AGRICULTURAL METHODS

Abstract

A method for introducing a plant growth mediating entity or substance into plants and in particular a nitrogen-fixing bacteria into plant cell, said method comprising administering said plant growth mediating entity or substance into plants to a plant in combination with a strain of Terribacillus.

Compositions and bacteria for use in the method form a further aspect of the invention.

Claims

1. A method for introducing a plant growth mediating entity or substance into plants, said method comprising administering said plant growth mediating entity or substance to a plant in combination with a strain of Terribacillus.

2. The method of claim 1 wherein the strain of Terribacillus is a Terribacillus saccharophilus.

3. The method of claim 1 wherein the strain of Terribacillus saccharophilus comprises any one of SEQ ID NOs 1-4.

4. The method of claim 1 wherein the plant growth mediating entity or substance is a nitrogen-fixing bacteria.

5. The method of claim 4 wherein the nitrogen-fixing bacteria is a bacteria which forms a levan coat.

6. The method of claim 4 or claim 5 wherein the nitrogen fixing bacteria is Gluconacetobacter diazotrophicus (Gd).

7. The method of claim 6 wherein the Terribacillus is intimately associated with the Gd prior to administration.

8. The method of claim 1 wherein a combination of Terribacillus and a plant growth mediating entity or substance is administered to a growing plant.

9. The method of claim 8 wherein the plant is subjected to a wounding process prior to administration of said combination.

10. The method of claim 1 wherein a combination of Terribacillus and nitrogen-fixing bacteria is administered to a seed.

11. An agricultural composition comprising a strain of Terribacillus.

12. The agricultural composition of claim 11 which further comprises a nitrogen-fixing bacteria.

13. The agricultural composition of claim 12 wherein the nitrogen fixing bacteria is Gluconacetobacter diazotrophicus (Gd).

14. The agricultural composition of claim 11 wherein the strain of Terribacillus is a Terribacillus saccharophilus, Terribacillus halophilus, Terribacillus goriensis or Terribacillus aidingensis.

15. The agricultural composition of claim 13 wherein the strain of Terribacillus saccharophilus comprises any one of SEQ ID NOS 1-4.

16-26. (canceled)

27. A plant or seed which is colonised by Terribacillus.

28. The plant or seed of claim 27 which is a progeny of a plant or seed which is colonized by Terribacillus.

29. The plant or seed of claim 28 which further comprises a nitrogen fixing bacteria.

30. The plant or seed of claim 29 wherein the nitrogen fixing bacteria is Gluconacetobacter diazotrophicus (Gd).

31. The agricultural composition of claim 12 which is obtained by co-culturing a strain of Terribacillus and a strain of nitrogen fixing bacteria together in a medium which supports the growth of both strains.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0058] The invention will now be particularly described by way of example with reference to the accompanying diagrammatic drawings in which:

[0059] FIG. 1A shows PCR Fingerprints of samples of DNA extracted using a QIAamp DNA Stool kit from the nitrogen fixing bacteria Gd, where the PCR was carried out using the BOX primer.

[0060] Lanes left to right represent 1=100 bp ladder, 2,3=LN rep 1; 4,5=LN rep2; 6,7=LN rep 3; 8,9=LN rep 4; 10,11=LN rep 5, 12,13=FD rep A1, 14,15=FD rep A2, 16,17-FD rep A3; 18=control; 19=100 bp ladder; where LN=liquid nitrogen, FD=Freeze dried.

[0061] FIG. 1B shows PCR Fingerprints of samples of DNA extracted using a QIAamp DNA Stool kit from the nitrogen fixing bacteria Gd, where the PCR was carried out using the BOX primer. Lanes left to right represent 1=100 bp ladder, 2,3=FD rep A4; 4,5=FD rep A5; 6,7=LN rep 1; 8,9=LN rep 2; 10,11=LN rep 3, 12,13=LN rep 4, 14,15=LN rep 5, 16,17=unpreserved; 18=control; 19=100 bp ladder; where LN=liquid nitrogen, FD=Freeze dried.

[0062] FIGS. 2A-D show PCR fingerprints of clonal single colony samples of DNA extracted from bacteria using a Qiagen DNeasy Tissue kit, and amplified using BOX primers wherein in all sets, the lanes running left to right represent: 100 bp ladder; colony 1 triplicate reactions; colony 2 triplicate reactions; colony 3 triplicate reactions; colony 4 triplicate reactions; No DNA control; 100 bp latter. FIG. 2A, day 1. FIG. 2B, day 2. FIG. 2C, day 3. FIG. 2D, day. 4.

[0063] FIG. 3 shows multiple sequence alignments of the organism found in this work and Terribacillus species where DD2BR, DD1BR, DD3BR and DD4BR are the sequences isolated in this work and represent SEQ ID NOS 1-4 respectively, and where AB243849 is the corresponding Terribacillus halophilus sequence (SEQ ID NO 9) and AB243845 is the corresponding Terribacillus sacchrophilus sequence (SEQ ID NO 10).

[0064] FIG. 4 shows a comparison of growth curve for Gd (A, low oxygenation; , high oxygenation) and Ts () in PDB, along 52 hours. Cell number for Gd culture with high oxygenation as log.sub.10 (cfu/ml) (). OD as the average of duplicate samples. Cell number at each time is represented as the average of three dilutions plated by duplicate. Error bars as standard deviation from every plates.

[0065] FIG. 5 shows a growth curve for Ts in LB. Cell number for Ts culture as log.sub.10 (cfu/ml) (). OD is the average of duplicate samples (). Cell number at each time is represented as the average of three dilutions plated by duplicate. Error bars as standard deviation from every plates.

[0066] FIG. 6 is a graph showing a comparison of the growth of Gd, Ts and a mixed culture of Gd+Ts in different media, with and without inclusion of Herbst's artificial seawater (H's).

[0067] FIGS. 7A-D show the results of competence assays over 120 hours. FIG. 7A shows growth in MYP as OD at 600 nm. FIG. 7B shows growth in SYP as OD at 600 nm. FIG. 7C shows viability of cells grown in MYP as colony forming units per ml (cfu/ml). FIG. 7D shows viability of cells grown in SYP as colony forming units per ml (cfu/ml).

[0068] FIG. 8 shows the correlation between ethylene production by Gd single cultures and Gd+Ts mixed cultures, in different proportions of both strain Gd:Ts as compared to their OD at 600 nm, as follows: () Gd:Ts 2:1; () Gd:Ts 1:1; () Gd:Ts 1:2. Both, trendline as regression equation are labelled with the pattern shown by the legend.

[0069] FIG. 9 shows levan production by single (G) and mixed cultures (GT) in two different media, MYP and SYP, along 15 days post-inoculation (dpi). Average of two repetitions. Error bars represent standard deviation for the average value. Groups of significant difference between each sample along time (a, b and , ) and between different samples at each sampling time (A, B, C, D) are labelled.

[0070] FIG. 10 shows nifH and lsdA relative expression for mixed cultures (Gd+Ts) in relation to single culture (Gd) in two different media (SP and MP). Gene expression was quantified using RT-qPCR. 23S was used as reference gene for data normalization. For each condition, C.sub.T was measured from three repetitions.

[0071] FIG. 11 is a graph showing the results of an experiment to determine the ability of Gd, Ts and mixed cultures of both strains, in different proportions, to attach an artificial surface. Data are the average of five independent samplesstandard errors (se) of the means. Values with common letter are not significantly (F0.05) different according to LSD test.

[0072] FIG. 12 is a graph showing the number of cells recovered from the surface of seeds after inoculation, as log.sub.10 of cfu per seed. Data are the average of three replicatesse of the means. Values with common letter are not significantly (F0.05) different according to LSD test.

[0073] FIG. 13 is a graph showing colonisation of OSR plants by Gd, 2 weeks post-inoculation and after incubation of the plants in a plant growth chamber. In this figure OREP represents oilseed rape root epiphytic result, OREN represents oilseed rape root endophytic result and OLEN represents oilseed rape leaf endophytic results. Data are the average of cfu counted in four plates per sample (as log.sub.10 of cfu/ml)se of the means. Values with common letter are not significantly (F<0.05) different according to LSD test.

[0074] FIG. 14 is a graph showing colonisation of OSR plants by Ts, after 2 weeks post-inoculation and incubation of the plants in a plant growth chamber (Fitotron). Data are the average of cfu counted in four plates per sample (as log.sub.10 of cfu/ml)se of the means. Values with common letter are not significantly (F<0.05) different according to LSD test.

[0075] FIGS. 15A-C graphically show results of an experiment measuring plants weights after various treatments: FIG. 15A shows the wet weight measured, FIG. 15B shows the dry weight measured and FIG. 15C shows the derived weight of water. Data are the average of seven plants per samplese of the means. Values with common letter are not significantly (a and c, F<0.05; b, F=0.1) different according to LSD test.

[0076] FIG. 16 is a graph showing levan quantification for Gd, Ts and mixed cultures in SYP after 3 days. Data are the average of three repetitionsse of the means. Values with common letter are not significantly (F<0.001) different according to LSD test.

[0077] FIG. 17 is a graph showing the effect of Terribacillus on the production of the plant hormone, IAA, on Gd. The values show the average between three repetitions and the error bars show the standard error (se). Samples with the same letter has no significant differences according with LSD test (F<0.001)

EXAMPLE 1

Identification of Combination

[0078] Work was undertaken to characterise strains of Gd. The strains used included a test strain which had been cultured from IMI 501986, (now IMI 504998) available from Azotic Technologies Ltd and CABI UK, over a long period of time and which was known to have good plant cell colonisation properties.

[0079] The strains were grown initially on ATGUS medium ([0.8% (w/v) agar, yeast extract (2.7 g L.sup.1), glucose (2.7 g L.sup.1), mannitol (1.8 g L.sup.1), MES buffer (4.4 g L.sup.1), K.sub.2HPO.sub.4 (0.65 g L.sup.1), pH 6.5], and incubated at 25 C. It was found however that the use of ATGUS broth improved the quality and reproducibility of the strains.

[0080] PCR fingerprinting was undertaken on pre- and post-preservation samples (including cryopreserved and freeze dried samples of the strains. DNA extraction was carried out using the Qiagen DNeasy Plant, DNeasy tissue or QIAamp DNA Stool kits, used in accordance with the manufacturers' instructions. The DNA concentration of each sample was assessed by spectrophotometer (GeneQuant, GE Healthcare, UK). Concentration was then standardised for each sample to 10 ng/l.

[0081] PCR fingerprinting using bacterial repeat unit BOX & ISSR TGT primers was undertaken and gel electrophoresis used to separate the DNA fragments. All PCR reactions were undertaken in duplicate and repeated.

[0082] DNA was successfully extracted from all preserved and wild-type samples. However, the PCR results with both the Box and TgT primers for the test strain were inconsistent. It was expected that the banding patterns should be 100% similar across the separate replicates, but although some bands appeared common to all samples, some were distinctly different with extra bands appearing. The intensity of like bands also differed between different treatments. The results obtained with using Box primer were generally better and contained up to 8 distinct bands.

[0083] The work was repeated using DNA extracted using the QIAamp DNA Stool extraction kit. This gave fingerprints of better quality and more consistency. However, although there were common bands appearing in all samples, there were clear differences in banding profile between some replicates of the same sample (FIGS. 1A-1B). Profiles should be identical across replicates, but especially so within duplicate reactions of the same replicate. This was not the case as is clear from a comparison of the banding patterns in boxes which are associated with replicates of the same sample.

[0084] Partial 16S rDNA sequence analysis of the strains carried out at this time indicated that the organism was predominantly Gd.

[0085] Due to the unexpected irreproducibility of the profiles, the methods were repeated in a modified manner and using four single colonies derived from a single line of the Gd in order to ensure that the samples were identical. The colonies were grown in ATGUS broth and DNA extracted from each of the four colonies using the Qiagen DNeasy tissue kit Gram positive bacteria protocol in order to remove any problems due to any potential inhibitory or protective effects of Gd levan. The samples were subjected to BOX PCR (using triplicate reactions and over four separate days) and reproducible and uniform profiles were observed (FIGS. 2A-2D).

[0086] Surprisingly, subsequent partial 16S rDNA sequence analysis of the monocolony-derived samples gave >99% matches to Terribacillus spp., most closely matching T. saccharophilus (see FIG. 3)

EXAMPLE 2

Separation of Strains

[0087] Selective culturing of the test strain at different pH and NaCl concentrations were attempted in order to try to isolate both organisms. However, nothing grew on the NaCl media but growth was observed on the pH 5.5 agar and two morphotypes were noted on the pH 9.5 media. Information from the literature suggested that Gd would not grow at the higher pH and so it was thought that the secondary organism had been isolated. However, partial 16S rDNA sequence analysis revealed that both isolates were Gd.

[0088] Subsequent Gram staining was not conclusive, but revealed the predominance of a gram negative rod (i.e. Gd) with possible traces of a gram positive rod.

[0089] The more acidic pH conditions should favour the growth of Gd while Terribacillus should tolerate a higher pH (i.e. more alkaline) conditions more readily. It was not possible in this work to separate the Terribacillus from the Gd as indicated by the sequencing work. The fact that Gd remained the predominant strain suggests that the Terribacillus is present in the overall population of cells in very low numbers. However, in this experiment, the Gd appeared to tolerate more alkali conditions than previously described. This may be due to the presence of the Terribacillus which is acting in a mutualistic way to achieve this. In a similar way, the Terribacillus is believed to be responsible for the enhanced plant cell colonisation activity of this particular strain of Gd.

[0090] Furthermore, the difficulties encountered in separating the strains under certain circumstances is indicative of the very close association formed between these two strains. Without being bound by theory, it seems possible or even probable that the sugar-loving property of the Terribacillus means that it can become attached to or incorporated into the levan of the Gd. Alternatively, the Gd may by using the levan of the Terribacillus as a carbon source.

EXAMPLE 3

Co-Culture of Strains of Terribacillus and Gd

[0091] The strains used in these experiments were Gluconacetobacter diazotrophicus, (IMI504958CABI (UK)) and Terribacillus saccharophilus, strain provided by CBS Biodiversity Centre (AZ0008), Terribacillus goriensis (AZ0007), Terribacillus halophilus (AZ0009) and Terribacillus aidingensis (AZ0010). A range of media were used to grown the strains either individually or in combination.

[0092] A range of media, in particular those known to support the growth of Gd were tested to see whether they may also support Ts growth.

[0093] These included: [0094] 1. Potato Dextrose Broth/Agar (PDB/PDA, Fluka). [0095] 2. ATGUS (2.7 g L.sup.1 glucose, 0.65 KH.sub.2PO.sub.4, 4.8 g L.sup.1K.sub.2HPO.sub.4, 1.8 g L.sup.1 mannitol, 4.4 g L.sup.1 2-(N-morpholino)ethanesulfonic acid (MES hydrate), 2.7 g L.sup.1 yeast extract) (Cocking, et al., 2006.) [0096] 3. Marine Agar (Difco)

[0097] In each case, starter cultures were prepared in 100 ml of broth (PDB for Gd, LB for Ts) from colonies in fresh plates and incubated at 28 C. with shaking at 150 rpm. 10 l of the normalized cultures at OD.sub.600 0.1 were inoculated into centrifuge tubes with 5 ml of sterile medium, in duplicate, and incubated in the same conditions. Single tubes were used per sample, and were discarded after each use. Samples were taken at 0, 4, 8, 16, 20, 24, 28, 32, 40, 44, 48, and 52 hours after inoculation, OD.sub.600 was measured and serial dilutions plated on PDA or Marine Agar to count number of viable cells (cfu/ml).

[0098] The results after 4 days are shown in the following Table in which different levels of growth are showed as: +++ very high growth, ++ (high growth), + (growth), (+) (not clear growth), (no growth).

TABLE-US-00001 Strain ATGUS Marine Agar PDA AZ0007 ++ (+) AZ0008 ++ (Pale yellow + (White colonies) colonies) AZ0009 ++ (+) AZ0010 ++ IMI 504958 +++ +++

[0099] It was clear that ATGUS would not support Terribacillus growth whilst Marine Agar supported Terribacillus but not Gd. Growth of Terribacillus in PDA was not clear or reliable.

[0100] A growth curve in PDB was then constructed over 52 hours of incubation under the conditions described above and this is shown in FIG. 4. The growth of Gd was affected by the oxygenation of the culture, being slower at lower oxygenation level.

[0101] However, Ts was not able to grow in PDB after several days of incubation. It did however grow successfully in Luria-Bertani Broth (LB) (Sambrook, J. & Russel, D., 2001. Molecular Cloning: A Laboratory Manual. New York: Cold Spring Harbour Laboratory Press), as shown in FIG. 5.

[0102] Gd growth was tested in LB. Also, growth in Nutrient Agar (NA) was checked for both strains. After one week Gd did not grow on NA, nor in Marine Agar Gd grew poorly on LB.

[0103] Attempts were made to modify the media to see if they would support additional strain growth. In particular, sucrose, reported to be essential for Gd growth was added to LB. Similarly, PDB was modified by addition of sodium chloride to try to ensure that it would support Ts growth. However, the results were not successful in so far as Ts showed no ability to grow in PDB, even with a high concentration of salt, while it showed growth in LB modified with sucrose after 48 hours of incubation. Gd was affected by the salt concentration in PDB, and it was able to grow at 1% (w/v) NaCl, but not at higher salinity. It was not able to grown in LB supplemented with sucrose, suggesting a lack of compatibility in these media.

[0104] Different media were tested to see if Gd and Ts would grow together. The media were:

[0105] MYP (25 g L.sup.1 d-mannitol, 3 g L.sup.1 yeast extract, 3 g L.sup.1 peptone; pH 6.2) (Yamada, Y. et al., 1997. Biosiciences, Biotechnology and Biochemistry, Volume 61, pp. 1244-1251); and

[0106] SYP (10 g L.sup.1 sucrose, 3 g L.sup.1 yeast extract, 1 g l.sup.1 K.sub.2HPO.sub.4, 3 g L.sup.1 KH.sub.2PO.sub.4; pH 5.8) (Silva-Froufe, L., et al., (2009) Brazilian Journal of Microbiology, 40 (4), pp. 866-878).

[0107] Different carbon sources were tested with these two media, sucrose (SYP) and mannitol (MYP). To improve the growth of Ts, all suspension media were supplemented with 0.5 Herbst's artificial seawater, as recommended by An, S. et al., 2007 International Journal of Systematic and Evolutionary Microbiology, Volume 57, pp. 51-55. Media were tested with and without Herbst's solution for both strains Ts and Gd. The same was made for PDB.

[0108] Culture growth was measured with respect to the OD at =600 nm. Also viable cells were checked by plating on ATGUS and/or Marine Agar and Gd and Ts detected. In the case of cultures with Gd, the cultures were plated also onto LGIP to test the ability of nitrogen fixing from these cultures.

[0109] The graph in FIG. 6 shows that both strains grew well in SYP and MYP. Ts was unable to grow in PDB, except after addition of the salt solution and after a long incubation, but in this case Gd did not grow. Gd did not show growth in any media modified with artificial seawater. The mixed culture presented higher growth than single cultures without salt solution, which is likely due to the growth of both strains in the medium, as is clear from the cell viability tests, in which cells were plated on optimal growth medium (ATGUS for Gd, and Marine Agar for Ts) and incubated for 4 days. The results are presented in Table 1 below.

TABLE-US-00002 TABLE 1 Cell viability from cultures in modified media for 24 hours and for 48 hours. In mixed cultures, the viability of the cells are represented as Gd/Ts. Different levels of growth are showed as: +++ (very high growth), ++ (high growth), + (growth), (+) (not clear growth), (no growth). Growth media PDB + SYP + MYP + PDB SYP MYP H's H's H's Gd 24 h + + ++ + + + Gd 48 h +++ +++ +++ (+) + + Ts 24 h ++ ++ ++ ++ ++ Ts 48 h + ++ + ++ +++ ++ Gd + Ts 24 h ++/ ++/++ ++/++ +/(+) +/++ +/++ Gd + Ts 48 h ++/ +++/ +++/+ (+)/+ +/++ +/+

[0110] The results show that both media, SYP and MYP, are suitable as growth media for mixed cultures of Gd and Ts. When in single culture, Ts growth was improved when these media were supplemented with Herbst's artificial seawater. However, in both media, growth of Ts slowed compared with Gd growth over time. This trend could be observed in more detail following competence assays.

[0111] In these assays, separate cultures of Gd and Ts were incubated in appropriate media for 24-48 hours at 28 C. with agitation at 150 rpm. Cultures were washed twice with 0.9% (w/v) NaCl solution to remove the used medium and OD.sub.600 was checked and adjusted to 0.2 (10.sup.8 cfu/ml). To a final volume of 5 ml of the selected media (MYP and SYP), 1 ml of each culture (Gd and Ts) was inoculated, with one tube per sample. Samples were taken 0, 6, 24, 36, 48, 54, and 120 hours after inoculation, OD.sub.600 measured and serial dilutions plated by duplicate on ATGUS, LGIP and Marine Agar as selective media. Single cultures of Gd and Ts were inoculated and incubated using the same conditions as control.

[0112] The results are shown in FIGS. 7A-7D. Both strains were able to grow in these media, but Ts grew at higher concentrations as single culture in SYP than in MYP. In the mixed culture, the behaviour of the strains were similar in SYP and MYP. This suggests that the presence of Gd provides some benefit for Ts growth in MYP. Gd growth did not seem to be affected by the presence of Ts in terms of cell viability.

[0113] As a result, these media are suitable for producing a viable co-culture of Gd and Ts.

EXAMPLE 4

Nitrogen Fixation In-Vitro

[0114] The reduction of acetylene (C.sub.2H.sub.2) to ethylene (C.sub.2H.sub.4) is an indirect method of measuring nitrogenase activity in natural samples (Cojho, et al., Volume 1993. 106 pp 341-346).

[0115] Cultures of Gd and Ts were prepared in MYP and SYP and incubated at 28 C., with shaking at 150 rpm. After 2 days of incubation the cultures were centrifuged (4,000 rpm, 15 min) and washed with 0.9% (w/v) NaCl solution. The resultant pellets were re-suspended in a modified media, one of which was nitrogen free; specifically the MYP and SYP media described above but without adding yeast extract (hereinafter referred to as MP and SP respectively). The OD.sub.600 of Gd cultures were adjusted to 1 and to 0.5 for Ts. The samples were prepared by inoculation of 1 ml of Gd, for single cultures, and 1 ml Gd+1 ml Ts for mixed samples, in a final volume of 5 ml of broth in the following proportions Gd:Ts (taking as reference the OD.sub.600 of the cultures): 1:1, 2:1 and 1:2. Tubes were sealed and 10% of air volume was replaced with acetylene. Ethylene production in the samples was analysed by gas chromatography (GC) after incubation for 1, 2, 4, and 8 days at 28 C. as static cultures. Also OD.sub.600 and cell number in the samples were tested. Every samples were prepared in triplicate.

[0116] The ethylene production from the single Gd culture and mixed Gd+Ts cultures grown in SP and MP media were measured by gas chromatography (GC). No acetylene reduction was detected in samples grown in MP along all the experiment; neither in SP before 4 days of incubation, even in the single cultures. After 4 and 8 days of incubation ethylene production was detected in samples in medium SP, in single and mixed cultures.

[0117] The peak of ethylene production measured by GC showed a high variation, but there was a constant correlation between the values, as is showed in FIG. 8. The ethylene production for mixed samples when the relation Gd:Ts was 2:1 was almost double than for single cultures. The correlation between both types of samples changes when the proportion between strains change, being 25% lower for mixed cultures when Ts is in higher concentration than Gd (1:2), than for single cultures. When Gd and Ts are mixed in the same proportion there is no variation in the ethylene production in regard with Gd single culture. These results show that the presence of Ts is able to modify the activity of the enzyme nitrogenase, and that this effect is related to the relative proportions of Gd and Ts.

EXAMPLE 5

Effects of Ts on Levan Production by Gd

[0118] Flasks with 100 ml of MYP and SYP were inoculated with 1 ml Gd+1 ml Ts of overnight cultures of both strains with OD.sub.600 0.2 and incubated for 14 days. Samples were collected every two days and 1 ml of each culture was stored in duplicate at 20 C. prior to analysis for levan production. OD.sub.600 was measured and serial dilutions plated onto ATGUS and Marine Agar to estimate the cell number for both strains at the sampling times. Single cultures of Gd in MYP and SYP were inoculated and incubated in the same conditions as a control. To extract and quantify levan in the samples, samples were precipitated with 3 volumes of methanol and placed into a heat block set to 60 C. until dried and volume was reduced to approximately 1 ml. The precipitates were dissolved in 1 ml of H.sub.2O. Acid hydrolysis was carried out by adding 0.01M H.sub.2SO.sub.4 and incubating 1 h at 121 C. After the incubation, 500 l of DNSA reagent (100 ml aqueous solution contains 1 g 3,5-dinitrosalicylic acid, 0.4M NaOH and 30 g KNa tartrate tetrahydrate) was added and mixed. Tubes were placed into boiling water for 15 minutes and 500 l of 40% (w/v) KNa tartrate was added and mixed. The absorbance of the samples was measured at 540 nm and the glucose concentration was estimated using a calibration line from a range of glucose standards.

[0119] The results are shown in FIG. 9. The levan production was higher for cultures in SYP than in MYP in the first days. The levan production along time in samples grown in MYP did not show significant differences, while there are statistical significant differences in samples grown in SYP. If the values are compared in relation to the sampling times, there are significant differences between the four samples 3 days post-inoculation and after 8 days of incubation.

EXAMPLE 6

Expression of Nitrogenase and Levansucrose Genes

[0120] Since differences in the nitrogenase activity were detected between samples of single Gd cultures and mixed cultures Gd+Ts grown in SP after 8 days of incubation, RNA was extracted from these samples and gene expression was assessed by qPCR. With real-time PCR or qPCR, the applicants investigated whether there were any differences on the enzymes nitrogenase and levansucrose genetic expression under different culture conditions.

[0121] Samples were prepared as described before in Example 4. After adjusting the OD.sub.600 to 1, 5 ml of Gd for single cultures, and 5 ml Gd+5 ml Ts for mixed cultures, a final volume of 50 ml of MP and SP were inoculated and incubated at 28 C. as static cultures. After 8 days of incubation, 1 ml of samples was stored and RNA was extracted for qPCR. The RNA was isolated with Trizol according to the manufacturer's protocol. RNA purity and integrity appropriate for downstream RT-qPCR applications were confirmed by measuring the ratio of absorbance at 260 nm/280 nm and 260 nm/230 nm. DNase digestion of samples was carried out using DNase I kit (Sigma) according to the manufacturer's instructions, to ensure that no DNA was present in the samples. First-strand cDNA synthesis was carried out using the Superscript III Reverse Transcriptase kit (Invitrogen) performed according to the manufacturer's manual. Reverse transcription reactions were carried out with 1 g of RNA. The cDNA was diluted 1:5 and then used in RT-qPCR.

[0122] qPCR.

[0123] RT-qPCR was performed on the CFX96 Real-Time System (BioRad) using the iTap Universal SYBR Green Supermix (BioRad). The primers were used at 3.75 M and 10 ng of cDNA per reaction. Thermocycling was performed as follows: 2 min at 95 C., 40 cycles of 5 sec at 95 C. and 15 sec at 60 C. The melt curve was performed to check the liability of the primers as follow: increment of 0.5 C. every 5 sec, from 60 C. to 95 C. All RT-qPCR assays were carried out using three technical replicates and non-template control. Primers to 23S were used to amplify reference genes (Galisa, et al., J. Microbiol Methods (2012) October; 91(1):1-7) and primers for nifH (forward: 5-TCGACGACCTCCCAGAATAC-3(SEQ ID NO 5); reverse: 5-CCTTGTAGCCGATCTTCAGC-3) (SEQ ID NO 6) and lsdA (forward: 5-ACGCCGATCAGTTCAAGCTAT-3 (SEQ ID NO 7); reverse: 5-CCTGGTTCGTGTAGGTCTGG-3) (SEQ ID NO 8) genes were used to target genes for the enzymes nitrogenase and levansucrose respectively. All primers' amplification efficiency was tested, and the efficiency value, calculated in relation to the slope got from the standard curve generated from serial dilutions of the template. Amplification efficiency was calculated as a percentage of template that was amplified in each cycle (% Efficiency=(E1)*100; E=10.sup.1/slope). % E in the range 90%-105% indicate high amplification efficiency.

[0124] The results of this experiment are shown in FIG. 10. It appears that there is a trend for an increase in genetic expression of nifH and lsdA in mixed cultures, as compared to single Gd cultures, especially in samples grown in SP.

[0125] The study of nitrogenase and levansucrose genes expression by qPCR shows a trend to increase the expression for the two enzymes. These qPCR experiments show an effect of the presence of Ts in nitrogenase expression, which seems to increase as a response to the conditions generated by a mixed culture of Gd plus Ts especially in SP, but also in MP to some extent.

[0126] The levansucrose gene similarly shows a tendency to increase its expression in the presence of this second microorganism, which could be related with nitrogen fixation, since it appears that there is a correlation between nitrogen fixation and levan production, probably due to the protected environment created by levans to nitrogenase, avoiding oxygen diffusion, which could inhibit the enzyme activity. It has been demonstrated that the mucilaginous matrix produced by Gd is used by the bacteria to keep an anaerobic environment (Dong, et al., Applied Environmental Microbiology, 2002, pp 1754-1759). Since levan is the major exopolysaccharide secreted by Gd, it could be involved in the nitrogenase protection from oxygen besides in tolerance to other abiotic stresses such as NaCl, sucrose and desiccation (Velazquez-Hernandez, et al., Archives of Microbiology 2011 vol 193, 139-149).

EXAMPLE 7

Effect of Ts on Adherence of Gd

[0127] Assessment of the adhesion capacity of the mixed culture Gd+Ts in comparison with Gd single culture was carried out by measuring the ability of cultures to attach to artificial surface (centrifuge tubes), as described (Favre-Bont et al. Biomed. Central Microbiology (2007) 7 (33) ppl-12). Starter cultures of Gd and Ts were prepared in SYP (da Silva-Froufe, et al., 2009 supra.) and SYP modified with Herbst's artificial seawater (An, et al., 2007 supra.), respectively. After 48 hours growing at 28 C. at 150 rpm, cultures were washed twice with PBS (4000 rpm, 15 minutes), and the pellets were re-suspended with PBS at OD.sub.600 1 and 0.5. The samples were prepared by inoculation of 1 ml of Gd or Ts, for single cultures, and 1 ml Gd+1 ml Ts for mixed samples, in a final volume of 10 ml of SYP in the following proportions Gd:Ts (taking as reference the OD.sub.600 of the cultures): 1:1, 2:1 and 1:2. Five repetition per treatment were prepared by distributing 1.5 ml of samples in 15 ml centrifuge tubes, and incubated as static culture at 28 C., during 8 days.

[0128] To check the adherence of the cells to the surface, the medium was removed, the tubes were washed 3 times with 1.5 ml of PBS, and cells were fixed with 10 ml of methanol for 15 minutes. The alcohol was removed and the tubes were air dried. The biofilm formed was stained with crystal violet 0.1% w/v for 5 minutes. The dye was rinsed with distilled water and air dried. Later, the dye was re-suspended with 10 ml of glacial acetic acid 33%. The absorbance of the samples was measured at 595 nm. As control, 5 tubes with 1.5 ml of SYP were incubated and treated in the same way. The value from the blank was taken as background and it was subtracted from the values obtained for the rest of the samples.

[0129] The results are shown in FIG. 11. This experiment shows a tendency of the cell cultures to attach to an artificial surface, which could be an indication of the ability of Gd to form biofilm in different conditions. Formation of biofilm would be expected to impact on plant colonization properties. This ability appears to be enhanced by the combination of Ts as a mixed culture, in particular in a ratio of 1:1.

EXAMPLE 8

Attachment and Seed Germination

[0130] Oilseed rape (OSR) seeds were sterilized by soaking them in 100% ethanol, vortexing and allowing them to stand for 2 minutes, after which they were thoroughly washed and vortexed with sterile de-ionised water (SDW) three times. After that, seeds were soaked in 70% (v/v) bleach with 1% (v/v) Tween 80, vortexed and allowed to stand for 30 minutes. After that they were again thoroughly washed and vortexed with SDW at least more five times.

[0131] To prepare the inoculum, fresh cultures of Gd and Ts were incubated for 24 hours in optimal media. OD was adjusted to 0.35 for Gd (10.sup.8 cfu/ml) and to 0.2 and 0.1 for Ts (10.sup.8 cfu/ml and 10.sup.6 cfu/ml respectively). Cultures were diluted in an aqueous composition comprising 3% (v/v) sucrose, 0.1% (v/v) Tween80 and 0.3% (v/v) gum Arabic. The treatments were as follow: a) control, just water; b) Gd, 2.5.Math.10.sup.5 cfu/ml; c) Ts, 2.5.Math.10.sup.5 cfu/ml; d) Gd+Ts 1:1, both strains at 2.5.Math.10.sup.5 cfu/ml; e) Gd+Ts 2:1, Gd 2.5.Math.10.sup.5 cfu/ml and Ts 2.5.Math.10.sup.3 cfu/ml. The actual inoculum concentrations were as follows:

TABLE-US-00003 TREATMENT Gd (cfu/ml) Ts (cfu/ml) Gd 3.75 .Math. 10.sup.5 Ts 2.45 .Math. 10.sup.5 Gd + Ts 1:1 4.90 .Math. 10.sup.5 2.15 .Math. 10.sup.5 Gd + Ts 2:1 3.3 .Math. 10.sup.5 6.52 .Math. 10.sup.4 Control

[0132] The sterilized seeds were soaked in the different treatments for 30 minutes.

[0133] After the treatment of the seeds, the solution was discarded and seeds were placed onto an inverted Petri dish with a glass fibre paper moistened with 3 ml of SDW. Plates were protected from light by covering with aluminium foil and incubated at 21 C. To test the attachment of the bacteria to the seeds, three groups of 20 seeds per treatment, were washed in 5 ml of PBS and shaken vigorously at 20 C. for 2 hours. The solution obtained was serially diluted and plated onto LGIP and Marine Agar media. Plates were then incubated for 4 days at 28 C. The number of cells recovered from the surface of the seeds were counted and the results are shown in FIG. 12.

[0134] After 24 hours of incubation, germinated seeds were counted to calculate germination rate and placed in conical tubes with 5 ml with Murashige and Skoog (MS) basal medium (Sigma M0404). Plants were incubated in a plant growth chamber (cycle: 12 hours of dark at 15 C. and 60% RH/12 hours of light at 28 C. and 60% RH) at least until two true leaves had grown. 15 plants per treatment were incubated.

[0135] The results show that the attachment of Gd and Ts to the seeds showed differences for the different treatments. It was not possible to re-isolate Ts from the seeds, despite that this strain was detected in plant extracts, suggesting that this becomes well adhered to the seed surface. The mixed cultures show reduced recovery and so enhanced adherence as compared to the single Gd culture.

[0136] After inoculation and germination of the seeds, the germinated seedlings were incubated in the plant growth chamber for a further 15 days. Seven seedlings from each treatment were processed to get extracts from the roots and the leaves, after sterilization of the surface.

[0137] For the re-isolation of the epiphytic microorganisms from the root surface, the procedure was similar to that used for seed attachment assay. In particular, roots were also washed in an isotonic buffer to re-isolate the epiphytic microorganisms colonizing the surface of the roots.

[0138] For endophytic colonisation, roots and leaves surface were sterilized by immersion for 10 minutes in 10% (v/v) bleach, pH adjusted to 8.0, plants were rinsed with SDW and macerated in 1 ml of PBS (O'Callagham, et al., Applied and Environmental Microbiology, 2000, 66(5) 2185-2191). The plant extract was serial diluted and plated in duplicate onto LGIP and Marine Agar. The results are shown in FIGS. 13 and 14 respectively.

[0139] Uninoculated plants did not show contamination, and no cross contamination between the different treatments was found. Except in plants inoculated just with Ts, it was not possible to recover Ts in endophytic association with the plant, but this bacterium was found on the surface of the roots. Nevertheless Ts colonized the root in lower number when it was inoculated in association with Gd (FIG. 14).

[0140] The recovery of Gd showed important differences between the treatments (FIG. 13). While in the plants inoculated with a single culture of Gd there were essentially no differences in the number of re-isolated microorganisms as endophytes between roots and leaves, the distribution of Gd inside the plant changed in the different parts of the plant (roots versus leaves) when they were inoculated with the mixed cultures, and also with respect to those plants inoculated just with Gd. Plants inoculated with Gd+Ts, in a proportion 2:1 showed improved endophytic colonization, despite the fact that the colonisation on the surface of the plant was the same than with the single inoculation. Furthermore, it seemed that the treatment with the combined inoculum affected the migration of the bacteria inside the plant, favouring localisation to the leaves. The number of isolated from inside of the aerial part of the plant was ten-fold higher for plants inoculated with the mixed culture in regard with the plants inoculated with pure culture of Gd.

[0141] In addition, once plants were sufficiently grown, some were taken out from the tubes, and wet and dried weights were measured. The results are shown in FIGS. 15A,B.

[0142] The measure of the plant parameters do not show differences between the diverse treatments with Gd in terms of the LSD, but it shows a tendency of the plants treated with Gd+Ts (1:1) to a higher dry weight (DW), and there is a significant difference between the DW of this mixed treatment on regard with the control plants, which shows a tendency of the plants treated with Gd+Ts in a mixture of 1:1 to increase the biomass, despite the wet weight (FIGS. 15A-C).

EXAMPLE 9

Levan Production and Quantification

[0143] Flasks with 100 ml of SYP were inoculated with 1 ml Gd+1 ml Ts of overnight cultures of both strains with OD.sub.600 0.2 and incubated during 3 days. Samples were collected every day for 3 days and triplicate 1 ml aliquots of each culture was stored at 20 C. to analyse the levan production. OD.sub.600 was measured and serial dilutions plated on ATGUS and Marine Agar to get the cfu for both strains. Single cultures of Gd and Ts in SYP were inoculated and incubated in the same conditions as control.

[0144] To extract and quantify levan in the samples, the cultures were firstly centrifuged (10,000 rpm, 20 min, 4 C.) and the supernatants were decanted and placed into new tubes. Samples were precipitated as described in Example 5, with 3 volumes of methanol and placed into a heat block (Thermomixer) set to 60 C. until dried and volume was reduced to approximately 1 ml. The precipitates were dissolved in 1 ml of water. Acid hydrolysis was carried out adding 0.2% of H.sub.2SO.sub.4 5 M and incubating for 1 hour at 121 C. After the incubation, 500 l of DNSA reagent (in 100 ml, 1 g 3,5-dinitrosalicylic acid dissolved in 50 ml of H2O, 20 ml NaOH 2M, 30 g KNa tartrate tetrahydrate) was added and mixed. Tubes were placed into boiling water for 15 min and 500 l of 40% KNa tartrate was added and mixed. The absorbance of the samples was measured at 540 nm and the glucose concentration was estimated using a calibration line from a range of glucose standards.

[0145] Levan from Ts was also quantified since recent research showed the presence of levansucrose and levanase enzymes for Terribacillus species (Lu, et al., Genome Announcements 2015 3 (2), pp e00126-15).

[0146] The results are shown in FIG. 16. While there were no significant differences in the levan concentration in the mixed culture in proportion 1:1, there was a large decrease in the quantification of levan detected in the medium when both strains were present with a higher concentration of Gd. In this case, the levan concentration in the medium was following a similar trend to that of the single Gd culture but with a more intense effect on the levan concentration. Without being bound by theory, this may be due to the consumption of sucrose and the degradation of levan by both strains.

EXAMPLE 10

Effect of Terribacillus on Production of Plant Hormone by Gd

[0147] To check the ability of Gd to produce the auxin IAA on its own or in the presence of Ts, cultures of Gd, Ts, Gd+Ts (1:1) and Gd+Ts (2:1) were prepared in SYP and incubated at 28 C., 150 rpm for one week. Samples (1 ml) of all the cultures were taken in triplicate daily during the first three days, and on the seventh day, of incubation. Samples were spun at 13000 rpm for 5 minutes and the supernatant recovered in glass tubes. The supernatants were mixed with 4 volumes of Salowski's reagent (150 ml of H.sub.2SO.sub.4 96%, 250 ml of H.sub.2O and 7.5 ml of FeCl.sub.3 0.5 M) per volume of sample. Development of a pink colour indicated IAA production.

[0148] OD.sub.535 was read using a spectrophotometer. The concentration of IAA was estimated by a standard IAA graph. The final value was calculated as the average of the three replicate for every samples, and the error was calculated with the standard error.

[0149] The results are shown in FIG. 17. It was clear that whilst Ts did not produce IAA to any significant extent, in the presence of Ts, the production of IAA by Gd increased overall, suggesting that the Ts was modifying the behaviour of Gd.