Abstract
Paenibacillus azotofixans is a nitrogen-fixing bacterium often found in soil and in the rhizospheres of different grasses. In this study, two Brazilian clay soils were planted with cross-hybrid maize (BR-201) and four stages of plant growth were analyzed to characterize the P. azotofixans populations present in the rhizoplanes, rhizospheres, and non-root-associated soils (herein called nonrhizospheres). A total of 106 strains were isolated and identified as P. azotofixans with an API 50CH kit, by classical biochemical tests, and via the use of specific primers based on the 16S rRNA gene in PCRs. To compare the isolated strains, phenotypic characteristics were determined and three different probes were used in hybridization experiments: two nif probes and one probe comprising a 0.58-kb fragment cloned from the P. azotofixans C3L4 genome. These results were used to construct a dendrogram, in which two main clusters could be observed. One cluster contained exclusively strains from Várzea soil, and the other contained the majority of strains from Cerrado soil. The 60 strains from Várzea soil and the 46 strains from Cerrado soil were further analyzed with REP and BOX primers, respectively. Based on the patterns obtained, it was possible to identify 21 different groups among strains from Várzea soil and 4 different groups among strains from Cerrado soil. These different patterns were tested by multivariate analysis of variance, and differences in the populations of P. azotofixans during the four stages of plant growth were demonstrated. Moreover, strains isolated from the rhizoplanes, rhizospheres, and nonrhizospheres of maize planted in Cerrado and Várzea soils were shown to be statistically different; the diversity of P. azotofixans strains was affected by the soil type.
In recent years, the interest in soil microorganisms has increased because they play an important role in the maintenance of soil fertility. A major challenge for the development of sustainable agriculture is the use of nitrogen-fixing bacteria which are able to assimilate gaseous N2 from the atmosphere. Many different N2-fixing bacteria, including symbionts, such as root-nodulating Rhizobium spp. (14, 26) and different free-living rhizobacteria, such as Azospirillum spp. (16), Bacillus spp. (8, 9, 36), and Paenibacillus spp. (46), have already been described. When used as seed inoculants, some of these free-living N2-fixing bacteria show beneficial effects on plant growth, and hence they are called plant growth-promoting rhizobacteria (11, 23).
Strains belonging to the species Paenibacillus azotofixans were shown to be efficient nitrogen fixers prevalent in the rhizospheres of maize, sorghum, sugarcane, wheat, banana, and forage grasses (38, 44, 46, 48). Some strains are able to produce antimicrobial substances (50) and solubilize organic phosphates (38). These characteristics can be considered to be very important for the establishment of P. azotofixans in plant rhizospheres.
Growth promotion may occur when a plant is inoculated with a bacterium with which it coexisted previously (7, 8), and thus the diversity among populations of P. azotofixans associated with a variety of different gramineous plants was investigated (38). The results showed that the plants studied did not select a specific phenotypic or genotypic subpopulation of P. azotofixans. However, more data are needed to elucidate the diversity of the populations in individual plant samples during different stages of plant growth and in different soil types.
In this study, we aimed to determine the diversity among populations of P. azotofixans associated with maize by their phenotypic and genetic characteristics. Bacterial populations isolated from three different compartments (rhizosphere, rhizoplane, and nonrhizosphere [i.e., non-root-associated soil]) in two agriculturally important Brazilian soils were compared. Moreover, the influence of plant development on the diversity of the P. azotofixans populations was investigated. These approaches will help to elucidate whether arbitrary diversity exists among strains or whether a plant selects specific bacterial populations to coexist with it.
MATERIALS AND METHODS
Maize genotype, soils, and experimental conditions.
Two soils (Cerrado and Várzea) were planted with maize (Zea mays L. BR-201, single cross-hybrid) in Sete Lagoas, Brazil. Cerrado soil is present over 1.8 million km2, corresponding to about 20% of the territory of Brazil (12); it is a dark-red distrophic latosol with a clayey texture. Várzea soil is another dominant soil in Brazil; it is a low-humus eutrophic gley with a clayey texture. Both soils are representatives of soils commonly cultivated with maize and were supplied with 80 kg of nitrogen per ha 40 days after being sown with nontreated maize seeds.
Strains of P. azotofixans were isolated from the rhizoplanes and the rhizospheres of maize plantlets 10, 30, 60, and 90 days after being sown in these soils, by using a modification of the method described by Baldani and Döbereiner (2). Five plantlets were harvested, and their roots were shaken to remove the loosely attached soil. One gram of the adhering soil, considered the rhizosphere, was mixed with 9 ml of distilled water. Roots were washed 5 to 10 times by gentle shaking with 100 ml of distilled water and weighed, and 1 g was macerated in 9 ml of water. Also, 1 g of non-root-associated soil (from an area 0 to 10 cm in depth) used for the isolation procedure was collected from a maize plant. The different suspensions were pasteurized (10 min, 80°C), and twofold serial dilutions were plated onto thiamine-biotin agar (45) and incubated anaerobically in GasPak jars (80% N2, 10% CO2, 10% H2) for 3 days. Further isolation and purification were performed as described before (45).
Identification and maintenance of P. azotofixans isolates.
All presumptive P. azotofixans strains (white, convex, and mucous colonies isolated in thiamine-biotin solid medium, after incubation under anaerobic conditions) were identified by the biochemical tests proposed by Gordon et al. (19). Cellular morphology, forms and positions of spores, and swelling of the sporangia were observed by microscopy of crystal violet-stained smears. Three other carbohydrates (sorbitol, dulcitol, and starch) were used to group the strains into one of five groups based on fermentation patterns described previously (38, 48). Whenever necessary, the API 50CH microtube system (bioMérieux, Marcy l’Etoile, France) was used in addition to conventional tests. The API test galleries were prepared and read as described in the work of Seldin and Penido (48). In addition to these tests, a molecular method for identification of P. azotofixans was used. It consisted of PCR amplification of part of the variable regions V1 to V4 of the 16S rRNA gene with two primers, which were BAZO1 and BAZO2, followed by hybridization with a specific 18-bp oligonucleotide probe homologous to part of the intervening region. The PCR product generated with P. azotofixans strains was 565 bp long and was specifically detected by a P. azotofixans-specific probe, BAZOP (37). Strains identified as P. azotofixans were stored aerobically at room temperature on GB agar slants (45) supplemented with 1% CaCO3 (wt/vol).
Nitrogen fixation.
Acetylene reduction was measured by assessing the ethylene production of cultures in 18-ml vials as described previously (46).
Solubilization of organic phosphate.
Aliquots (10 μl) of young cultures of P. azotofixans were spot inoculated onto calcium phytate agar plates (38) and incubated for 5 days at 30°C. A positive result was considered the formation of a clear zone around the growth spot.
Detection of inhibitory substances.
The activities of P. azotofixans strains inhibiting a strain of Corynebacterium fimi were assayed by the lawn-spotting technique described by Seldin and Penido (50).
Resistance to heavy metals and to antibiotics.
Resistance to the heavy metals and antibiotics listed here was determined by spot inoculating 18-h-old cultures of P. azotofixans on TBN agar plates (48) supplemented with CuSO4 (250 μg/ml), NiSO4 (263 μg/ml), HgCl2 (20 μg/ml), CoSO4 (240 μg/ml), chloramphenicol (10 μg/ml), erythromycin (10 μg/ml), or rifampin (10 μg/ml).
Preparation of genomic DNAs.
The DNAs were isolated by a modification of the method described by Marmur (27). Cells from 30-ml cultures were concentrated in 2 ml of Tris-EDTA-NaCl buffer (47) and treated with 1 mg of lysozyme per ml for 30 min at 37°C and with sodium dodecyl sulfate (final concentration, 1%). The preparations were incubated for 10 min at 65°C. Purification steps were done as described in the work of Seldin and Dubnau (47). DNAs were quantified spectrophotometrically (GeneQuant apparatus; Pharmacia Biotech, Uppsala, Sweden). Chromosomal DNA was routinely digested with 10 to 20 U of EcoRI per μg of DNA for 20 h at 37°C. Agarose gel electrophoresis of restricted DNA samples was performed with 0. 8% agarose (wt/vol) in Tris-borate-EDTA buffer (42) at 2 V cm−1 for 16 h at room temperature. Bacteriophage lambda DNA digested with HindIII and labeled with digoxigenin (DIG; Boehringer Mannheim Biochemicals [BMB], Mannheim, Germany) was routinely used as the molecular weight marker.
Cloning and nucleotide sequencing of the 0.58-kb fragment from the strain C3L4 genome and cloning of part of the nifH gene of strain P3L5T.
The degenerate primers described by Zehr and McReynolds (57) were used to amplify nifH DNA from P. azotofixans P3L5T for the cloning experiments. Furthermore, a fragment of 0.58 kb from the P. azotofixans C3L4 genome was obtained by PCR with the random primer P5 (38). The PCR products were gel purified with a GeneClean II kit (Bio 101, La Jolla, Calif.) and cloned into the cloning vector pCRII (Invitrogen, Leek, The Netherlands), after which competent Escherichia coli Inv-alfa cells were transformed (TA cloning kit) according to the protocol of the manufacturer (Invitrogen). Plasmids with the 0.58-kb insert, to be used as templates in sequencing reactions, were purified by using Wizard resin spin columns (Promega, Madison, Wis.), and a Thermo Sequenase fluorescently labeled-primer cycle sequencing kit, with 7-deaza-dGTP (Amersham Life Science, Inc., Arlington Heights, Ill.), was used with this DNA. Both strands of the cloned amplification products were sequenced with an automatic sequence analyzer (ALF DNA sequencer; Pharmacia). The DNA sequence data were analyzed by using the sequence analysis software package developed by the University of Wisconsin’s Genetics Computer Group (for its Wisconsin Package) via the CAOS/CAMM Center (University of Nijmegen, Nijmegen, The Netherlands).
Isolation of plasmid DNAs.
E. coli strains harboring different plasmids were maintained at −20°C in Luria-Bertani broth (43) with 20% glycerol. Whenever necessary, this broth was supplemented with 10 μg of ampicillin per ml (i.e., for plasmids pCRII nifH or pCRII 0.58). Plasmids pSA30 (Klebsiella pneumoniae nifKDH gene cluster cloned into pACYC184) (6), pCRII nifH (38), and pCRII 0.58 (this work) were isolated by the alkaline lysis method of Birnboim and Doly (5). The preparations were purified in CsCl-ethidium bromide gradients as described by Sambrook et al. (42).
Preparation of plasmid DNA probes, blotting, and hybridization procedures.
The probes of the three plasmids were random-primer labeled with DIG by using a DIG labeling kit (BMB). Chromosomal DNAs digested with EcoRI were blotted from gels to positively charged nylon membranes (BMB) as described by Sambrook et al. (42). Prehybridization and hybridization conditions with DIG-labeled probes were those described by the BMB manual for the DIG nucleic acid detection kit.
PCR amplifications.
Amplification reactions with primers BOXA1R (25) and REPI and REPII (54) were performed in the following mix: 50 ng of target DNA; 2.5 μl of 10× PCR buffer (20 mM Tris-HCl [pH 8.4], 50 mM KCl); 4 mM MgCl2; 250 μM concentrations of each deoxynucleoside triphosphate; 2 μmol of the primer BOXA1R or REPI-REPII; and 1.25 U of Taq polymerase (Gibco BRL, Gaithersburg, Md.) in a 25-μl final volume. The reaction mixtures were overlaid with 20 μl of mineral oil (Sigma, St. Louis, Mo.). The cycle applied was 1 course of 90 s at 94°C, 3 min at 47°C, and 3 min at 72°C; 35 courses of 30 s at 94°C, 30 s at 53°C (with BOX primers) or 47°C (with REP primers), and 1 min at 72°C; and 10 min at 72°C. With the primers BAZO1 and BAZO2, the PCR mixtures were run for 1 course of 1 5 min at 94°C; 35 courses of 1 min at 94°C, 1 min at 55°C, and 2 min at 72°C; and 1 course of 10 min at 72°C. The mix used contained 20 mM Tris-HCl (pH 8.3), 50 mM KCl, 200 μM concentrations of each deoxynucleoside triphosphate, 3 mM MgCl2, 0.2 μM concentrations of each primer, 100 ng of target DNA, and 1.25 U of Taq polymerase. Agarose gel electrophoresis of PCR products was performed with 1.6% agarose in Tris-borate-EDTA buffer at 10 V/cm for 2 h at room temperature.
Statistical analyses.
The results of phenotypic and hybridization analyses were collected into a matrix indicating the presence or absence (scored as 1 or 0, respectively) of the parameters studied. A dendrogram was obtained on the basis of the data, following cluster analysis with minimum-variance criteria (Ward) and Euclidean distance (35). The data of the matrix were analyzed for principal components with Euclidean distance. The computational calculations of principal components and cluster analysis were done with routines written by A. Nobrega with the software MATHEMATICA (Wolfram Research Inc., Champaign, Ill.). Data from hybridization experiments were clustered according to similarities between individual isolates taken in pairs. Simple matching coefficients were calculated with the NTSYS software package for personal computer (version 1.8; Exeter Software, Setauket, N.Y.). Multivariate analysis of variance (MANOVA) of data from rep-PCR genomic fingerprinting was done with the statistical package NTSYS.
Nucleotide sequence accession number.
The sequence obtained in this study was deposited in the EMBL database with the accession no. AJ005254.
RESULTS
Isolation and identification of P. azotofixans strains.
Sixty strains identified as P. azotofixans were isolated from nonrhizospheres (MSV strains), rhizospheres (PV or LV strains), or rhizoplanes (MRV strains) of maize planted in Várzea soil. Forty-six strains were isolated from nonrhizospheres (CSM strains) or rhizospheres (CRiP or CRiL strains) of maize planted in Cerrado soil. Strains were isolated 10, 30, 60, and 90 days following maize sowing, and they were numbered based on when they were isolated: 10 days (strains with numbers 1 to 49), 30 days (strains with numbers 50 to 99), 60 days (strains with numbers 100 to 149), and 90 days (strains with numbers 150 to 200). A total of 106 strains (Table 1) isolated in this study were analyzed for their phenotypic and genetic characteristics.
TABLE 1.
Designations of P. azotofixans strains isolated from maize plantings at different stages of growth in Cerrado and Várzea soils
| Sampling (no. of days after planting) | Designationsa of strains from Várzea and Cerrado soils isolated from:
|
||
|---|---|---|---|
| Rhizoplanes | Rhizospheres | Nonrhizospheres | |
| S1 (10) | PV9, PV18, PV19, PV21, PV23, LV10, LV12, LV15, LV16, LV17, CRiP9, CRiP15, CRiP21, CRiL33, CRiL35, CRiL36 | MSV5, MSV6, MSV7, MSV15, MSV23, MSV25, MSV32, CSM27 | |
| S2 (30) | PV54, PV55, PV56, PV63, LV53, LV56, CRiL56, CRiP53, CRiP57, CRiP58 | MSV54, MSV55, MSV62, MSV63, MSV66, MSV68, CSM50, CSM53, CSM56, CSM63, CSM64 | |
| S3 (60) | MRV100, MRV105, MRV106, MRV108, MRV111, MRV117 | PV104, LV106, LV107, LV114, LV115, CRiP103, CRiP104, CRiP105, CRiP115, CRiP125, CRiP126, CRiP127, CRiP128, CRiP129, CRiL123, CRiL124, CRiL125 | MSV101, MSV107, MSV108, CSM100, CSM106, CSM107, CSM108, CSM116, CSM121 |
| S4 (90) | MRV153, MRV154, MRV156 | PV150, PV151, PV155, PV156, PV159, PV163, CRiP151, CRiP153, CRiP158, CRiP165, CRiL151 | MSV151, MSV153, MSV155, MSV156, MSV157, MSV158, MSV159, MSV166, CSM152, CSM154, CSM157, CSM159, CSM163, CSM165, CSM169 |
MSV and CSM, strains isolated from Várzea and Cerrado soils, respectively; PV and LV, strains isolated from the rhizospheres of maize planted in Várzea soil. CRiP and CRiL, strains isolated from the rhizospheres of maize planted in Cerrado soil; MRV, strains isolated from roots of maize planted in Várzea soil. Numbers from 1 to 50, strains isolated from maize after 10 days of growth; numbers from 51 to 100, strains isolated from maize after 30 days of growth; numbers from 101 to 150, strains isolated from maize after 60 days of growth; numbers from 151 to 200, strains isolated from maize after 90 days of growth.
All strains were identified as P. azotofixans based on the characteristics previously described for the species (38, 46, 48) and with specific primers and a probe, which were constructed based on three variable regions of the 16S rRNA gene sequence of P. azotofixans (37). The PCR product generated with all target strains was 565 bp long and was specifically detected with the 18-mer probe (data not shown). All strains were facultatively anaerobic, were catalase positive and oxidase negative, were Voges-Proskauer test positive, and did not reduce nitrate, as evidenced by the methods described by Gordon et al. (19). Also, all strains isolated here effectively reduced acetylene to ethylene in assays for acetylene-reducing activity and were able to solubilize calcium phytate (organic phosphate). When P. azotofixans strains were tested by using the API system (API 50CH), they all produced the pattern previously described for the species (38, 48). Abilities to ferment three carbohydrates (starch, sorbitol, and dulcitol) varied among strains, and these results could be used to divide strains into groups, as proposed in the work of Rosado et al. (38). Strains isolated from Cerrado soil could be divided into three groups of related strains (Table 2). The largest group (61%) was formed by strains not able to use any of these three carbohydrates. Strains isolated from Várzea soil formed four different groups of related strains (Table 2). The largest group (70%) was made up of strains able to metabolize starch but not sorbitol or dulcitol. Only one Várzea soil strain (MSV158) showed the same phenotype as the majority of strains from Cerrado soil (inability to metabolize starch, sorbitol, and dulcitol). Also, only one strain (MSV6) was able to metabolize dulcitol.
TABLE 2.
Phenotypic characteristics of strains of P. azotofixans isolated from maize planted in Cerrado and Várzea soils
| Várzea and Cerrado soil strain(s) | Ability to metabolize:
|
Ability to produce antimicrobial substancea | Ability to solubilize phosphateb | ||
|---|---|---|---|---|---|
| Starch | Sorbitol | Dulcitol | |||
| MSV15, MSV32, PV18, PV23, PV54, PV55, PV56, MSV62, MSV101, MSV107, LV114, MRV111, MSV157, MSV159, PV151, CSM50, CSM53, CSM152, CSM165, CSM169 | − | + | − | − | + |
| PV63, CRiP126 | − | + | − | + | + |
| MSV6 | − | − | + | + | + |
| MSV5, MSV7, MSV23, MSV25, PV9, PV19, PV21, LV10, LV12, LV15, LV16, LV17, MSV54, MSV55, MSV63, MSV66, MSV68, LV53, LV56, MSV108, LV106, MRV153, MRV154, MRV156, CRiP21, CRiL56, CSM56, CRiP57, CSM116, CRiL151, CRiP165 | + | − | − | − | + |
| LV107, LV115, PV104, MRV100, MRV105, MRV106, MRV108, MRV117, MSV151, MSV153, MSV155, MSV156, MSV166, PV150, PV155, PV156, PV159, PV163, CSM121, CRiL124, CRiP151, CRiP153, CRiP158 | + | − | − | + | + |
| MSV158, CRiP15, CRiL33, CRiL35, CRiL36, CSM27, CSM63, CRiP53, CRiP58, CSM106, CSM107, CSM108, CRiL125 | − | − | − | + | + |
| CRiP9, CSM64, CSM100, CRiL123, CRiP103, CRiP104, CRiP105, CRiP115, CRiP125, CRiP127, CRiP128, CRiP129, CSM154, CSM157, CSM159, CSM163 | − | − | − | − | + |
Growth inhibition of Corynebacterium fimi.
Calcium phytate.
Strains were also tested for their capacity to produce antimicrobial substances and for their resistance to different antibiotics and heavy metals. Among the Várzea soil isolates, 21 strains (35%) were able to inhibit the growth of the indicator strain, and among the Cerrado soil isolates, 18 strains (39%) showed this characteristic. None of the strains tested was able to grow in media containing antibiotics (chloramphenicol, erythromycin, and rifampin) or heavy metals (CuSO4, NiSO4, HgCl2, and CoSO4).
When the phenotypic characteristics (fermentation of carbohydrates, antimicrobial substance production, and solubilization of phosphate) were considered, six groups of strains could be formed with the Cerrado soil isolates and with the Várzea soil isolates (Table 2). However, if we consider all strains together, seven groups of strains could be formed.
Sequence analysis of a 0.58-kb random fragment from strain C3L4.
A fragment of 0.58 kb was obtained from strain C3L4 by randomly amplified polymorphic DNA-PCR with primer P5 (38), cloned, and sequenced. This fragment was shown to be present in 34% of the P. azotofixans strains tested, and this fragment was proposed for use as a probe to detect this subgroup of strains directly in soil or to identify them in taxonomic studies (38). The DNA sequence of the 0.58-kb fragment of C3L4 was compared with sequences present in the GenBank and EMBL databases, by using FASTA (18) or sequence alignments and by determining percentages of homology. The results obtained showed that the sequence has a low level of similarity to any one of the sequences present in the databases. The best scores occurred with a 518-bp sequence that overlaps a fragment of Bacillus subtilis chromosomal DNA corresponding to a replication origin (63.3% identity) and with an ATPase from B. subtilis Marburg 168.
DNA homology to nifKDH, to part of nifH, and to a fragment of 0.58 kb from P. azotofixans C3L4.
All strains were tested for the presence of homology to the K. pneumoniae nifKDH gene cluster (nitrogenase structural genes cloned in pACYC184-originating plasmid pSA30) and to part of the nifH gene from P. azotofixans P3L5T (cloned in pCRII) by Southern hybridization. Total DNA from all strains (digested with EcoRI) showed the same pattern of homology to plasmid pSA30 as to plasmid pCRII nifH. These hybridizations allowed the separation of strains from Várzea soil into 15 different groups and strains from Cerrado soil into only 2 groups of related strains (Table 3). Among the strains from Várzea soil, 53.3% showed a hybridization pattern similar to that shown by strains from Cerrado soil (Table 3; Fig. 1).
TABLE 3.
Hybridization of DNAs from P. azotofixans strains isolated from maize in different stages of plant growth with K. pneumoniae nifKDH or part of nifH from P. azotofixans P3L5T
| Groupa | Sizes of EcoRI fragments (kb) | SM (%)b
|
||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n1 | n2 | n3 | n4 | n5 | n6 | n7 | n8 | n9 | n10 | n11 | n12 | n13 | n14 | n15 | ||
| n1 | 4.0, 3.4, 2.9 | 100 | ||||||||||||||
| n2 | 4.0, 2.9 | 94.1 | 100 | |||||||||||||
| n3 | 8.4, 4.0 | 82.4 | 88.2 | 100 | ||||||||||||
| n4 | 4.0, 3.4, 1.9 | 88.2 | 82.4 | 82.4 | 100 | |||||||||||
| n5 | 9.6, 4.2, 4.0 | 76.5 | 82.4 | 82.4 | 76.5 | 100 | ||||||||||
| n6 | 9.6, 5.1, 4.0, 1.1 | 70.6 | 76.5 | 76.5 | 70.6 | 82.4 | 100 | |||||||||
| n7 | 10.5, 8.4, 4.0 | 76.5 | 82.4 | 94.1 | 76.5 | 76.5 | 70.6 | 100 | ||||||||
| n8 | 4.0, 2.9, 2.7 | 88.2 | 94.1 | 82.4 | 76.5 | 76.5 | 70.6 | 76.5 | 100 | |||||||
| n9 | 8.9, 7.6, 4.0 | 76.5 | 82.4 | 82.4 | 76.5 | 76.5 | 70.6 | 76.5 | 76.5 | 100 | ||||||
| n10 | 8.9, 4.0, 2.7 | 76.5 | 82.4 | 82.4 | 76.5 | 76.5 | 70.6 | 76.5 | 88.2 | 88.2 | 100 | |||||
| n11 | 7.6, 5.4, 4.6 | 64.7 | 70.6 | 70.6 | 64.7 | 64.7 | 58.8 | 64.7 | 64.7 | 76.5 | 64.7 | 100 | ||||
| n12 | 5.4, 4.6, 2.9 | 76.5 | 82.4 | 70.6 | 64.7 | 64.7 | 58.8 | 64.7 | 76.5 | 64.7 | 64.7 | 88.2 | 100 | |||
| n13 | 8.4, 6.3, 5.1, 2.9 | 70.6 | 76.5 | 76.5 | 58.8 | 58.8 | 64.7 | 70.6 | 70.6 | 58.8 | 58.8 | 58.8 | 70.6 | 100 | ||
| n14 | 10.5, 2.9, 1.4 | 76.5 | 82.4 | 70.6 | 64.7 | 64.7 | 58.8 | 76.5 | 76.5 | 64.7 | 64.7 | 64.7 | 76.5 | 70.6 | 100 | |
| n15 | 10.5, 7.6, 6.3, 5.1 | 58.8 | 64.7 | 64.7 | 58.8 | 58.8 | 64.7 | 70.6 | 58.8 | 70.6 | 58.8 | 70.6 | 58.8 | 76.5 | 70.6 | 100 |
Groups of Várzea and Cerrado soil strains: n1, PV18, PV21, PV23, MSV15, MSV23, PV54, PV55, PV56, MRV111, MSV101, MSV107, LV114, MSV157, MSV159, PV150, PV151, CRiP9, CRiP21, CRiL33, CRiL35, CRiL36, CSM53, CSM56, CSM64, CRiP57, CRiP58, CRiL56, CSM106, CSM107, CSM108, CRiP103, CRiP104, CRiP105, CRiP115, CRiL123, CRiP125, CRiP126, CRiP127, CRiP128, CRiP129, CSM154, CSM157, CSM163; n2, PV63, MSV54, MSV62, MSV66, MRV100, MRV105, MRV106, MRV108, MRV117, LV106, PV104, LV107, LV115, PV155, PV156, PV163, CRiP15, CSM27, CSM50, CSM63, CRiP53, CSM100, CSM116, CSM121, CRiL124, CRiL125, CSM152, CSM159, CSM165, CSM169, CRiP151, CRiP153, CRiP158, CRiP165, CRiL151; n3, PV9, PV19; n4, LV10, LV17; n5, LV15, LV16; n6, MSV25; n7, MSV5, LV53, LV56, MSV108, MSV151; n8, MSV6; n9, MSV7, LV12, MSV158; n10, MSV32; n11, MSV68; n12, MSV55, MSV63; n13, MSV153, MSV155, MSV156, MSV166; n14, PV159, n15, MRV153, MRV154, MRV156.
SM, simple matching coefficients.
FIG. 1.
Southern hybridization of DNAs from P. azotofixans strains isolated from maize planted in Cerrado and Várzea soils with the following probes: nifKDH from K. pneumoniae or part of nifH from P. azotofixans (A) and the 0.58-kb fragment from P. azotofixans C3L4 (B). Only one representative strain of each major group is presented here. Lanes: 1, pattern of 1-kb ladder (Gibco BRL); 2, pattern observed in 58.7% of strains from Cerrado soil and 26.7% of strains from Várzea soil; 3, pattern observed in 41.3% of strains from Cerrado soil and 26.7% of strains from Várzea soil; 4, pattern observed in 100% of strains from Cerrado soil and 58.3% of strains from Várzea soil.
Strains were also tested for the presence of DNA homology to the 0.58-kb fragment from P. azotofixans C3L4 cloned in pCRII and sequenced in this study. Genomic DNAs of all strains digested with EcoRI showed homology to plasmid pCRII 0.58 but not to the cloning vector. Five different patterns of hybridization could be observed in the strains from Várzea soil. One of these groups is made up of 58.3% of the strains. A 5.1-kb EcoRI fragment was detected in all strains isolated from Cerrado soil, and this pattern was the same as that observed for the prevalent group of strains from Várzea soil (Table 4; Fig. 1).
TABLE 4.
Hybridization of DNAs from P. azotofixans strains isolated from maize in different stages of plant growth with a genomic fragment of 0.58 kb from P. azotofixans C3L4 cloned in pCRII
| Groupa | Size(s) of EcoRI fragment(s) | SM (%)b
|
||||
|---|---|---|---|---|---|---|
| h1 | h2 | h3 | h4 | h5 | ||
| h1 | 1.1 | 100 | ||||
| h2 | 5.1 | 60 | 100 | |||
| h3 | 2.9, 1.1 | 80 | 40 | 100 | ||
| h4 | 4.0, 1.1 | 80 | 40 | 60 | 100 | |
| h5 | 3.1 | 60 | 60 | 40 | 40 | 100 |
Groups of Várzea and Cerrado soil strains: h1, PV9, PV19, LV15, LV16, MSV5, MSV7, LV12, LV53, LV56, MSV108, MSV151, MSV153, MSV155, MSV156, MSV158, MSV166, PV163; h2, LV10, LV17, PV18, MSV6, MSV15, MSV23, PV21, PV23, PV54, PV55, PV56, PV63, MSV54, MSV62, MSV66, MRV100, MRV105, MRV106, MRV108, MRV111, MRV117, PV104, MSV101, MSV107, LV106, LV107, LV114, LV115, MSV159, PV150, PV151, PV155, PV156, PV159, MSV157, and all strains from Cerrado soil; h3, MSV25, MSV32, MRV153; h4, MSV55, MSV63, MSV68; h5, MRV154, MRV156.
SM, simple matching coefficients.
Cluster analysis based on phenotypic and hybridization data.
Simple matching coefficients were calculated with hybridization data and are presented in Tables 3 and 4.
Phenotypic and hybridization results were used in a matrix indicating the presence or absence (scored as 1 or 0, respectively) of the parameters studied. The data were clustered according to minimum-variance criteria (Ward) with Euclidean distance. The dendrogram obtained is shown in Fig. 2. Two main clusters, denoted A and B, were observed. Cluster A was made up predominantly of strains from Cerrado soil, while cluster B was made up exclusively of strains from Várzea soil. Cluster A comprised 16 groups harboring 82 strains, while cluster B comprised 13 groups harboring 24 strains.
FIG. 2.
Dendrogram based on data from phenotypic characteristics and from hybridization studies, followed by cluster analysis with minimum-variance criteria (Ward) and Euclidean distance. Strains in each group are as follows: group 1, CRiP57, CRiL56, CRiP21, PV21, MSV23, and CSM56; group 2, PV150; group 3, LV17 and LV10; group 4, CRiL151, CRiP165, MSV66, MSV54, CSM116, and LV106; group 5, CSM159 and CSM100; group 6, CRiL123, CRiP129, CRiP128, CRiP127, CRiP115, CRiP105, CRiP104, CRiP103, CSM64, CRiP9, CSM163, CSM157, and CSM154; group 7, PV63; group 8, CRiP126; group 9, CSM27, CRiP15, CRiP53, CRiL125, and CSM63; group 10, MSV6; group 11, CSM108, CSM107, CSM106, CRiP125, CRiP58, CRiL36, CRiL35, and CRiL33; group 12, LV114, MRV111, MSV159, MSV157, PV151, CSM53, PV56, PV55, PV54, MSV15, PV23, PV18, MSV107, and MSV101; group 13, CSM50, MSV62, CSM169, CSM165, and CSM152; group 14, MRV117, MRV108, MRV106, MRV105, MRV100, CRiL124, LV115, LV107, CSM121, CRiP158, CRiP153, CRiP151, PV156, PV155, and PV104; group 15, PV163; group 16, PV159; group 17, MSV63 and MSV55; group 18, MSV68; group 19, MSV166, MSV156, MSV155, and MSV153; group 20, MSV7 and LV12; group 21, MSV158; group 22, MSV32; group 23, LV16 and LV15; group 24, MSV25; group 25, MSV108, LV56, LV53, and MSV5; group 26, MSV151; group 27, PV19 and PV9; group 28, MRV156 and MRV154; group 29, MRV153.
Data from phenotypic characteristics and hybridization experiments were also subjected to principal-component analysis (Fig. 3). Factors 1 and 2 were found to explain 56% of the total variance. The remaining variance was more evenly distributed among factors, with factors 3 to 9 representing 36% of variance. The division of the data into two main clusters, as shown in Fig. 3, was again present in the factorial plane. The dichotomy was caused mainly by the distribution of the data in the first factor. The contribution of each variable to the first factor is shown in Fig. 3.
FIG. 3.
Principal-component analysis of phenotypic and hybridization data from the 106 P. azotofixans strains (Cerrado and Várzea soils). Factors 1 and 2 reflect 56% of the total variance. Numbers correspond to the groups defined in Fig. 2.
PCR analyses.
Primers corresponding to conserved DNA sequences of rep elements, BOXA1R and REPI-REPII, when used with P. azotofixans genomic DNA, generated different fingerprints. The fingerprints obtained from Várzea soil strains with the BOX primer and those from Cerrado soil strains with the REP primers could not be used in this study because of the excessive number of bands present in the fingerprints. With the optimized parameters for rep-PCR described previously (38), amplification of template DNA of each of the 60 strains from Várzea soil and of the 46 strains from Cerrado soil was achieved with the REP primers and the BOX primer, respectively. Comparison of the sizes of the amplification products with molecular size markers via gel electrophoresis allowed for an estimation of their lengths. Figure 4 shows the different P. azotofixans groups based on these PCR patterns. From the variation in the numbers and sizes of bands, it was possible to identify similarities among strains, resulting in 21 different groups being identified among strains from Várzea soil and in 4 different groups being identified among strains from Cerrado soil.
FIG. 4.
Grouping of P. azotofixans strains from maize planted in Várzea and Cerrado soils by their amplification patterns by PCR with REP and BOX primers, respectively. (A) Várzea soil strains. Strains in each group are as follows: group 1, MSV5, MSV7, LV10, LV15, LV16, LV17, PV9, and PV19; group 2, MSV6; group 3, LV12; group 4, MSV15, MSV23, PV18, PV21, PV23, PV54, PV55, PV56, MRV111, MSV101, MSV107, LV114, MSV157, MSV159, PV151, and PV163; group 5, MSV25; group 6, MSV32; group 7, MSV55 and MSV63; group 8, MSV54 and MSV62; group 9, MRV153; group 10, MRV154 and MRV156; group 11, MSV151; group 12, MSV153, MSV155, and MSV166; group 13, MSV156; group 14, MRV100, MRV105, MRV106, MRV108, MRV117, LV107, LV115, PV104, PV155, and PV156; group 15, PV150; group 16, PV159; group 17, MSV66; group 18, MSV68; group 19, LV106; group 20, LV53, LV56, and MSV108; and group 21, PV63 and MSV158. (B) Cerrado soil strains. Strains in each group are as follows: group 1, CSM50, CRiP53, CSM63, CRiL124, CRiL125, CSM107, and CSM152; group 2, CSM53, CRiP58, CSM100, CSM106, CSM108, CRiP103, CRiP105, CRiP115, CRiP125, CRiP127, CRiP128, CRiP129, and CRiL123; group 3, CRiP9, CRiP15, CRiP21, CRiL33, CRiL35, CRiL36, CSM27, CRiL56, CRiP57, CSM56, CSM64, CRiP104, CRiP126, CSM116, CRiL151, CRiP165, CSM163, and CSM169; and group 4, (not shown) CSM121, CSM157, CSM159, CSM165, CSM154, CRiP151, CRiP153, and CRiP158. The underlined strains correspond to the samples on the gel. Lane numbers correspond to group numbers. Lane a contains a 1-kb ladder (Gibco BRL).
All P. azotofixans strains isolated in this study were classified into groups according to different criteria: (i) P. azotofixans populations isolated from rhizoplanes, rhizospheres, and nonrhizospheres and (ii) P. azotofixans populations obtained from four different samplings carried out 10, 30, 60, and 90 days after planting (sampling 1 [S1], S2, S3, and S4, respectively). Patterns obtained by rep-PCR of the different groups were tested for identity by MANOVA. The resulting scores of comparisons between groups are shown in Table 5. The data showed the following. (i) Strains from Várzea soil isolated from rhizoplanes, rhizospheres, and nonrhizospheres differed significantly (P = 0.006). Two-by-two group tests also were performed, and strains isolated from rhizoplanes and from nonrhizospheres showed the most significant difference (P = 0.019) when compared to strains isolated from rhizoplanes and rhizospheres (P = 0.813) and those isolated from rhizospheres and nonrhizospheres (P = 0.189). (ii) Strains from Cerrado soil isolated from rhizospheres and nonrhizospheres were not significantly different (P = 0.938). (iii) Strains from S1, S2, S3, and S4 isolated from Várzea soil were statistically distinct with high significancy (P < 0.002). (iv) Strains from S1, S2, S3, and S4 isolated from Cerrado soil were also statistically distinct, with high significance (P = 0.007). MANOVA was also done for all the possible combinations of two by two and three by three with S1, S2, S3, and S4 strains. The results showed that, among strains from Várzea soil, all three-by-three groupings that included S1 strains were very significantly different by MANOVA (P < 0.02), suggesting that the S1 group of strains deviates from the other groups. Among strains from Cerrado soil, the tests that included combinations of two and three groups showed that S3 and S4 strains contribute more than S1 and S2 strains to the heterogeneity of the whole population.
TABLE 5.
MANOVA of 106 P. azotofixans strains from Várzea and Cerrado soils with REP and BOX patterns, respectively
| Soil | Canonical variate analysis | Statistica | F statistic | Pb |
|---|---|---|---|---|
| Várzea | Rhizoplane × rhizosphere × nonrhizosphere | 0.22543 | 1.949 | 0.006 |
| Rhizoplane × rhizosphere | 0.50370 | 0.657 | 0.813 | |
| Rhizosphere × nonrhizosphere | 0.49325 | 1.419 | 0.189 | |
| Rhizoplane × nonrhizosphere | 0.13087 | 3.479 | 0.019 | |
| S1 × S2 × S3 × S4 | 0.05254 | 2.893 | <0.002 | |
| S1 × S2 × S3 | 0.10453 | 1.993 | 0.015 | |
| S1 × S2 × S4 | 0.08651 | 2.628 | <0.002 | |
| S1 × S3 × S4 | 0.13132 | 2.095 | 0.006 | |
| S2 × S3 × S4 | 0.19547 | 1.202 | 0.281 | |
| S1 × S2 | 0.26632 | 0.918 | 0.595 | |
| S1 × S3 | 0.28943 | 1.052 | 0.495 | |
| S1 × S4 | 0.26139 | 1.615 | 0.197 | |
| S2 × S3 | 0.36851 | 0.326 | 0.961 | |
| S2 × S4 | 0.31455 | 0.726 | 0.734 | |
| S3 × S4 | 0.46893 | 0.485 | 0.917 | |
| Cerrado | Rhizosphere × nonrhizosphere | 0.92883 | 0.354 | 0.938 |
| S1 × S2 × S3 × S4 | 0.31690 | 2.069 | 0.007 | |
| S1 × S2 × S3 | 0.51488 | 1.181 | 0.317 | |
| S1 × S2 × S4 | 0.38025 | 1.399 | 0.197 | |
| S1 × S3 × S4 | 0.28600 | 2.936 | 0.002 | |
| S2 × S3 × S4 | 0.43962 | 1.842 | 0.047 | |
| S1 × S2 | 0.64647 | 0.479 | 0.838 | |
| S1 × S3 | 0.41665 | 2.800 | 0.038 | |
| S1 × S4 | 0.57577 | 0.921 | 0.537 | |
| S2 × S3 | 0.81776 | 0.501 | 0.839 | |
| S2 × S4 | 0.60317 | 0.822 | 0.601 | |
| S3 × S4 | 0.46381 | 3.035 | 0.020 |
Multivariate test of significance (Wilk’s lambda value).
P, probability of occurrence of extreme variance.
DISCUSSION
The potential for bacterial adaptation to highly heterogeneous and fluctuating environments like the rhizoplane, rhizosphere, or nonrhizosphere depends on their genetic diversity. A combination of phenotypic and genetic measures is known to provide a good estimate of soil bacterial diversity (51); however, studies of dominant diazotrophs or plant growth-promoting rhizobacteria associated with maize roots and rhizospheres have been performed only with isolates of Azospirillum spp. (15, 52, 55, 56), Enterobacter cloacae (34), Azotobacter spp. (22), Bacillus circulans, Klebsiella terrigena, Rahnella aquatilis (4), and Burkholderia cepacia (13). Strains of P. azotofixans in the maize rhizosphere have also been described, but no data were provided on the putative selection of a specific phenotypic or genotypic subpopulation of P. azotofixans in this environment (38). This fact may be explained by the approach used, as the populations in individual plant samples during the different stages of plant growth were not assessed. Therefore, in this study, two dominant Brazilian clay soils were planted with maize and four stages of plant growth were analyzed to characterize the P. azotofixans populations. The analysis performed on the phenotypic features of 60 strains from Várzea soil and 46 strains from Cerrado soil showed that their identification as P. azotofixans was unambiguous and that they could be divided into six groups for each soil (Table 2). The major group of Várzea soil isolates (70% of the strains), and also 12 strains from Cerrado soil, were capable of metabolizing starch. Hence, starch metabolism might play an important role in the establishment of rhizobacteria in the vicinity of maize roots, as proposed by Mavingui et al. (28) for sorbitol metabolism in wheat roots. Other sugars (glucose, galactose, inositol, arabinose, and fucose) were shown to be present in the exudates of several plants (1), but some of them are not metabolized by P. azotofixans strains (48).
The method developed by Pace et al. (32) to determine species diversity with ribosomal DNA was extended by applying it to a functionally important gene, nifH, to determine the importance and diversity of nitrogen-fixing bacteria associated with rice roots (53). DNA probes containing K. pneumoniae nifKDH have been used to identify related sequences from several nitrogen-fixing organisms (31, 33, 41, 49). Different hybridization patterns have been reported for different strains. In this study, 15 different hybridization patterns were observed in 60 P. azotofixans strains from Várzea soil and only 2 patterns were observed in 46 strains from Cerrado soil when the plasmids pSA30 and pCRII nifH were used as probes (Table 3). All strains isolated from Cerrado soil showed a characteristic fragment of 4.0 kb, demonstrating their great homogeneity with respect to nif structural genes. Previous studies have also demonstrated the presence of various hybridization patterns in P. azotofixans strains (38, 49).
Data from phenotypic and hybridization analyses were used to construct a dendrogram. Although this dendrogram does not imply phylogenetic relationships, it is potentially useful for defining whether a population is clonal or not, as suggested by different authors (13, 29). All strains could be distributed in 29 groups. Some groups contained more than one strain, indicating that they are genetically closely related and that they may have originated from a common ancestor. The strains isolated from Cerrado soil were less heterogeneous than the strains from Várzea soil, and a possible reason for this major difference may be the soil type. Although both soils have a clay texture, Cerrado soil is a dark-red latosol distrophic soil and Várzea is a low-humus eutrophic gley. However, it is still unknown whether the difference in the constitutions of the soils used in this study is responsible for the differences demonstrated by the heterogeneity between Cerrado and Várzea soil populations. Berge et al. (4) have also shown that the compositions of the populations of B. circulans varied in the rhizospheres of maize depending on the soil in which the maize was planted.
Total DNA from all strains from Várzea soil was amplified with the primers REPI and REPII, and total DNA from all strains from Cerrado soil was amplified with the primer BOXA1R. The use of rep-PCR has previously been described (3, 24, 25, 40). Only four PCR fingerprints were observed in Cerrado soil isolates with the BOX primer, suggesting a considerable degree of DNA conservation over the genomes of all 46 strains. Figure 4 shows these similar amplification patterns, with consistent bands appearing in the different groups. By PCR with the REP primers, the strains isolated from Várzea soil could be separated into 21 groups.
Data from rep-PCR analyses were used for MANOVA. Our study showed that plant development significantly affected the diversity of P. azotofixans populations associated with maize planted in Várzea and Cerrado soils. Particularly for Várzea soil, it was found that strains isolated during the first stage of maize growth may play an important role in the difference observed in P. azotofixans populations during the maize life cycle. This difference could be explained by the unstable ecosystem of young plants compared to the more stable ecosystem of mature plants. It is well known that production and diffusion of root exudates, which represent nutritional sources for rhizosphere microorganisms, are affected by plant development (10, 39). Hence, it is expected that the microflora changes while the exudation patterns of roots vary as plants grow (20). Di Cello et al. (13) and McArthur et al. (30) working with Burkholderia cepacia isolated from maize also found that the Burkholderia cepacia populations changed while plants grew.
Concerning the populations of P. azotofixans isolated from rhizoplanes, rhizospheres, and nonrhizospheres during the four stages of maize growth, it can be observed that among strains from Várzea soil the most significant difference is observed between strains isolated from rhizoplanes and those isolated from nonrhizospheres. This suggests that a subpopulation of P. azotofixans may have been selected by the plant. Similar results were described by Mavingui et al. (28) for wheat roots with Paenibacillus polymyxa isolates. No significant difference was observed between strains isolated from Cerrado soil rhizospheres and nonrhizospheres. Comparisons among rhizoplanes, rhizospheres, and nonrhizospheres from Cerrado soil could not be done because P. azotofixans strains were not isolated from rhizoplanes. This inability to isolate P. azotofixans might have been due to a problem with sampling the rhizoplanes or to a weak rhizosphere effect, influenced by the soil type. A similar situation was previously reported by Hekman et al. (21) for Burkholderia cepacia.
In this study, differences in the populations of P. azotofixans during four stages of maize growth were demonstrated. Also, it was shown that the degree of diversity of P. azotofixans isolates was affected by the soil type and that populations from rhizoplanes, rhizospheres, and nonrhizospheres of Várzea soil were statistically different. These observations should be taken into account in the selection of P. azotofixans strains for use as maize inoculants. Furthermore, since Garcia de Salomone et al. (17) suggested that the populations of Azospirillum spp. in maize are dependent on the plant genotype, additional data which consider the populations of different cultivars of maize samples, cultivated in Várzea and Cerrado soils, are needed in our work.
ACKNOWLEDGMENTS
This work was supported by FINEP and by the National Research Council of Brazil (CNPq).
Thanks are due to colleagues from EMBRAPA-CNPMS who provided maize samples.
REFERENCES
- 1.Ankenbauer R G, Nester E W. Sugar-mediated induction of Agrobacterium tumefaciens virulence genes: structural specificity and activities of monosaccharides. J Bacteriol. 1990;172:6442–6446. doi: 10.1128/jb.172.11.6442-6446.1990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Baldani V L, Döbereiner J. Host-plant specificity in the infection of cereals with Azospirillum spp. Soil Biol Biochem. 1980;12:433–439. [Google Scholar]
- 3.Bassam B J, Caetano-Anollés G, Gresshoff P M. DNA amplification fingerprinting of bacteria. Appl Microbiol Biotechnol. 1992;38:70–76. doi: 10.1007/BF00169422. [DOI] [PubMed] [Google Scholar]
- 4.Berge O, Heulin T, Achouak W, Richard C, Bally R, Balandreau J. Rahnella aquatilis, a nitrogen-fixing enteric bacterium associated with the rhizosphere of wheat and maize. Can J Microbiol. 1991;37:195–203. [Google Scholar]
- 5.Birnboim H C, Doly J. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 1979;7:1513–1523. doi: 10.1093/nar/7.6.1513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cannon F C, Riedel G E, Ausubel F M. Overlapping sequences of Klebsiella pneumoniae nif DNA cloned and characterized. Mol Gen Genet. 1979;174:59–66. doi: 10.1007/BF00433306. [DOI] [PubMed] [Google Scholar]
- 7.Chanway C P, Nelson L M, Holl F B. Cultivar-specific growth promotion of spring wheat (Triticum aestivum L.) by coexistent Bacillus species. Can J Microbiol. 1988;34:925–929. [Google Scholar]
- 8.Chanway C P, Holl F B, Turkington R. Genotypic coadaptation in plant growth promotion of forage species by Bacillus polymyxa. Plant Soil. 1988;106:281–284. [Google Scholar]
- 9.Chanway C P, Nelson L M. Field and laboratory studies of Triticum aestivum L. inoculated with co-existent growth-promoting Bacillus strains. Soil Biol Biochem. 1990;22:789–795. [Google Scholar]
- 10.Chiarini L, Bevivino A, Tabacchioni S. Proceedings of the Third International Workshop on Plant Growth-Promoting Rhizobacteria. Adelaide, Australia: CSIRO Australia; 1994. Factors affecting the competitive ability in rhizosphere colonization of plant-growth promoting strains of Burkholderia cepacia; pp. 204–206. [Google Scholar]
- 11.Davison J. Plant beneficial bacteria. Biotechnology. 1988;6:282–286. [Google Scholar]
- 12.Dianese J C, Santos L T P, Furlanetto C. New Meliola species on Salacia crassifolia. Fitopatol Bras. 1994;19:314. [Google Scholar]
- 13.Di Cello F, Bevivino A, Chiarini L, Fani R, Paffetti D, Tabacchioni S, Dalmastri C. Biodiversity of a Burkholderia cepacia population isolated from the maize rhizosphere at different plant growth stages. Appl Environ Microbiol. 1997;63:4485–4493. doi: 10.1128/aem.63.11.4485-4493.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Djordjevic M A, Gabriel D W, Rolfe B G. Rhizobium—the refined parasite of legumes. Annu Rev Phytopathol. 1987;25:145–168. [Google Scholar]
- 15.Döbereiner J, Marriel I E, Nery M. Ecological distribution of Spirillum lipoferum Beijerinck. Can J Microbiol. 1976;22:1464–1473. doi: 10.1139/m76-217. [DOI] [PubMed] [Google Scholar]
- 16.Fages J. Azospirillum inoculants and field experiments. In: Okon Y, editor. Azospirillum plant associations. Boca Raton, Fla: CRC Press; 1994. pp. 87–109. [Google Scholar]
- 17.Garcia de Salomone I E, Döbereiner J, Urquiaga S, Boddey R M. Biological nitrogen fixation in Azospirillum strain-maize genotype associations as evaluated by the 15N isotope dilution technique. Biol Fertil Soils. 1996;23:249–256. [Google Scholar]
- 18.Genetics Computer Group. Program manual for the Wisconsin Package. Madison, Wis: Genetics Computer Group; 1994. [Google Scholar]
- 19.Gordon R E, Haynes W C, Pang H-N. The genus Bacillus. Agriculture handbook 427. Washington, D.C: Agricultural Research Service, U.S. Department of Agriculture; 1973. [Google Scholar]
- 20.Hamlen R A, Lukezic F L, Bloom J R. Influence of age and stage of development on the neutral carbohydrate components in root exudates from alfalfa plants grown in a gnotobiotic environment. Can J Plant Sci. 1972;52:633–642. [Google Scholar]
- 21.Hekman W E, Heijnen C E, Burgers S L G E, van Veen J A, van Elsas J D. Transport of bacterial inoculants through intact cores of two different soils as affected by water percolation and the presence of wheat plants. FEMS Microbiol Ecol. 1995;16:143–158. [Google Scholar]
- 22.Hussain A, Arshad M, Hussain A, Hussain F. Response of maize (Zea mays) to Azotobacter inoculation under fertilized and unfertilized conditions. Biol Fertil Soils. 1987;4:73–77. [Google Scholar]
- 23.Kloepper J, Lifshitz R, Zablotowicz R M. Free-living bacterial inocula for enhancing crop productivity. Trends Biotechnol. 1989;7:39–44. [Google Scholar]
- 24.Laguerre G, Mavingui P, Allard M-R, Charnay M-P, Louvrier P, Mazurier S-I, Rigottier-Goit L, Amarger N. Typing of rhizobia by PCR DNA fingerprinting and PCR restriction fragment length polymorphism analysis of chromosomal and symbiotic gene regions: application to Rhizobium leguminosarum and its different biovars. Appl Environ Microbiol. 1996;62:2029–2036. doi: 10.1128/aem.62.6.2029-2036.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Louws F J, Fulbright D W, Stephens C T, de Bruijn F J. Specific genomic fingerprints of phytopathogenic Xanthomonas and Pseudomonas pathovars and strains generated with repetitive sequences and PCR. Appl Environ Microbiol. 1994;60:2286–2295. doi: 10.1128/aem.60.7.2286-2295.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lynch J M. Beneficial interactions between microorganisms and roots. Biotechnol Adv. 1990;8:335–346. doi: 10.1016/0734-9750(90)91069-s. [DOI] [PubMed] [Google Scholar]
- 27.Marmur J. A procedure for the isolation of deoxyribonucleic acid from microorganisms. J Mol Biol. 1961;3:205–218. [Google Scholar]
- 28.Mavingui P, Laguerre G, Berge O, Heulin T. Genetic and phenotypic diversity of Bacillus polymyxa in soil and in the wheat rhizosphere. Appl Environ Microbiol. 1992;58:1894–1903. doi: 10.1128/aem.58.6.1894-1903.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Maynard Smith J, Smith N H, O’Rourke M, Spratt B. How clonal are bacteria? Proc Natl Acad Sci USA. 1993;90:4384–4388. doi: 10.1073/pnas.90.10.4384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.McArthur J V, Kovacic D A, Smith M H. Genetic diversity in natural populations of a soil bacterium across a landscape gradient. Proc Natl Acad Sci USA. 1988;85:9621–9624. doi: 10.1073/pnas.85.24.9621. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Oliveira S S, Seldin L, Bastos M C F. Identification of structural nitrogen-fixation genes in Bacillus polymyxa and Bacillus macerans. World J Microbiol Biotechnol. 1993;9:387–389. doi: 10.1007/BF00383088. [DOI] [PubMed] [Google Scholar]
- 32.Pace N R, Stahl D A, Lane D J, Olsen G J. The analysis of natural microbial populations by ribosomal RNA sequences. Adv Microb Ecol. 1986;9:1–55. [Google Scholar]
- 33.Possot O, Henry M, Sibold L. Distribution of DNA sequences homologous to nifH among archaebacteria. FEMS Microbiol Lett. 1986;34:173–177. [Google Scholar]
- 34.Raju P N, Evans H J, Seidler R J. An asymbiotic bacterium from the root environment of corn. Proc Natl Acad Sci USA. 1972;69:3474–3478. doi: 10.1073/pnas.69.11.3474. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Rencher A C. Methods of multivariate analysis. New York, N.Y: Wiley Interscience; 1995. Principal component analysis; pp. 415–438. [Google Scholar]
- 36.Rennie R J, Larson R I. Dinitrogen fixation associated with disomic substitution lines of spring wheat. Can J Bot. 1979;57:2771–2775. [Google Scholar]
- 37.Rosado A S, Seldin L, Wolters A C, van Elsas J D. Quantitative 16S rDNA-targeted polymerase chain reaction and oligonucleotide hybridization for the detection of Paenibacillus azotofixans in soil and the wheat rhizosphere. FEMS Microbiol Ecol. 1996;19:153–164. [Google Scholar]
- 38.Rosado A S, de Azevedo F S, da Cruz D W, van Elsas J D, Seldin L. Phenotypic and genetic diversity of Paenibacillus azotofixans strains isolated from the rhizoplane or rhizosphere soil of different grasses. J Appl Microbiol. 1998;84:216–226. doi: 10.1128/aem.64.10.3860-3868.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Rovira A D. Plant root exudates and their influence upon soil microorganisms. In: Baker K F, Snyder W C, editors. Ecology of soil-borne plant pathogens—prelude to biological control. Berkeley: University of California Press; 1965. p. 170. [Google Scholar]
- 40.Rowadi G, Lemercier B, Roulland-Dussoix D. Application of an arbitrarily-primed polymerase chain reaction to mycoplasma identification and typing within the Mycoplasma mycoides cluster. J Appl Bacteriol. 1995;78:586–592. doi: 10.1111/j.1365-2672.1995.tb03103.x. [DOI] [PubMed] [Google Scholar]
- 41.Ruvkun G B, Ausubel F M. Interspecies homology of nitrogenase genes. Proc Natl Acad Sci USA. 1980;77:191–195. doi: 10.1073/pnas.77.1.191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Sambrook J, Fritsch E F, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1989. [Google Scholar]
- 43.Schleif R, Wensink P C. Practical methods in molecular biology. New York, N.Y: Springer-Verlag; 1981. [Google Scholar]
- 44.Seldin L. Primary characterization of the bacteriophage BA-4 from a nitrogen-fixing Bacillus azotofixans strain. Microbios. 1992;71:167–177. [Google Scholar]
- 45.Seldin L, van Elsas J D, Penido E G C. Bacillus nitrogen fixers from brazilian soils. Plant Soil. 1983;70:243–255. [Google Scholar]
- 46.Seldin L, van Elsas J D, Penido E G C. Bacillus azotofixans sp. nov., a nitrogen-fixing species from brazilian soils and grass roots. Int J Syst Bacteriol. 1984;34:451–456. [Google Scholar]
- 47.Seldin L, Dubnau D. DNA homology among Bacillus polymyxa, Bacillus macerans, Bacillus azotofixans, and other nitrogen-fixing Bacillus strains. Int J Syst Bacteriol. 1985;35:151–154. [Google Scholar]
- 48.Seldin L, Penido E G C. Identification of Bacillus azotofixans using API tests. Antonie Leeuwenhoek. 1986;52:403–409. doi: 10.1007/BF00393468. [DOI] [PubMed] [Google Scholar]
- 49.Seldin L, Bastos M C F, Penido E G C. Identification of Bacillus azotofixans nitrogen fixation genes using heterologous nif probes. In: Skinner F A, Boddey R M, Fendrik I, editors. Nitrogen fixation with non-legumes. Dordrecht, The Netherlands: Kluwer Academic Publishers; 1989. pp. 179–187. [Google Scholar]
- 50.Seldin L, Penido E G C. Production of a bacteriophage, a phage tail-like bacteriocin and an antibiotic by Bacillus azotofixans. An Acad Bras Cienc. 1990;62:85–94. [PubMed] [Google Scholar]
- 51.Torsvik V, Salte K, Sorheim R, Goksoyr J. Comparison of phenotypic diversity and DNA heterogeneity in a population of soil bacteria. Appl Environ Microbiol. 1990;56:776–781. doi: 10.1128/aem.56.3.776-781.1990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Tyler M E, Milam J R, Smith R L, Schank S C, Zuberer D A. Isolation of Azospirillum from diverse geographic regions. Can J Microbiol. 1979;25:693–697. doi: 10.1139/m79-100. [DOI] [PubMed] [Google Scholar]
- 53.Ueda T, Suga Y, Yashiro N, Matsuguchi T. Remarkable N2-fixing bacterial diversity detected in rice roots by molecular evolutionary analysis of nifH gene sequences. J Bacteriol. 1995;177:1414–1417. doi: 10.1128/jb.177.5.1414-1417.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Versalovic J, Schneider M, de Bruijn F J, Lupski J R. Genomic fingerprinting of bacteria using repetitive sequence-based polymerase chain reaction. Methods Mol Cell Biol. 1994;5:25–40. [Google Scholar]
- 55.Vlassak K, Reynders L. Association of free-living nitrogen-fixing bacteria with plant roots in temperate regions. In: Loutit M W, Milles J A R, editors. Microbial ecology. Berlin, Germany: Springer; 1978. pp. 307–309. [Google Scholar]
- 56.Von Bulow J F W, Dobereiner J. Potential for nitrogen-fixation in maize genotypes in Brazil. Proc Natl Acad Sci USA. 1975;72:2389–2393. doi: 10.1073/pnas.72.6.2389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Zehr J, McReynolds L. Use of degenerate oligonucleotide primers for amplification of the nifH gene from marine cyanobacterium Trichodesmium thiebautii. Appl Environ Microbiol. 1989;55:2522–2526. doi: 10.1128/aem.55.10.2522-2526.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]