Biology Reference
In-Depth Information
codons, phase 1 if the intron lies between the first and second nucleotides of a
codon, and phase 2 if it lies between the second and third nucleotides of a codon.
The requirement for phase compatibility dictates that only symmetrical exons of
class 1-1, class 2-2 and class 0-0 [i.e. exons flanked by introns of the same phase
(1, 2, or 0)] are suitable for duplication or insertion. It is therefore not surprising
that genes encoding mosaic proteins (proteins which comprise multiple domains
and which are likely to have evolved by exon shuffling) contain a disproportion-
ate number of class 1-1 exons encoding a wide variety of different modules, for
example EGF-like, calcium-binding, fibronectin-like finger, kringle, complement
B, LDL receptor, von Willebrand, thyroglobulin, C-type lectin etc (Patthy 1991a,
1994, 1996). Class 1-1 introns predominate in the immunoglobulin, T-cell recep-
tor and HLA genes, the Thy-1 glycoprotein ( THY1 ; 11q22.3-q23) gene and the
2-microglobulin ( B2M ; 15q21-q22) gene, consistent with exon shuffling having
been important in their evolution (Patthy et al ., 1987). Further examples of this
phenomenon include the exons encoding the complement B-type domains in the
human C4b-binding protein
gene ( C4BPA ; 1q32; Hillarp et al ., 1993), the
immunoglobulin-like and fibronectin type III modules in the human Axl onco-
gene ( AXL ; 19q13.1-13.2; Schulz et al ., 1993) gene and four of six throm-
bospondin modules of the human properdin gene ( BF ; 6p21.3; Nolan et al ., 1992).
Genes with predominantly class 0-0 exons include those encoding the type III
collagens (see Chapter 4) and
-casein ( CSN2 ; 4pter-q21) whilst class 2-2 exons
are exclusively found in the glucagon ( GCG ; 2q36-q37) gene (Patthy, 1987).
Using a large database of eukaryotic genes, Long et al . (1995) demonstrated there
to be an excess of symmetrical exons over expectation and estimated that at least
19% of exons had been involved in exon shuffling. Why the preponderance of class
1-1 exons? Patthy (1994) suggested that since 92% of introns flanking signal pep-
tide domains were phase 1, modularization of exported proteins with secretory sig-
nal peptide domains is likely to have employed class 1-1 introns. According to
Patthy (1994), modularization proceeds in different stages: (i) insertion of introns
of identical phase at boundaries of the protein fold, (ii) tandem duplication of the
symmetrical module by intronic recombination, and (iii) module transfer to a
novel location ( Figure 3.6 ). Nonsymmetrical exons (those flanked by introns of dif-
ferent phase) are potentially much less versatile and may be expected to have been
utilized much less often. Tomita et al . (1996) noted that when introns are inserted
so as to disrupt codons, the site of insertion occurs much more frequently between
the first and second bases than between the second and third bases. The reason for
this remains unclear.
There are however various instances which do not conform to the above exon
shuffling rules. In some cases, the original exon-intron organization of genes
has become eroded with the passage of evolutionary time. Thus, although in the
human perlecan ( HSPG2 ; 1p35-p36; Cohen et al ., 1993) gene, the LDL receptor
modules and the immunoglobulin-like modules are still flanked by phase 1
introns, the original introns have been lost from the regions of the gene encod-
ing laminin A, laminin B and epidermal growth factor-like modules. Similarly,
in the human complement 6 ( C6 ; chromosome 5; Hobart et al ., 1993) gene,
phase 1 introns are found only at the boundaries of one of the complement B
modules having been lost from the boundaries of the class 1-1 thrombospondin,
 
Search WWH ::




Custom Search