Transposon induced colour variation in maize.

Transposons, Barbara McClintock's 'jumping genes', are the ultimate parasite. Intimately entwined with their host's DNA, they passively replicate along with it. Less benignly, they can spread in a genome, disrupting and mutating genes as they go. Two recent studies on Arabidopsis demonstrate that epigenetic modification can prevent this destructive jumping by showing that demethylating otherwise dormant transposons activates and mobilizes them. It had previously been known, both in Arabidopsis and mice, that transposons become transcriptionally active in methylation-deficient backgrounds. These new studies go further by showing that demethylation also mobilizes transposons.

Both studies looked at the activities of transposons in ddm1 (decrease in DNA methylation) Arabidopsis mutants. DDM1 encodes a protein that is similar to the chromatin-remodelling factor SWI2/SNF2 and is essential for silencing many methylated and repeated genes in Arabidopsis — ddm1 mutants have hypomethylated genomes.

Taking a bioinformatics approach, Martienssen and colleagues searched the recently completed Arabidopsis genome sequence and identified 22 transposons related to the Mutator transposable elements from maize. In wild-type Columbia ecotype plants, these transposons are methylated and dormant, but in ddm1 plants they are hypomethylated and transcribed. In addition, more than one in eight ddm1 plants had extra Mutator-like transposons, indicating that complete transposition had occurred.

Conversely, Kakutani and colleagues began by investigating the developmental abnormalities of ddm1 plants. One dwarfing ddm1 mutant phenotype was found to be caused by a DNA insertion that disrupted a gene on chromosome 3, which is involved in synthesis of the plant hormone brassinosteroid. The insertion, when sequenced, seemed to have been caused by the transposition of an 8,479-bp region from chromosome 2. This region, named CAC1 , resembled a CACTA family transposon, the members of which have previously been identified in both maize and Antirrhinum. Searching the Arabidopsis genome turned up only four sequences related to CAC1, but in 11 out of 12 ddm1 lines investigated there were many more copies scattered throughout all 5 chromosomes, presumably produced by transposition of 'wild-type' CAC transposons.

The precise mechanisms involved are yet to be established, but these results show that epigenetic control of gene expression has been co-opted to guard against the chaos caused by unfettered transposon activity. Or, given that maintenance of a stable genome is essential for any organism, perhaps this defence function came first.