Intron positions can be lost, notably the most 3' ones



The YFL037w scaffold gene contains 6 introns. This picture is consistent with that observed in S. pombe where the YFL037w gene exhibit five intron positions, scattered all along the ORF sequence. This multi-intronic picture is no longer valuable in the hemiascomycetous yeast species, where most YFL037w homologues are intron-lacking or mono-intronic genes. When intron positions are missing, it is usually the most 3' that are affected, like i3, i4, i5 and i6. Introns located near the 5'end, like i1 and i2 are less frequently affected.

This picture tends to be a common rule since the vast majority of intron-containing hemiascomycetous genes are mono-intronic and nearly all introns are located in the first few codons of the 5' end. Note that only 5% or so of the nuclear hemiascomycetous yeast genes are intronic genes, the remaining in their vast majority being intron-lacking genes (see Common features). If ancient ascomycetes, like S. pombe [Wood, 2002, Nature, 415, 871-880], had many introns scattered all along the length of ORFs, this suggests that introns were massively erased from nearly 95% of hemiascomycetous genes [Bon et al., 2003].

The model (F2_page2) proposed by Fink [Fink, 1987, Cell, 49, 5-6] provides a plausible explanation for intron ?disappearance? and such 5' bias in intron distribution. It speculates that yeast genes lost their introns due to homologous recombination with reverse-transcribed cDNAs initiated at the 3' end of the message. The source of such endogenous RT activity is supposed to originate from non-LTR retrotransposable elements (LINEs) [Eickbush 2000, Nature, 404, 940-943] which are known to produce pseudogenes in humans [Dhellin et al., 1997, EMBO J., 16, 6590-6602; Esnault et al., 2000, Nat. Genet., 24, 363-367], but LINE-independent intron removal may also exist [Bon et al., 2003].

The Fink model




Last modified: Tue Feb 10 11:11:31 CET 2004