Metabolic Footprints of Burkholderia Sensu Lato Rhizosphere Bacteria Active against Maize Fusarium Pathogens

Abstract

Consistent with their reported abundance in soils, several Burkholderia sensu lato strains were isolated from the rhizosphere of maize plants cultivated at different sites in central México. Comparative analysis of their 16S rRNA gene sequences permitted their separation into three distinctive clades, which were further subdivided into six other clusters by their close resemblance to (1) Trinickia dinghuensis; (2) Paraburkholderia kirstenboschensis, P. graminis, P. dilworthii and P. rhynchosiae; (3) B. gladioli; (4) B. arboris; (5) B. contaminans, or (6) B. metallica representative species. Direct confrontation assays revealed that these strains inhibited the growth of pathogenic Fusarium oxysporum f. sp. radicis-lycopersici, and F. verticillioides within a roughly 3-55% inhibition range. The use of a DIESI-based non-targeted mass spectroscopy experimental strategy further indicated that this method is an option for rapid determination of the pathogen inhibitory capacity of Burkholderia sensu lato strains based solely on the analysis of their exometabolome. Furthermore, it showed that the highest anti-fungal activity observed in B. contaminans and B. arboris was associated with a distinctive abundance of certain m/z ions, some of which were identified as components of the ornbactin and pyochelin siderophores. These results highlight the chemical diversity of Burkholderia sensu lato bacteria and suggest that their capacity to inhibit the Fusarium-related infection of maize in suppressive soils is associated with siderophore synthesis.

Keywords: Burkholderia; DIESI-MS; Fusarium; Paraburkholderia; biocontrol; siderophores.

Figure 1

Maximum likelihood phylogenetic tree of rhizosphere Burkholderia sensu lato strains. The tree was constructed based on a comparative analysis of 16S rRNA gene sequences. Shown is the phylogenetic position of isolated Burkholderia sensu lato strains in relation to Burkholderia-type strains. The numbers at the nodes represent percentage levels of bootstrap support from 1000 replications. The sequence of Ralstonia solanarum LMG 2299, used as an outgroup, roots the tree.

Figure 2

Antagonism of rhizosphere Burkholderia sensu lato strains against phytopathogenic Fusarium. Antagonism vs. (a) F. verticillioides MF-257 and (b) F. oxysporum f. sp. radicis-lycopersici was determined by the capacity of rhizosphere Burkholderia strains to inhibit their radial hyphal growth when placed in direct confrontation. The single plates in the left end of the image correspond to the negative controls. The colored circles below the plates represent the different Burkholderia sensu lato groups: green = Clade I; yellow = Clade II; and orange-red = Clade III. The results shown are representative of experiments that were performed with five (F. verticillioides) and three (F. oxysporum) replicates per bacterial isolate.

Figure 3

Antifungal activity of Burkholderia sensu lato strains against Fusarium pathogens. The colored bars represent the percentage of inhibition produced against (a) F. verticillioides MF-257 and (b) F. oxysporum f. sp. radicis-lycopersici by Burkholderia sensu lato bacteria isolated from the rhizosphere of maize plants. The bar colors represent their different phylogenetic classifications: green = Clade I; yellow = Clade II; and orange-red = Clade III. Data are the mean values (±std. error) representative of experiments that were performed with five (F. verticillioides) and three (F. oxysporum) biological replicates per bacterial isolate. Different letters alongside the inhibition percent values represent statistically significant differences (one-way ANOVA, Tukey test, p < 0.05).

Figure 4

Principal component analysis (PCA) obtained from DIESI-MS analysis of culture extracts of the rhizosphere Burkholderia sensu lato strains. Different colored symbols represent the phylogenetic clustering resulting from previous 16S rRNA analysis: green squares (clade I), yellow circles (clade II) and orange-red triangles (clade III). Black asterisks correspond to uninoculated M9 medium controls.

Figure 5

Metabolic heat map based on DIESI-MS data of 50 discriminant ions selected by using the random forest model. The Euclidean distance and Ward’s algorithm were used for the hierarchical classification of rhizosphere Burkholderia sensu lato strains, listed on the bottom end of the chart, based on the respective abundance of the discriminant ions selected. Some of these were assigned an identity, as shown for ions corresponding to pyochelin and ornibactin C8 and C6. The colors at the top of the chart are representative of the grouping of the different Burkholderia sensu lato strains: clade I (green), clade II (yellow) and clade III (orange-red). The colored bars (in the blue-red range) indicate the abundance of each metabolite ion, where red indicates higher signal intensity.

Figure 6

Quantification of siderophore accumulation in the culture media of rhizosphere Burkholderia sensu lato strains. The box plots represent the normalized intensities of the ion signals identified as part of the pyochelin (m/z 325.2), ornibactin C6 (709.6) and ornibactin C8 (737.8) siderophores. These were detected in the culture media used to grow different rhizosphere Burkholderia sensu lato strains. Different letters over the box plots represent statistically significant differences (one-way ANOVA, Tukey test, p < 0.05). The box plot colors are representative of the grouping of the different Burkholderia sensu lato strains: clade I (green), clade II (yellow) and clade III (orange-red).