120
4 Molecular Evolution
tion in selection can lead to marked differences in
the use of synonymous codons [244], and that
molecular evolution is retarded by the existence
of codon preference [210, 212].
Amongst the mutations leading to amino acid
substitution, the stabilizing effect of selection
favours conservative exchanges, i.e. those that
change only minimally the physicochemical properties that are important for protein structure and
function. Because the spatial structure of proteins
is generally maintained during evolution, even
when most of the amino acids have been
exchanged, the properties of the amino acids that
determine the spatial structure clearly playa significant role. These properties include hydrophobicity, charge, influence on the secondary
structure of the polypeptide chains and the spatial
requirements of the side-chains. On the basis of
such criteria, various indices of chemical similarity have been formulated and the amino acids
classified in different ways (see footnotes to Table 4.3). It has been shown in many investigations
that conservative exchanges' are more frequent
than would be expected for random substitution
[87, 210, 212]. For example, in the distribution
into polarity classes of eight protein families,
88 % of substitutions in the inner regions of the
molecules are conservative compared with 76 %
for amino acids on the surface, i.e. there were no
changes of polarity class; non-conservative
exchanges would have led internally to a decrease
and exter~ally to an increase in polarity. According to the genetic code, each amino acid substitution requires one, two or three nucleotide substitutions, and each amino acid pair thus corresponds to a "minimal mutation distance"
(MMD). If the genetic code were itself optimized
for minimal mutational effects, then conservative
exchange would be more frequent after one-step
than after two- or three-step mutations; however
this is not the case. Even conservative amino acid
substitutions can have lethal consequences. The
substitution 104-Glu to Asp in human triosephosphate isomerase makes the enzyme thermolabile; the resulting enzyme deficiency in homozygous carriers of the allele causes haemolytic
anaemia and neuromuscular dysfunction [83].
The exchange of a glycine codon for a serine
codon in the gart gene complex of Drosophila
melanogaster leads to a complete breakdown of
purine synthesis [168].
Nucleotide substitutions that place stop codons
within coding sequences (nonsense mutations)
are unquestionably always selectively negative.
For example, ~-thalassaemia, a heritable disease
in humans which leads to reduced synthesis of the
haemoglobin ~-chain, can be caused by mutation
of the 39-Gln codon CAG to the stop codon TAG
[144]. The Adh nB mutant of Drosophila melanogaster codes for a 2-kDa smaller alcohol dehydrogenase that has only 1 % of the normal activity,
due mostly to enhanced degradation; the cause
here is the mutation 235-TGG (Trp) to TGA
(stop) [260]. Amongst the more than 100 known
mutations of the unc-54 gene of the nematode
Caenorhabditis elegans there are four mutations
of the sort CAA (GIn) to TAA (stop), and one
mutation of CAG (GIn) to TAG (stop) [91]. The
fatal consequences of a nonsense mutation can be
mitigated at least temporarily, by the presence of
a nonsense-suppressor tRNA which can read
through the stop codon. The deletion or insertion
of one or two base pairs in a coding sequence
leads to a shift in the reading frame (frame-shift
mutation), usually with far-reaching consequences. Thus, the b-globin gene of the anubis ape,
Papio doguera, is not expressed because the insertion of a nucleotide in codon 55 shifts the reading
frame to the extent that a stop codon appears at
position 59 [207]. The deletion or insertion of one
or more complete triplets leaves the reading
frame unaltered and is more often than not without negative consequences.
Nucleotide substitutions, deletions or insertions in non-coding sequences are often selectively neutral, but they may have far-reaching
phenotypic consequences. For example, two Adh
mutants in Drosophila melanogaster show a nucleotide substitution at the intronlexon border of
the second intron; instead of GT ... AG, one has
GT ... GG and the other has GC ... AG. Consequently the mutants have no alcohol dehydrogenase (ADH) due to deficient mRNA maturation.
One case of ~-thalassaemia was the result of a
mutation in the TATA box; ATACAA was found
instead of ATAAAA, and the transcription of the
~-globin gene was drastically reduced [399].
Another heritable human disease, the persistence of the foetal Ay haemoglobin in adults,
could be traced to a mutation in the upstream
CCAAT box; the nucleotide substitution AACCAAT to GACCAAT changes the expression of
the Ay haemoglobin gene [129]. The substitution
rate in coding and non-coding DNA regions
influences the base composition and, conversely,
is influenced by it. The AT-rich isochores of the
cold-blooded vertebrates have changed to GCrich ones in the warm-blooded birds and mammals [26]. A comparison of 42 protein-coding
genes between man and the rat demonstrated a
4 Molecular Evolution
tion in selection can lead to marked differences in
the use of synonymous codons [244], and that
molecular evolution is retarded by the existence
of codon preference [210, 212].
Amongst the mutations leading to amino acid
substitution, the stabilizing effect of selection
favours conservative exchanges, i.e. those that
change only minimally the physicochemical properties that are important for protein structure and
function. Because the spatial structure of proteins
is generally maintained during evolution, even
when most of the amino acids have been
exchanged, the properties of the amino acids that
determine the spatial structure clearly playa significant role. These properties include hydrophobicity, charge, influence on the secondary
structure of the polypeptide chains and the spatial
requirements of the side-chains. On the basis of
such criteria, various indices of chemical similarity have been formulated and the amino acids
classified in different ways (see footnotes to Table 4.3). It has been shown in many investigations
that conservative exchanges' are more frequent
than would be expected for random substitution
[87, 210, 212]. For example, in the distribution
into polarity classes of eight protein families,
88 % of substitutions in the inner regions of the
molecules are conservative compared with 76 %
for amino acids on the surface, i.e. there were no
changes of polarity class; non-conservative
exchanges would have led internally to a decrease
and exter~ally to an increase in polarity. According to the genetic code, each amino acid substitution requires one, two or three nucleotide substitutions, and each amino acid pair thus corresponds to a "minimal mutation distance"
(MMD). If the genetic code were itself optimized
for minimal mutational effects, then conservative
exchange would be more frequent after one-step
than after two- or three-step mutations; however
this is not the case. Even conservative amino acid
substitutions can have lethal consequences. The
substitution 104-Glu to Asp in human triosephosphate isomerase makes the enzyme thermolabile; the resulting enzyme deficiency in homozygous carriers of the allele causes haemolytic
anaemia and neuromuscular dysfunction [83].
The exchange of a glycine codon for a serine
codon in the gart gene complex of Drosophila
melanogaster leads to a complete breakdown of
purine synthesis [168].
Nucleotide substitutions that place stop codons
within coding sequences (nonsense mutations)
are unquestionably always selectively negative.
For example, ~-thalassaemia, a heritable disease
in humans which leads to reduced synthesis of the
haemoglobin ~-chain, can be caused by mutation
of the 39-Gln codon CAG to the stop codon TAG
[144]. The Adh nB mutant of Drosophila melanogaster codes for a 2-kDa smaller alcohol dehydrogenase that has only 1 % of the normal activity,
due mostly to enhanced degradation; the cause
here is the mutation 235-TGG (Trp) to TGA
(stop) [260]. Amongst the more than 100 known
mutations of the unc-54 gene of the nematode
Caenorhabditis elegans there are four mutations
of the sort CAA (GIn) to TAA (stop), and one
mutation of CAG (GIn) to TAG (stop) [91]. The
fatal consequences of a nonsense mutation can be
mitigated at least temporarily, by the presence of
a nonsense-suppressor tRNA which can read
through the stop codon. The deletion or insertion
of one or two base pairs in a coding sequence
leads to a shift in the reading frame (frame-shift
mutation), usually with far-reaching consequences. Thus, the b-globin gene of the anubis ape,
Papio doguera, is not expressed because the insertion of a nucleotide in codon 55 shifts the reading
frame to the extent that a stop codon appears at
position 59 [207]. The deletion or insertion of one
or more complete triplets leaves the reading
frame unaltered and is more often than not without negative consequences.
Nucleotide substitutions, deletions or insertions in non-coding sequences are often selectively neutral, but they may have far-reaching
phenotypic consequences. For example, two Adh
mutants in Drosophila melanogaster show a nucleotide substitution at the intronlexon border of
the second intron; instead of GT ... AG, one has
GT ... GG and the other has GC ... AG. Consequently the mutants have no alcohol dehydrogenase (ADH) due to deficient mRNA maturation.
One case of ~-thalassaemia was the result of a
mutation in the TATA box; ATACAA was found
instead of ATAAAA, and the transcription of the
~-globin gene was drastically reduced [399].
Another heritable human disease, the persistence of the foetal Ay haemoglobin in adults,
could be traced to a mutation in the upstream
CCAAT box; the nucleotide substitution AACCAAT to GACCAAT changes the expression of
the Ay haemoglobin gene [129]. The substitution
rate in coding and non-coding DNA regions
influences the base composition and, conversely,
is influenced by it. The AT-rich isochores of the
cold-blooded vertebrates have changed to GCrich ones in the warm-blooded birds and mammals [26]. A comparison of 42 protein-coding
genes between man and the rat demonstrated a