Interspecific interactions facilitate keystone species in a multispecies biofilm that promotes plant growth

Abstract

Microorganisms colonizing plant roots co-exist in complex, spatially structured multispecies biofilm communities. However, little is known about microbial interactions and the underlying spatial organization within biofilm communities established on plant roots. Here, a well-established four-species biofilm model (Stenotrophomonas rhizophila, Paenibacillus amylolyticus, Microbacterium oxydans, and Xanthomonas retroflexus, termed as SPMX) was applied to Arabidopsis roots to study the impact of multispecies biofilm on plant growth and the community spatial dynamics on the roots. SPMX co-culture notably promoted root development and plant biomass. Co-cultured SPMX increased root colonization and formed multispecies biofilms, structurally different from those formed by monocultures. By combining 16S rRNA gene amplicon sequencing and fluorescence in situ hybridization with confocal laser scanning microscopy, we found that the composition and spatial organization of the four-species biofilm significantly changed over time. Monoculture P. amylolyticus colonized plant roots poorly, but its population and root colonization were highly enhanced when residing in the four-species biofilm. Exclusion of P. amylolyticus from the community reduced overall biofilm production and root colonization of the three species, resulting in the loss of the plant growth-promoting effects. Combined with spatial analysis, this led to identification of P. amylolyticus as a keystone species. Our findings highlight that weak root colonizers may benefit from mutualistic interactions in complex communities and hereby become important keystone species impacting community spatial organization and function. This work expands the knowledge on spatial organization uncovering interspecific interactions in multispecies biofilm communities on plant roots, beneficial for harnessing microbial mutualism promoting plant growth.

Keywords: interspecies interactions; keystone species; multispecies biofilms; mutualism; plant growth; spatial organization.

Figure 1

Impacts of SPMX on Arabidopsis root growth and shoot fresh weight over time; (A) growth phenotype of seven-day-old Arabidopsis seedlings inoculated with SPMX (a four-species consortium composed of S. rhizophila, P. amylolyticus, M. oxydans, and X. retroflexus) co-cultured for 5, 10, and 15 days (D5, D10, and D15) (scale bar = 1 cm); (B) boxplot showing main root length (cm) of Arabidopsis seedlings incubated with SPMX co-culture at D5, D10, and D15 (n = 9); (C) boxplot presenting shoot fresh weight (mg) of Arabidopsis seedlings incubated with SPMX co-culture at D5, D10, and D15 (n = 9); data in each group are based on nine biological replicates (n = 9) from three independent experiments; statistical significance indicates the difference between SPMX-treated and control plants based on Wilcoxon rank-sum test (Wilcoxon test).

Figure 2

SPMX co-culture and monocultures colonization of Arabidopsis roots; (A) representative images captured by confocal laser scanning microscopy (CLSM) showing bacterial colonization of an Arabidopsis root after five days (D5) of incubation with SPMX or one of four individual strains of M. oxydans (Mo), S. rhizophila (Sr), X. retroflexus (Xr), and P. amylolyticus (Pa); the bacterial colonization was visualized by staining with SYTO9; the root was stained with calcofluor white (CFW) (scale bar = 50 μm); (B) bar chart showing biomass volumes of mixed SPMX and four individual species colonizing the roots quantified by 3D quantitative analysis of confocal images. Letters above bars indicate statistical differences; (C) bar chart showing the number of total bacterial cells that colonize the whole root by SPMX co-culture and monocultures, quantified by colony forming units (CFU) assay and normalized by average root length (mm); letters above bars indicate statistical differences.

Figure 3

FISH-confocal laser scanning microscopy (FISH-CLSM) micrographs of multispecies biofilm formed by SPMX on the root surface; (A) FISH-CLSM images showing the cell shape and size of each species used in this study, and all four strains (SPMX) in a mixed-species community; the four left images show individual species and one on the far right shows SPMX-mixed community. Pa (P. amylolyticus), Mo (M. oxydans), Xr (X. retroflexus), and Sr (S. rhizophila) are labeled with Cy5, Cy3, Pacific blue, and FAM, respectively (scale bar = 5 μm); (B) representative FISH-CLSM images, from SPMX-root co-cultivation samples and non-bacteria inoculated root samples (control), visualizing multispecies biofilm formation over time (D5, D10, and D15), captured from root side of samples; the multispecies biofilm formed on the roots is indicated with a white dashed box and arrow, and the root region is denoted in the images; illustration in the left panel indicates the location of root side where micrographs are captured; micrographs from the root side contain four-channel separate and merged images (scale bar = 10 μm); (C) bar chart showing the estimated total cell numbers in the 5-, 10-, and 15-day multispecies biofilm formed on the whole root (D5, D10, and D15), determined by qPCR and normalized by average root length (mm); data are based on three biological replicates with three root samples in each independent experiment (n = 9); Kruskal–Wallis test was used to test the statistical differences of total cell numbers on the roots over time: “ns” refers to “not significant (P >.05)”, and ^*** indicates P < .001 (FDR adjusted), ^** indicates P < .01 (FDR adjusted).

Figure 4

Relative abundance of each species in a four-species biofilm formed on roots over time; (A) principal coordinate analysis (PCoA) showing global differences in composition of the four-species community (SPMX) composed of M. oxydans (Mo), S. rhizophila (Sr), X. retroflexus (Xr), and P. amylolyticus (Pa), generated by using the Bray-Curtis distance based on the four species abundance across the three different time points (n = 3 for each time point; PERMANOVA, correlation coefficient R² = 0.85, P = .003); the community composition is estimated by 16S rRNA gene sequencing data; (B) bar chart showing differential composition of relative abundance in four species colonizing the Arabidopsis root, and abundant shifts of each species over time at D5, D10, and D15 (n = 3) based on 16S rRNA gene sequencing data; Kruskal–Wallis test was used to test the statistical differences in relative abundance among three different time points: “ns” refers to “not significant (P >.05)”, and ^** indicates P < .01 (FDR adjusted), ^* indicates P < .05 (FDR adjusted).

Figure 5

Differences in spatial organization and function between four- and three-species biofilm communities on plant roots; (A) three-dimensional (3D) view displaying the structure of the four-species (SPMX), three-species (SMX) biofilm community established by M. oxydans (Mo), S. rhizophila (Sr), X. retroflexus (Xr) with and without P. amylolyticus (Pa) on root surfaces at 5-, 10-, and 15-day post inoculation (D5, D10, and D15), captured by FISH-CLSM with z-stacks; (B) quantification (μm³, biomass volume) of each species throughout the SPMX and SMX biofilm at D5, D10, and D15; Kruskal–Wallis test was used to test the statistical difference of the bio-volume of Pa growing at D5, D10, and D15; P = .004 in SPMX treatment (FDR adjusted) (n = 3); P = .056 in SMX treatment (FDR adjusted) (n = 6); (C) bar chart showing the estimated total cell numbers in three-species (SMX) biofilm formed on the whole root over time (D5, D10, and D15), determined by qPCR and normalized by average root length (mm) (n = 9); Kruskal–Wallis test was used to test the total cell numbers on the root over time: “ns” refers to “not significant (P >.05)”; (D) growth phenotype of 7-day-old Arabidopsis seedlings inoculated with SMX (a three-species consortium composed of S. rhizophila, M. oxydans, and X. retroflexus in the absence of P. amylolyticus), incubated for 5, 10, and 15 days (D5, D10, and D15) (scale bar = 1 cm); (E) boxplot showing main root length (cm) of Arabidopsis seedlings incubated with SMX co-culture at D5, D10, and D15 (n = 9); (F) boxplot presenting shoot fresh weight (mg) of Arabidopsis seedlings incubated with SMX co-culture at D5, D10, and D15 (n = 9); data in each group are based on nine biological replicates (n = 9) from three independent experiments; statistical significance indicates the difference between SMX-treated and control plants based on Wilcoxon rank-sum test (Wilcoxon test).

Figure 6

Co-localization and cell aggregate analysis of the multispecies biofilm community on the root surface; (A) co-localization of three biofilm members M. oxydans (Mo), S. rhizophila (Sr), and X. retroflexus (Xr) relative to P. amylolyticus (Pa) at D5, D10, and D15 with corresponding observations in magnified confocal images (bar = 10 μm); solid lines are means and shaded regions are SEM of the biological replicates, each with five times repeated analysis; Inserted images show representative results analyzed by 3D co-localization at three time points; (B) microscale cell aggregates of Pa in SPMX multispecies biofilm forming on the roots over time were analyzed by joining neighboring pixels in aggregates at the scale of 202.83 × 202.83 μm, with those larger than 10 μm³; point sizes are scaled by average aggregate volume and each point corresponds to one biological replicate; data from three biological replicates at each time point; (C) 3D-view images display simulated Pa cell aggregates identified as larger than 10 μm³ on the root surfaces, corresponding to the three time points (D5, D10, and D15); colors of the aggregates are recycled and indicate all different aggregates formed by Pa at the scale of 202.83 × 202.83 μm.