122
4 Molecular Evolution
The duplication of a complete gene, including
the transcription signals, gives two identical genes
or, in the case of repeated duplication, a series of
consecutively arranged identical genes (a tandem
cluster), all of which initially form the same gene
product. Processes like this, which lead to the
formation of mUlti-gene families have occurred
frequently in molecular evolution. If only the
coding region, or a part of it, is reproduced (an
internal duplication), then the resulting protein
displays an internal periodicity with two or more
homologous sections; this situation is detectable
by the previously described methods even when
the sequence agreement has been partly eliminated by amino acid substitution. The symmetry
and periodicity in the spatial structure of many
proteins also points to evolution through gene
multiplication, although convergence by the
occurrence of similar secondary structures cannot
be excluded [307]. Internal periodicities of proteins are wide spread (Table 4.5). In some cases, it
has been shown that the multiply occurring protein regions themselves have internal periodicities
produced by the multiplications of even shorter
sequences. This observation leads to the conclusion that the original (primordial) polypeptides
were short, and only by multiple duplications
were the greater lengths of the present proteins
reached. This is particularly clearly seen in collagens, which are largely made up of several
hundred tripeptide sequences.
4.2.3 Gene Fusion and Exon Shuming
If, between two genes, the region containing the
termination and initiation signals is removed by
deletion then a fusion product is formed. Good
examples of gene fusion as a mechanism of molecular evolution are to be found in the enzymes of
fatty acid synthesis. This always requires seven
catalytic and one acyl-carrier function. E. coli and
higher plants accordingly have seven monofunctional proteins; in yeast, the same functions
are performed by two multi-functional proteins,
and in vertebrates they are performed by just one
protein. The genes for the multi-functional proteins are unquestionably the result of the fusion of
the genes for the mono-functional molecules; the
different orders of the domains indicate that independent fusions occurred in the fungi and the vertebrates [268]. The fusion of neighbouring genes
of a gene family is by no means a rare phenomenon. For example, a series of haemoglobin
anomalies have arisen by the fusion of different
globin chains: Hb-Lepore from 6 and Band HbKenya from Ay and B; there is also a form of athalassaemia that involves fusion of the two aglobin genes [275].
The fact that domains with coincident structures are often found in different proteins has led to
the idea of new combinations of gene segments
arising by exon shutlling. An oft-quoted example
is the nucleotide-binding "Rossmann fold" which
is found, for example, in various dehydrogenases [glyceraldehyde-3-phosphate dehydrogenase
(GAPDH), lactate dehydrogenase (LDH), malate
dehydrogenase (MDH), and ADH] , phosphoglycerate kinases and glycogen phosphorylases
[276]. Despite great similarity in their spatial
structure, no significant sequence agreement is
found between the nucleotide-binding regions; in
this case the possibility of convergent evolution
can certainly not be excluded. There are, however, many clear examples of exon shuffling.
Thus, a sequence of 45 amino acids, which is
repeated eight times in the epidermal growth factor (EGF), is found in many other proteins with
completely different functions, e.g. in the complement component C9 and urokinase, in the
blood-clotting factors VII, IX and X and in
thrombomodulin, in the LDL receptor and the
thyroid gland peroxidase, and in the product of
the notch locus of Drosophila and the lin-12 gene
of Caenorhabditis. "Kringle" structures and
"zinc" fingers are also widely found [95]. The best
examples of exon shuffling are provided by the
plasma proteases of blood clotting, fibrinolysis
and the complement cascade of mammals (see
Fig.3.6, p.91). Here, the non-catalytic regions
are made up of different combinations of
"kringle", "zinc finger", growth factor and Ca 2 +binding domains, homologues of which are found
in fibronectin, EGF precursors and the LDL
receptor. During the evolution of the plasma proteases, these sequences were inserted between
the signal sequence and the activation peptide of
a primitive trypsin-like proteinase and were subsequently partly duplicated and translocated.
4.2.4 Transposition of DNA Sequences
Systematic investigations of the transposition of
mobile elements and the phenotypic consequences of such genetic events have been carried out,
in particular, with Drosophila and the nematode
Caenorhabditis elegans [74, 367]. The spread of
transposition variants in a population can be
described by particular mathematical models
4 Molecular Evolution
The duplication of a complete gene, including
the transcription signals, gives two identical genes
or, in the case of repeated duplication, a series of
consecutively arranged identical genes (a tandem
cluster), all of which initially form the same gene
product. Processes like this, which lead to the
formation of mUlti-gene families have occurred
frequently in molecular evolution. If only the
coding region, or a part of it, is reproduced (an
internal duplication), then the resulting protein
displays an internal periodicity with two or more
homologous sections; this situation is detectable
by the previously described methods even when
the sequence agreement has been partly eliminated by amino acid substitution. The symmetry
and periodicity in the spatial structure of many
proteins also points to evolution through gene
multiplication, although convergence by the
occurrence of similar secondary structures cannot
be excluded [307]. Internal periodicities of proteins are wide spread (Table 4.5). In some cases, it
has been shown that the multiply occurring protein regions themselves have internal periodicities
produced by the multiplications of even shorter
sequences. This observation leads to the conclusion that the original (primordial) polypeptides
were short, and only by multiple duplications
were the greater lengths of the present proteins
reached. This is particularly clearly seen in collagens, which are largely made up of several
hundred tripeptide sequences.
4.2.3 Gene Fusion and Exon Shuming
If, between two genes, the region containing the
termination and initiation signals is removed by
deletion then a fusion product is formed. Good
examples of gene fusion as a mechanism of molecular evolution are to be found in the enzymes of
fatty acid synthesis. This always requires seven
catalytic and one acyl-carrier function. E. coli and
higher plants accordingly have seven monofunctional proteins; in yeast, the same functions
are performed by two multi-functional proteins,
and in vertebrates they are performed by just one
protein. The genes for the multi-functional proteins are unquestionably the result of the fusion of
the genes for the mono-functional molecules; the
different orders of the domains indicate that independent fusions occurred in the fungi and the vertebrates [268]. The fusion of neighbouring genes
of a gene family is by no means a rare phenomenon. For example, a series of haemoglobin
anomalies have arisen by the fusion of different
globin chains: Hb-Lepore from 6 and Band HbKenya from Ay and B; there is also a form of athalassaemia that involves fusion of the two aglobin genes [275].
The fact that domains with coincident structures are often found in different proteins has led to
the idea of new combinations of gene segments
arising by exon shutlling. An oft-quoted example
is the nucleotide-binding "Rossmann fold" which
is found, for example, in various dehydrogenases [glyceraldehyde-3-phosphate dehydrogenase
(GAPDH), lactate dehydrogenase (LDH), malate
dehydrogenase (MDH), and ADH] , phosphoglycerate kinases and glycogen phosphorylases
[276]. Despite great similarity in their spatial
structure, no significant sequence agreement is
found between the nucleotide-binding regions; in
this case the possibility of convergent evolution
can certainly not be excluded. There are, however, many clear examples of exon shuffling.
Thus, a sequence of 45 amino acids, which is
repeated eight times in the epidermal growth factor (EGF), is found in many other proteins with
completely different functions, e.g. in the complement component C9 and urokinase, in the
blood-clotting factors VII, IX and X and in
thrombomodulin, in the LDL receptor and the
thyroid gland peroxidase, and in the product of
the notch locus of Drosophila and the lin-12 gene
of Caenorhabditis. "Kringle" structures and
"zinc" fingers are also widely found [95]. The best
examples of exon shuffling are provided by the
plasma proteases of blood clotting, fibrinolysis
and the complement cascade of mammals (see
Fig.3.6, p.91). Here, the non-catalytic regions
are made up of different combinations of
"kringle", "zinc finger", growth factor and Ca 2 +binding domains, homologues of which are found
in fibronectin, EGF precursors and the LDL
receptor. During the evolution of the plasma proteases, these sequences were inserted between
the signal sequence and the activation peptide of
a primitive trypsin-like proteinase and were subsequently partly duplicated and translocated.
4.2.4 Transposition of DNA Sequences
Systematic investigations of the transposition of
mobile elements and the phenotypic consequences of such genetic events have been carried out,
in particular, with Drosophila and the nematode
Caenorhabditis elegans [74, 367]. The spread of
transposition variants in a population can be
described by particular mathematical models