Frankia from root nodules of nine different species of Ceanothus were characterized. DNA was amplified directly from nodular material using the polymerase chain reaction (PCR). The amplified region includes the 3' end of the 16S rRNA gene, the intergenic spacer (IGS), and a large portion of the 23S rRNA gene. Restriction enzyme digestions of PCR products allowed us to designate PCR-RFLP groups among the Ceanothus-infective Frankia tested. The groupings did not follow the taxonomic lines of the Ceanothus host species. Instead, the Frankia strains present followed a geographical pattern. This information was used to choose representative Ceanothus-infective Frankia for phylogenetic analysis. Full-length 16S rDNA sequences were amplified directly from the nodules of two Ceanothus species using the PCR. Sequences were determined using an automated sequencer, compared against GenBank, and assembled into consensus sequences. The sequences were aligned with other full-length Frankia 16S rDNA sequences available from the database. Phylogenetic trees were obtained from three different algorithms: neighbor joining, parsimony, and the maximum-likelihood method. These Ceanothus microsymbionts appear to be most closely related to the microsymbiont associated with Dryas drummondii using all three methods. / Graduation date: 1998

The evolutionary relationship between Frankia and actinorhizal plants was evaluated by reconstructing molecular phylogenetic trees from nifH, 16S rDNA, and rbcL nucleotide sequences. Subgroupings in Frankia phylogenetic trees reconstructed from nifH and from 16S rDNA sequences were consistent in terms of plant origins of Frankia strains. Although the branching order of Frankia 16S rDNA and plant rbcL trees were different, subgroupings of Frankia and of plants correlated well in terms of symbiotic partnership. Tree matching, estimated divergence times, and molecular clock hypothesis tests indicated that Frankia clades diverged more recently than plant clades and that actinorhizal symbioses originated more than three times after the plant clades diverged. A phylogenetic tree of Ceanothus species, which are symbiotic partners of Frankia, was reconstructed using ndhF gene sequences. The analysis identified two main clades corresponding to two subgenera: Ceanothus and Cerastes. The analysis also suggested that three monophyletic clades within the subgenus ceanothus can be delimited on the basis of vegetative characters. Based on rbcL sequences, the two subgenera diverged 18-39 million years ago whereas species within each subgenus diverged more recently. These results support the current division of Ceanothus into two monophyletic subgenera and agree with the postulated recent divergence of many species within each subgenus. Specificity between Ceanothus species and their Frankia microsymbionts was evaluated by analysis of DNA in nodules collected from three copopulations of Ceanothus species. Sequencing of the intergenic spacer region between 16S and 23S rRNA genes suggested that Ceanothus-microsymbiont Frankia are closely related. Nodules were further analyzed by genomic fingerprinting using repetitive sequences and PCR (rep-PCR). A newly designed, direct repeat sequence and a BOX sequence showed that Ceanothus-microsymbiont Frankia exhibited less diversity within each copopulation than among copopulations. Furthermore, geographic separation was a more important factor for divergence of Ceanothus-microsymbiont Frankia than host plant. The population of Ceanothus-infective Frankia in soils under stands of Ceanothus velutinus and Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), a non-host plant, were compared. The population sizes were measured using plant bioassay methods with C. velutinus, C. sanguineus, and C. integerrimus as trap plants. Population size in soil under C. velutinus was about 10 times higher than that under the Douglas-fir. Nodulation capacities of the three trap plants were not significantly different. The diversity of Frankia nodulating trap plants was examined using rep-PCR. Results suggested that infective Frankia is not species-specific with regard to the three Ceanothus species used as trap plants and that although the degree of diversity was similar in both soils, the two populations consisted of different Frankia. / Graduation date: 1998