Effects of growth-promoting endophytic Methylobacterium on development of Citrus rootstocks

1 Department of Genetics, University of São Paulo, ESALQ/USP, Av. Pádua Dias, 11, 13418-900, Piracicaba, SP, Brazil. 2 Institute of Biodiversity and Forests, Federal University of Western Pará, IBEF/UFOPA, Rua Vera Paz s/n, 68035-110, Santarém, PA, Brazil. 3 LABMEM/NAP-BIOP, Department of Microbiology, Institute of Biomedical Sciences, University of São Paulo, Av. Prof. Lineu Prestes, 1374 -Ed. Biomédicas II, Cidade Universitária, 05508-900, São Paulo, SP, Brazil.

The establishment of Methylobacterium in association with the host is thought to directly influence seed germination and plant growth by mechanisms, such as cytokinin and auxin synthesis (Omer et al., 2004;Tani et al., 2012;Eevers et al., 2015), nitrogen fixation (Sy et al., 2001, Madhaiyan et al., 2009;Madhaiyan et al., 2014) and plant protection (Ardanov et al., 2012;Yim et al., 2014).In this context, Methylobacterium seems to be an important bacterial group to be employed for improvement of crop productivity.
Nowadays, the availability of genetic tools such as the genome sequencing and studies of central metabolism have attracted attention to the genetic manipulation of Methylobacterium for the production of various bioproducts (Fitzgerald and Lidstrom, 2003;Choi et al., 2008, Sonntag et al., 2015) as well as for symbiotic control of phytopathogens (Gai et al., 2009;Ferreira Filho et al., 2012).The occurrence of new pests and diseases has affected the yield and production cost of several crops around the world.However, the economic importance of crops such as Citrus requires changes in the management and treatment of plants and use of new technologies to improve the productivity and quality of fruits (Duenhas et al., 2002;Donadio, 2011).
Therefore, bacteria associated with plants play an important role in the adaptation of their hosts, either in natural environments or in stressful conditions, and may be used as a strategy to promote better association between some crops and the environment (Hallmann et al., 1997;Sturz and Nowak, 2000;Bhattacharyya and Jha, 2012).In this way, the use of plant growth-promoting bacteria could be a possibility to improve the productivity and the quality of citrus orchards, reducing the use of chemical inputs, generating economic and ecological benefits, since these products might often affect the environment negatively.
In citrus, one possibility to improve the productivity of the orchards could be through the inoculation of plant growth promotion endophyte (PGPE) into the rootstocks thereby, accelerating growth and consequently reducing the time that the plant remains in nursery.Therefore, in the present study we analyzed the effects of Methylobacterium species on seed germination and growth of C. limonia and C. sunki under commercial nursery conditions.

Evaluation of the effects of Methylobacterium spp. on seedlings growth of Citrus rootstocks
For seedlings production, the bark seeds (stored for 4 months in cold chamber at 5/8°C) was removed and the seeds treated with bacteria (10 8 CFU mL -1 ) or CHOI medium (control 1), for 5 h at 28°C with agitation by an orbital shaker at 80 rpm, and planted in dibble tubes (3 × 12cm; 50 cm 3 ) containing commercial substrate pH 6.0 (Golden Mix 11, Amafibra, Holambra, SP, Brazil).Each treatment was composed of 4 replicates, where each replicate consisted of 10 seedlings (40 seedlings per treatment).The germination rate was estimated as the number of seedling growth per replicate.
After germination, only one seedling was replanted per dibble tube containing substrate pH 6.0 (Golden Mix 11, Amafibra, Holambra, SP, Brasil) and maintained in greenhouse under controlled relative humidity (70%) and temperature ranging from 19 to 40°C.
The effects of the endophytic Methylobacterium spp. on seedling growth were compared to the effects of the controls (CHOI and without inoculation).The seed germination rate was evaluated 30 days after inoculation of bacterial strains.The plant height was evaluated after 30, 90 and 120 days after germination, while the seedlings biomass (shoot and root) were evaluated only after 120 days.

Isolation of Methylobacterium from seeds and inoculated Citrus rootstocks
The bacterial communities inside citrus seeds were evaluated from fresh seeds and those stored for 4 months in cold chamber at 5/8°C, the latter usually used for rootstocks production.For this, 100 seeds from each citrus species were surface disinfected (70% ethanol for 1 min, sodium hypochlorite solution (2% available Cl -) for 3 min, 70% ethanol for 30 s, two rinses in sterilized distilled water for 1 min) and triturated in sterile PBS solution containing (g L −1 ) Na2HPO4, 1.44; KH2PO4,0.24;KCl,0.20;NaCl,8.00;pH 7.4.Serial dilution was plated on CHOI medium amended with 50 µg mL -1 of the fungicide carbendazim and incubated at 28°C for 15 days.After growth, colonies were picked out, purified by streaking on CHOI agar medium and stored for further evaluation.
The citrus rootstocks are used for grafting 120 days after seed planting.Therefore, after this time we evaluated the presence of inoculated Methylobacterium, in order to assess the potential of these bacteria to be transmitted aerial plant from inoculated rootstock, increasing the plant growth promotion.For this 10 seedlings per treatment were sampled, roots and shoots were separated and surface disinfection was performed as following: 70% ethanol for 1 min, sodium hypochlorite solution (2% available Cl -) for 3 min, 70% ethanol for 30 s, two rinses in sterilized distilled water for 1 min.
Both roots and shoots tissues were cut in small pieces, triturated in sterile phosphate buffered saline (PBS) solution and incubated at 28°C for 1 h under agitation (100 rpm).Appropriated dilutions (10 -1 , 10 -2 , 10 -3 ) were plated onto CHOI medium supplemented with 50 µg mL -1 of the fungicide carbendazim and incubated at 28°C for 15 days.The disinfection process was checked by plating aliquots of the sterile distilled water, used in the last wash, on CHOI and incubated under the same conditions.Colonies were purified by streaking in CHOI medium and stored in 70% glycerol solution at -80°C for further analysis.

Amplified ribosomal DNA restriction analysis (ARDRA) and molecular identification of endophytic bacteria
The bacterial community isolated from seeds was identified by 16S rRNA gene sequencing using colony-PCR.After growth on culture medium, the isolates were transferred to a tube containing 200 L of sterilized ultra pure water.The bacterial suspension was used as the source of DNA in PCR reactions.Primers PO27F (5´-GAGAGTTTGATCCTGGCTCAG -3´) and 1387R (5´-CGGTGTGTACAAGGCCCGGGAACG -3´) (Heuer and Smalla, 1997) were used in 50 µl PCR reaction containing 10 X buffer (10 mM KCl, 10 mM Tris-HCl, pH 8.3) (Fermentas Life Sciences, Brazil), 0.2 mM dNTP, 3.75 mM MgCl2, 2.5 U Taq DNA polymerase (Fermentas Life Sciences, Brazil), 0.2 µM of primers and 2 µl of the boiled cells (15 min at 90°C in 80 µl H2O).Amplifications were performed in thermal cycler (PTC 200, MJ Research -USA) programmed for an initial denaturation (94°C for 4 min) followed by 35 cycles of 94°C for 30 s, 62.5°C for 1 min, 72°C for 1 min and a final extension of 10 min at 72°C.A negative control (PCR mixture without bacterial DNA) was included in all analysis.The PCR fragments were purified with polyethylene glycol (PEG) (20% PEG 8000; 2.5 mM NaCl) and sequenced at Human Genome Research Center (HGRC), (Institute of Biosciences, University of São Paulo, São Paulo, SP, Brazil).The nucleotide sequences were compared with sequences information available in the GenBank database, by BLASTn.
In order to compare with inoculated bacteria, the isolates obtained from seedlings rootstock shoot were identified by ARDRA technique and 16S rRNA gene sequencing.For this, the 16S rRNA gene was amplified by Methylobacterium colony-PCR using PO27F and 1387R primers as described above.The PCR products were digested with 2 units of AluI restriction enzyme (Life Technologies, Brazil) according to the manufacture´s recommendations.The reaction mixture was incubated at 37°C for 1 h.The restriction patterns were examined using 2.5% (w/v) agarose gel and stained with ethidium bromide.

Auxin production
Auxin production (IAA) was evaluated by the colorimetric method as described by Gordon and Weber (1951) with modifications.For this, the bacteria were grown in 10% TSB medium (10 mL) amended with 5 mM of L-tryptophan and incubated in the dark at 28°C for 72 h.Cells were harvested by centrifugation (8 000 x g for 5 min), the supernatant (900 l) was treated with 600 l of Salkowski reagent (50 mL of perchloric acid (35%) and 1 mL of FeCl3 solution (0.5 M) and incubated for 30 min at room temperature in the dark.IAA was quantified using a spectrophotometer (Pharmacia Biotech Ultroespec 3000) at 530 nm of absorbance.The readings were normalized by using the standard curve with different concentrations of IAA (g mL -1 ).All analysis was performed in triplicate.

Biological nitrogen fixation
The ability to fix N2 in Methylobacterium strains was evaluated in strains able to improve plant growth.The strains were inoculated in nitrogen-free semi-solid NFb medium (Döbereiner et al., 1995) and, after growth the bacterial cells were re-inoculated in nitrogen-free NFb medium.This procedure was repeated three times and strains able to form a growth pellicle in this medium were considered able to fix nitrogen.

Phosphorus solubilization
The ability to solubilize phosphorus was assayed according to Verma et al. (2001) with modifications.Ten microliters containing about 10 8 CFU mL -1 of bacterial strains grown in CHOI medium were inoculated onto agar medium containing inorganic phosphate (agar, 15 g; glucose, 10 g; NH4Cl, 5.0 g; NaCl, 1.0 g; MgSO4.7H2O, 1 g; Ca3(HPO4)2, 0.8 g) in a final volume of 1 L filled with distilled water, pH 7.2.Bacteria were incubated at 28°C for 7 days.The ability to solubilize inorganic phosphate was characterized by a clear halo around bacterial colonies.All analysis was performed in triplicate using a Burkholderia seminalis (TC3.4.2R3) as a positive control.

Screening of genes associated to nitrogen fixation, IAA biosynthesis and ACC deaminase in Methylobacterium mesophilicum SR1.6/6
M. mesophilicum SR1.6/6 genome was published as announcement (Almeida et al., 2013) under access ANPA01000000 in NCBI (www.ncbi.nlm.nih.gov/)thus some genomic re-annotation of genes that encode enzymes related to nitrogen fixation, IAA biosynthesis and ACC deaminase were made in order to confirm the physiological results approached in this present work.Artemis software (Rutherford et al., 2000) was used to overview the genome annotation as well as KEGG: Kyoto Encyclopedia of Genes and Genomes (http://www.genome.jp/kegg/)and Biocyc (http://biocyc.org/)to study target pathways, such as IAA biosynthesis and ACC deaminase.

Statistical analysis
The ESTAT (Barbosa et al., 1992) statistical package was used for ANOVA followed by 2×2 factorial design (plant height × month) and Tukey's test at 5% probability level for means comparison.

Checking the occurrence of Methylobacterium spp. in rootstocks seeds
No bacteria were recovered from fresh seeds indicating that in citrus cultivable endophytic bacteria are not

Effects of Methylobacterium spp. on seed germination and seedlings growth of C. limonia
The seed germination ranged from 47.85 to 73.5%.The inoculation of M. extorquens AR1.6/2 reduced the seed germination, while the others strains have no effect on this plant growth parameter (Table 1).However, the presence of this M. extorquens AR1.6/2 strain has no negative effect on plant growth (Table 1) indicating that this bacterium could be inoculated only after seed germination.
M. extorquens AR1.6/11 and M. mesophilicum SR1.6/6 inoculation resulted in plant growth promotion.The strain AR1.6/11 was able to increase plant height after 90 and 120 and root fresh biomass at 120 days, while the strain SR1.6/6 increased root dry and fresh biomass at 120 days (Table 1).The TP4/2 strain promoted a negative effect on height plant 120 days after inoculation (Table 1).Based on these results, the strains AR1.611, AR1.6/2, SR1.66 and TP 4/2 were selected for further analysis in C. sunki.The colonization of rootstocks seedlings by AR1.611, AR1.62 and SR1.66 was evaluated 120 days after inoculation.Except for AR1.6/11 that was recovered only from C. limonia roots, all strains were re-isolated from shoots and roots tissues of both rootstocks at 10 2 CFU g -1 of plant tissue.The identity of the strains was successfully confirmed by ARDRA technique with AluI restriction enzyme.

Effects of Methylobacterium spp. on seed germination and seedlings growth of C. sunki
Unlike our finds in C. limonia, the strain AR1.6/2 has no negative effect on germination of C. sunki seeds (Table 2).All treatments induced seed germination when compared with the control, but this effect was due to the CHOI medium, since the germination rate between CHOI medium and bacterial inoculated seeds were not significantly different (Table 2).The strains SR1.6/6, AR1.6/2 and AR1.6/11 increased the plant height after 120 days.In addition, AR1.6/2 strain enhanced shoot fresh biomass (Table 2).C. sunki inoculated with TP4/2 strain no presented negative effect on height plant as observed to C. limonia (Table 2).

In vitro plant growth promoting traits
Bacterial strains that exhibited positive effect on seedlings growth were evaluated in vitro for their ability to synthesize indole-3-acetic acid (IAA), phosphorus solubilization and N 2 fixation.The strains SR 1.6/6, AR 1.6/2, and AR 1.6/11 produced IAA at 2.3, 2.1 and 1.8 mg ml -1 (Table 2), respectively.A clear zone halo around the colonies was not observed, indicating that these strains are not able to solubilize inorganic phosphorus.Although a growth had been observed in nitrogen-free medium after the first bacterial inoculation, it was not observed after successive inoculation in this medium, indicating that these strains were not able to fix nitrogen.
M. mesophilicum SR 1.6/6, which had the genome published (Almeida et al., 2013) was one of the most effective strain in promoting height plant and root biomass.We screened this genome and found at least 22 genes potentially related to IAA synthesis.According to these information, there are three possible pathways in M. mesophilicum SR 1.6/6 for the production of indole-acetic acid, (1) from idole-3-acetonitrile, (2) from indole-3-acetamide, (3) tryptamine (Figure 1).In the first route, the Indole-3-acetonitrile is catalyzed to Indole-acetic acid by Nitrilase/cyanide hydratase and Nacyltransferase.Through the second route, Indole-3-acetamide can be synthetized from Indole-3-acetonitrile by a nitrile-hydratase (subunit alpha and beta), and then to IAA by amidase.In the third route, IAA could be synthetized via tryptamine and indoleacetaldehyde.We did not find any Trp decaborxylase and tryptophan-2monooxygenase enzymes to metabolize L-Tryptophan in M. mesophilicum SR 1.6/6 genome.
In addition, an evaluation of the SR1.6/6 genome showed that nif and fix gene clusters are not present in this strain, but a gene that encodes an ACC deaminase is present in the M. mesophilicum SR 1.6/6 genome, indicating that the suppression of stress response could be a mechanism present in these endophytic bacteria related to plant growth promotion.

DISCUSSION
In sustainable agriculture the application of growth-promoting bacteria may increase productivity and quality of the crop, reducing the costs associated with the use of chemicals and the environmental impacts.Previously studies had shown that Methylobacterium species have the ability to induce shifts in physiological traits of their hosts, promoting positive, negative or any effect on seed germination and/or on plant development (Holland andPolacco, 1994, Abanda-Nkpwatt et al., 2006;Lee et al., 2011;Pohjanen et al., 2014).The specificity of plantbacteria interaction has been considered an important factor for the generation of these different effects on plant host and the success of this approach for plant production.Our results revealed that some Methylobacterium strains originally isolated from Citrus sinensis promoted significantly the height and biomass of C. limonia and C. sunki, as already described in other plant species with native Methylobacterium (Madhaiyan et al., 2005;Lee et al., 2006;Madhaiyan et al., 2015).In fact, some authors emphasize the need to employ native isolates or isolates adapted to their hosts, justifying the higher capacity of colonization and lower risk in the introduction of exogenous microorganisms in the plants (Enebak et al., 1998;Khalid et al., 2004).However, as we observed, bacterial strains isolated from Citrus, such as AR1.6/2 induced negative effects on seed germination of C. limonia, but had no effect on plant growth of both C. limonia and C. sunki.In addition, this strain promoted the plant height and enhanced the shoot biomass.
During plant colonization, the cross-talking between bacteria and the host seems to has a key function (Rosenblueth and Martinez-Romero, 2006), which suggest that plants can communicate specifically to attract microorganisms for their own ecological and evolutionary benefits (Compant et al., 2005).Thus, endophytes may differentially express genes required to colonize the host and modulate plant genes stimulating plant growth (Rosenblueth and Martinez-Romero, 2006).In many cases, the plant responses to endophytes seem to be conditioned by the plant genotype (Nowak, 1998;Rosenblueth and Martinez-Romero, 2006).Besides, the environmental conditions and balance between other microorganisms can also be decisive for the final effect (Azevedo, 1998;Andrews and Harris, 2000;Montesinos et al., 2002).
Plant growth promotion by Methylobacterium isolates has been attributed to the nitrogen fixation and production of phytohormones such as auxins and gibberellins (Sy et al., 2001;Madhaiyan et al., 2014;Eevers et al., 2015).
In the present study, all Methylobacterium strains with potential to promote plant growth were able to synthetized IAA, but could not to solubilize phosphorus and growth in nitrogen-free medium.This result suggested that, although many factors, such as speciesspecific recognition may be related to the ability of bacteria to promote plant growth, the citrus growth promotion, based on height and biomass analysis seems to be only related to IAA production and ACC deaminase.These results were confirmed by the analysis of the genome of SR1.6/6 strain.Besides, a gene that encodes an ACC deaminase is present in the M. mesophilicum SR1.6/6 genome, indicating it could act in the suppression of the stress response resulting in plant growth promotion.The IAA plant hormone, under biotic and abiotic stress conditions, is able to activate the enzyme 1-aminocyclopropane 1-carboxylate (ACC) synthase, which synthetize ACC that is converted to ethylene by the enzyme ACC oxidase (Wang et al., 2002).The enzyme ACC deaminase, present in many endophytes, is known to compete with ACC oxidase, modulating the ethylene levels in plants, reducing the stress response triggered by ethylene and promoting plant growth under stress conditions (Hardoim et al., 2008).The inoculated Methylobacterium strains were not present in seeds, but were able to colonize roots and shoots endophytically after seed inoculation, as evident by their recovery from seedlings obtained from inoculated seeds.Generally, in experiments carried out in greenhouses the inoculated bacteria are subject to competition with other microorganisms present in the soil or inside the plant, as described previously (Omer et al., 2004), suggesting that the evaluated Methylobacterium strains present competitiveness to establish inside the plant even under this environment condition.The AR 1.6/11 was recovered only from root of inoculated plants, suggesting that the ability to colonize the plant shoot is not necessary for promoting the plant growth, since this strain was able to increase the plant height.
Our results obtained by seed inoculation highlights the potential use of these bacteria to stimulate seed germination, plant height and biomass production of C. limonia and C. sunki under commercial nursery conditions.We observed that the non-specificity of bacteria-plant interaction could lead to an undesired effect on the plant.Thus, is very important to conduct a selection of strains in a breeding program and being careful to not extrapolate the results produced on one plant to another host.Future proposals may be the development of a bacterial consortium with specific strains that could be effective in citrus growth promotion and protection generating a low-cost and environmentally safe product.

Table 1 .
Effect of Methylobacterium strains on seed germination and growth of Citrus limonia under commercial nursery conditions.
*Means followed by the same letter show no significant differences by the Tukey's test at p ≤ 0.05.

Table 2 .
Effect of Methylobacterium strains on seed germination and growth of Citrus sunki under commercial nursery conditions.
*Means followed by the same letter show no significant differences by the Tukey's test at p ≤ 0.05.