Contractions and expansions in gene size and number - Human Gene Evolution - page 354

which appears to contain two fewer repeats than baboon semenogelin II
(Ulvsbäck and Lundwall, 1997). The primate semenogelin genes arose by dupli-
cation of an ancestral gene about 60 Myrs ago (Lundwall, 1996) and the human
semenogelin I gene appears to have lost two 60 amino acid repeats by comparison
with the paralogous semenogelin II gene (Ulvsbäck and Lundwall, 1997). The
semenogelin I protein from cotton-top tamarin ( Saguinus oedipus ) possesses three
additional repeats as compared to human and, intriguingly, possesses 14 potential
glycosylation sites not present in the human protein (Lundwall, 1998).
Most known genes manifesting internal duplication exhibit strong conserva-
tion of intron positions between the duplicated domains (Chapter 3, sections 3.1
and 3.6.2). One notable exception, however, is the rabbit phosphofructokinase
gene (Lee et al ., 1987). Sequence homologies between bacterial and rabbit phos-
phofructokinases and between the amino and carboxy terminal ends of the rabbit
enzyme are consistent with the origin of this gene being via a process of internal
duplication and divergence. However, intron positions are not conserved even
between the two halves of the rabbit gene.
Exon duplication may potentiate alternative splicing. Thus, the human keto-
hexokinase ( KHK ; 2p23) gene contains two very similar 135 bp exons (termed 3a
and 3c) which are mutually exclusively spliced into KHK mRNA (Hayward and
Bonthron, 1998). Both the exon-intron structure and the pattern of alternative
splicing are conserved between human, rat and mouse, consistent with the exis-
tence of two evolutionarily conserved KHK isoforms. This alternative splicing
event is also tissue-specific since, in both rat and human, those tissues that
express high levels of KHK incorporate exon 3c whereas other tissues incorporate
exon 3a. Interestingly, a shift in splicing choice from exon 3a to 3c appears to
occur during development between the human fetus and adult (Hayward and
Bonthron, 1998).

8.6.3 Intra-exonic duplications

Internal gene duplications may sometimes be quite subtle. Thus, the 3 untrans-

lated regions of the human and monkey cytochrome c oxidase subunit II
( MTCO2 ; mitochondrial genome) genes have been generated by duplication
events involving a 13 bp region that occurred during primate evolution
(Ramharack and Deeley, 1987). The RNA derived from the MTCO2 gene has the
potential to form stable stem-loop structures in a region immediately preceding
the duplication site and these inverted repeat sequences may have played a role in
promoting the duplicational events.
In some cases, homogenization of internal repeats has occurred as for example
with the complement control protein repeats of the baboon and human comple-
ment receptor type 1 ( CR1 ; 1q32; Clemenza et al ., 1997; Hourcade et al ., 1990)
genes. This process has been termed horizontal or concerted evolution and the mech-
anism involved is likely to be either unequal crossing over or gene conversion.
Whichever mechanism is responsible, the following example suggests that intra-
genic homogenization may be a less efficient process than intergenic homoge-
nization. The human small proline-rich (SPRR) proteins, induced during
keratinocyte differentiation, are encoded by two SPRR1 genes (A and B), approx-
imately seven SPRR2 genes and a single SPRR3 gene, closely linked on chromo-