Fungal, yeast, and bacterial strains and culture conditions
Rahnella aquatilis strain 36 (Ra36) was previously isolated from chickpea roots23. Escherichia coli and Ra36 were routinely grown at 37 °C or 28 °C, respectively, in Luria Bertani (LB) medium (5 g/L yeast extract, 10 g/L peptone, 10 g/L NaCl, pH 7.0) containing kanamycin sulfate (50 µg/mL), ampicillin (250 µg/mL) or carbenicillin (100 µg/mL), where appropriate. Saccharomyces cerevisiae was grown in yeast extract peptone dextrose (YPD; 1% Bacto yeast extract, 2% bacto peptone, and 2% dextrose), except for the selection of recombinants which was carried out in yeast nitrogen base medium (Sigma-Aldrich, Madrid, Spain) containing the appropriate amino-acid supplements.
F. oxysporum f. sp. lycopersici race 2 isolate 4287 (FGSC9935; Fol) or previously reported Fol mutants either lacking the ste2 gene6 or constitutively expressing the GFP7 were used throughout the work. For microconidia production, cultures were grown in potato dextrose broth (PDB) at 28 °C with shaking at 170 rpm38. For phenotypic analysis of Fol colony growth, serial dilutions (106, 105, 104, and 103 mL−1) of freshly obtained microconidia were spotted onto YPD agar plates supplemented or not with 0.8% (w/v) GlcA (Sigma-Aldrich), incubated at 28 °C for 3 d and imaged. Experiments included three replicates and were performed at least three times with similar results. Microconidial and bacterial cell suspensions were routinely stored at −80 °C with 30% glycerol.
Bioinformatic analysis
Identification of gcd and fliC gene orthologs from different R. aquatilis strains was performed with the BLAST algorithm39, using the R. aquatilis HX2 gcd and fliC genes as baits. Nucleotide sequences encompassing the entire gcd and fliC genes and comprising 1.5-kb upstream and downstream of the retrieved ORFs were aligned using the BioEdit 7.0.0 software (Ibis Bioscience, Carlsbad, CA, USA). Primer pairs Gcd7/Gcd8 and Flic7/Flic8 were designed on conserved upstream and downstream gene regions and used for amplification and sequencing of the two genes from Ra36. A complete list of the primer sequences used in this study is provided in Supplementary Table 1.
Nucleic acid manipulations
Isolation of genomic and plasmid DNA from yeast and bacterial cells was performed with the Puragene Yeast/Bact. Kit B and QIAprep Spin Miniprep Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Total genomic DNA from tomato seedlings was extracted by using the DNeasy Plant Mini Kit (Qiagen) according to the manufacturer’s instructions. Routine nucleic acid manipulations were performed according to standard protocols40.
Rahnella aquatilis mutant generation
The Ra36 gcd− and fliC− mutants were generated by targeted replacement with the kanamycin resistance marker. DNA fragments flanking the gcd and fliC coding regions were amplified from gDNA of Ra36 using primer pairs Gcd1+Gcd2, Gcd3+Gcd4 and Flic1+Flic2, Flic3+Flic4, respectively. The kanamycin resistance marker was amplified from the pET28-c (+) plasmid (Novagen, Merck, Darmstadt, Germany) using primers Kanf+Kanr. The S. cerevisiae recombination cloning method41 was used to assemble the complete deletion cassette in the pRS246 vector42 linearized with EcoRI/HindIII, by co-transformation with 1 µg of each of the four DNA fragments into the yeast strain FY83443. The resulting plasmids pRS426.1 and pRS426.2 were isolated from yeast and cloned and propagated in the E. coli TOP10 strain (Invitrogen, Barcelona, Spain) using standard protocols40. The gcd− and fliC− deletion alleles were amplified from pRS426.1 and pRS426.2, respectively, with primers PRS426f+PRS426r and used to transform cells of the Ra36 wild-type strain. Kanamycin-resistant transformants showing homologous insertion in the gcd and fliC coding regions were identified by PCR of gDNA with primers Gcd5+Gcd6 and Flic5+Flic6, respectively. Allelic replacement of the gcd and fliC genes was further confirmed by DNA sequencing of a PCR fragment amplified from gDNA of the transformants with primers Gcd7+Gcd8 and Flic7+Flic8, respectively.
For fluorescence microscopy analysis, a red fluorescent strain of Ra36 was obtained by transforming bacterial cells with 5 µg of plasmid FPB-31-444 (ATUM Bio, Newark, CA, USA) carrying the red fluorescent protein Rudolph-RFP and an ampicillin resistance cassette, using a Bio-Rad Gene Pulser electroporator (Bio-Rad, Madrid, Spain) set at 1.8 kV, 200 W and 25 μF and 0.1 cm cuvettes.
Extracellular pH determination and GlcA quantification
For visual evaluation of Ra36 acidifying activity, 5 µL drops containing 2.5 × 106Ra36 or Ra36 gcd− colony forming unit (CFU) were spot-inoculated onto plates containing minimal medium with urea (MMU; 15 g L−1 glucose, 0.5 g L−1 MgSO4·7(H2O), 0.5 g L−1 KCl, 1.0 g L−1 KH2PO4, 1.5 g L−1 urea, 20 g L−1 agar) adjusted to pH 7. After 2 days of incubation at 28 °C, plates were overlaid with a 0.833 μM solution of the pH indicator bromocresol green or purple and imaged. Experiments were performed at least four times, with two replicates each.
To assess a possible inhibitory effect of Ra36 on Fol colony growth and to visualize Ra36-mediated alkalinisation of Fol colony margins, 5 µL drops containing 5 × 104Fol microconidia and 2.5 × 106 CFU mL−1Ra36 or Ra36 gcd− were spot-inoculated at a distance of 30 mm on MMU plates adjusted to pH 7.0. Visual analysis of pH change was performed by supplementing media with bromocresol purple (0.833 μM). Plates were incubated at 28 °C for 4–7 d before imaging. All fungus–bacteria confrontation assays were performed in three or more independent experiments with three replicates each.
For plate visualization of rhizosphere pH dynamics12 during the Fol-Ra36–plant interaction, 2-week-old tomato seedlings (Solanum lycopersicum cv. Moneymaker; used throughout the study) were either left untreated or inoculated by dipping roots into a suspension of Ra36 or Ra36 gcd− cells in water (1 × 109 CFU mL−1). Seedlings were then transferred onto 0.5% water agar plates supplemented with bromocresol purple (0.833 μM) and containing either Fol microconidia (1.25 × 106 mL−1) or water (negative control). Plates were incubated at 25 °C with 14 h light and 10 h dark up to 3 d before imaging. Experiments included three replicate plates and were performed at least three times with similar results.
To record pH dynamics in liquid media, Fol microconidia (5 × 105 conidia mL−1) were pre-germinated 14 h in MMU medium with shaking (170 rpm) at 28 °C, washed with sterile ddH2O and shifted to fresh MMU medium adjusted to pH 7. Then, 5 × 106 CFU mL−1Ra36 or Ra36 gcd− were added and cultures were incubated at 28 °C and 170 rpm. Each hour, aliquots of culture supernatant were withdrawn, and pH was measured in a pH meter. Cultures containing either Fol or Ra36 alone were used as controls. Experiments were performed on three to four independent occasions with three replicates each.
To monitor pH dynamics in liquid media in the presence of the plant, 5 × 105 mL−1Fol microconidia and 5 × 106 CFU mL−1Ra36 or Ra36 gcd− were inoculated in sterile H2O. Then, a 2-week-old tomato plant was added to keep only the root submerged in the liquid. The culture was incubated, aliquots of culture supernatant were collected, and pH was measured as described above. Cultures containing either Fol, Ra36, or tomato plants alone were used as controls. Experiments were performed on three to four independent occasions with three replicates each.
To investigate rhizosphere pH dynamics in soil, 3-week-old tomato seedlings were dip-inoculated with a suspension of Ra36 or Ra36 gcd− cells (1 × 109 CFU mL−1), transferred to non-sterilized or sterilized (autoclaved 20 min at 120 °C) horticultural soil obtained from a tomato field located in Riccia, Molise (Italy) and maintained in a growth chamber (14⁄10 h light⁄dark cycle) at 28 °C. After 2 weeks, 10 plants per treatment were collected, roots were cleaned from the soil, submerged in 5 mL sterile ddH2O for 12 hours at 28 °C and extracellular pH was measured in a pH meter. Uninoculated plants were used as controls. Experiments were performed on four independent occasions with three replicates each.
D-GlcA concentrations in culture supernatants or tomato root exudates described above was determined using the K-GATE detection kit (Megazyme International, Bray, Ireland) according to the manufacturer’s instructions. For all experiments, GlcA measurements were performed on three to four independent occasions with at least three replicates per treatment.
Virulence-related assays
Assays for the evaluation of Fol invasive growth on living plant tissue7 were performed using tomato fruits (cultivar Rovente). In brief, the epidermis of surface sterilized tomato fruits was punctured with a pipette tip and 10 µL of bacterial suspension (109 CFU mL−1) or water (control) was injected into the fruit tissue. After 30 min, 10 µL of a suspension of Fol microconidia (5 × 108 mL−1) or water (control) was injected into the same inoculation site. Fruits were incubated at 28 °C in a humid chamber and infection sites were imaged 5 days after inoculation (DAI). Experiments were performed on three independent occasions with three replicates each.
Cellophane invasion assays9 were performed on MMU plates. In brief, 100 µL of a Ra36 cell suspension (OD600 of 0.1) or water (control) was uniformly spread on the culture medium before placing an autoclaved cellophane sheet on the plate (colorless; Manipulados Margok, Zizurkil, Spain). Next, 2 × 105Fol microconidia were spot-inoculated at the center of the plate. After 4 days at 28 °C, plates were imaged (before), then the cellophane membrane was carefully removed, and plates were incubated for one additional day at 28 °C and imaged (after). MMU was either unbuffered or buffered to pH 5 or 7 with 100 mm 2-(N-morpholino) ethanesulfonic acid (MES). Experiments included three replicate plates and were performed three times with similar results.
For analysis of hyphal aggregation, fungal (2.5 × 106 conidia mL−1) and bacterial (108 CFU mL−1) strains were co-cultivated for 48 h in liquid MMU at 28 °C and 170 rpm. Experiments included four replicates and were performed at least three times with similar results.
For root adhesion assays7, entire roots of 2-week-old tomato seedlings were dip-inoculated with Ra36 as described above, placed in Erlenmeyer flasks containing a suspension of 107 mL−1Fol microconidia and incubated 3 days at 28 °C and 120 rpm. Fungal aggregates in liquid medium or on tomato roots were imaged in a Leica binocular microscope (Leica Microsistemas S.L.U., Barcelona, Spain) using a Leica DC 300 F digital camera. Experiments included four replicates and were performed at least three times with similar results.
Plant infection assays
Roots of 2-week-old tomato seedlings were dipped for 30 min into a suspension of Fol microconidia in water (5 × 106 conidia mL−1). For co-inoculation assays, plant roots were immersed for 2 h in a suspension of Ra36 cells (1 × 109 CFU mL−1) and then dip-inoculated with Fol microconidia (5 × 106 conidia mL−1). To test the effect of inoculum density, suspensions of Ra36 cells at different concentrations (1 × 1010; 1 × 109; 1 × 108; 1 × 107, and 1 × 106 CFU mL−1) were used. After root inoculation, tomato seedlings were planted in vermiculite or in non-sterilized or sterilized horticultural soil (see above), maintained in a growth chamber (14⁄10 h light⁄dark cycle) at 28 °C and irrigated either with unbuffered water or with a solution of 1 mM MES buffer adjusted to pH 5.0 or pH 7.0.
Soil inoculation was performed by mixing non-sterilized or sterilized horticultural soil with Ra36 cells and/or Fol microconidia at concentrations of 1 × 108 CFU and 4 × 105 conidia per g of soil, respectively.
Plant survival was recorded daily up to 55 days, calculated by the Kaplan–Meier method, and compared among groups using the log-rank test. All infection assays included fifteen plants per treatment and were performed at least three times with similar results. Data were plotted using the GraphPad Prism 5 software (GraphPad Software, La Jolla, CA, USA).
Detection of Fol and Ra36 in tomato plant tissue
Qualitative assessment of fungal and bacterial colonization of tomato plants was performed 2 weeks after dip-inoculation of 2-week-old tomato roots with a suspension of Ra36 cells (1 × 109 CFU mL−1) and/or Fol microconidia in water (5 × 106 conidia mL−1) as described above. Seedlings were removed from the vermiculite and the roots and stems were surface sterilized by submerging them into 2% sodium hypochlorite for 2 min. Plant tissues were rinsed twice with sterile distilled water and cut into 1 cm long sections, which were transferred onto MMU plates with or without 0.833 μM of the pH indicator bromocresol purple. Fungal or bacterial growth was recorded after 5 days of incubation at 28 °C. Experiments included three replicate plates per treatment and were performed at least three times with similar results.
Quantification of fungal and bacterial biomass in Fol- and/or Ra36-inoculated tomato plants grown in vermiculite was carried out by real-time qPCR19, using total gDNA extracted from tomato roots 15 days after dip-inoculation with a suspension of Ra36 cells (1 × 109 CFU mL−1) and/or Fol microconidia (5 × 106 conidia mL−1). To measure the effect of Fol in soil on the colonization ability of Ra36, 2-week-old tomato seedlings were planted in horticultural soil containing 2 × 104Fol microconidia g−1 soil. After 3 days, 50 µl of a suspension of Ra36 cells containing 8 × 108 CFU mL−1 were added at a distance of 5 cm from the hypocotyl. After 10 days, total gDNA was extracted from the tomato roots and biomass of Ra36 was measured by real-time qPCR.
Real-time qPCR reactions were performed with the SYBR® Premix Ex Taq™ (Takara Bio, Inc., Otsu, Japan) in a Eppendorf Mastercycler ep gradient S system (Eppendorf, Milan, Italy), using the primer pairs Gcd9 + Gcd10 (located outside of the deleted region of the Ra36 gcd gene), ACT2 + ACTQ6 (Fol actin gene) and GADPH1 + GADPH2 (tomato gadph gene) (Supplementary Table 1). Relative amounts of fungal and bacterial genomic DNA were calculated by comparative ΔΔCt with the tomato gadph gene. DNA concentrations in each sample were extrapolated from standard curves obtained by plotting the logarithm of known concentrations (10-fold dilution series from 10 ng to 1 pg/25 μL reaction) of Fol and Ra36 gDNA against the Ct values. To normalize the serially diluted DNA samples, 100 ng gDNA from non-inoculated plants was added to each sample of the dilution series. Real-time qPCR data represent the mean ± SE from three independent experiments, each with five plants per treatment.
Collection of exudates from tomato roots and Fol hyphae
To obtain tomato root exudate6, uninoculated or Ra36-inoculated tomato roots (1 × 109 CFU/mL) were placed in sterile ddH2O in the absence or presence of 5 × 105 microconidia mL−1 of Fol. After 48 h at 25 °C, the supernatant was sterilized by filtration through a 0.22-μm membrane (Merck Millipore) and stored at −20 °C until use.
To obtain Fol hyphal exudate4, 1 × 107 microconidia mL−1 were pre-germinated for 16 h in 50 mL diluted PDB (1:50; v:v in H2O) at 28 °C with shaking at 170 rpm. Germlings were washed twice with sterile ddH2O and incubated for 48 h in 5 mL sterile ddH2O at 28 °C and 170 rpm. The supernatant was sterilized by filtration through a 0.22-μm membrane (Merck Millipore) and stored at −20 °C until use. When required, the pH was adjusted to 5.8 with 0.1 N HCl.
Assays for fungal chemotropism and bacterial chemotaxis
Plate preparation, chemoattractant application, and scoring of Fol germ tube redirectioning toward gradients of root exudates obtained from untreated or Ra36-inoculated plants were performed by using a hyphal chemotropism assay6. In brief, 2.5 × 106Fol microconidia mL−1 were embedded in 4 ml water agar (WA; 0.5%, w/v) (Oxoid) and poured into a 9 cm Petri dish. Then, two parallel wells, each at 5 mm distance from the scoring line, were filled with 40 μL of the test compound or the solvent control solution. Scoring was done on five independent batches of cells (n = 100 cells per batch) for each test compound. Experiments were performed at least three times with similar results.
Chemotaxis capillary assays32 were carried out as follows. Ra36 wild-type, Ra gcd−, and Ra fliC− strains were grown in LB overnight at 28 °C, washed either with sterile ddH2O, PBS, tomato root exudate or Fol hyphal exudate depending on the experiment, and diluted in the same medium to an OD600 of 0.1. Aliquots of 250 μL bacterial suspension were added to individual wells of a 96-well microtiter plate together with a 10-µL capillary containing the test compound or the solvent control. For competing gradient assays, two capillaries containing the different test compounds were added to the well. Plates were incubated for 60 min at 28 °C, capillaries were carefully lifted, the content was serially diluted and plated onto LA medium, and CFUs were counted 48 h after incubation at 28 °C. The following chemoattractant compounds and concentrations were tested: glutamine (Gln), tryptophan (Trp), all at 295 mM; glucose (Gluc), galactose (Gal), all at 50 mM. The chemotaxis ratio was calculated by dividing the number of bacteria in the tube containing the test compound by the number of bacteria in the tube containing the solvent control. All experiments included two replicates and were performed at least three times with similar results.
Flagellum-dependent swimming motility assays44 were performed as follows. In brief, LB plates (0.3% w/v agarose) were spot-inoculated in the center with 5 μl of an overnight culture of the Ra36 wild-type or fliC− strain, incubated 24 h at 30 °C and colony radial growth was imaged. Experiments were performed three times, with four replicates each.
Fluorescence microscopy
For microscopic observation in tomato roots of GFP-tagged Fol and RFP-tagged Ra36 strains, 2-week-old tomato seedlings were dip-inoculated into suspensions of Fol microconidia (5 × 106 conidia mL−1) and/or Ra36 cells (1 × 109 CFU mL−1), planted in moist vermiculite and maintained in a growth chamber (14⁄10 h light⁄dark photoperiod) at 28 °C. Two and four DAI, roots were gently washed to remove the adhering vermiculite, incubated in a 95% perfluorodecalin solution (Sigma-Aldrich), stained for 2 min with 0.005% (w/v) calcofluor white (CFW; Sigma-Aldrich) to visualize the plant cell wall and imaged. Experiments were performed three times, with two replicates each.
To observe Ra36 swarming toward Fol hyphae, 2 µL drops containing 5 × 104Fol-GFP microconidia or 2.5 × 106 CFU Ra36-RFP were spot-inoculated at a distance of 5 mm on a 1 × 1 cm square pad of soft (0.25% w/v) agarose MMU medium placed on top of a microscope glass slide. Images were recorded every 30 min up to 4 h post inoculation. Experiments were performed three times, with two replicates each.
To study bacterial movement along fungal hyphae, 2 µL drops containing 5 × 104Fol-GFP microconidia or 2.5 × 106Ra36-RFP CFU were spot-inoculated at a distance of 30 mm on MMU medium plates. Microscopic observation of Fol hyphae was performed 48 h after fungal and bacterial colonies had merged. Experiments were performed four times, with two replicates each.
To determine the movement of Ra36 and Fol hyphae across a medium-free space, 5 × 104Fol-GFP microconidia and 2.5 × 106Ra36-RFP CFU were spot-inoculated either individually or together on soft (0.7% w/v) agarose MMU plates. Next, a 2-day-old tomato seedling was placed at a distance of 15 mm, and a 5-mm wide medium-free gap was created between the inoculation point and the tomato root by removing the medium with a sterile spatula. Microscopic observation was performed daily up to 3 DAI to follow Ra36 and Fol dynamics over time. To qualitatively assess the presence of Ra36 and Fol on the plant roots, tomato seedlings were gently removed from the plate at 2 DAI, rinsed under sterile H2O and stained with CFW, as described above. Experiments were performed at least three times on two or more separate days.
Low-resolution imaging was performed using a SteREO Lumar.V12 fluorescence stereomicroscope (Zeiss, Barcelona, Spain). Wide-field fluorescence imaging was performed using a Zeiss Axio Imager M2 Dual Cam microscope (Zeiss) equipped with a Photometrics Evolve EM512 digital camera (Photometrics Technology, Tucson, AZ, USA). Examination using epifluorescence (×400 magnification) was performed with the following filter blocks: CFW staining (G 365, FT 395, LP 420), RFP (BP 560/40, FT 585, BP 630/75), GFP (BP 450/490, FT 510, LP 515). Images were captured and processed using Axiovision 4.8, ZEN lite 2.3 (both from Zeiss) or ImageJ (v1.52)45.
Statistical analysis
Percentage of plant survival was compared among treatment groups using the log-rank test. For multiple-group comparisons, a one-way analysis of variance was used for testing no differences among the group means. Post hoc comparisons were adjusted using Dunnett’s or Tukey’s corrections (chemotaxis and real-time qPCR data). Comparisons between two groups were carried out using a two-tailed unpaired Student’s t test (chemotaxis data). A Yates’ corrected Chi-squared test (two-sided) was used to determine significant differences between the observed frequencies of fungal germ tubes pointing toward the chemoattractant or the solvent control (chemotropism data).
Real-time qPCR data are presented as mean ± SE. Data from chemotaxis and chemotropism experiments and from pH and GlcA measurements are presented as mean ± SD.
Statistical analyses were performed using GraphPad Prism 5 software (GraphPad Software). In all cases a value of P < 0.05 was considered statistically significant.
Reporting summary
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
Source: Ecology - nature.com
