Figure 5. Mutualisms emerge when a nutrient-deficient mutant is the sole provider of extracellular matrix.

(A) Extracellular matrix is not required for alrA rescue (third column), and rescue is enhanced when the alrA full knockdown is combined with a matrix-deficient parent strain (fourth column). In the left column, the parent-GFP strain is false-colored in green, and in the other columns the alrA knockdown strain is green. The parent-RFP strain is false-colored in magenta. In the first two columns, both strains express matrix proteins. In the third column, both strains lack epsH and tasA, which encode key matrix components. In the fourth column, epsH and tasA are deleted from the parent-RFP strain while the alrA knockdown produces matrix. Images were acquired after 48 hr of growth. Scale bar: 5 mm. (B) Competitive fitness of the alrA full knockdown decreased over time when both strains or neither produced matrix, while fitness remained 1 (equal proportions of the two strains) and stable over time in co-cultures when the parent was matrix-deficient. The matrix-deficient parent-RFP+alrA knockdown (purple) and the parent-RFP+parent-GFP control (blue) data were not significantly different (p>0.2 at all time points); nor were the parent-RFP+alrA knockdown (orange) and the matrix-deficient parent-RFP+matrix-deficient alrA knockdown (yellow) data (p>0.07, Student’s t-test). The matrix-deficient parent-RFP+alrA knockdown (purple) fitness data were significantly higher than the parent-RFP+alrA knockdown (orange) at all time points after 0.5 days (p<0.014). Data were normalized to fitness at 0.5 days. Curves are means, and error bars represent one standard deviation (n = 2–3 biological replicates). Statistical analysis: Student’s unpaired t-test with a Benjamini–Hochberg multiple-hypothesis correction. (C) Cells with full knockdown of alrA were outcompeted by the parent-RFP strain during two-dimensional growth. Co-cultures were spotted on a MSgg-xylose agar pad and limited to growth in a layer with thickness one cell by applying a cover slip over the cells. Images were acquired after 24 hr of growth. Scale bar: 0.4 mm. (D) Design of screen to identify mutants that exhibit an increase in fitness when co-cultured with a matrix-deficient parent on MSgg-xylose agar. The sacA::gfp library was used in this screen; plate 3 from the 16 hr time point is shown. The distance between the centers of each colony is 9 mm. White box, controls; white dashed box, titration row. (E) Full knockdowns that exhibited mutualism generally involve genes related to nutrient sharing, translation, and DNA binding. Top: competitive fitness of co-cultures with the wild-type parent-RFP (purple) and matrix production-deficient parent (red) at 16, 24, and 48 hr. Numbers in the top right indicate the mutualism score. Bottom: merged images of biofilm colony co-cultures on MSgg-xylose agar at 48 hr. Scale bar: 5 mm. (F) Full knockdowns that exhibited mutualism in three-dimensional biofilm colonies were outcompeted by the matrix-deficient parent-RFP strain when growth was limited to two dimensions as in C. Images were acquired after 24 hr of growth. Scale bar: 0.4 mm. For (A, C–F), GFP is false-colored in green and RFP is false-colored in magenta.

Figure 5—figure supplement 1. A mutualism screen reveals full knockdowns with improved growth in co-culture when the parent is deficient in production of extracellular matrix.

(A) The matrix-deficient parent-RFP+matrix-proficient parent-GFP co-culture did not form sectors in a three-dimensional colony (left), but did at the edge of a two-dimensional colony (right). Merges show the parent-GFP and parent-RFP strains false-colored in green and magenta, respectively. Left scale bar: 5 mm; right scale bar: 0.4 mm. (B) The leading edge of a co-culture grown between agar and a coverslip is one cell thick. One microliter of cell culture was spotted onto an MSgg agar pad, and a coverslip was applied to limit growth to two dimensions. The culture was incubated for 24 hr, and the colony edge was imaged. Scale bar: 50 μm. (C) Results from a mutualism screen comparing the competitive fitness of knockdown strains co-cultured with a matrix-proficient (left) or matrix-deficient (right) parent. Control parent-RFP+parent-GFP co-cultures are located on the right and left edges of the library, and the titration row is shown on the top and bottom rows. The distance between the centers of neighboring colonies is 9 mm. GFP (from parent-GFP or knockdown strains) and RFP (from parent-RFP) fluorescence signals are false-colored in green and magenta, respectively. (D) The library exhibited a wide range of mutualism scores, with 11 full knockdowns exhibiting a mutualism score >2 standard deviations higher than the mean across all strains (>0.22). The library is shown in blue, and controls are shown in gray. (E) Representative controls of the parent-RFP strain grown with the parent-GFP strain showing the final composition of RFP and GFP in the colonies. In merged images, the parent-RFP and parent-GFP are false-colored in magenta and green, respectively. These controls are from the mutualism screen at 48 hr (and are the same as the controls shown in Figure 5E). (F) Five non-essential gene knockdowns exhibited mutualism. Top: competitive fitness of co-cultures with the wild-type-like parent-RFP (magenta) and the matrix production-deficient parent (red) at 16, 24, and 48 hr. Numbers in top right indicate the mutualism score. Bottom: merged images of biofilm colony co-cultures on MSgg-xylose agar at 48 hr. Scale bar: 5 mm.