Until recently, we believed that it was the job of proteins to regulate gene expression. More recently it has become clear that RNA is another important player in this process. Using mouse oocytes and preimplantation embryos, Barbara Knowles and co-workers now show that retrotransposons have an important role in gene regulation — possibly by controlling chromatin states.

That mobile elements can affect gene expression is not news — it has been known for a long time that transposon–promoter sequences can induce endogenous gene expression by transcriptional read-through. As far back as the 1950s, when Barbara McClintock first described transposons, she referred to them as controlling elements. What was previously unappreciated, however, was the scale on which this might occur.

The authors' discovery began with the analysis of cDNA libraries from mature mouse oocytes, 2-cell-stage embryos and blastocysts. They observed that two types of retrotransposons were strikingly abundant among transcripts in oocytes and in 2-cell-stage embryos. But as well as these transcripts, the authors found abundant chimeric transcripts, made up of host sequences spliced downstream of alternative 5′ retroelement sequences.

Aligning these chimeric sequences to the mouse genome showed that all of them came from genes that harboured the transposable elements, either within the gene locus or upstream of it.

Wanting to look into the potential roles of these chimeric transcripts, the authors compared their expression dynamics with that of the corresponding wild-type gene. The fact that the chimeric transcripts disappear after the 8-cell stage indicated that they are maternal transcripts, and the authors suggest that the long terminal repeats of these retrotransposons might behave as oocyte-specific promoters. Although some ORFs of these chimeric transcripts do not differ from the wild-type transcripts, others do, and as the authors show that at least some are translated, it is therefore possible that they might encode novel, stage-specific functions.

Support for the functional importance of this retrotransposon-mediated transcription comes from phylogenetic conservation — the positioning of some of the retrotransposons associated with chimeric transcripts is conserved between two distantly-related strains of mice and, in some cases, between rat and mouse. Unfortunately, what their function is remains an open question. The authors propose that sequential activation and silencing of retrotransposons might underlie stage-specific, potentially RNAi-mediated, chromatin remodelling at specific genomic locations at early stages of development and during oogenesis — an attractive model that urgently needs further testing.