IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

Exon-Intron boundaries

Arlin Stoltzfus arlin at thidwick.biochem.dal.ca
Fri Oct 9 10:41:10 EST 1998

Thorsten Burmester wrote:
> Dear evolutionists,
> Is there any paper on the frequencies of the phases (relative to the
> reading frame of the exon) in which the introns are inserted.

There are a number of papers that address this topic.  Probably 
the first to document an asymmetry in phase distribution was 

  Hickey, et al., 1989. J. Theor. Biol. 137:41

Since then there have been several more comprehensive and 
quantitative surveys of data.  For animal and plant genes 
(which are most of the intron-containing genes in databases), 
40-55% of introns are phase 0, 30-40% are phase 1, and about
20% are phase 2.  See: 

  Federov, et al., 1992. Nucl. Acids Res. 20: 2553
  Gelfand, 1992. J. Mol. Evol. 35:239
  Fichant, 1992.  Human Mol. Genet. 1:259.

Gelfand's paper uses an unusual sample of genes with a 
larger fraction of phase 1 introns.  

Quite probably, the assymetric distribution is due to 
addition of introns to target sequences (something like 
AG^G) that are themselves non-randomly distributed with 
respect to phase.  In particular, Federov et al (Tab. 2), 
Gelfand (p. 248) and Long, et al: 

  Long, et al. 1998.  Proc. Natl. Acad. Sci. 95:219.

(Tab. 3) all show that phase 0 introns are expected to be 
the most common if introns are added non-randomly at a 
sequence like AG^G, due to the distribution of AGG in 
coding regions.  

There have also been several papers on the subject of intron 
phase autocorrelation, that is, the tendency for neighboring 
introns to have the same phase: 

  Long, et al., 1995.  Proc. Natl. Acad. Sci. 92:12495
  Tomita, et al., 1996. Mol. Biol. Evol. 13:1219.
  Federov, et al., 1998.  J. Mol. Evol. 46:263.
The causes of this pattern, and whether it is an ancient 
(as opposed to eukaryote-specific) phenomenon, are topics 
presently under study by these research groups.  All three 
groups tend to believe that the autocorrelation is due 
to exon shuffling, and it is clearly partly due to exon 
shuffling, though there is a good deal of room for
interpretations about additional causes.  Hope this helps. 

Arlin Stoltzfus, Ph.D. (arlin at is.dal.ca)
Department of Biochemistry, Dalhousie University
Halifax, Nova Scotia B3H 4H7 CANADA
phone: 902-494-2968     fax: 902-494-1355

More information about the Mol-evol mailing list

Send comments to us at biosci-help [At] net.bio.net