INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Frankia strains are nitrogen-fixing actinomycetes that form symbiotic relationships with 24 genera of woody dicotyledonous plants. The focus of this paper is the Ceanothus-Frankia symbiosis. Although this symbiotic relationship is of ecological and practical importance, very little is known about Ceanothus-infective Frankia strains. The main reason for this is the fact that no isolates that can reinfect the host after isolation have been recovered from Ceanothus spp. Consequently, we studied the diversity of the Ceanothus-infective Frankia strains by using molecular techniques. Prior studies in which the PCR and restriction fragment length polymorphism (RFLP) have been used have been useful in discriminating among closely related Frankia strains obtained from members of other host genera (8, 13, 15, 17). In this study, we used PCR-RFLP techniques to characterize the Frankia strains that infect nine Ceanothus species.

The Ceanothus species studied represent the taxonomic diversity and geographic distribution of the genus (4, 9). The genus Ceanothus, which is indigenous only to North America, contains 55 species. Two subgenera of the genus, subgenera Ceanothus and Cerastes, are recognized. The subgenus Ceanothus, the more ancient subgenus, includes evergreen species, such as C. cordulatus, C. velutinus, and C. thyrsiflorus, and deciduous species, such as C. integerrimus, C. sanguineus, and C. americanus. The subgenus Cerastes includes the evergreen species C. cuneatus, C. prostratus, and C. pumilus. Because we included the diversity of the host plants in our study, we expected that the breadth of the diversity of the Frankia strains that infect Ceanothus spp. would be represented.

In less comprehensive surveys, other workers have found a considerable amount of diversity in the Frankia-Ceanothus symbiosis (2, 14). Baker and Mullin (2) assessed the Frankia strains inhabiting C. americanus nodules at seven different locations in a 70-mile radius in eastern Tennessee by using restriction endonuclease digestion of genomic DNAs in combination with genetic probes. A total of 25 to 30 nodules were collected per site. These authors found differences in RFLP patterns between sites and among different plants collected from a single site. It is unclear whether they found more than one pattern per nodule.

Murry et al. (14) used repetitive extragenic palindromic (REP) PCR methods to characterize the diversity of the Frankia strains associated with six Ceanothus species located in the coastal region of southern California. The study area was an area that had a 10-mile radius and contained seven sites. These authors sampled two to five mature nodule clusters per species. In addition, they compared nodules collected from mature and immature plants. The results varied. Murry et al. found variation within and between sites for some species and found no variation within or between collection sites for other species. Also, they found more than one pattern per nodule cluster. No evidence of contaminating DNA was found, however. The conclusion of Murry et al. was that we need more collections of Ceanothus species to further elucidate the diversity of the Frankia strains that inhabit these plants.

Our study was more extensive than previous studies since we included a greater number of sites, host species, and environmental conditions. Therefore, we predicted that we would find even higher levels of diversity than were found previously.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Sample collection. Nodule, soil, and shoot samples were collected from each of the nine Ceanothus species examined (Table 1). Samples were obtained from 19 sites in Oregon (one site for each of the seven Oregon species and 12 sites for C. velutinus, which is more widespread than the other species assayed) and from one site in Tennessee. At least two plants were collected per site (Table 1). The samples consisted of at least five distinct nodule clusters per site (with the exception of the C. americanus samples [Table 1]). Each collection was made within the geographic range of the host plant species. All samples were transported to the laboratory on ice. Upon arrival, nodule samples were rinsed in deionized water, washed with Tween 80, surface sterilized with ethanol, and stored at 20°C until they were used.

DNA extraction. Genomic DNA was prepared by using a protocol adapted from the protocol of Baker and Mullin (2). Individual lobe tips were excised with a scalpel and frozen in a dry ice-ethanol bath (70°C). DNA was extracted from a minimum of 15 lobes per species (at least three replicates per nodule cluster). No composites were prepared to increase the chance of capturing the total genetic diversity of the Frankia strains that infect individual plants. Individual lobe tips were homogenized with a mortar and pestle by using 600 µl of cetyltrimethylammonium bromide extraction buffer (2% cetyltrimethylammonium bromide, 100 mM Tris [pH 8.0], 20 mM EDTA, 1.4 M NaCl) and were incubated at 65°C for 30 min. The DNA was extracted twice with an equal volume of chloroform-isoamyl alcohol (24:1) and was precipitated with 1 volume of ice-cold isopropanol. The precipitated DNA was resuspended in 50 µl of TE (10 mM Tris [pH 8.0], 0.1 mM EDTA) (10:0.1), reprecipitated by adding 0.25 volume of 10 M ammonium acetate and 1 volume of ice-cold isopropanol, and washed with 70% ethanol. DNA pellets were resuspended in 50 µl of TE (10:0.1). Samples were further purified by adding 1 volume of a polyethylene glycol-NaCl mixture (20% polyethylene glycol 8000, 2.5 M NaCl). Samples were kept at 37°C for 15 min and centrifuged. The DNA pellets were washed twice with 80% ethanol, dried, resuspended in 50 µl of TE (10:0.1), and stored at 20°C until they were used.

PCR amplification. Purified genomic DNA (5 to 10 ng) was used to amplify a 2,098-bp region of the 16S and 23S rRNA genes by the PCR. The amplified portion included the 3' end of the 16S rRNA gene, the intergenic spacer (IGS), and a large portion of the 23S rRNA gene. All reactions were performed in 50-µl (final volume) reaction mixtures containing 1.5 µl of template DNA, 5 µl of GeneAmp 10× PCR buffer II (Perkin-Elmer Corp., Branchburg, N.J.), 1.5 mM MgCl₂, each deoxynucleoside triphosphate (Perkin-Elmer) at a concentration of 0.2 mM, 0.2 µM forward primer 1649F (5'-GATTGGGACGAAGTCGT-3') (20), 0.2 µM reverse primer 2309R (5'-ATCGCATGCCTACTACC-3') (6), and 2 U of AmpliTaq DNA polymerase (Perkin-Elmer). An initial denaturation at 95°C for 2 min was followed by 35 cycles consisting of denaturation at 95°C for 45 s, annealing at 53°C for 45 s, and extension at 72°C for 1.5 min (Coy Thermocycler, Ann Arbor, Mich.). There was a final extension step consisting of 72°C for 5 min. The PCR products were loaded onto a 1% agarose gel (Sea Kem; FMC BioProducts, Rockland, Maine) and electrophoresed. After electrophoresis, bands were excised, purified with GeneClean II (BIO 101, Vista, Calif.), resuspended in 18 µl of TE (10:0.1), and stored at 20°C until they were used.

RFLP analysis. Purified PCR products were digested with restriction endonucleases for at least 4 h by using the suppliers' instructions. The following 12 restriction enzymes were used: HaeIII, HhaI, HinFI, TaqI, RsaI, NciI, NdeII, MvnI, AluI, MspI, BstUI, and Sau96I. The digested products were electrophoresed on chilled 3 to 5% Metaphor gels (FMC BioProducts). Gels containing higher percentages of Metaphor were used for smaller fragments. Each digestion was repeated at least twice to verify the patterns obtained.

Phylogenetic analysis. The sizes and numbers of bands were determined manually, and the results were used to calculate similarity indices for each strain. Relationships were inferred from restriction patterns by using both parsimony (PAUP) (18) and distance-based (NT-SYS) methods.

Sequencing. Amplification reactions were performed by using the primers used for the RFLP analysis. The PCR products were ethanol precipitated, resuspended in deionized water, and subjected to double-stranded sequencing. Primers 1649F and 23S12R (5'-TCCACCGTGTGCCCTTA-3'; synthesized for this study) were used to determine the sequence of the IGS region. Taq dye terminator chemistry was used to determine sequences with an ABI cycle sequencer (Center for Gene Research and Biotechnology, Central Services Laboratory, Oregon State University, Corvallis).

Sequence data analysis. All of the sequences obtained were compared with sequences in the GenBank database by using BLAST (1). The sequences were aligned by using the ASSEMBLE CONTIGS option of the Genetic Data Environment (version 2.2) sequence analysis software (provided by Steven Smith, Millipore Corp., Marlborough, Mass.). The alignment was optimized by manual adjustment.

Nucleotide sequence accession numbers. The nucleotide sequences determined in this study have been deposited in the GenBank database under accession no. AF050760 to AF050768.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

The data suggested that there were some geographic and taxonomic relationships between Ceanothus spp. and the microsymbiont genus Frankia. These relationships were not as clear as we anticipated, however.

PCR-RFLP analysis. PCR amplification was successful with nodular microsymbionts of the nine Ceanothus species assayed. Only one 2,098-bp band was present after most amplification reactions. Twelve different restriction enzymes were utilized. HaeIII digestion resolved four distinct RFLP groups (Fig. 1). Digestion with HhaI, HinFI, TaqI, RsaI, NciI, NdeII, MvnI, AluI, MspI, BstUI, and Sau96I separated the microsymbionts into only two RFLP groups (data not shown). Only one RFLP group was identified for all of the individual lobe tips from each plant analyzed. Actually, except for sites A, D, and L, all of the lobe tips from a given site assayed produced the same RFLP pattern (Fig. 2).

View larger version (109K): [in this window] [in a new window]   FIG. 1.   PCR fingerprints generated after HaeIII digestion of amplified sequences from uncultured microsymbionts of Ceanothus spp. Lanes 1 and 13, 100-bp ladder; lanes 3 to 12, symbiotic Frankia strains obtained from C. velutinus (site J), C. velutinus (site E), C. thyrsiflorus, C. sanguineus, C. pumilus, C. prostratus, C. integerrimus, C. cuneatus, C. cordulatus, and C. americanus nodule DNA, respectively. Lane 2 contained the marker pBR322:MspI. There are four different band patterns. The pattern in lane 12 is the RFLP group I pattern; the pattern in lane 10 is the RFLP group II pattern; the pattern in lanes 3 and 5 is the RFLP group III pattern; and the pattern in lanes 4, 6 to 9, and 11 is the RFLP group IV pattern. All band patterns were verified by repeating each digestion experiment a minimum of five times.

View larger version (13K): [in this window] [in a new window]   FIG. 2.   Map of Ceanothus sites as identified by RFLP group. RFLP group I Frankia strains were collected in Tennessee (site 1). Site 6 is the only place where RFLP group II Frankia strains were collected. The shaded region in the Willamette Valley of Oregon (sites AII, DI, I, J, K, LII, and 7) contained Ceanothus species exhibiting the RFLP group III Frankia pattern. RFLP group IV Frankia strains were collected from Ceanothus species growing in the unshaded region of Oregon.

View larger version (19K): [in this window] [in a new window]   FIG. 3.   Dendrogram (as determined by the NT-SYS method) showing the relationships of amplified 16S ribosomal DNA-23S ribosomal DNA sequences based on the results of a PCR-RFLP analysis performed with seven restriction endonucleases (HaeIII, HinFI, TaqI, NciI, NdeII, AluI, and Sau96I). Unless otherwise noted, the Ceanothus species are members of the subgenus Ceanothus.

View larger version (19K): [in this window] [in a new window]   FIG. 4.   Relationship between elevations of collection sites and RFLP groups. Data were compiled for each of the 19 sites (RFLP group II, 1 site; RFLP group III, 7 sites; RFLP group IV, 14 sites); 3 sites had representatives belonging to two groups. The data for samples from each RFLP group were averaged, and standard errors were calculated.

Sequencing. Because RFLP group IV contains Frankia strains obtained from six of the nine Ceanothus species sampled, we sequenced the IGS between the 16S rRNA and 23S rRNA genes to confirm the homogeneity of this group. Almost all of the members of RFLP group IV had identical sequences; the only exceptions were the C. pumilus-infective Frankia strains. The C. pumilus-infective Frankia strains differed from the rest of the strains in the group at 2 of 450 bases (level of similarity, 99.6%). The exception of the C. pumilus-infective Frankia strains is not surprising since C. pumilus grows only in a chemically unique environment, serpentine soils.