perfect alignment; 0, orthogonal). Cartoon, early NFB with maximal neighbor alignment near the radial center and ~20–40 µm from the center. (C) Tangential orientation of rods radially in NFBs over time. Orientation is the cosine of the angle of a rod with the circumference of the NFB (1, tangent to the circumference; 0, orthogonal). Cartoon, early NFB with maximum tangential orientation ~20–40 µm from the center.

Fig 3

Radial patterns of wild type (WT) and csgA mutant cells in co-developed mixtures at different times poststarvation. In the experiment described in the Fig. 1 and 2 legends, segmented cells from z-stacks were classified as rods, transitioning cells, or spores. The results for WT green- and csgA red-labeled cells are shown separately. (A) Combined density of all cell classes from the center (0 µm) to the edge (60 µm) of nascent fruiting bodies (NFBs). (B) Proportion of WT and csgA mutant rods and spores (relative to the combined total number of cells in all classes for both strains) radially in NFBs over time. Line, median. Shaded region, 90% credible interval.

Fig 4

Ratios of csgA to wild type (csgA/WT) cell classes and rates of disappearance and sporulation in co-developed mixtures over time. In the experiment described in Fig. 1 and 2 legends, segmented cells from z-stacks were classified as rods, spores, or transitioning cells (transitioning). Cells within 60 µm of the radial center of nascent fruiting bodies were analyzed. (A) Ratios of csgA/WT cell classes at indicated times poststarvation (PS). Bold color, major cell class. Pale color, minor cell class. Dashed line, initial ratio. (B) Rates of csgA and WT disappearance and sporulation during 6-h intervals PS. Time, interval halfway point. Dot, median. Line, 90% credible interval.

Fig 5

Ratios of csgA to wildtype (csgA/WT) cells at different times and locations early in development. WT and csgA cells were mixed at ratios indicated at the top. The mixtures contained the strains described in the Fig. 1 legend, but each labeled strain was premixed 1:3 (18- and 24-h time points) or 1:1 (30-h time point) with its unlabeled parent. Vanillate (0.5 mM) was added, and the mixtures were starved under submerged culture conditions. z-stacks were collected as described in the Fig. 2 legend as follows: distal from any mound (and prior to the formation of most) at 18 h (designated “biofilm”) or at three locations (within a nascent fruiting body [designated “NFB”], “proximal” to the same NFB [one field of view outside], and “distal” from any NFB) at 24 and 30 h. Segmented cells from z-stacks of at least five biological replicates were classified as rods, transitioning cells, or spores. Ratios are for totals of all cell classes and were indistinguishable from ratios for only rods, the major cell class. Dot, median. Line, 90% credible interval. Dashed line, initial ratio.