Biology Reference
In-Depth Information
(Cavalier-Smith, 1991; Palmer and Logsdon, 1991). At its extremes, the debate
between these two viewpoints is reminiscent of the old controversies between selec-
tionists and neutralists. However, perhaps not altogether surprisingly, the debate
has generated a significant amount of evidence supporting both theories (Gilbert
et
al
., 1997; Rogers 1990).
One prediction of the introns-early hypothesis is that intron location should
tend to demarcate structural or functional domains within proteins (Blake, 1978;
Branden
et al
., 1984; Craik
et al
., 1983; Go, 1987). This question has proved highly
contentious (Traut, 1988). The extreme introns-late view is that no such correla-
tion exists and that introns have been inserted quasi-randomly into the structure
of genes. Consistent with this standpoint, several genes encoding evolutionarily
ancient proteins (actins, alcohol dehydrogenase, carbonic anhydrase II, pyruvate
kinase, globins) appear not to exhibit any correlation between intron position and
structural features (Stoltzfus
et al
., 1994; Weber and Kabsch, 1994). However,
more recently, a computer program designed to predict precisely the locations of
module boundaries in proteins has helped to demonstrate that there is indeed a
strong correlation between intron position and structural elements at least in the
sample of 32 ancient proteins examined (de Souza
et al
., 1996). The introns
involved were invariably phase zero (de Souza
et al
., 1998; see Section 3.6.2).
As we have noted in Section 3.2, the conservation of intron location is a common
finding. That intron location is often conserved in evolutionarily ancient genes such
as glyceraldehyde-3-phosphate dehydrogenase (
GAPD
; 12p13; Kersanach
et al
.,
1993), carbamoylphosphate synthetase (
CPS1
; 2q35; van den Hoff
et al
., 1995) and
triose phosphate isomerase (
TPI1
; 12p13; Gilbert and Glynias 1993; Marchionni
and Gilbert 1986; McKnight
et al
., 1986; Straus and Gilbert 1985) is certainly com-
patible with the introns-early hypothesis. Most informative in terms of the introns-
early/late debate, however, are discordant cases, examples of evolutionarily related
genes from different taxa in which either intron position varies from between several
nucleotides to several codons or where the intron is either present or absent (Rogers,
1989). The introns-early theorists have explained these discrepancies by invoking
intron sliding
and deletion (Craik
et al
., 1983) but despite the occasional convincing
example [e.g. intron 8 of the histidyl-tRNA synthetase (
HARS
; chromosome 5)
gene; Brenner and Corrochano, 1996;
Figure 3.3
; intron 2 of the glucose-dependent
insulinotropic peptide (
GIP
; 17q21.3-q22) gene which results in an 8 amino acid
deletion of the prepropeptide of the rat protein as compared to human; Higashimoto
and Liddle, 1993], the evidence for the widespread occurrence of intron sliding is
still rather weak (Yuasa
et al
., 1997; Stoltzfus
et al
., 1997). The introns-late view is that
intron sliding is inherently improbable because such a process would almost
innevitably involve an intermediate stage that would alter the reading frame thereby
leading to the loss or inactivation of the protein product. Therefore introns-late
devotees have regarded discordant intron location, as found in the evolutionarily
ancient
-tubulin (
TUBA1
, 2q;
TUBA2
, 13q11;
TUBB
, 6p21-pter; Dibb and
Newman, 1989), aldehyde dehydrogenase (
ALDH1
, 9p21;
ALDH2
, 12q24;
ALDH3
,
17;
ALDH5
, 9;
ALDH6
, 15q26;
ALDH9
, 1;
ALDH10
, 17p11.2; Rzhetsky
et al
.,
1997) and triose-phosphate isomerase (
TPI1
; 12p13; Kwiatowski
et al
., 1995;
Logsdon
et al
., 1995) genes as evidence for the occurrence of multiple independent
insertional or deletional events.
- and
