All modern rosids originated from a common hexapolyploid ancestor, and the genomes of some rosids have undergone one or more cycles of paleopolyploidy. After the duplication of the ancient genome, wholesale gene loss and gene subfunctionalization has occurred. Using the extensin super-gene family as an example, we tracked the differential retention and expansion of ancestral extensin genes in four modern rosids, Arabidopsis, Populus, Vitis and Carica, using several analytical methods. The majority of extensin genes in each of the modern rosids were found to originate from different ancestral genes. In Arabidopsis and Populus, almost half of the extensins were paralogous duplicates within the genome of each species. By contrast, no paralogous extensins were detected in Vitis and Carica, which have only undergone the common γ-triplication event. It was noteworthy that a group of extensins containing the IPR006706 domain had actively duplicated in Arabidopsis, giving rise to a neo-extensin around every 3 million years. However, such extensins were absent from, or rare in, the other three rosids. A detailed examination revealed that this group of extensins had proliferated significantly in the genomes of a number of species in the Brassicaceae. We propose that this group of extensins might play important roles in the biology and in the evolution of the Brassicaceae. Our analyses also revealed that nearly all of the paralogous and orthologous extensin-pairs have been under strong purifying selection, leading to the strong conservation of the function of extensins duplicated from the same ancestral gene. Our analyses show that extensins originating from a common ancestor have been differentially retained and expanded among four modern rosids. Our findings suggest that, if Arabidopsis is used as the model plant, we can only learn a limited amount about the functions of a particular gene family. These results also provide an example of how it is essential to learn the origination of a gene when analyzing its function across different plant species.

Background: Intronless genes are a significant characteristic of prokaryotes. Systematic identification and annotation are primary and crucial steps for determining the functions of intronless genes and understanding their occurrence in eukaryotes. Description: In this paper, we describe the construction of the Intronless Genes Database in Dicots (IGDD; available at http://bio.njfu.edu.cn/igdd/), which contains data for five well-annotated plants including Arabidopsis thaliana, Carica papaya, Populus trichocarpa, Salix suchowensis and Vitis vinifera. Using highly visual settings, IGDD displays the structural and functional annotations, the homolog groups, the syntenic relationships, the expression patterns, and the statistical characteristics of intronless genes. In addition, useful tools such as an advanced search and local BLAST are available through a user-friendly and intuitive web interface. Conclusion: In conclusion, the IGDD provides a comprehensive and up-to-date platform for researchers to assist the exploration of intronless genes in dicot plants.