Polyploidization-driven transcriptomic dynamics in Medicago sativa neotetraploids: mRNA, smRNA and allele-specific gene expression.

Whole genome duplication (WGD) is a powerful evolutionary mechanism in plants. Autopolyploids have been comparatively less studied than allopolyploids, with sexual autopolyploidization receiving even less attention. In this work, we studied the transcriptomes of neotetraploids (2n = 4x = 32) obtained by crossing two diploid (2n = 2x = 16) plants of Medicago sativa that produce a significant percentage of either 2n eggs or pollen. Diploid progeny from the same cross allowed us to separate the transcriptional outcomes of hybridization from those of WGD. This material can help to elucidate events at the base of the domestication of cultivated 4x alfalfa, the world's most important leguminous forage. Three 2x and three 4x progeny plants and 2x parental plants were used for this study. The RNA-seq data revealed that WGD did not dramatically affect the transcription of leaf protein-coding genes. The two parental genotypes did not contribute equally to the progeny transcriptomes, and genome-wide expression level dominance of the male parent was observed. A large majority of the genes whose expression level changed due to WGD presented increased expression, indicating that the 4x state requires the upregulation of approximately 2.66% of the protein-coding genes. Overall, we estimated that 3.63% of the protein-coding genes were transcriptionally affected by WGD and may contribute to the phenotypic novelty of the neotetraploid plants. Pathway analysis suggested that WGD could affect secondary metabolite biosynthesis, which in turn may influence forage quality. We found four times as many transcription factor genes among the polyploidization-affected genes than among those affected only by hybridization. Several of these belong to classes involved in stress response. Small RNA-seq revealed that very few miRNAs were significantly associated with WGD, but they target several hundred genes, and their role in the WGD response may be relevant. Integrated network analysis led to the identification of putative miRNA: mRNA interactions potentially involved in transcriptome reprogramming. Allele-specific expression analysis indicated that parent-of-origin bias was not a significant outcome of WGD, but we found that parentally biased RNA editing may be a significant source of variation in neopolyploids.