Evaluation of plant-growth promoting properties of Gluconacetobacter diazotrophicus and Gluconacetobacter sacchari isolated from sugarcane and tomato in West Central region of Colombia

1 Biological Research Group, Research Institute of Microbiology and Agro-industrial Biotechnology, Universidad Católica de Manizales, Carrera 23 No. 60-63, 170002, Manizales, Colombia. 2 Research Group on Food and Agro-industry, Department of Engineering, Universidad de Caldas, Calle 65 No. 26-10, 170004, Manizales, Colombia. 3 Research Group on Chromatography and Related Techniques, Department of Chemistry, Universidad de Caldas, Calle 65 No. 26-10, 170004, Manizales, Colombia.


INTRODUCTION
The increase of world population has led to the excessive use of chemical fertilizers causing soil erosion and loss of its physicochemical and microbiological equilibria provoking negative environmental impacts (White and Brown, 2010).For these reasons, the need to produce alternative environmentally friendly agricultural supplies contributing to enhance the crop yield to meet the global food demand has arisen.An important amount of bacteria exhibit plant-growth promoting properties and are being employed for production of biofertilizers alternative to chemical fertilizers.However, the production volume and availability of these bio-inputs are still small so the research in the isolation, identification, and characterization of new or already discovered bacterial strains is of paramount importance to develop new biotechnological products and processes relevant for the world agriculture.
Plant-growth promoting rhizobacteria are soil microorganisms growing in the surface of the root and in the space surrounding it.They are directly or indirectly related to the growth promotion of the plants (Ahemad and Kibret, 2014).The direct-action mechanisms include nitrogen biological fixation, phosphate solubilization, iron sequestering, and production of phytohormones and 1aminocyclopropane-1-carboxylate deaminase.Main indirect mechanism is related to their function as biocontrol agents, which is evidenced through the competition for nutrients and space, induction of systemic resistance and production of siderophores or antifungal metabolites (Glick, 2012).Biological nitrogen fixation associated with vascular plants is the most important process for nitrogen input to the natural ecosystems (Monteiro et al., 2012).This process converts the atmospheric nitrogen into ammonia.Nitrogen-fixing bacteria have an enzyme, called nitrogenase, which catalyzes the breakdown of the triple covalent bond of the nitrogen molecule at ambient temperature and pressure (Eskin et al., 2014).Phosphorus is the second inorganic nutrient necessary for all life forms.It is an essential element in molecules such as RNA, DNA, and ATP, as well as phospholipids (Ahemad and Kibret, 2014).In several tropical soils, phosphorus predominates in the form of inorganic insoluble phosphates.These insoluble forms are not absorbed by the plants (Bashan et al., 2013).Inorganic phosphate can be divided into two groups: Calcium phosphates and iron and aluminum phosphates.On the other hand, phytohormones are important metabolites promoting the plant growth.The auxins are a group of phytohormones regulating the plant development and having a direct effect on the growth, cell division, and formation of the roots.They are applied during the in vitro cultivation of plant material and in plantations (Castillo et al., 2005).
In this sense, bacteria of the genus Gluconacetobacter have a huge potential.This genus was described by Yamada and Yukphan (2008) and includes gram negative bacteria.Gluconacetobacter bacteria, belonging to the class Alphaproteobacteria, order Rhodospirillales and family Acetobacteraceae, do not form endospores, and are obligate aerobes.G. liquefaciens, G. azotocaptans, G. diazotrophicus, G. hansenii, G. johannae and G. sacchari, among others are the main species of the genus (Sievers and Swings, 2005).Gluconacetobacter diazotrophicus is an endophytic plantgrowth promoting bacterium first reported by Cavalcante and Döbereiner (1988) associated with sugarcane in Brazil.This bacterium has been isolated from sugarcane cultivars in different countries (Bellone et al., 1997;Fuentes et al., 1993;Li and Macrae, 1991).Likewise, it has been recovered from other crops like coffee (Jiménez et al., 1997), pineapple (Tapia et al., 2000), carrot (Madhaiyan et al., 2004), and rice (Muthukumarasamy et al., 2005).G. diazotrophicus is a subject of great interest among nitrogen-fixing bacteria, as it fixes nitrogen under aerobic growth conditions (Eskin et al., 2014).This property allows it to perform the process in nonleguminous plants.Additionally, nitrates do not inhibit nitrogenase activity; this is because G. diazotrophicus lacks the nitrate reductase enzyme (Cavalcante and Döbereiner, 1988).Another feature of interest that has been attributed to this bacterium is its ability to solubilize phosphates (Stephen et al., 2015) and zinc (Crespo et al., 2011;Natheer and Muthukkaruppan, 2012).This solubilization ability allows G. diazotrophicus to access insoluble nutrients and make them available for absorption by plants.G. diazotrophicus produces auxins such as 3-indoleacetic acid (IAA) (Eskin et al., 2014), which highlights its potential even more.IAA has control over many important developmental processes in plants, including cell division and growth, vascular tissue differentiation, root initiation, apical dominance, phototrophic and gravitropic responses, flowering, fruit ripening, leaf senescence and leaf and fruit abscission (De Souza et al., 2015;Hernández-Rodríguez et al., 2010;Malekpoor Mansoorkhani et al., 2014).
G. sacchari has been isolated from the pods of sugarcane leaves and the pink sugar-cane mealy bug attacking this crop.G. sacchari does not fix nitrogen as G. diazotrophicus does (Franke et al., 1999).It is a bacterium efficient for cellulose production when grown in glucose-based media and the yield and quality of the cellulose obtained is comparable to those of other *Corresponding author.E-mail: osanchez@ucaldas.edu.co.
Author(s) agree that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License cellulose-producing bacteria (Trovatti et al., 2011).
No studies have been published concerning the presence and potential use of bacteria belonging to the genus Gluconacetobacter under the agro-climatic conditions of Colombia.The aim of this study was to determine the plant-growth promoting properties of bacteria of the genus Gluconacetobacter obtained from tomato and sugarcane cultivars in West Central region of Colombia and identify potential isolates for the development of biofertilizer-type inoculants for economically important crops.

Sampling location
Samples from sugarcane (Saccharum officinarum L.) and tomato (Solanum lycopersicum) were collected in the West Central region of Colombia.Sampling was performed at sites where no chemicals had been recently applied.A total of 10 plants were collected at each site.A bromatological analysis of each soil sample was made.
The sugarcane sampling was carried out in crops at three farms with different planting times (270, 360 and 450 days) located at 919, 1001 and 1121 m.a.s.l., respectively; from each plant sample, leaves, stems and roots were collected.Additionally, samples from a greenhouse-grown tomato crop were taken at the "Tesorito" farm at the University of Caldas.Sampling was performed three times at different stages of the production cycle (50, 100 and 150 days).Stem and root samples were collected.

Culture media and growing conditions
A nitrogen-free medium that allows for the enrichment of samples and the selection of diazotrophic bacteria (LGI-P semisolid medium) was initially used (Cavalcante and Döbereiner, 1988).Isolation was performed using LGI-PNs agar.LGI-PNs agar has the same composition per liter as the semisolid LGI-P medium, with the addition of 50 mg of yeast extract and 15 g of agar-agar.Bacteria of interest were purified using potato dextrose agar (PDA) (Oxoid ® , United Kingdom).Modified DYGS broth (Rodrigues Neto et al., 1986) was used to obtain the biomass needed for polymerase chain reaction (PCR) assays and the evaluation of indolic compound production and phosphate solubilization.Phosphate solubilization was evaluated using National Botanical Research Institute's Phosphate (NBRIP) medium, which was individually supplemented with Ca3(PO4)2 and AlPO4 at 5 g×L -1 (Nautiyal, 1999).
Cultures were incubated at 30°C for 6 to 7 days.The following criteria were used to select isolates with typical characteristics of Gluconacetobacter spp.(Cavalcante and Döbereiner, 1988): Gas production, formation of a yellow-orange film on the medium surface, and a color change.Cultures matching these criteria were grown again in semisolid LGI-P medium.Once it was verified that samples had the same characteristics as the original culture, isolation in LGI-NPs medium was performed.The colonies were purified on PDA.The strain of G. diazotrophicus PAL5 (ATCC 49037) was used as a positive control.

Biochemical characterization of isolates
The biochemical test API 20NE (BioMérieux, France) was performed as directed by the manufacturer.Additionally, gram stains and oxidase and catalase tests were completed (Cavalcante and Döbereiner, 1988;Dong et al., 1995;Gillis et al., 1989).Isolates were stored at -80°C in a 15% glycerol solution.

PCR identification of isolates
Preserved isolates were reactivated.One colony from each plate was picked and transferred to 5 mL modified DYGS broth (Rodrigues Neto et al., 1986).Cultures were incubated at 30°C under stirring at 150 rpm for 72 h.Cultures were centrifuged for 10 min at 2500 rpm, and pellets were washed twice with sterile 0.85% saline solution.
Extraction and purification of genomic DNA was performed using a PureLink® Genomic DNA kit (Invitrogen ® , USA) and the recommended protocol for gram-negative bacteria.The quality of extraction was verified by 1% agarose-ethidium bromide electrophoresis using λ/Hind III phage as a molecular weight marker.
The 16S rDNA subunit was amplified via PCR from the previously extracted genomic DNA.For this reaction, the universal primers RB and RM (Madhaiyan et al., 2004) were used; the expected fragment was approximately 800 bp.The final volume of the PCR reaction was adjusted to 25 µL (1X PCR buffer, 0.2 mM dNTPs, 1 mM MgCl2, 0.5 μM primers, 1 U of polymerase taq DNA recombinant polymerase by Invitrogen and 100 ng of bacterial DNA).The amplification was performed using a T100 thermocycler (Bio-Rad, USA) according to the following amplification program: initial denaturation at 95°C for 10 min, then 35 cycles at 92°C for 1 min, 57°C for 30 s and 72°C for 1 min.The PCR products were analyzed via 1% agarose-ethidium bromide electrophoresis.Sequencing of the PCR products was performed by Macrogen ® (Republic of Korea).

Bioinformatics analysis
Quality analysis of the obtained sequences was performed using the CLC ® Main Workbench program (Qiagen Company, USA).Sequences with a Phred value less than 30 per base and a size less than 600 bases were rejected.
The sequences were annotated using BLASTN version 2.2.27+ (Altschul et al., 1990).Sequences were compared against the nt database of the National Center for Biotechnology Information (NCBI), with an e-value threshold of 1×10 -5 .These results were filtered with 80% similarity and 80% coverage of the region amplified with RM/RB primers.Once the BLASTN descriptions were obtained, a database of sequences belonging to the genus Gluconacetobacter was built and validated using data from RDP (Ribosomal Database Project).Clustal W version 2.1 (Larkin et al., 2007) was used for alignments.Phylogenetic tree was built using the maximum likelihood method based on the Hasegawa-Kishino-Yano model (Hasegawa et al., 1985).The percentage of trees in which the associated taxa clustered together is shown next to the branch (bootstrap value).The analysis involved 14 nucleotide sequences.All positions containing gaps and missing data were eliminated.There is a total of 577 positions in the final dataset.

Assessment of growth-promoting properties
The evaluation of the growth-promoting properties of isolates was carried out using inocula obtained by transferring colonies to 5 mL tubes containing DYGS medium (pH 6.5).These inoculates were incubated at 30°C with stirring at 150 rpm for 4 days.The cells were washed three times with 0.85% isotonic saline solution, and the pellet was resuspended to a cell concentration of 10 8 cells×mL -1 (OD600= 0.9 -1).

Nitrogenase activity
Bacterial nitrogenase activity was determined using the nitrogenfree semisolid LGI-P medium and the acetylene-reduction assay (ARA) as described by Muthukumarasamy et al. (1999), Videira et al. (2007) and Boddey et al. (2007) with some modifications.A suspension of 20 μL at cell concentration of 10 8 cells×mL -1 was used to inoculate 5 mL of LGI-P medium.Cultures were incubated at 30°C for 48 h.Subsequently, 10% v/v of the culture vial atmosphere was replaced with acetylene gas.After incubation at 30°C for 24 h, 500 µL gas samples were analyzed on a GCMS-QP2010 Plus (Shimadzu®) gas mass chromatography flame ionization detector using a GS-Alumina column (Agilent Technologies ®).Each sample was evaluated six times.The results were expressed as nmol reduced acetylene×mL -1 ×h -1 and nmol reduced acetylene×10 8 cells×h -1 .

Assessment of indolic compounds
Total indolic compounds were determined by spectrophotometry using Salkowski reagent (Glickmann and Dessaux, 1995;Gordon and Weber, 1951;Sarwar and Kremer, 1995).A recovered colony grown in PDA medium was transferred to the DYGS medium and incubated at 30°C under stirring at 150 rpm for 24 to 48 h.Two washes were performed with sterile distilled water, and the inoculum was prepared at a final concentration of approximately 10 8 cells×mL -1 .
Indolic compounds production was induced by adding Ltryptophan (1.01%) (Panreac ® , USA) to the DYGS medium.Cultures were incubated at 30°C and 150 rpm for 60 h.The determination of indolic compounds was made in the supernatant using Salkowski reagent and by reading absorbance at a wavelength of 540 nm.Each sample was analyzed in triplicate.The levels of indolic compounds were estimated from a standard curve of 150 ppm of indolic acetic acid (Merck ® , USA) and were expressed in µg×mL -1 .

Phosphate solubilization
Phosphate solubilization was assessed using NBRIP medium with bromocresol green (22 mg×L -1 ) (Nautiyal, 1999), which was supplemented individually with 5 g×L -1 of Ca3(PO4)2 or AlPO4.To each plate, 20 µL volume of the inoculum was added in triplicate.Readings at 24, 48 and 72 h were performed.The diameter of the colony and the size of solubilization halo were determined in order to calculate the solubilization index SI (Kumar and Narula, 1999) and the percentage of solubilization efficiency SE (Nguyen et al., 1992).Additionally, the production of organic acids was evaluated by noting the changes in the pH of the culture medium.

Statistical analysis
Analysis of variance was performed using the GLM procedure of SAS (SAS Institute, Cary, USA) to determine the occurrence of significant differences among treatments for all evaluated variables.The comparison of means was made using the Duncan test (pvalue <0.05).

Isolation,
biochemistry, characterization and molecular identification of Gluconacetobacter spp.
Soils used for evaluated crops were rich in organic matter (between 13.0 and 20.4%), with pH values between 4.4 and 5.9 and nitrogen levels between 0.40 and 0.68%, according to analyses performed by the Laboratory of Chemistry and Soil Fertility at the Universidad de Caldas.Morphotypes matching the selection criteria, that is, the colonies obtained from a yellow-orange surface film that caused clearance on LGI-PN culture medium and formed brown colonies in PDA medium (Figure 1), were considered as presumptive Gluconacetobacter spp.A total of 16 of the 21 isolated colonies from sugarcane presented the characteristics described previously.Only four of the 33 isolates from tomato presented this pigmentation in LGI-P, LGI-PN and PDA culture media.
PCR with the RB/RM universal primers resulted in the amplification of products of the expected size (800 bp) for 54 presumptive isolates.Through a BLASTN analysis of the 54 PCR products amplifying the 16S rRNA gene, two isolates were identified as G. diazotrophicus and three isolates as G. sacchari.The two G. diazotrophicus isolates from sugarcane were obtained from root (coded as GIBI025) and stem (coded as GIBI029) samples with a BLAST identity (BI) of 99.37% in both cases and global alignment identity (GI) of 98.42 and 98.24%, respectively.Two of the three G. sacchari isolates were recovered from sugarcane: GIBI014 (BI 100% and GI 98.78%) and GIBI031 (BI 100% and GI 98.95%); the third isolate was recovered from tomato and coded as GIBI141 (BI 99.83% and GI 99.65%).For greater identification reliability, phylogenetic analyses supported by the scientific literature were performed with reported reference sequences.An alignment was performed over 543 bp that included the sequences obtained and 8 representative sequences from different species from the genus Gluconacetobacter.The tree was rooted using the 16S rRNA sequence of Burkholderia cepacia NR 114491 as shown in Figure 2. In this figure, two groups among Gluconacetobacter isolates can be observed.One group includes G. johanae, G. azotocaptans, and G. diazotrophicus.The other cluster is formed by G. liquefaciens and G. sacchari.These findings are consistent with the classification of the Gluconacetobacter genus described by Yamada and Yukphan (2008).
Isolates were positive for catalase and indolic compounds and negative for oxidase and the liquefaction   of gelatin.Additionally, isolates grew at pH 5.5 and 30°C.
The biochemical characteristics of G. diazotrophicus isolates are similar to the reference strain ATCC 49037 (Sievers and Swings, 2005).

Nitrogenase activity
Table 1 shows the test results for nitrogenase activity for the Gluconacetobacter spp.isolates compared to the characteristics of the reference strain.The GIBI141 strain presented higher nitrogenase activity than the other isolates and the reference strain while the GIBI031 isolate exhibited values lower than the reference strain.These isolates showed statistically significant differences at 5% significance level.

Phosphate solubilization capacity
Phosphate solubilization halos were not present in plates with aluminum phosphate; only color change in halos was present, which indicates acid production (Table 2).In this test, the rate of acid production was calculated by comparing the halo diameter for the color change to the diameter of the bacterial colony.To determine the efficiency of the tricalcium phosphate solubilization, the percentage of solubilization efficiency was evaluated after 72 h of incubation.Only one G. sacchari isolate (GIBI031) showed better solubilization efficiency than the reference strain with a value statistically equal to the best G. diazotrophicus isolate (GIBI029).In the case of aluminum phosphate, GIBI031 and GIBI029 isolates exhibited a performance statistically equal to that of the reference strain (p-value <0.05).

DISCUSSION
The strains evaluated in this study were isolated from the roots (GIBI025), leaves (GIBI014) and stems (GIBI029, GIBI031) of a 270-day-old sugarcane crop; GIBI141 isolate was recovered from tomato stems.This is consistent with reported studies in which G. diazotrophicus bacteria were recovered from sugarcane in Brazil (Cavalcante and Döbereiner, 1988), Australia (Li and Macrae, 1991), Mexico (Fuentes et al., 1993) and Argentina (Bellone et al., 1997).The source of the G. sacchari isolates (GIBI014 and GIBI031) corresponds to that reported by Franke et al. (1999).In addition, this species was also recovered from tomato stems (GIBI141) of a 150-day-old crop, for which there is no any work reporting this antecedent in the available scientific literature.It is worthy to highlight that the isolation of G. diazotrophicus and G. sacchari has not been previously reported in Colombia.G. diazotrophicus and G. sacchari are endophytic bacteria whose occurrence has been proven in the sugarcane tissues by Beneduzi et al. (2013) and Franke-Whittle et al. (2005), respectively.
The high content of assimilable sugars favors the presence of G. diazotrophicus.In natural environments, these bacteria use the organic acids in sugarcane juice and other sugars present (like glucose) in the aboveground part of the plant as energy sources.This is evidenced in vitro by the growth of this species in culture media with 10% sucrose at pH 5.5.These conditions are similar to those of sugarcane juice (Baldani et al., 2014).Additionally, the isolates of the genus Gluconacetobacter from sugarcane were obtained from a 270-day-old crop.Samples from such plants have lower percentage of lignified roots and stems compared to samples taken from older sugarcane crops.In coffee crops, Jiménez et al. (1997) observed greater recovery of endophytic bacteria from young plant tissues.This was because processing those samples facilitates tissue homogenization and the release of associated bacteria.In the present study, this aspect could have influenced the failure to obtain isolates from 360-and 450-day-old sugarcane crops.
The isolation of G. diazotrophicus and G. sacchari from tomato crops has not been previously reported in the literature.The only work relating G. diazotrophicus to tomato crops corresponded to the evaluation of colonization and yield promotion of tomato by the inoculation of G. diazotrophicus (Luna et al., 2012).This represents a great potential for production of biofertilizers for crops other than sugarcane for which commercial inoculants are being developed in Brazil and India.
Orange colonies obtained from LGI-PN medium and brown colonies formed in PDA (sugar 10%) are typical of the bacteria under study.Such colonies were described by Cavalcante and Döbereiner (1988) and Baldani et al. (2014).The recovered isolates presumably formed exopolysaccharides (EPS), which were evidenced by the mucoid and shiny appearance of the colonies.Cultures in liquid media also increased in viscosity between 60 and 72 h of incubation as described by Meneses et al. (2011).Exopolysaccharide biosynthesis is required for the biofilm formation during plant colonization by the endophytic bacterium G. diazotrophicus (De Souza et al., 2015).
For the isolates obtained, macroscopic and microscopic features and the results of the biochemical tests were consistent with the most parsimonious tree built using the 16S rRNA gene.Clustering of the genus Gluconacetobacter (bootstrap> 50) was evidenced when compared to sequences from different genera (data not shown).The tree containing Gluconacetobacter spp.(Figure 2) showed clustering between G. diazotrophicus and taxonomically related bacteria such as G. azotocaptans and G. johannae.These findings were supported by a bootstrap value greater than 50 (Matsutani et al., 2010) and also were consistent with the similarities found among the groups with some GI values below 98.65%, the threshold suggested by Kim et al. (2014) to be considered within the same species.In the cluster of G. sacchari and G. liquefaciens isolates, the grouping and bootstrap values are in agreement with the identity percentage, which was above 98.78% for the three isolates.This provides a greater certainty on the taxonomic allocation of G. sacchari isolates.Despite the results of the tree were not conclusive for the genomic region used, the BLAST results (identities greater than 98.65% in the 16S rRNA gene) let us infer that the isolates can be associated to the G. diazotrophicus species with a high degree of certainty.They were associated with reference sequences of the species G. azotocaptans and G. diazotrophicus.Although the 16S ribosomal gene provides evolutionary information, it should be noted that only a comparison of complete genomes could precisely define evolutionary relationships between closely related microorganisms (Mende et al., 2013).The use of the 16S rRNA gene facilitates the identification of microorganisms from a broader perspective.
The nitrogenase activity of the isolates was assessed by the reduction of acetylene to ethylene.Reference strain ATCC 49037 and strains GIBI014 and GIBI025 showed lower enzyme activity than the strains GIBI029 and GIBI141.For this assay, cultures were started for each isolate from a density of 10 8 cells×mL -1 . The nitrogenase activity of G. diazotrophicus is influenced under in vitro and in vivo conditions by various factors, such as the presence of inorganic nitrogen sources (Fuentes et al., 1993), the aeration of the culture medium used for biomass production (Fuentes et al., 1993;Muthukumarasamy et al., 2002), and the plant genotype.It has been estimated that G. diazotrophicus can fix up to 150 kg N×ha -1 ×year -1 in sugarcane (Boddey et al., 1991).
Therefore, for future research, the effect of adding the isolates found on the interaction with sugarcane plantlets under regional conditions should be assessed.G. sacchari is not considered a nitrogen-fixing bacterium since it does not exhibit any nitrogenase activity measured by ARA (Franke-Whittle et al., 2005).However, in this work, G. sacchari GIBI141 isolate presented the highest acetylene-reduction activity compared to the other isolates of the same species with an activity statistically equal to that of the G. diazotrophicus GIBI029 isolate.In the phylogenetic tree, GIBI141 isolate is found in a group separate of the other two G. sacchari isolates (Figure 2).This could explain its differential behavior.
The recovered isolates had the ability to produce indolic compounds.Works performed by Chaves et al. (2015) and Ríos et al. (2016) also demonstrated the ability of G. diazotrophicus to produce auxin compounds.Isolate GIBI029 produced an amount of indolic compounds similar to that produced by the reference strain ATCC 49037.It is recommended that the specific determination of indolic acetic acid be performed via chromatographic techniques such as HPLC since spectrophotometric determination only allows for the quantification of total indolic compounds.G. sacchari isolates produced indolic compounds in an amount equal or higher than that of G. diazotrophicus GIBI025.It is necessary to assess this behavior in field to determine the potential of these isolates for promoting plant growth.In this way, the application spectrum of this bacterium can be broadened considering that most reports on G. sacchari are aimed at synthesis of bacterial cellulose (Trovatti et al., 2011).
Phosphate solubilization was evaluated using two sources of phosphorus, tricalcium phosphate and aluminum phosphate.The NBRIP medium was supplemented with bromocresol green to improve the visualization of the test.After 72 h, the solubilization of tricalcium phosphate was observed in both of the native isolates of G. diazotrophicus and in G. sacchari.This finding demonstrated the potential of the recovered isolates with the exception of the GIBI141 isolate, which had the lowest values of tricalcium phosphate solubilization and acid production in the presence of aluminum phosphate.Phosphorus solubilization has been observed in several G. diazotrophicus isolates recovered from sugarcane.Delaporte-Quintana et al. (2016) reported the solubilization of insoluble phosphate (dicalcium phosphate, tricalcium phosphate, hydroxyapatite, basis slag, and ferric phosphate) by four different strains of G. diazotrophicus (PAL5 and its Tn5derivative mutants K416, 16D10 and 16G6 defective in organic acid production) in NBRIP solid minimal medium.The PAL5 strain solubilized all the sources of insoluble phosphate with SI of 9.25 (dicalcium phosphate), 6.84 (tricalcium phosphate), 2.82 (hydroxyapatite), 1.11 (basis slag), and 1.20 (ferric phosphate).These authors indicate that K416 and 16G6 mutants were able to solubilize some of the insoluble phosphate sources tested although with less ability, compared to PAL5.16D10 mutant was completely unable to solubilize any insoluble phosphate source tested.G. diazotrophicus and G. sacchari isolates used in this study, ATCC 49037 (PAL5), GIBI029, GIBI025, GIBI014, GIBI031, and GIBI 141, exhibited SI with tricalcium phosphate lower than those reported for PAL5 in the above-cited work (3.10, 3.62, 3.33, 2.59, 3.55 and 1.61, respectively); NBRIP medium and shorter incubation period of 72 h was used.Aluminum phosphate solubilization was not observed.However, the production of acids was evidenced by the change in the medium color.It is important to perform laboratory tests in liquid media with other sources of phosphorus before performing field assessments.It is also important to complete the characterization of the acids involved in solubilization (Bashan et al., 2013;Restrepo et al., 2015).In the case of G. sacchari, reports on the evaluation of phosphate solubilization were not found.Nevertheless, G. sacchari GIBI031 isolate presented the highest values of solubilization among the phosphorus sources (SI: Ca 3 (PO 4 ) 2 , 3.55 and AlPO 4 , 15.87) evaluated in this work within this species.This performance should be confirmed through phosphate solubilization in liquid medium as mentioned previously.

Conclusion
In this work, two native strains of G. diazotrophicus and three strains of G. sacchari were isolated from the roots and stems of sugarcane.G. sacchari strains were recovered from tomato crop.This is the first report on the isolation of G. diazotrophicus and G. sacchari in Colombia.G. diazotrophicus promotes plant growth and has a great potential as an active principle of biofertilizers.The isolates from sugarcane recovered in this study (GIBI029 and GIBI031) showed growthpromoting properties, such as indolic compounds production and phosphate solubilization, with values above or close to those of the reference strain ATCC 49037.For G. sacchari, the highest acetylene-reduction activity was observed in GIBI141 isolate that is an important finding considering that this activity has not been reported before; the activity value achieved was statistically equal to that of the G. diazotrophicus GIBI029 isolate.The evaluation of phosphate solubilization in Gluconacetobacter isolates reported in this study showed statistically significant differences being G. diazotrophicus GIBI029 and G. sacchari GIBI031 the isolates with the best phosphate solubilization.All these isolates are therefore promising for the development of commercial inoculants.
The specific determination of indolic acetic acid by chromatographic techniques such as HPLC is recommended, because spectrophotometric determination only allows for the quantification of total indolic compounds.The continued evaluation of phosphate solubilization by G. diazotrophic and G. sacchari using liquid culture media with other sources of phosphate is also recommended.This will enable a more efficient evaluation of solubilization properties in tomato crops at the seedling level (seeder) and in greenhouse-grown tomatoes.
The development of biotechnology-based inoculants from these bacteria would allow for a long-term reduction in chemicals used in agriculture and promote the development of environmentally friendly agronomic strategies.However, the success of these bacterial inoculants depends on the selection of native strains efficient in this soil type.Such strains must have the ability to colonize the rhizosphere and maintain biological activity.In this sense, the isolates obtained have great potential for the subsequent commercial development of an innovative biotechnological inoculant for crops of tomato and even other vegetables.

Figure 1 .
Figure 1.(a) Orange-yellow surface film and clearance on the LGI-PN culture medium.(b) Orange colonies in the LGI-PN medium incubated during 8 days.(c) Brown colonies from potato dextrose medium incubated during 8 day.

Table 2 .
Indolic compound production after 60 h of incubation and phosphate solubilization after 72 h of incubation.
Values obtained are averages of three replicates.Identical letters indicate the absence of statistically significant differences.API: Acid production index; SI: Solubilization index; SE: Solubilization efficiency; Std dev: Standard deviation.