Phyllosphere symbiont promotes plant growth through ACC deaminase production

Abstract

Plant growth promoting bacteria can confer resistance to various types of stress and increase agricultural yields. The mechanisms they employ are diverse. One of the most important genes associated with the increase in plant biomass and stress resistance is acdS, which encodes a 1-aminocyclopropane-1-carboxylate- or ACC-deaminase. The non-proteinogenic amino acid ACC is the precursor and means of long-distance transport of ethylene, a plant hormone associated with growth arrest. Expression of acdS reduces stress induced ethylene levels and the enzyme is abundant in rhizosphere colonizers. Whether ACC hydrolysis plays a role in the phyllosphere, both as assembly cue and in growth promotion, remains unclear. Here we show that Paraburkholderia dioscoreae Msb3, a yam phyllosphere symbiont, colonizes the tomato phyllosphere and promotes plant growth by action of its ACC deaminase. We found that acdS is required for improved plant growth but not for efficient leaf colonization. Strain Msb3 readily proliferates on the leaf surface of tomato, only occasionally spreading to the leaf endosphere through stomata. The strain can also colonize the soil or medium around the roots but only spreads into the root if the plant is wounded. Our results indicate that the degradation of ACC is not just an important trait of plant growth promoting rhizobacteria but also one of leaf dwelling phyllosphere bacteria. Manipulation of the leaf microbiota by means of spray inoculation may be more easily achieved than that of the soil. Therefore, the application of ACC deaminase containing bacteria to the phyllosphere may be a promising strategy to increasing plant stress resistance, pathogen control, and harvest yields.

Fig. 1. Tomato colonization and growth promotion.

a Tomatoes were grown for 60 days in pots after inoculation with either heat killed or live Msb3. Two thirds of the plants were fertilized with a commercial plant fertilizer (+Wuxal) while one third did not receive any fertilization (−Wuxal). Depicted is the dry weight of the shoots 60 DPI (n = 18, 17, 9, 9). Significance was determined between treatments via a two-sided Students’ T-test; the stars correspond to the level of significance. b In two contrasting plant cultivation systems P. dioscoreae strain Msb3 can be found on above ground tissues. Both panels display species specific copy numbers of the single copy marker gene gyrB detected through qPCR. After seven days within a gnotobiotic growth system strain Msb3 is very abundant (n = 5). P. phytofirmans strain PsJN was included as a positive control (n = 5). Negative controls were non-inoculated tomato seedlings (n = 5). Msb3 could also be detected on the topmost tomato leaves of plants grown in soil 60 days post inoculation (DPI) (n = 4). Controls were inoculated with heat killed bacteria (n = 4). Significance was determined within treatments via ANOVA; letters correspond to a Tukey post hoc test.

Fig. 2. Microscopic analysis of Msb3 population dynamics after leaf inoculation.

Visualization of bacteria by DOPE-FISH/CLSM microscopy. The FISH probe Burkho was applied to leaves of tomato. The lower epidermis is depicted at different timepoints: a 24, b 48 and c 72 hours post inoculation with a highly diluted bacterial suspension. Plant autofluorescence was visualized through the cyan channel, the Burkho probe through the magenta channel and the nucleic acid stain SYBR-Safe through the yellow channel. Panels a–f all show overlays of all three channels. Panels d, e and f show high-resolution images of the sections highlighted with a yellow frame in a, b and c, respectively.

Fig. 3. Quantitative analysis of Msb3 and PsJN population dynamics on different tissues via qPCR.

Tomato plants were grown in vitro in gnotobiotic systems. One third of the plants was inoculated with Msb3 (left panels), one third with PsJN (middle panels) and one third was not inoculated functioning as control (right panels). Each week (x-axes) some plants were harvested, divided into phyllosphere (upper panels) and rhizosphere (lower panels) sections and submitted to DNA extractions. qPCRs were performed on the single copy marker gene gyrB with different, species specific primers for both Msb3 and PsJN. Control samples were subjected to amplification with Msb3 specific primers. Shown are the log10 copy numbers per ng of total DNA (y-axes). Significance was determined between all timepoint and treatments via ANOVA; letters correspond to a Tukey post hoc test.

Fig. 4. Colonization patterns of strain Msb3 on tomato.

Visualization of bacteria by DOPE-FISH/CLSM microscopy. Plant autofluorescence was visualized in cyan, signals representing the Burkho probe in magenta and that of the nucleic acid stain SYBR-Safe in yellow. a, d and f are composite images of all three channels, b, c and e are overlays of the magenta and the yellow channel only, as the yellow channel was usually sufficient to visualize epidermal autofluorescence in leaf samples. a, b, c and e are images of the lower leaf epidermis, d and f of roots. a and b show stomata colonized by Msb3, c a trichome. d shows a root crack filled with labelled bacteria, f shows a non-colonized 1st degree side root with root hairs that are colonized at the tips.

Fig. 5. Phenotypic changes in tomato and Arabidopsis in response to Msb3 or Msb3ΔacdS.

a Number of 1st degree side roots of tomato seedlings grown in Johnson medium (JM) 7 DPI with Msb3 or the acdS deficient mutant of Msb3 and a non-inoculated control (n = 16). Significance was determined within each treatment via ANOVA; letters correspond to a Tukey post hoc test. b number of 1st degree lateral roots of tomato seedlings grown in Johnson medium (JM) that contained 400 nM of ACC with the same conditions and treatments as in a (n = 8). Significance was determined within each treatment via ANOVA; letters correspond to a Tukey post hoc test. c Cumulative length of the 1st degree side roots of non-inoculated samples in JM or JM containing 400 nM of ACC. Significance was determined within each treatment via ANOVA; letters correspond to a Tukey post hoc test. d CFUs recovered from whole tomato shoots inoculated with fluorescently labelled derivative strains of either Msb3 wt or the acdS deficient mutant. Two Msb3 derivative strains (n = 3 each) make up the red bar, three Msb3ΔacdS derivative strains (n = 5 each) make up the yellow bar. The jittered shapes correspond to the respective derivative mutant: pluses: Msb3::eGFP2; checked boxes: Msb3::mScarlet; circles: Msb3ΔacdS::eGFP2.1; triangles: Msb3ΔacdS::eGFP2.2; squares: Msb3ΔacdS::mScarlet. e Lengths of different phenotypic parameters of tomato seedlings grown in Johnson medium (JM) 7 DPI with Msb3 or the acdS deficient mutant of Msb3 and a non-inoculated control (n = 8). Significance was determined within each treatment and between all parameters via ANOVA; letters correspond to a Tukey post hoc test. f Binarized images of representative seedlings of tomato (left) and Arabidopsis (right) to illustrate which phenotypic parameters were considered. g Primary root elongation of Arabidopsis seedlings, either the Arabidopsis wt Col-0 or the ethylene insensitive mutant line ein2-1, grown with RGI-inducing hormonal treatments (100 nM ACC), individually (none) or with Msb3 or the acdS deficient mutant of Msb3 (n = 8). Significance was determined within each treatment via ANOVA; letters correspond to a Tukey post hoc test.

Fig. 6. Effect of Msb3 and Msb3ΔacdS on tomato in the presence of a native microbial community and inoculated via foliar spray.

a Hypocotyl length and b, weight of tomato seedlings grown in non-treated soil and exposed to the environment at all times. Measurements were taken 7 DPI with Msb3 or the acdS deficient mutant of Msb3 (ΔacdS). Two modes of inoculation were compared. Seedlings were either dipped into a suspension of bacterial cells (OD₆₀₀ = 0.01) in 10 mM MgCl₂ during transplanting or were sprayed with it after transplanting. Controls were dipped into an MgCl₂ solution without inoculant (n = 20 for each mode of inoculation and each condition, respectively). Significance was determined within each treatment and between all parameters via ANOVA; letters correspond to a Tukey post hoc test. Red dots represent the mean.