Figure 1.

A mechanism for a 1-nt intron shift involving the ancestral existence of an alternatively spliced [D1,2]/[A2,3] stwintron intermediate. (A) Structure and nomenclature of stwintrons relevant to this study. Intronic nt are in lower case letter: nt of the internal intron (light gray) in red, and nt of the external intron (dark gray) in blue lettering. The conserved terminal sequences essential for spliceosomal intron excision are indicated: D, donor sequence (gt-) at the 5′ splice site, and A, acceptor sequence (-ag) at the 3′ splice site. In a [D1,2] stwintron, an internal U2 intron interrupts the donor element of an external U2 intron between the first and the second nt. In a [A2,3] stwintron, the internal intron is nested in the (3-nt) acceptor element of the external intron between the second and the third nt. The G upstream the most 5′ donor is a characteristic feature of the [D1,2] stwintron but is exonic for the [A2,3] stwintron; The G downstream the most 3′ acceptor is a characteristic feature of the [A2,3] stwintron but is exonic for the [D1,2] stwintron. (B) A mechanism for the evolution of neighboring discordant intron positions. An alternatively spliced [D1,2]/[A2,3] stwintron occurs. When subsequently its [A2,3] internal intron is lost, the remaining canonical intron is localized at the 3′ of neighboring, alternative exon fusion sites utilized by the ancestor [D1,2]/[A2,3] stwintron. However, if its [D1,2] internal intron is lost, the de novo canonical intron is localized at the 5′ of those exon fusion sites. The two U2 introns resulting from mutually exclusive internal intron loss events at the DNA level are shown respectively at the top and the bottom, the two intron positions being called 5′ and 3′.