The Type VI Secretion Systems in Plant-Beneficial Bacteria Modulate Prokaryotic and Eukaryotic Interactions in the Rhizosphere

Abstract

Rhizosphere colonizing plant growth promoting bacteria (PGPB) increase their competitiveness by producing diffusible toxic secondary metabolites, which inhibit competitors and deter predators. Many PGPB also have one or more Type VI Secretion System (T6SS), for the delivery of weapons directly into prokaryotic and eukaryotic cells. Studied predominantly in human and plant pathogens as a virulence mechanism for the delivery of effector proteins, the function of T6SS for PGPB in the rhizosphere niche is poorly understood. We utilized a collection of Pseudomonas chlororaphis 30-84 mutants deficient in one or both of its two T6SS and/or secondary metabolite production to examine the relative importance of each T6SS in rhizosphere competence, bacterial competition, and protection from bacterivores. A mutant deficient in both T6SS was less persistent than wild type in the rhizosphere. Both T6SS contributed to competitiveness against other PGPB or plant pathogenic strains not affected by secondary metabolite production, but only T6SS-2 was effective against strains lacking their own T6SS. Having at least one T6SS was also essential for protection from predation by several eukaryotic bacterivores. In contrast to diffusible weapons that may not be produced at low cell density, T6SS afford rhizobacteria an additional, more immediate line of defense against competitors and predators.

Keywords: GacS/GacA; PGPB (plant growth-promoting bacteria); Pseudomonas; T6SS; bacterivores; competition; rhizosphere ecology.

FIGURE 1

Pseudomonas chlororaphis 30–84 T6SS gene clusters. (A) The T6SS-1 cluster of genes. (B) The T6SS-2 cluster of genes. Red arrows designate genes encoding conserved proteins involved in T6SS structure and function. Blue arrows designate genes encoding hypothetical proteins. Yellow arrows designate genes encoding proteins associated with the TagQRST system in T6SS-1. *Refers to a gene encoding a T6SS protein with a PAAR domain.

FIGURE 2

Rhizosphere persistence over repeat harvests. Bacterial populations (log10 CFU/g root dry weight) in sterile and field soil after the first and fifth harvest. Strains tested included 30–84 WT, the single T6SS mutants ΔTssA1 and ΔTssA2, and the double mutant ΔTssA1/2. Data are the mean and standard errors (bars may be too small to see for some treatments) from three replicate experiments (n = 10/per replicate). Lettering indicates significant differences. Data were analyzed using a one-way ANOVA and Tukey’s tests and significant differences are indicated, p < 0.05.

FIGURE 3

In vitro competition assays. The competitive fitness of 30–84 WT, the single T6SS mutants, ΔTssA1 and ΔTssA2, and the double mutant, ΔTssA1/2 were evaluated by comparing their populations when grown separately or in 50:50 mixtures with other Pseudomonas rhizosphere colonizing bacteria in liquid media. Data are expressed as competitive ratios (population in mixture/population when grown separately). Gray scale bars (left) indicate performance of 30–84 WT and derivatives, and colored bars (right) indicate performance of competitors, including P. fluorescens 2–79 (yellow color, having no T6SS), P. protegens Pf-5 (green color, having a T6SS-1 homolog), and P. fluorescens Q2–87 (purple color, having three T6SS, including T6SS-1 and T6SS-2 homologs). Individual bacterial cultures or mixture cultures were spotted onto nitrocellulose filters on LB plates and incubated at 28°C, 5 h. Bacterial cells were washed from filters, collected via centrifugation, and populations were enumerated after 48 h via serial dilution plating. Data are the mean competitive ratios of five replicates pooled across three experiments (n = 15) and standard errors are indicated.

FIGURE 4

Rhizosphere competition assays. The competitive fitness 30–84 WT, the single T6SS mutants, ΔTssA1 and ΔTssA2, and the double mutant, ΔTssA1/2 were evaluated by comparing their populations when grown separately or in 50:50 mixtures with other Pseudomonas rhizosphere colonizing bacteria in the rhizosphere. Data are expressed as competitive ratios (population in mixture/population when grown separately). Gray scale bars (left) indicate performance of 30–84 WT and derivatives, and colored bars (right) indicate performance of competitors, including P. fluorescens 2–79 (yellow color, having no T6SS), P. protegens Pf-5 (green color, having a T6SS-1 homolog), and P. fluorescens Q2–87 (purple color, having three T6SS, including T6SS-1 and T6SS-2 homologs). After 28 days, bacterial populations from the entire root system and loosely adhering soil were collected and enumerated via serial dilution plating. Populations were standardized to root dry weight. Data are the mean competitive ratios of at least eight replicates pooled across three experiments (n = 24) and standard errors are indicated.

FIGURE 5

Aggregation behavior of Dictyostelium discoideum grown with different bacterial strains as a food source. Bacteria used as prey in the feeding assay included 30–84 WT, ΔTssA1, ΔTssA2, ΔTssA1/2, 30–84 GacA, 30–84 I/I2, or 30–84 ZN and E. coliΔB (used as a preferred prey in the lab). D. discoideum without prey bacteria was used as a negative control. D. discoideum cells were grown in 24-well plates in low nutrient PBM media for 24 h and aggregation behavior was observed using DIC microscopy (100X oil). The D. discoideum control (without prey) showed high levels of aggregation caused by stress due to the lack of nutrition in the media. D. discoideum growing with 30–84 WT, ΔTssA1, ΔTssA2, 30–84 I/I2, and 30–84 ZN displayed a similar level of aggregation, indicating that D. discoideum cannot eat these strains. D. discoideum grown with E. coliΔB showed little to no aggregation. D. discoideum growing with ΔTssA1/2 and 30–84 GacA treatments showed similarly low levels of aggregation. Two replicate experiments were performed, and representative images from the same replicate are presented.

FIGURE 6

Levels of two starvation markers in Dictyostelium discoideum. D. discoideum AX2 cells were incubated in PBM with the indicated prey strains (30–84 WT, ΔTssA1, ΔTssA2, ΔTssA1/2, 30–84 GacA, 30–84 I/I2, or 30–84 ZN and E. coli ΔB) or without prey (negative control) for 24 h. Western blots were stained with anti-Discoidin I, anti-CsA, or anti-beta-Actin antibodies. Graphs show the levels of (A) Discoidin I or (B) CsA normalized to total actin. Values are mean and standard error of 2 independent experiments. * Indicates p < 0.01 compared to D. dictyostelium AX2 control (Unpaired t-tests, Welch’s correction).

FIGURE 7

Bacterial plate clearing by Dictyostelium discoideum. Bacteria used as prey in the feeding assay included 30–84 WT, ΔTssA1, ΔTssA2, ΔTssA1/2, 30–84 GacA, 30–84 I/I2, or 30–84 ZN and E. coliΔB (used as a preferred prey in the lab). The diameter of bacterial lawn cleared by D. discoideum was measured in cm after 72 h. Lettering indicates significant differences. D. discoideum colonies grown on 30–84 WT, both single mutants, 30–84 I/I2, and 30–84 ZN showed little to no clearing, indicating low to no bacterial cell death. D. discoideum growing on E. coliΔB showed significant levels of clearing. When grown on 30–84 GacA, levels of clearing similar to the control were observed. When grown on ΔTssA1/2, less clearing was observed than on the control, but levels were still significantly higher than clearing on 30–84 WT and the single mutants, indicating a decrease in bacterial fitness when lacking at least one functional T6SS. Data are the means and standard error (may be too small to see for some treatments) of four biological replicate experiments (n = 4). Data were analyzed using one-way ANOVA and Student t-tests, p < 0.05.

FIGURE 8

Tetrahymena mating assay. T. thermophila CU427 and CU330 were grown separately overnight in Tris Buffer (pH = 7.4) to induce starvation. Populations were standardized to a cell density of 1.5 – 2 × 10⁵/10 mL (via direct counts) in fresh Tris Buffer and mixed together with only the bacterial strain as a food source. Bacteria used as prey included 30–84 WT, ΔTssA1, ΔTssA2, ΔTssA1/2, 30–84 GacA, 30–84 I/I2, or 30–84 ZN, and the T. thermophila pairs without prey was used as a negative control. After 4 h, the treatments were viewed using a Leitz (Epivert) microscope (100X magnification), and the frequency of mated cell pairs (number of mated pairs/total number of observations) determined. Mean frequencies and standard errors are shown. This experiment was repeated three times (n = 3). Letters denote significant differences. Data were analyzed using one-way ANOVA and Student t-tests, p < 0.05.

FIGURE 9

Percentage of C. elegans that reach maturity after 72 h. Five adult C. elegans were transferred at 1-h intervals to new prey-containing plates to facilitate egg laying (a total of four successive transfers), and then the percentage of nematodes maturing to adults were measured every 24 h over a 72-h period. Plates contained either 30–84 WT, ΔTssA1, ΔTssA2, ΔTssA1/2, 30–84 GacA, 30–84 I/I2, 30–84 ZN, or E. coli OP50 (control) as a food source. Data are the mean and standard errors of three replicate experiments (four plates/replicate). Data were analyzed using one-way ANOVA and Tukey tests. Letters indicate significant differences, p < 0.01.