Joined: May 2002
This is a good review article about the origin of alternative splicing, and provides evidence of non-coding intronic sequences being transcribed as part of the protein. In other words, more or less random non-coding DNA being adapted to be part of a functional protein.
Trends Genet. 2003 Mar;19(3):115-9.
Evolution of alternative splicing: deletions, insertions and origin of functional parts of proteins from intron sequences.
Kondrashov FA, Koonin EV.
|Alternative splicing is thought to be a major source of functional diversity in animal proteins. We analyzed the evolutionary conservation of proteins encoded by alternatively spliced genes and predicted the ancestral state for 73 cases of alternative splicing (25 insertions and 48 deletions). The amino acid sequences of most of the inserts in proteins produced by alternative splicing are as conserved as the surrounding sequences. Thus, alternative splicing often creates novel isoforms by the insertion of new, functional protein sequences that probably originated from noncoding sequences of introns.|
Some relevant text:
|From the evolutionary standpoint, inserted alternative sequences could be expected to be short if they evolved from noncoding sequences because in-frame stop codons are likely to occur in long noncoding sequences. None of the inserted alternative sequences showed significant similarity to any protein sequences except for their counterparts in orthologs from other species. Furthermore, inserted alternative sequences never included more than one exon (Fig. 2). These observations are compatible with the origin of inserts in LDAS [length difference alternative splicing] from noncoding sequences, most likely from a part of the intron separating the adjacent constitutive exons. On four occasions, this was supported by more direct observations whereby the inserted alternative sequence comprised either an entire intron or a portion of intron joining the adjacent exon.|
Thus, in addition to straightforward exon skipping, a major route for origin of LDAS is insertion of new exons, which encode new, functionally important protein sequences, thus creating functionally distinct isoforms of the respective proteins. Such new exons might in some cases have evolved by tandem duplication of adjacent exons, but more often, they appear to have evolved de novo from noncoding intron sequences. Most of these intron sequences apparently have been recruited to become new coding sequences relatively recently; for example, only in mammals (supplementary material at http://archive.bmn.com/supp/tig/march_Kondrashov_supply.pdf). These observations suggest that evolution of new coding sequences from noncoding ones is an active, ongoing process in eukaryotes. It seems probable that we uncovered only the tip of the proverbial iceberg. Indeed, we detected insertion of new exons that probably originate from intron sequences only for LDAS and only for those cases where yeast and/or prokaryotic orthologs were readily detectable. The actual contribution of intron sequences to the emergence of new protein sequences in eukaryotes is probably substantially greater.