Plant conserved non-coding sequences and paralogue evolution

Genome duplication is a powerful evolutionary force and is arguably most prominent in plants, where several ancient whole-genome duplication events have been documented. Models of gene evolution predict that functional divergence between duplicates (subfunctionalization) is caused by the loss of regulatory elements. Studies of conserved non-coding sequences (CNSs), which are putative regulatory elements, indicate that plants have far fewer CNSs per gene than mammals, suggesting that plants have less complex regulatory mechanisms. Furthermore, a recent study of a duplicated gene pair in maize suggests that CNSs are lost in a complementary fashion, perhaps driving subfunctionalization. If subfunctionalization is common, one expects duplicate genes to diverge in expression; recent microarray analyses in Arabidopsis thalinia suggest that this is the case. Plant genomes are relatively complex on a genomic level because of the prevalence of whole-genome duplication and, paradoxically, subfunctionalization after duplication can lead to relatively simple regulatory regions on a per gene basis.

Introduction

Gene duplication is a major determinant of the size and gene complement of eukaryotic genomes. Perhaps the most spectacular method of gene duplication is whole-genome duplication via polyploidization, which has a major role in the evolutionary history of eukaryotes. For example, the yeast Saccharomyces cerevisiae contains numerous duplicated genes and chromosomal regions that are attributed to a polyploidy event ∼100 million years ago (mya) 1, 2. Similarly, the human genome retains vestiges of duplication events that occurred between 350 and 650 mya and that might be attributable to at least one polyploid event [3].

Genome duplication is particularly prominent in plants. Arabidopsis thaliana has experienced at least three ancient polyploid events 4, 5, 6; rice (Oryza sativa) contains duplicated chromosomal regions that are attributable either to ancient segmental duplications [7] or to a paleopolyploid event 8, 9; genetic maps of maize (Zea mays ssp. mays) contain evidence of several large-scale duplication events [10]. More recently, Blanc and Wolfe [11] used EST data to investigate the number and relative age of duplicated genes in 14 plant species. Nine of these species contained evidence of ancient large-scale duplication events, reflecting at least seven paleopolyploid events in the phylogenetic history of the sample. At least 16 polyploid events have been documented during the evolutionary history of a relatively small sample of angiosperm taxa (Figure 1).

The inescapable conclusion is that the organization and evolution of plant genomes have been shaped by many recent and ancient polyploid events. By contrast, vertebrates have probably undergone only one or two large-scale genome duplication events throughout their ∼500 million year history 3, 12, 13. With a few exceptions (such as amphibians [14]), extant polyploids are also rare. In mammals, for example, the only known polyploid is the tetraploid red viscacha rat of Argentina [15].

The relative frequency of genome duplication in plant and animal lineages affects their genome organization but might also have profound effects on gene function and regulation. Gene duplication has long been considered a crucial step in ‘freeing’ single-copy genes from selective constraint, enabling them to evolve new functions [12], but a pair of duplicated genes can also diverge in function as a result of changes in regulatory elements [16]. These observations raise a number of questions: What are the consequences of genome duplication for the regulatory complexity of plant genomes? Are differences in regulatory motifs – more specifically conserved non-coding sequences (CNSs) – evident between plant and animal genomes? How does genetic duplication affect regulatory elements, and what are the consequences for the evolution of promoter regions between paralogous genes? We address these questions in this article.