164
4 Molecular Evolution
years; this also agrees with the geological data on
the separation of the two continents.
The relative constancy of the rate of evolution,
together with the great extent of genetic polymorphism, was one of the main arguments in the
development of the neutral theory of evolution.
In connection with the controversy between neutralism and selectionism, the concept of a molecular clock was, therefore, questioned from the very
beginning. Criticism was also levelled at certain
results obtained using the molecular clock
approach, in particular those related to the evolution of man and the anthropoids (p.169). As
arguments against the claimed constancy in the
rates of molecular evolution, examples have been
quoted where varying rates were observed in different evolutionary lines or at different moments
in the evolution of a line. Up to fivefold differences in the rates of synonymous substitution
were cited for different animal groups, with the
lowest value of about 1.3 . 10- 9 for the primates
and birds, and the highest value of about 6.6 . 10- 9
for the rodents, sea urchins and Drosophila
[38, 59, 373]. The rates of synonymous substitution can vary by as much as twofold within a species [373].
There are also many results showing variable
rates of evolution within an evolutionary line. For
example, the molecular phylogenetic trees constructed by Goodman and colleagues by use of
the maximum parsimony method, consistently
show strongly variable rates of amino acid
exchange, with an extreme difference of 1 :32
[141, 142]. Kimura, however, challenged these
results, quoting deficiencies in the maximum parsimony method and errors in the choice of separation time used for calibration of the phylogenetic
trees [210,212]. Goodman also found that individual proteins were not usable measures of time,
but considered that using the combined data from
several proteins, "the hypothesized protein clock
does not perform too badly" [141]. A comparison
of the globin genes of sheep and goats showed
that the rates of amino acid-exchanging substitutions increased after each gene duplication; this
increase is especially distinct in functionally less
critical parts of the sequence, and is therefore
related more to the reduced effect of eliminating
selection than to an increase in positive selection
pressure [242]. An increased rate of evolution due
to positive selection is shown by the ruminant
lysozymes, which have assumed the function of a
digestive enzyme [194]. A variation in the substitution rate may therefore be explained by
changes in the fraction of nucleotide positions
that are variable without negative consequences
during different periods of evolution [313]. The
rate of substitution is dependent upon the base
composition of the DNA region (isochore) and is
lower in GC-rich regions [444]. Thus, constant
rates of substitution are only to be expected so
long as the gene remains in a stable state, i.e. the
base composition of its surroundings does not
change. Non-stable genes show significantly
higher rates of substitution [356].
Because the molecular clock is a probabilistic
and not a metronomic clock, there is already theoretically a minimum standard deviation of size
yM, where M is the number of evolutionary
events used for measuring time, e.g. the number
of amino acid exchanges or nucleotide substitutions. In reality, the standard deviation of the
molecular clock is found to be two- to fourfold
larger [210, 212, 440]. This increased variability is
also compatible with the neutral theories [404].
First, the rare, significantly selection-positive
substitutions will be fixed at an increased rate.
Second, in the course of, and as the result of, the
evolution of a protein, the proportion pn of the
neutral evolutionary changes can vary. Finally,
the further developed neutral theory assumes
that whether a certain substitution behaves neutrally is dependent upon the population size Ne
(p. 140); changes in the population size, which of
course occur frequently during evolution, will
therefore lead to changes in the rate of evolution
[210, 212]. Because, according to the neutral theory, the greater and more variable the influence
of selection, the greater will be the variation in
the rate of molecular evolution, synonymous substitutions confer a particular reliability on a molecular clock; however, due to their high rates they
may be used only in the most recent phases of
evolution.
4.7 Some Results of Molecular Research
into Evolutionary Relationships
Biochemical characters vary in their suitability
for the different tasks of biological systematics.
They are not basically superior to other characters for species diagnosis, i.e. for determining
whether an animal belongs to a particular species
or subspecies, but are often the last resort if, for
example, insufficient morphological characters
are available. It is mostly only possible with the
help of molecular characters to recognize whether
4 Molecular Evolution
years; this also agrees with the geological data on
the separation of the two continents.
The relative constancy of the rate of evolution,
together with the great extent of genetic polymorphism, was one of the main arguments in the
development of the neutral theory of evolution.
In connection with the controversy between neutralism and selectionism, the concept of a molecular clock was, therefore, questioned from the very
beginning. Criticism was also levelled at certain
results obtained using the molecular clock
approach, in particular those related to the evolution of man and the anthropoids (p.169). As
arguments against the claimed constancy in the
rates of molecular evolution, examples have been
quoted where varying rates were observed in different evolutionary lines or at different moments
in the evolution of a line. Up to fivefold differences in the rates of synonymous substitution
were cited for different animal groups, with the
lowest value of about 1.3 . 10- 9 for the primates
and birds, and the highest value of about 6.6 . 10- 9
for the rodents, sea urchins and Drosophila
[38, 59, 373]. The rates of synonymous substitution can vary by as much as twofold within a species [373].
There are also many results showing variable
rates of evolution within an evolutionary line. For
example, the molecular phylogenetic trees constructed by Goodman and colleagues by use of
the maximum parsimony method, consistently
show strongly variable rates of amino acid
exchange, with an extreme difference of 1 :32
[141, 142]. Kimura, however, challenged these
results, quoting deficiencies in the maximum parsimony method and errors in the choice of separation time used for calibration of the phylogenetic
trees [210,212]. Goodman also found that individual proteins were not usable measures of time,
but considered that using the combined data from
several proteins, "the hypothesized protein clock
does not perform too badly" [141]. A comparison
of the globin genes of sheep and goats showed
that the rates of amino acid-exchanging substitutions increased after each gene duplication; this
increase is especially distinct in functionally less
critical parts of the sequence, and is therefore
related more to the reduced effect of eliminating
selection than to an increase in positive selection
pressure [242]. An increased rate of evolution due
to positive selection is shown by the ruminant
lysozymes, which have assumed the function of a
digestive enzyme [194]. A variation in the substitution rate may therefore be explained by
changes in the fraction of nucleotide positions
that are variable without negative consequences
during different periods of evolution [313]. The
rate of substitution is dependent upon the base
composition of the DNA region (isochore) and is
lower in GC-rich regions [444]. Thus, constant
rates of substitution are only to be expected so
long as the gene remains in a stable state, i.e. the
base composition of its surroundings does not
change. Non-stable genes show significantly
higher rates of substitution [356].
Because the molecular clock is a probabilistic
and not a metronomic clock, there is already theoretically a minimum standard deviation of size
yM, where M is the number of evolutionary
events used for measuring time, e.g. the number
of amino acid exchanges or nucleotide substitutions. In reality, the standard deviation of the
molecular clock is found to be two- to fourfold
larger [210, 212, 440]. This increased variability is
also compatible with the neutral theories [404].
First, the rare, significantly selection-positive
substitutions will be fixed at an increased rate.
Second, in the course of, and as the result of, the
evolution of a protein, the proportion pn of the
neutral evolutionary changes can vary. Finally,
the further developed neutral theory assumes
that whether a certain substitution behaves neutrally is dependent upon the population size Ne
(p. 140); changes in the population size, which of
course occur frequently during evolution, will
therefore lead to changes in the rate of evolution
[210, 212]. Because, according to the neutral theory, the greater and more variable the influence
of selection, the greater will be the variation in
the rate of molecular evolution, synonymous substitutions confer a particular reliability on a molecular clock; however, due to their high rates they
may be used only in the most recent phases of
evolution.
4.7 Some Results of Molecular Research
into Evolutionary Relationships
Biochemical characters vary in their suitability
for the different tasks of biological systematics.
They are not basically superior to other characters for species diagnosis, i.e. for determining
whether an animal belongs to a particular species
or subspecies, but are often the last resort if, for
example, insufficient morphological characters
are available. It is mostly only possible with the
help of molecular characters to recognize whether