There is relatively little published information on the evolution of gene expression in polyploids, particularly outside of model groups such as maize, wheat, Brassica, and Arabidopsis. Our work on the systematics of allopolyploids in the wild perennial relatives of soybean (Glycine subgenus Glycine) provides us with the knowledge of genome origins that is required for testing hypotheses of gene and genome evolution. We have sought evidence of gene silencing in these recently formed polyploids. Initially we adopted a gene-by-gene approach, choosing classes of genes we felt were candidates for silencing. The 18S-26S nuclear ribosomal gene family was chosen because nucleolar dominance is well-documented from model allopolyploids. Ribosomal genes are differentially expressed in at least some Glycine allopolyploids, both in cases where there are great differences in copy number between homoeologous repeats and where homoeologous copy number is approximately equal. The chloroplast-expressed nuclear-encoded isozyme of glutamine synthetase was chosen as an example of a nuclear gene functioning in the maternally-inherited chloroplast. However, we found no evidence of silencing. This led us to adopt a genomics approach to searching for silenced genes. We isolated a potentially methylated DNA fraction from allopolyploids and their diploid progenitors, and used this fraction to probe filters containing 2,300 unique Glycine max cDNA clones. Hybridization patterns were very similar both among diploids from different genome groups and from polyploids, suggesting that few, if any, homoeologous copies of genes are methylated in Glycine allopolyploids.

Key words: gene expression, genomics, Glycine subgenus Glycine, homoeologous gene silencing, nuclear ribosomal genes, polyploidy