The minimal medium used in the EE should render the bacteria dependent on root exudates and dead plant material to survive.To test this, we used a modified version of the minimal medium employed during the EE. Instead of 0.05% glycerol, the MSN medium was supplemented with 0.5% cellobiose . Cellobiose is a disaccharide and a product of partial hydrolysis of cellulose, found in plant cell walls . In addition, MSNc was supplemented with 0.5% xylan. The ancestor showed a growth profile typical of bacterial growth under planktonic conditions . In contrast, several evolved isolates displayed distinct growth profiles, including 3.2, 7.1, 7.2, and 7.3, which showed no decline phase, but instead displayed a pro-longed stationary phase. When analyzing the carrying capacity , several isolates showed significantly increased carrying capacity , whereas few isolates showed significantly decreased carrying capacity . While cellobiose and xylan do not completely represent the plant compounds present in the selective environment, these results suggest that adaptation to the plant root could also be facilitated through the altered utilization of plant compounds.During the EE, B. subtilis was adapted to the plant root alone – in the absence of other microbes. This selective environment is far from its natural habitat in the rhizosphere, where B. subtilis encounters other microbial residents. In fact, the ancestor DK1042 is a derivate of the wild strain NCIB 3610, originally isolated from hay infusion . To this end, we wondered how the pro-longed adaptation of B. subtilis to the plant root environment in the absence of other microbial species affected the ability to colonize the root in the presence of soil microbes. The ancestor and Ev7.3 were tested for their ability to colonize A. thaliana roots in the presence of a synthetic, soil-derived community . This community comprises four bacterial species, Pedobacter sp., Rhodococcus globerulus, Stenotrophomas indicatrix and Chryseobacterium sp. that were previously isolated from soil samples that also contained B. subtilis, thereby representing bacterial soil inhabitants that B. subtilis would normally encounter in nature.
The isolate Ev7.3 was chosen for this test since it was highly adapted to the selective environment, i.e. the isolate displayed significantly increased individual root colonization and out competed the ancestor during competition on the root, where it formed a robust biofilm along the root . To capture any potential difference in the establishment on the root,dutch bucket for tomatoes here defined as root colonization after 48 h, between B. subtilis ancestor and Ev7.3 in the presence of the community, the ancestor or Ev7.3 was co-inoculated with the community in four different ratios: 0.1:1, 1:1, 10.1, and 100:1 of B. subtilis and community, respectively. When B. subtilis was initially under-represented or highly in excess, i.e. inoculation ratio 0.1:1 and 100:1, respectively, no significant difference was observed in the establishment on the root between B. subtilis ancestor and Ev7.3 within the same inoculation ratio . In contrast, when co-inoculated with the community in intermediate ratios, i.e. 1:1 and 10:1, isolate Ev7.3 showed significantly enhanced establishment on the root compared with the ancestor. Since Ev7.3 displayed increased carrying capacity in MSNc + xylan in monoculture compared with the ancestor , we wondered whether the enhanced establishment on the root by Ev7.3 in the presence of the community could be partly attributed to improved utilization of plant compounds. Indeed, growth profiles in MSNc + xylan of the ancestor or Ev7.3 in co-culture with the community revealed that Ev7.3 displayed a significantly increased carrying capacity at inoculation ratio 1:1, 10:1, and 100:1 compared with the ancestor . Finally, in vitro confrontation assays on LB agar showed no major difference in the inhibition of the community members by Ev7.3 compared with the ancestor . Taken together, these results show that even though B. subtilis was adapted to the plant root alone, isolate Ev7.3 displayed increased root colonization also in the presence of a synthetic, soil-derived community under certain inoculation ratios, possibly mediated by robust biofilm formation on the root and enhanced utilization of plant compounds.
A recent study investigated the adaptive response of the PGPR Pseudomonas protegens to the A. thaliana rhizosphere in a sand system, which revealed mutations in genes encoding global regulators and genes related to motility and cell surface structure across independent populations and during such adaptation, the initially plant-antagonistic P. protegens bacterium evolved into mutualists . Furthermore, Lin et al. observed that adaptation of Bacillus thuringiensis to A. thaliana roots under hydroponic conditions led to the evolution of multicellular aggregating phenotypes, which, surprisingly, in certain lineages were accompanied by enhanced virulence against the Galleria mellonella larvae. Here, we employed experimental evolution to study the adaptation of B. subtilis to A. thaliana roots under hydroponic conditions. Our initial hypothesis was that B. subtilis would adapt to the plant roots by acquiring mutations that would provide the bacteria with a fitness advantage over the ancestor during root colonization. We could demonstrate that B. subtilis rapidly adapted to the plant roots, as observed by evolved isolates displaying improved root colonization relative to the ancestor already after 12 transfers and the detection of genetic changes in evolved isolates from transfer 12, 18, and 30. In addition, competition between the ancestor and two selected evolved isolates from the final transfer on the root revealed that both evolved isolates had a fitness advantage over the ancestor during root colonization, thereby confirming our hypothesis.Further phenotypic characterization of the evolved isolates from the final transfer revealed that most isolates across independent populations developed more robust biofilms in response to the plant polysaccharide xylan compared with the ancestor. Except for isolate Ev3.3, the robust biofilm formers tended to be increased in individual root colonization, indicating that robust biofilm formation is associated with adaptation to the plant root.
Motility represents an important trait for many bacteria as it allows them to explore the environment for nutrients and escape unfavorable conditions. Of relevance to the adaptation of B. subtilis to plant roots, motility has been shown to be important for root colonization of different plant species under different conditions. For example, a B. subtilis Dhag mutant was shown to be delayed or reduced in A. thaliana root colonization under hydroponic conditions as well as in tomato root colonization under vermiculites pot conditions . Yet, we found that five out of six isolates from two independent populations were impaired in both swimming and swarming motility, indicating that motility is not important for root colonization in the selective environment of the EE, i.e. under hydroponic, shaking conditions. Indeed, this was verifified in a competition experiment between a non-motile Dhag mutant and the WT, revealing that motility is not required for root colonization under shaking conditions. In contrast to our observations, Li et al. observed several evolved isolatesIn B. subtilis, motility and biofilm formation are incompatible processes: B. subtilis can exist as single, motile cells or in chains of sessile, matrix-producing cells which is regulated by an epigenetic switch involving SinR . The enhanced biofilm formation and impaired motility of isolates from populations 6 and 7 could thereby indicate a possible biofilm-motility trade-off. An inverse evolutionary trade-off between biofilm formation and motility was observed when the opportunistic pathogen Pseudomonas aeruginosa was subjected to repeated rounds of swarming that lead to the evolution of hyper-swarmers that were impaired in biofilm formation . Considering that the ability to form robust biofilm in vitro was shown to positively correlate with root colonization in B. subtilis , and the demonstration that motility is not important for root colonization under shaking conditions, a possible biofilm-motility trade-off could provide B. subtilis with enhanced fitness during root colonization in the selective environment. Indeed, isolate Ev7.3, which developed hyper-robust biofilms in LB + xylan and was impaired in motility, significantly out competed the ancestor during root colonization.
Re-sequencing of selected evolved isolates revealed that Ev7.1, Ev7.2, and Ev7.3 all harbored a single nucleotide polymorphism two base pairs upstream from the start codon of the sinR gene , encoding a transcriptional repressor of matrix genes . This SNP is located in the spacer region between the Shine Dalgarno sequence and the start codon in the ribosome binding site. Interestingly, the nucleotide composition of the spacer sequence has been shown to influence translation efficiency . The SNP upstream from sinR might thereby potentially affect the translation efficiency from the mRNA transcript, resulting in reduced levels of SinR. Reduced levels of SinR could in turn result in increased expression of matrix genes. This is supported by Richter et al. who demonstrated that a DsinR mutant shows increased matrix gene expression, and by Subramaniam et al. reporting that SinR translation and therefore protein level affects matrix gene expression. Potential increased matrix production caused by this mutation could contribute to the Snow-type colony morphology, as this colony morphology was exclusively observed for isolates harboring this mutation. Furthermore, in accordance with the robust biofilm formation and increased root colonization observed for Ev7.1, Ev7.2 and Ev7.3, a DsinR mutant was shown to form a hyper-robust biofilm in biofilm-inducing medium as well as on tomato roots , supporting the possible relevance of this mutation for the observed phenotypes of these isolates. Based on these previous studies, we therefore speculate that the mutation upstream from sinR results in increased matrix gene expression, which in turn enables more robust biofilm formation and increased root colonization as observed for the three isolates in population 7. These three isolates did not harbor mutations in motility-related genes. However, besides a possible effect of reduced SinR levels on the epigenetic switch that could lock the cells in a sessile, matrix-producing stage, a potential reduction in SinR levels leading to over-expression of the eps operon could possibly reduce motility owing to the EpsE clutch . Such mutation and the observed corresponding phenotypes could be responsible for the biofilm-motility trade-off,blueberry grow pot and be an example of antagonistic pleiotropy in which the same mutation is beneficial in one environment, i.e. during root colonization under shaking conditions, but disadvantageous in another, i.e. where motility is required for survival.
In addition, this mutation affecting a biofilm regulator could possibly explain why Ev7.1, Ev7.2, and Ev7.3 show improved biofilm formation also in the absence of xylan. Isolate Ev6.1 and Ev6.3 harbored a non-synonymous point mutation in the fliM gene, which was not present in Ev6.2. This gene encodes a flagellar motor switch protein, part of the basal body C-ring controlling the direction of flagella rotation . Interestingly, Ev6.1 and Ev6.3 were impaired in both forms of motility, whereas Ev6.2 showed similar swimming as the ancestor and was less affected in swarming. We speculate that the R326I substitution affects the function of FliM and consequently the flagellar machinery, resulting in hampered motility in these two isolates. Since we showed that motility was not important for root colonization under shaking conditions, a mutation hampering motility could provide the bacterium a fitness advantage during the adaptation to A. thaliana roots owing to the reduced cost of this apparently redundant trait. However, we do not expect the mutation in fliM to result in reduced cost; it merely changes an amino acid in a protein part of the flagellar machinery. Other mutations in the population six isolates must explain the robust biofilm formation by Ev6.2 and Ev6.3 and the fitness advantage of Ev6.1 over the ancestor during root colonization. For example, isolate Ev6.1 and Ev6.3 harbor a mutation in kinA encoding a two-component sensor kinase which once activated initiates the phosphorelay leading to phosphorylation of the master regulator Spo0A . The isolate from population one at transfer 30 harbored a frame shift mutation in the rsiX locus, encoding an anti-sigma factor controlling the activity of SigX . Inconsistent with the smooth morphology and reduced root colonization of this isolate, an DrsiX mutant was shown to have increased eps expression . However, Ev1.1 additionally harbored several mutations in gtaB encoding a UTP-glucose-1-phosphate uridylyltransferase involved in the biosynthesis of a nucleotide sugar precursor for EPS biosynthesis . A study conducted by Reverdy et al. showed that acetylation of GtaB is important for biofilm formation of B. subtilis and that a gtaB mutant was reduced in pellicle formation. In addition, Xu et al. showed that a DgtaB mutant of Bacillus velezensis SQR9 was significantly decreased in colonization of cucumber roots compared with the WT, although the effect of gtaB on root colonization may be species-dependent.