Linking Comparative Genomics of Nine Potato-Associated Pseudomonas Isolates With Their Differing Biocontrol Potential Against Late Blight

Abstract

For plants, the advantages of associating with beneficial bacteria include plant growth promotion, reduction of abiotic and biotic stresses and enhanced protection against various pests and diseases. Beneficial bacteria rightly equipped for successful plant colonization and showing antagonistic activity toward plant pathogens seem to be actively recruited by plants. To gain more insights into the genetic determinants responsible for plant colonization and antagonistic activities, we first sequenced and de novo assembled the complete genomes of nine Pseudomonas strains that had exhibited varying antagonistic potential against the notorious oomycete Phytophthora infestans, placed them into the phylogenomic context of known Pseudomonas biocontrol strains and carried out a comparative genomic analysis to define core, accessory (i.e., genes found in two or more, but not all strains) and unique genes. Next, we assessed the colonizing abilities of these strains and used bioassays to characterize their inhibitory effects against different stages of P. infestans' lifecycle. The phenotype data were then correlated with genotype information, assessing over three hundred genes encoding known factors for plant colonization and antimicrobial activity as well as secondary metabolite biosynthesis clusters predicted by antiSMASH. All strains harbored genes required for successful plant colonization but also distinct arsenals of antimicrobial compounds. We identified genes coding for phenazine, hydrogen cyanide, 2-hexyl, 5-propyl resorcinol and pyrrolnitrin synthesis, as well as various siderophores, pyocins and type VI secretion systems. Additionally, the comparative genomic analysis revealed about a hundred accessory genes putatively involved in anti-Phytophthora activity, including a type II secretion system (T2SS), several peptidases and a toxin. Transcriptomic studies and mutagenesis are needed to further investigate the putative involvement of the novel candidate genes and to identify the various mechanisms involved in the inhibition of P. infestans by different Pseudomonas strains.

Keywords: Pseudomonas; beneficial microorganisms; biocontrol; comparative genomics; de novo genome assembly; genotype-phenotype correlation; phyllosphere; rhizosphere.

FIGURE 1

Phylogenetic tree based on 107 core genes from nine sequenced Pseudomonas strains (this study; marked in red), 48 Pseudomonas reference genomes and one outgroup (Azotobacter vinelandii DJ). Bootstrap values obtained from 100 bootstrap runs are shown for each node. The scale at the bottom indicates the number of amino acid substitutions per site.

FIGURE 2

Circular genome view with R32 strain as reference. Moving inward, tracks 1–8 visualize the BLAST results (presence/absence) of all CDSs from each of the sequenced genomes, against the reference. Track9 (mauve) shows the CDS for R32. The tenth (black) and eleventh (purple) rings show the GC content and GC skew of the R32 genome respectively. The innermost ring shows the genomic coordinates of strain R32.

FIGURE 3

Presence/absence table of known antibiotics, siderophores, cyclic lipopeptides (CLP), toxins, bacteriocins, and extracellular hydrolases with reported antibacterial, antifungal or anti-oomycetal properties (2,3bd, 2,3-butanediol; PAA, phenylacetic acid; IAA, indole-3-acetic acid, GABA, γ-aminobutyric acid; ACC, 1-aminocyclopropane-1-carboxylic acid). All genomes were also screened for genes implicated in chemotaxis, plant colonization, plant-bacteria interactions, type III secretion systems (T3SS) and type VI secretion systems (T6SS). Full sized dots indicate the presence of homologs for all necessary genes or complete gene clusters. Small dots indicate incomplete clusters and dot size is proportional to the number of homologous genes found. All genes and reference sequences, including the genes encoding for traits for which no homologous genes were found, can be found in Supplementary Table S4. Strains are grouped according to their phenotypical activity in vitro (R47, R84, R32, and S49) and on plant tissue (R32, S49, S35). The least inhibiting strains are at the bottom of the table (S34, S19, S04).

FIGURE 4

Effect of Pseudomonas strains on mycelial growth on rye glucose agar (top) and pea agar (bottom) after 7 days of incubation. Growth is expressed as percentage of the corresponding unexposed controls. Letters indicate significant differences between the strains and the unexposed control (Kruskal–Wallis’ test, p < 0.05; see section Materials and Methods).

FIGURE 5

Effects of bacterial strains on sporangia germination. (A) Observed germination phenotypes included regular germination (g), sporangia-like swellings (sl) and tip-swelling (ts). (B) Mean sporangia germ tube length was estimated. Different letters indicate significant differences between strains (Kruskal Wallis’ test, p < 0.05; see section Materials and Methods). (C) PCA of sporangia germination as a function of germ tube phenotype. Germ tube length (l) was added as a supplementary variable (dashed arrow).

FIGURE 6

(A) Effects of bacterial strains on zoospore release. (B) Effects of Pseudomonas strains on zoospore germination. Observed germination phenotypes included regular appressorium formation (a), long germ tube formation with appressorium (g), germ tube formation without appressorium (n), zoospore encystment (e), and non-encysted zoospore (z). (C) PCA of zoospore germination in function of germination phenotype.

FIGURE 7

(A) Epiphytical and (B) endophytical colonization of the bacterial strains on tuber-treated plants after 4 weeks. For endophytical colonization, assessed stems were sterilized. Per cultivar, three stems cuts from at 3 different heights from stems of two different plants were taken. Dot color corresponds to the cut’s location on the plant (top, light green; middle, green; base, dark green). Dot size is proportional to the number of stems cuts out of which bacterial colonies grew (1–3).

FIGURE 8

Phenotypical table summarizing the results obtained in the (A) in vitro bioassays, (B) plant colonization assays and (C) in planta inhibition assays. Inhibition or colonizing abilities were translated to a value of performance, by computing the percentages of corresponding controls (see section Materials and Methods) with maximum inhibition or maximum colonization corresponding to 100%. Mycelial growth (Pea), sporangia germination, germ tube germination and colonization were considered for correlation analysis between accessory genes and phenotype. Stars indicate the strains considered as active. (D) Clustering of known antimicrobial compounds for RDA analysis (FER, ferric enterobactin receptor; 2,3bd, 2,3-butanediol; rhiz. TonB recept., rhizobactin TonB receptor; WLIP, white-line-producing; PAA, phenylacetic acid; IAA, indole-3-acetic acid). (E) RDA of the results obtained in the bioassays (response variables) and the known genetic determinants table (explanatory variables).

FIGURE 9

(A) Number of genes positively correlated to mycelial growth inhibition and exclusively present in Pseudomonas strains with activity on mycelial growth. (B) Number of genes positively correlated to mycelial growth inhibition present in the active strains and R76.

FIGURE 10

List of the genes positively correlated to mycelial growth inhibition and exclusively present in at least four Pseudomonas strains with activity on mycelial growth. Genes are sorted according to their functional COG category.

FIGURE 11

List of genes positively correlated to mycelial growth inhibition present in the active strains and R76 or S34. Genes are sorted according to their functional COG category.