Reporting in Nature, Olivier Hachet and Anne Ephrussi reveal a new role for mRNA splicing — regulating the cytoplasmic localization of mRNA.

Until now, it was thought that the cytoplasmic localization of oskar mRNA to the posterior pole of the Drosophila melanogaster oocyte was specified by the 3′ untranslated region (UTR) alone. However, deposition of the human exon–exon junction complex (EJC) on mRNA is splicing-dependent and RNA-sequence independent, and two of the human EJC components are homologues of the D. melanogaster EJC proteins that are needed to localize oskar mRNA. This prompted the authors to investigate whether splicing regulates oskar mRNA localization.

Hachet and Ephrussi found that deleting all three oskar introns did not affect the transportation of this so-called oskΔi(1,2,3) mRNA into the oocytes of transgenic flies, in which no endogenous oskar mRNA was produced. However, by mid-oogenesis the localization of oskΔi(1,2,3) mRNA was defective (diffuse), with only a small amount present at the posterior pole by late oogenesis. A series of intron deletions showed that only intron 1 was required for the localization of oskar mRNA. The oskar-mRNA-dependent localization of the EJC protein Y14 to the posterior pole was also shown to depend only on the presence of intron 1. But is it the splicing of intron 1 or sequence-specific information in intron 1 that determines the localization of oskar mRNA?

To answer this question, the authors replaced the intron-1 sequence with the intron-3 sequence. This osk(i3 in i1) mRNA was correctly localized to the posterior pole, as was Y14. So the intron-1 sequence itself isn't required for localization, but splicing of the first exon–exon junction is.

Although oskar transcripts with no 3′ UTR fail to localize correctly, these results indicate that the 3′ UTR alone is not sufficient for the localization of oskar mRNA. Indeed, Hachet and Ephrussi found that the oskar 3′ UTR can only drive localization in the presence of (appropriately spliced) endogenous oskar mRNA, and that it is independent of Oskar protein, which indicates that other proteins are also involved. So, it seems that the 3′ UTR promotes the splicing-dependent assembly of oskar mRNA into an oskar mRNP complex. And as EJC components Y14 and Mago nashi are needed to localize oskar mRNA, it is the splicing-dependent deposition of the EJC at the first exon–exon junction that couples splicing to the correct cytoplasmic localization of oskar mRNA at the posterior pole of the oocyte.

The authors suggest that where the EJC is deposited on an mRNA molecule determines the architecture of the localization complex that is formed (by mediating interactions between factors that are bound to different regions of the mRNA). And, more broadly, they suggest this could explain why the EJC is not always involved in cytoplasmic-mRNA localization.