Intron loss? (mechanisms)

arlin at ac.dal.ca arlin at ac.dal.ca
Tue Nov 16 17:43:15 EST 1993


In article <2c8r78$hf5 at overload.lbl.gov>, deh at s27w007.pswfs.gov (Dave Harry) writes:>  stuff deleted....
>(stuff deleted)
>>is a straightforward answer to this.  Instead of incorporating into
>>the genome illegitimately, the 'processed cDNA' could recombine
>>homologously with its parent locus, and an intron-containing region 
>>in the parent locus could be replaced with a homologous but 
>>intron-less region in the cDNA.  The result would be the loss of 
>>one or several adjacent introns.  There are several testable 
>>implications that could conceivably be used to distinguish this 
>>RNA-mediated mechanism for intron loss from DNA-based mechanisms.
> 
> I agree in part.  Such a mechanism could account for the loss of any
> single intron.  Methinks, however, that recombination could occur with
> an extra-chromosomal cDNA, or with a pseudogene following its illegitmate
> incorpration into the genome from the cDNA copy.  My sense is that since
> gene insertion requires only 1 break-repair step, this might be relatively
> common.  A double-recombination event would be required to accomplish the
> homologous replacement of an intron-containg sequence with a non-intron
> containing sequence.  The latter process would seem to be relatively rare
> (but at least in yeast it seems to happen).  Might such a recombination
> event be more likely to occur between non-homologous chromosomes (ie
> between an intron-containing gene and a pseudogene) than with an
> extrachomosomal cDNA?  (Forgive me, but my ignorance is showing here....
(stuff deleted)
> 
> Dave Harry

It is quite true that in some organisms, non-specific incorporation of
introduced DNA is far more common than incorporation by homologous
recombination.  However, even if a double-crossover or gene conversion between
the cDNA and the parent locus is less common than random insertion, this is not
important, so long as the process is common enough to explain the observed rate
of intron loss.  

The mechanism you have suggested makes perfect sense, too, and is a good idea.
The intron loss could occur by gene conversion or double-crossover between i)
the parent locus and its processed cDNA or ii) the parent locus and an existing
processed pseudogene. The effect is the same in both cases, except that the
event involving the pre-existing pseudogene would tend to introduce sequence
changes into the parent gene.  It comes down to a question of whether the parent
locus is more likely to be converted by a free DNA fragment (the cDNA) or a
non-allelic locus (the processed pseudogene), given that there are X cDNAs
floating around at any point in time, and Y pseudogene loci elsewhere in the
genome.  Take your pick.

Arlin



More information about the Bioforum mailing list