POLYPHASIC CHARACTERIZATION OF RHIZOBIAL ISOLATES OBTAINED FROM DIFFERENT COMMON BEAN-PRODUCING REGIONS

In Brazil, the common bean is a crop with significant social and economic importance. The prospecting of N 2 fixing bacteria is crucial since biological nitrogen fixation (BNF) is an eco-friendly technique. This work aimed to obtain and characterize rhizobium isolates based on morpho-physiological, molecular, and symbiotic efficiency parameters, using the strains SEMIA 4077, SEMIA 4080, and SEMIA 4088 as references. The characteristics of the isolates and colonies, their tolerance to salinity and temperature, as well as their utilization of carbon sources, served as the basis for the morpho-physiological characterization. BOX-PCR, REP-PCR, and ERIC-PCR markers were used for genotypic characterization. Assessment of the symbiotic efficiency was carried out in a greenhouse, determining the number of nodules (NN), nodule dry weight (NDW), shoot dry weight (SDW), and total-N (Total-N) accumulation in the shoot. Among the isolates, those exhibiting: neutral culture medium pH, fast growth, colony diameter <2 mm, opaque transparency, homogeneous appearance, and cream color were predominant. Compared to temperature, salinity was the most restrictive factor to the growth of the isolates. Most of the isolates grew on sucrose (88.43%) and mannitol (87.28%). Genotypic analysis revealed that 90% of the isolates clustered in the same group as the reference strain SEMIA 4080. The TaMsG2R1 and BaDeG4R2 isolates showed higher Total-N in the shoot than the reference strains and should be evaluated in future studies under field conditions.


Introduction
Common bean (Phaseolus vulgaris L.) is of tremendous economic and social importance worldwide, particularly in developing countries, owing to its capacity to supply the populace with proteins, carbohydrates, and other essential nutrients.Additionally, it is thought to be a more affordable source of protein than foods of animal origin and can be included in a variety of food formulations (Santiago-Ramos et al. 2018).
With an average export of 25 kg N per ton of grain, whose supply is accomplished by the application of N-fertilizers.The common bean is a crop with a high demand for nitrogen (N), one of the most absorbed nutrients by the crop (Soratto et al. 2013) and the one that most influences crop growth (Nascente et al. 2017).However, like other leguminous plants, the common bean can establish an association with

Soil sampling and rhizobial isolation
Soil samples were collected at a depth of 0-20 cm in commercial production areas in the states of Bahia, Goiás, Minas Gerais, Paraná, and São Paulo.Details of the collection sites are outlined in Table 1.The physicochemical characteristics and granulometry of the samples were determined according to Embrapa (1997).
To obtain the rhizobial isolates, six common bean genotypes belonging to 3 different grain color groups, 1-carioca (BRS Aporé and Pérol), 2-black (Ouro Negro and BRS Grafite), and 3-mulatinho (BRS Agreste and Corrente) were used as trap plants.For each genotype, one 5 L pot containing the collected soil and sown with four seeds was used.The plantlets were thinned to two plants per pot five days postemergence.Before sowing, the seeds were superficially disinfected (70% alcohol, 1 min; 2% NaClO, 30 s) and five times washed with sterile water.At 35 days after emergence (DAE), the nodules were collected, and five nodules from each pot were randomly selected to perform the isolation of the rhizobia.
The nodules were disinfected (70% alcohol, 3 min; 2% NaClO, 1 min), and ten times washed with sterile water.Then, in a laminar flow hood, each nodule was placed in an Eppendorf microtube containing 0.5 mL of sterile water and macerated using a sterile glass rod.The supernatant was inoculated in a Petri dish containing Yeast Mannitol Agar (YMA) with bromothymol blue (Fred and Waksman, 1928), according to the methodology described by Hungria (1994).The Petri dishes were incubated at 28 °C and checked daily for one week to track the appearance of colonies, resulting in a total of 479 rhizobial isolates.

Assessment of salinity and temperature tolerance
From the grouping of 479 isolates according to their morphological characteristics, 305 isolates most similar to the type strains (SEMIAs) were selected to determine their tolerance to salinity and temperature (TST).

Assessment of carbon sources use by rhizobial isolates
From the salinity and temperature tolerance assessment of the 305 isolates, 173 were selected for characterization based on the use of 12 different carbon sources: sucrose, glycerol, D-xylose, methyl-β-Dxylopyranoside, dextrose, D-mannose, L-sorbose, inositol, mannitol, D-sorbitol, methyl-α-glucopyranoside, and D-maltose monohydrate.The assessments were according to the method described by Cardoso et al. (2017).

Genotypic characterization of rhizobial isolates
Based on the assessment of salinity and temperature tolerance and carbon source use, 63 isolates were selected for genotypic characterization.Genomic DNA was extracted according to Ausubel et al. (1999).DNA quantity was estimated by spectrophotometry (NanoDrop®, Thermo Scientific, Wilmington, USA), and DNA concentration was adjusted to 50 ng µL -1 for all samples.
PCR products were subjected to electrophoresis on an agarose gel 1% (50 V; 7 h) in TAE buffer 0.75x using 1 kb DNA Ladder® (Norgen) as a DNA band position marker.The agarose gel was stained with SYBR ® green (Life Technologies) and visualized with a MultiDoc-it ® system.

Symbiotic efficiency under greenhouse conditions
Based on genotypic characterization, 15 isolates were selected for symbiotic efficiency assessment under greenhouse conditions.In addition to the isolates, the treatments were composed of the reference strains (SEMIA 4077,SEMIA 4080,and SEMIA 4088), nitrogen fertilizer treatment (120 kg N ha -1 ), and one control treatment (without inoculation and without N).A 1:1 mixture of autoclaved (1 h; 1.5 kg cm -2 ; 127 °C) sand and vermiculite was placed in the upper part of the Leonard jar.Two seeds of common bean cv.Pérola were sown in autoclaved Leonard jars in a randomized block design with four replicates.At five days after emergence (DAE), plantlets were inoculated with a cell suspension containing 1×10 9 cell mL -1 of each isolate and reference strain.Once a week, 200 mL of nutritive solution without N was added.To the nitrogen fertilizer treatment, 2 mL of a solution containing 106.68 mg mL -1 of urea was added.Plants were harvested at 35 DAE.Roots were carefully washed and dried in a paper towel, afterwards, the nodules were detached and counted to determine the number of nodules (NN).Shoots and nodules were dried (65 °C; 72 h) to determine the shoot dry mass (SDM) and nodule dry mass (NDM).Subsequently, shoots of plants were milled to determine the total N (N-Total) using the Kjedahl method, as described by Silva and Queiroz (2009).

Statistical analysis
The rhizobial isolates were grouped by their morphology, salinity and temperature tolerance, and carbon source use traits.The grouping analyses were performed separately for traits.Genotypic characterization data (BOX, ERIC, and REP-PCR) were transformed into a consensus binary matrix.The matrix was used for the construction of a similarity matrix using the Jaccard coefficient.The Unweighted pair-group method (UPGMA) was applied to transform the similarity matrix into a similarity dendrogram using NTSYSpc ® software (Rohlf 1988).Data obtained from the greenhouse experiment were subjected to analysis of variance; when F was significant, the means were compared by Scott-Knott test (p<0.05)using the software SISVAR (Ferreira 2019).

Salinity and temperature tolerance assessment
As seen in Figure 1, in the absence of salinity (0% NaCl), 100% of the isolates grew at temperatures of 28 °C and 33 °C, about 98% of the isolates grew at 38 °C, while 80% of the isolates grew at 48 °C.On the other hand, when the growth under high salinity (6% NaCl) was evaluated, it was observed that about 40%, 20%, and 5% of the isolates were able to grow at 28 °C, 33 °C, and 38 °C, respectively.Besides, no isolate growth was observed at 43 °C and 48 °C.The 305 isolates evaluated for salinity and temperature tolerance were distributed into 116 groups.Group 1 was composed of the isolate LeNfG5R1, showing growth from the least restrictive condition (28°C; 0% NaCl) to the most restrictive condition (43°C; 6% NaCl).In addition to the LeNfG5R1 isolate, another 195 isolates, distributed between groups 2 and 72, deserve special attention because they were more tolerant to salinity and temperature than the three standard strains.The isolate LeNfG5R1, from soil collected in the State of Bahia, showed greater resistance to stresses caused by changes in salinity and temperature.

Carbon source use assessment
The three standard strains and 173 rhizobial isolates, selected from the salinity and temperature tolerance assay, were evaluated for their ability to use 12 carbon sources.Based on the behavior of the rhizobial isolates in relation to the use of carbon sources, the reference strains and the isolates were divided into 102 groups.Group 1 was composed of SEMIA 4080 and 31 isolates, corresponding to 17.9% of the evaluated isolates, showing growth in all carbon sources.Group 9, formed by the standard strains SEMIA 4077, SEMIA 4088, and 6 isolates, did not grow on L-Sorbose (CS7) as a carbon source.Besides, only 47.3% of isolates were able to grow using L-Sorbose (CC7) a carbon source.
On the other hand, sucrose (CS1) and mannitol (CS9) were the carbon sources in which the highest growth percentages were observed, with 88.43% and 87.28% of the total isolates, respectively, having the capacity to utilize them.

Genotypic characterization based on BOX, REP, and ERIC-PCR markers
BOX, ERIC, and REP-PCR markers were applied to determine the genetic diversity of 63 rhizobial isolates, selected from the carbon source use assay, using the three reference strains (SEMIAs 4077, SEMIA 4080, and SEMIA 4088) for the grouping of isolates.The BOX-PCR fingerprinting revealed the formation of at least 4 and at most 20 bands with different molecular weights (Figure 2A), demonstrating a high polymorphism among the isolates.Similarity was observed in the banding profile between BaDeG4R2 and BaDeG4R3 isolates; BaDeG5R5 and BaDeG6R3, LeNfG2R3 and LeNfG4R1.Despite the similarity found among some isolates, none of them presented an identical profile to the reference strains (SEMIAs 4077,SEMIA 4080,and SEMIA 4088).Similar results were observed by Cardoso et al. (2012).
With REP-PCR fingerprinting (Figure 2B) at least three and at most 12 bands were observed, demonstrating high polymorphism, represented by the predominance of single banding profiles.The reference strains SEMIA 4077 and SEMIA 4088 showed identical banding profiles.Although none of the isolates showed an identical profile to the reference strains, the SiGaG3R2 and SiGaG3R5 isolates showed a similar profile to each other.
By the ERIC-PCR fingerprinting (Figure 2C) a high polymorphism can be observed, represented by the predominance of single banding profiles.Among them, it was observed at least three and at most 15 bands, with the reference strains SEMIA 4077 and SEMIA 4088 presenting identical profiles.
The banding profiles from the BOX-PCR, REP-PCR, and ERIC-PCR markers were transformed into a consensus binary matrix, which was used to group the isolates using the UPGMA algorithm and the Jaccard coefficient (Figure 3).It was observed that the evaluated isolates, including the reference strains SEMIA 4077, SEMIA 4080 and SEMIA 4088, showed 70% similarity.Considering a similarity of 75%, seven groups were formed.the UPGMA algorithm, and the similarity matrix was determined using the Jaccard coefficient.

Evaluation of symbiotic efficiency in a greenhouse
From the cluster analysis based on molecular markers (Figure 3), 15 isolates were selected, 13 belonging to the G1 group, the group with the highest number of isolates and where the reference strain SEMIA 4080 was found, and the two isolates from the G7 group, which showed the lowest similarity (70%) with the reference strains, to be evaluated for symbiotic efficiency.A significant difference (p<0.05) was observed between treatments by the Scott-Knott test for the number of nodules (NN).Among the evaluated isolates, 11 of them presented NN statistically equal to the reference strains, SEMIA 4077, SEMIA 4080, and SEMIA 4088.The other 4 isolates (BaDeG4R3, BaDeG4R2, ItCaG1R4, and CaMaG4R3) presented NN statistically equal to the nitrogen (NfT) and absolute control (AC) treatments (Table 2).The NN ranged from 99.33 nodules plant -1 (PaJuG2R5) to 283 nodules plant -1 (PaPoG4R4), and the reference strains SEMIA 4077,SEMIA 4080 and SEMIA 4088 presented 170.00,190.66 and 119.33 nodules plant -1 , respectively (Table 2).
For the nodule dry weight (NDW), all the evaluated isolates presented NDW significantly superior to the reference strain SEMIA 4080 and equal to the reference strains SEMIA 4077 and SEMIA 4088 (Table 2).The TiIvG1R2 isolate showed the highest NDW, with 101 mg plant -1 , while the UnCaG6R4 isolate presented an NDW of 5 mg plant -1 , which was the lowest value found (Table 2).In this work, we found a 20.2-fold variation in NDW among the isolates (5 to 101 mg plant -1 ).
All isolates presented SDW values statistically equal to the values presented by the reference strains (SEMIAs), differing only with the NfT treatment (Table 2).The reference strains showed SDW values ranging from 0.74 (SEMIA 4088) to 1.59 g plant -1 (SEMIA 4077), while the isolates showed values ranging from 0.77 (TiIvG4R5) to 1.58 g plant -1 (PaPoG3R4).
By analyzing the Total-N, it was observed that the isolates TaMsG2R1 and BaDeG4R2 presented values higher than the reference strains, being statistically similar to the NfT treatment (Table 2).

Discussion
The morphological characterization of 479 isolates resulted in the formation of 62 morphological groups, representing a high diversity among the evaluated isolates.These results corroborate those of Sampaio et al. (2016), who used Cerrado soils and wild genotypes of the common bean as trap plants.
Although there is a difference regarding the used trap plants, both studies indicate a high morphological diversity among rhizobial isolates obtained from Brazilian soils.
High salinity (6% NaCl) and temperature (43 °C and 48 °C) inhibited the bacterial growth.Sharma et al. ( 2013) also highlighted the effect of salinity on rhizobial growth, however, they also found isolates capable of growing in a culture medium with high salinity.However, the isolate LeNfG5R1, from soil collected in the State of Bahia, showed greater resistance to stresses caused by changes in salinity and temperature.This result is similar to that observed by Sharma et al. (2013), which showed that the rhizobia isolated from the desert soils are able to survive, grow and effectively nodulate their leguminous hosts even at high salt concentrations.
By the carbon source analysis the isolates were grouped into 102 groups.Among the isolates from G1, those from the state of Minas Gerais showed greater metabolic diversity, being able to use a greater number of carbon sources.This result corroborates the studies by Fernandes Júnior et al. ( 2012), Patel and Dubey (2014).The data obtained by these authors indicate that rhizobia have the ability to use several sources of carbon, including those used in this study.
Literature data suggest that carbon sources, such as xylose (CS3), can influence the composition and production of exopolysaccharides, commonly called mucus (Razika et al. 2012).This mucus is extremely important for soil bacteria because it protects against abiotic stresses such as pH, temperature, UV radiation, and salinity, among other factors.In this regard, the findings of this kind of research are important not only for understanding the symbiotic relationship but also for evaluating the biotechnological potential of these bacteria.
Sucrose (CS1) and mannitol (CS9) provided the highest growth percentages among the isolates.These results are in concurrence with those obtained by Sampaio et al. (2016), who found that 89% of the 76 isolates studied were able to utilize sucrose and mannitol as carbon sources.
YMA culture medium, which has mannitol in its composition, is the standard medium for the isolation of bacteria from legume nodules.However, studies of the metabolic behavior of bacteria regarding the use of carbon sources are important in order to allow the adoption of other carbohydrates for the preparation of the medium, replacing mannitol, favoring in a more specific way the conditions for microbial development (Fernandes Júnior et al. 2012).
The evaluated isolates are too close on the genotypic point of view, since only seven groups were formed at a similarity level of 75%.In a study carried out by Youseif et al. (2014) using these same molecular markers with 11 rhizobial isolates, these isolates grouped into three clusters at about 35% level of similarity.
Under greenhouse condition, the NN ranged from 99.33 to 283 nodules plant -1 , while for NDW was found a 20.2-fold variation among the isolates (5 to 101 mg plant -1 ).A Similar variation in NDW (18.3-fold variation) was found by Cardoso et al. (2017), with values ranging from 21 to 384 mg plant -1 .Cardoso et al. (2012) found a much smaller variation between isolates (3.9-fold variation), with values ranging from 146 to 564 mg plant -1 .
Among the variables used in the evaluation of nodulation (NN and NDW), NDW was more useful for the selection of rhizobium isolates, due to the better correlation of NDW with symbiotic performance.Correlation analysis between NN and NDW with growth parameters, such as root length, root dry weight, plant height and shoot dry weight showed better adjustment for NDW over NN (Koskey et al. 2017;Samudin and Kuswantoro 2018).According to Yang et al. (2017), correlations were detected between fixed N2 and NN.However, NN is not always positively correlated with fixed N2 in plants, as they only found a significant correlation between fixed N2 and NN after inoculation with certain rhizobium strains.
The SDW of the plants inoculated with the reference strains and with the isolates were so similar, indicating the effectiveness of the isolates in fixing N, as pointed out by Cardoso et al. (2017), resulting in high Total-N amounts accumulated in the aerial part of the plants.

Figure 1 .
Figure 1.Growth percentage of rhizobial isolates under different salinity and temperature conditions.

Figure 3 .
Figure 3. Consensus dendrogram produced by combining the BOX, REP and ERIC-PCR data of 63 isolates obtained from different soil origins and cultivars of common bean.The dendrogram was generated usingthe UPGMA algorithm, and the similarity matrix was determined using the Jaccard coefficient.

Table 1 .
Details of the soil sampling locations for obtention of new rhizobial isolates.

Table 2 .
Nodule number (NN), nodule dry weight (NDW), shoot dry weight (SDW) and total nitrogen accumulation (Total-N) of the common bean inoculated with different rhizobial isolates.NfT-N-fertilizer treatment, AC-absolute control.Means followed by different letters are statistically different by the Skott-Knott test (p<0.05).