Darwinian alchemy: Human genes from noncoding DNA

At least since the publication of Susumu Ohno's Evolution by Gene Duplication (Ohno 1970), the conventional wisdom has been that, in the emergence of novel genes, “natural selection merely modified, while redundancy created.” In other words, new genes generally arise by the duplication of existing genes. While the notion that duplication plays a prominent role in the emergence of novel genes is perhaps most famously associated with Ohno, it actually traces back to the early days of the modern evolutionary synthesis (Bridges 1935; Muller 1936). Decades of modern sequence-based research have largely supported this general view (Graur and Li 2000). In recent years, the classic model of whole gene duplication and subsequent divergence has been enlarged to include phenomena such as exon shuffling, gene fusion and fission, retrotransposition, and lateral gene transfer (for review, see Long et al. 2003). Nevertheless, despite their additional complexity, these mechanisms remain essentially duplicative, in the sense that sequences encoding one or more protein-coding genes are copied, by one mechanism or another, and used as the starting point for a new gene sequence. (An exception is the exonization of noncoding transposable elements, such as Alus, but this process tends to generate individual exons rather than entire genes;Makalowski et al. 1994; Nekrutenko and Li 2001.) By contrast, the origination of protein-coding genes de novo from nonrepetitive, noncoding DNA has been thought to occur only as an exceptionally rare event during evolution. Indeed, the emergence of complete, functional genes—with promoters, open reading frames (ORFs), and functional proteins—from “junk” DNA would seem highly improbable, almost like the elusive transmutation of lead into gold that was sought by medieval alchemists