Molecular Systematics of Glycine

Glycine clandestina

The subgenus Glycine includes at least 21 named perennial species and several undescribed taxa. Australia and nearby islands (Papua New Guinea) is home to the bulk of the diversity in these perennial species. In addition a few polyploid lineages of the species G. tomentella and G. tabacina have spread further afield to islands in the western Pacific Islands (eg Fiji, Tonga, Vanuatu, New Caledonia, Philippines, Taiwan, Japan). The subgenus is of worldwide importance because of its affinity with soybean (Glycine max). Our prior research on wild Glycine includes studies on geographic distribution, interspecific hybridization, cytogenetics; isozyme polymorphism; rust resistance; photoperiod responses; population genetics and breeding systems; broadly aimed at improving strategies for conserving and using this biodiversity.

More recent research has been on molecular phylogenetic analysis and polyploid evolution in various groups of the species to aid a better understanding of the indigenous perennials as a genetic resource. One problem concerns the diploid species that belong to the B-genome group. These include four of the named species G. tabacina, G. latifolia, G. microphylla and G. stenophita. As a group, the B-genome diploids stretch from Tasmania to north Queensland along the east coast, and inland across the dividing ranges and slopes down to about the 500 mm annual rainfall isohyet. All species except G. stenophita possess adventitious roots.

In addition, the B-genome diploid species have related chloroplast genomes that are distinct from the other perennial diploid species. Collaborative phylogenetic research (with Professor J J Doyle, Cornell University) on this cpDNA variation within the B-genome group showed only limited congruence with the taxonomy of the group based on morphology. A comparable analysis of sequences at the single-copy nuclear gene, histone H3-D, which is a member of the histone H3 multigene family, in this group of species affords a fresh look at this problem. The phylogenetic analysis of H3-D sequences is congruent with the taxonomy and in particular highlights that G. stenophita is basal to the whole B-diploid group.

Polyploid G. tabacina are of two types as their genome formulae, AAB'B' and BBB'B' serve to indicate. Morphologically the allotetraploid types are distinct, being either non-stoloniferous or stoloniferous respectively. While their diploid relatives are essentially confined to Australia, both kinds of tetraploid G. tabacina occur outside Australia. A survey of RFLP variation in cpDNA for the stoloniferous BBB'B' group pointed to a multiple origin of these polyploids - 6 of 9 plastome types were identical to different diploid plastome types and implicated three different diploid B species as donors of the B genome. Migration from the continent north and east across the Pacific has apparently happened several times.

Recent DNA sequence information for the single-copy nuclear histone H3D locus has confirmed and extended earlier evidence from hybrid cytogenetic analyses, isozyme polymorphisms, 5S-RNA, chloroplast RFLPs and 18S-26S ribosomal gene (nrDNA) ITS sequences concerning the multiple origins of the G. tabacina complex. The two distinct reproductively isolated allotetraploid races share H3D B'-alleles that are most closely related to those of G. stenophita. The source(s) of the A-allele in the first race are closest to one diploid cytotype of G. tomentella. The sources of the three H3D B-alleles found in the other tetraploid races are two distinct B-species (G. tabacina and G. latifolia). This H3D sequence identity argues that for these lineages, both polyploidy and its subsequent spread to the western Pacific have happened within 30,000 years. The diploid B-species donors are interfertile, and their BBB'B' G. tabacina tetraploid derivatives have recombined to generate a diverse complex.

This situation differs from the G. tomentella polyploid complex. Diploid G. tomentella is polymorphic and polyphyletic with several clades well defined in gene trees using ITS and histone H3-D sequences. Artificial hybrids between these clades are sterile. Polyploid races of G. tomentella combine genomes of the diploid races and some other Glycine taxa, where each distinct combination is reproductively isolated. Some races have evidence of more than one origin, providing scope for lineage recombination. New H3D data have specified and revealed new combinations. H3-D sequence identity or close relationship between the alleles in tetraploid G. tomentella and their putative donors is the rule, suggesting recency of origin and spread. One further twist is that one race in each of the tetraploid species complexes (G. tabacina race AAB'B', and G. tomentella race T2) share the same putative donor (G. tomentella D4 for both), linking the two polyploid groups in a higher order complex.

Thus comparison of the two widespread polyploid complexes, their diversity, origins and distribution provides hypotheses on the evident success these allopolyploids have had in spreading to a diversity of habitats. The results support the idea that continental Australia is the origin of the multiple polyploid events in the tetraploid species, and that several migrations account for the Pacific distribution. This growing body of data from polyploids enables an understanding of the levels of diversity and its evolutionary origin, and points to a highly uneven distribution of diversity on a macro geographic scale.

Scientific Staff: Brown, Craven