Disruption of Firmicutes and Actinobacteria abundance in tomato rhizosphere causes the incidence of bacterial wilt disease

Abstract

Enrichment of protective microbiota in the rhizosphere facilitates disease suppression. However, how the disruption of protective rhizobacteria affects disease suppression is largely unknown. Here, we analyzed the rhizosphere microbial community of a healthy and diseased tomato plant grown <30-cm apart in a greenhouse at three different locations in South Korea. The abundance of Gram-positive Actinobacteria and Firmicutes phyla was lower in diseased rhizosphere soil (DRS) than in healthy rhizosphere soil (HRS) without changes in the causative Ralstonia solanacearum population. Artificial disruption of Gram-positive bacteria in HRS using 500-μg/mL vancomycin increased bacterial wilt occurrence in tomato. To identify HRS-specific and plant-protective Gram-positive bacteria species, Brevibacterium frigoritolerans HRS1, Bacillus niacini HRS2, Solibacillus silvestris HRS3, and Bacillus luciferensis HRS4 were selected from among 326 heat-stable culturable bacteria isolates. These four strains did not directly antagonize R. solanacearum but activated plant immunity. A synthetic community comprising these four strains displayed greater immune activation against R. solanacearum and extended plant protection by 4 more days in comparison with each individual strain. Overall, our results demonstrate for the first time that dysbiosis of the protective Gram-positive bacterial community in DRS promotes the incidence of disease.

Fig. 1. Differences in rhizosphere disease suppression between two adjacent tomato plants.

a Images of healthy and diseased tomato plants grown within a 30-cm distance in three different locations (Damyang, Yongin, and Gwangju) in South Korea. Red arrows indicate the wilted tomato plants infected by Ralstonia solanacearum. To prepare the microbial fraction, healthy rhizosphere soil (HRS) and diseased rhizosphere soil (DRS) were suspended in 2.5-mM MES buffer. Roots of 14-day-old tomato seedlings were dipped in the microbial fractions for 30 min. Severity of bacterial wilt disease caused by R. solanacearum was quantified. Data represent mean ± standard error of the mean (SEM; n = 12 plants per treatment). Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001). b Cell density of R. solanacearum in HRS and DRS fractions plated on casamino acid-peptone-glucose (CPG) agar medium containing 2,3,5-triphenyl tetrazolium chloride (TZC), 0 or 50-g/mL ampicillin (AP), and 0 or 5-g/mL vancomycin (Van). c Scoring of disease severity on a 0–5 scale. d Disease severity in HRS and DRS fraction-treated tomato plants at 10–14 days post inoculation (dpi). BTH, 0.5-mM benzothiadiazole treated tomato; control, 2.5-mM MES buffer-treated tomato.

Fig. 2. Comparison of soil community structure between HRS and DRS samples based on pyrosequencing of 16S rRNA amplicons.

a Relative abundance of rhizobacteria at the phylum level in HRS and DRS samples collected from greenhouses in Damyang, Yongin, and Gwangju in South Korea. b Two-dimensional principal coordinate analysis (PCoA) of Bray–Curtis dissimilarity. Significant differences in microbial community composition were detected between HRS and DRS samples in Damyang, Yongin, and Gwangju. c The absolute read numbers of Firmicute, Actinobacteria, Proteobacteria, Acidobacteria, and Bacteroidetes in HRS and DRS from fields in Damyang, Yongin, and Gwangju. Gram+ Gram-positive bacterial groups, Gram− Gram-negative bacterial groups. d Measurement of the ratio of viable Firmicutes and Actinobacteria to Gram-negative bacteria in HRS and DRS samples using the 3% KOH string test and based on the quantification of colony-forming unit (CFU) values of bacterial isolates on TSA medium containing 20-μg/mL polymyxin B (toxic to Gram-negative bacteria) and 5-μg/mL vancomycin (toxic to Gram-positive bacteria). Data represent mean ± SEM. Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001). e LefSe analysis of the Firmicutes and Actinobacteria community in HRS and DRS samples collected from Damyang, Yongin, and Gwangju. LefSe analysis was used to identify the most discriminating ASVs of Firmicutes and Actinobacteria phyla in HRS.

Fig. 3. Dysbiosis of Firmicutes and Actinobacteria in the tomato rhizosphere.

a Disease severity and viable Firmicutes ratio in HRS and DRS samples pretreated with or without 500-μg/mL vancomycin. After 3-h incubation at 30 °C, soil fractions were washed twice with 2.5-mM MES buffer. The prepared soil fractions pretreated with or without vancomycin were applied to tomato roots via the root-dipping method. b Severity of bacterial wilt disease in tomato plants treated with HRS and DRS fractions, with or without 500-μg/mL vancomycin pretreatment. HRS HRS fraction, HRS + vancomycin HRS pretreated with 500-µg/mL vancomycin, DRS DRS fraction, DRS + vancomycin DRS pretreated with 500-µg/mL vancomycin, 500-mg/mL vancomycin root-dipping treatment with 500 mg/mL vancomycin. Data represent mean ± SEM. Different letters indicate significant differences between treatments (P < 0.05; least significant difference [LSD] test). c Changes in R. solanacearum cell density in HRS and DRS fractions pretreated with or without 500-μg/mL vancomycin. d Changes in Firmicutes abundance in HRS and DRS fractions pretreated with or without 500-μg/mL vancomycin. The ratio of viable Firmicutes bacteria was measured using the 3% KOH string test and by the quantification of the CFU values of bacterial isolates grown on TSA medium containing 20-μg/mL polymyxin B or 5-μg/mL vancomycin.

Fig. 4. Identification of heat-stable Firmicutes and Actinobacteria in HRS and DRS samples.

a Isolation of spore-forming Firmicutes bacteria from HRS and DRS fractions treated with high temperature (80 °C) for 30 min. Heat-treated soil fractions were inoculated on TSA medium and incubated at 30 °C for 3 days. b Identification of HRS-specific Firmicutes bacteria present in soil from tomato fields in Damyang, Yongin, and Gwangju. A total of 326 colonies of spore-forming bacteria were randomly selected and identified by 16S rDNA sequencing, and 21 species were identified as HRS-specific bacteria. c Investigation of the distribution of selected bacteria in the tomato rhizosphere using 16S rRNA sequencing. d Severity of bacterial wilt in tomato seedlings treated with HRS-specific Firmicutes bacteria. Data represent mean ± SEM (n = 12 plants per treatment). Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig. 5. Activation of induced systemic resistance (ISR) against Ralstonia solanacearum in HRS samples by the spore-forming Firmicutes and Actinobacteria.

a Co-cultivation of HRS-specific Firmicutes bacteria and bacterial wilt pathogen R. solanacearum on TSA agar medium. Four selected bacterial strains (50 μL; OD₆₀₀ = 1), gentamycin (GM; 0.5 mg/mL; positive control), or sterile distilled water (Control; negative control) were dispensed on a lawn of R. solanacearum on TSA agar plates, and photographs were captured after 2 days. b Spatial separation system. A suspension of R. solanacearum (50 µL; OD₆₀₀ = 1) was injected into the tomato stem 7 days after the root system was treated with each of the four selected Firmicutes strains. Severity of bacterial wilt disease caused by the injection of R. solanacearum suspension into the stems of tomato plants treated with HRS-specific Firmicutes bacteria (c), or with HRS and DRS fractions (d). Data represent mean ± SEM (n = 12 plants per treatment). Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001).

Fig. 6. Activation of ISR by a SynCom comprising four HRS-specific Firmicutes strains (HRS1–4).

a Severity of bacterial wilt disease in tomato plants inoculated with Ralstonia solanacearum after treatment with different combinations of the four selected Firmicutes bacteria. HRS1 + HRS2 mixture of Brevibacterium frigoritolerans (HRS1) and Bacillus niacini (HRS2), HRS1 + HRS2 + HRS3 mixture of HRS1, HRS2, and Solibacillus silvestris (HRS3), HRS1 + HRS2 + HRS4 mixture of HRS1, HRS2, and Bacillus luciferensis (HRS4), HRS1 + HRS2 + HRS3 + HRS4 mixture of all four Firmicutes strains. Data represent mean ± SEM (n = 12 plants per treatment). Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001). b, c Relative expression levels of jasmonic acid (JA) signaling marker genes (b) and salicylic acid (SA) signaling marker genes (c) in systemic leaves of tomato plants treated with the SynCom comprising all four HRS-specific strains (HRS1–4) at 0 and 12-h post inoculation (hpi) with R. solanacearum. Different letters indicate significant differences between treatments (P < 0.05; LSD test). Data represent mean ± SEM. SynCom mixture of all four Firmicutes bacterial strains, BTH 0.5-mM BTH treatment, control 2.5-mM MES buffer treatment.