More than one way to mess up a gene

John Armour, University of Nottingham

Sometimes the relationship between a genetic disease phenotype and its causative mutation is very specific:
mutation of codon 6 of the beta-globin gene, changing glutamate (GAG) to valine (GTG), is the one way of making the allele causing sickle cell anaemia. In other cases – and especially where the mutation results in loss of function – there are all sorts of ways in which a mutation can make a disease-causing allele. At the cystic fibrosis gene CTFR, for example, the human mutation database lists hundreds of different ways of messing up the gene function, including nonsense (stop), missense, splicing, regulatory and deletion mutations. Sometimes the mutation is more exotic, but none the less catastrophic for the gene: for example, the
commonest allele causing beta-thalassaemia in Greece is a mutation from G to A 110 bases into intron 1: this creates a new splice acceptor site, incorporates additional sequence from intron 1 into the mature mRNA and consequently creates a loss-of-function allele.

An interesting addition to the possible ways of messing up a gene was reported in the April 2002 issue of Nature Genetics by Pagani et al. The structural basis of the mutation in the ATM gene was deletion of 4bp deep into intron 20, the net effect of which was to activate the inclusion of an additional 65bp cryptic exon. What is new and interesting about this mutation is that the inclusion of the 65bp cryptic exon is caused not by a mutation in splicing signals (as in the Greek beta-thalassaemia mutation above), but by a mutation inside the cryptic exon itself. The affected sequence is an Intron- Splicing Processing Element (ISPE), which normally interacts with U1 snRNP particles to mediate accurate removal of intron 20. When this element is disrupted by the mutation, intron 20 is not correctly recognised and allows the alternative splicing pathway which incorporates the cryptic 65bp “exon”.

We are just beginning to unravel what chromosomal elements outside (and sometimes very distant from) coding sequence contribute to controlling gene expression. In some of the best-characterised examples, we have some idea of how they work and thus can guess what the effect of mutation in them might be. Much analysis of genome evolution acknowledges one type of significant mutation: non-synonymous changes in codons. The work of Pagani et al. should highlight not only how significant mutations in human genes need not involve codons at all, but also indicate that the detail of how genes and genomes really evolve might sometimes be very subtle and complicated.

Reference
Pagani, F., et al. (2002). A new type of mutation causes a
splicing defect in ATM. Nature Genetics 30, 426-429.