Michael Freeling
Andrew Paterson
Alex Feltus



Michael Freeling Lab at UC Berkeley

Plant Genome Mapping Laboratory at the University of Georgia

Alex Feltus Lab at Clemson University

Plant Genome Duplication Database

GIFT Program

NSF 2010 collaborative projects MCB 0820821; 0821096; 0820345:
The evolution of gene position and function in Arabidopsis using outgroup genomes.

Project Abstract

Genes commonly move around in genomes, but the functional consequences of gene transpositions are unknown. Arabidopsis shares many of its 26,500 nontransposon, nontandem genes with its fellow Brassicales species, papaya, and also with its fellow rosid species, grape and poplar. Preliminary Results used a new top-down multi-alignment algorithm to infer gene order in the last common ancestors of Brassicales and eudicots, respectively, and found evidence for an ancient hexaploidy shared by all dicots and perhaps all plants. Additional Preliminary Data shows that about 40% of Arabidopsis genes are no longer ancestral, either because they have 1) diverged rapidly, or 2) exhibit a “high birth-and-death” behavior in the phylogenetic tree, or 3) somehow got positioned in rearrangement-prone regions of chromosome, or 4) transposed to new locations in the genome. A “flanking gene test”, employing papaya as an outgroup for Arabidopsis and grape as an outgroup for papaya, shows that some families—like NBS-LRR, MADS-box, B3-box and F-box genes—contain genes that are almost always newly transposed. We propose to greatly improve our deductions of common ancestral gene contents and orders, and to evaluate every gene in Arabidopsis in terms of the positional/behavioral categories enumerated above.

Many Arabidopsis genes are no longer "whole genes,” relative to their ancestors. Arabidopsis has undergone two sequential tetraploidies since diverging from papaya, and neither papaya nor grape have had a polyploidy since the ancient hexaploidy in a common ancestor. Accordingly, many papaya or grape genes have 2, 3 or 4 Arabidopsis homeologs. We show that conserved noncoding sequences sometimes subfunctionalize among homeologs so that no single Arabidopsis gene is expected to have all functions or regulatory sequences of an ancestral or papaya gene. There should be over 4,000 deeply conserved noncoding sequences (At-Cp CNSs) shared between papaya and Arabidopsis. These regulatory sites will be sorted to gene, made into a database and evaluated for subfunctionalization at the level of sequence, gene expression and fitness (reproductive output).

Even before this project’s proposed data are complete, several hypotheses involving gene ancestry/position and gene expression have emerged. For examples: Is a transposed gene compared to its closest ancestral gene paralog divergent in any measure of gene expression? Do genes that have been clustered together on the chromosome because of post-tetraploid fractionation tend to be co-regulated? We will test many hypotheses such as these in silico, using all microarray data normalized by The Arabidopsis Information Resource. Relationships between the spectrum of duplicate gene fates/expression patterns and fitness (reproductive output) will also be evaluated, using knockout mutations and phenotypic analyses.




 · Objectives
 · Timeline
 · Reference database
 · News




 Duplicated Arabidopsis genes




Web sites

Education/Outreach Tools