NOTE:
Some of the first section was already included in the notes for last lecture (9/30/99)

Introduction to:

THE NEUTRAL THEORY OF MOLECULAR 
EVOLUTION

Prior to the 1960s, allelic variation in populations was considered 
to be small and attributable to natural selection

MOTOO KIMURA was the first to point out the importance of 
alleles that were invisible to natural selection

Refs. Kimura 1968. Evolutionary rate at the molecular level. Nature 217: 624-628.
Kimura, M. 1983 The Neutral Theory of Molecular Evolution Cambridge 
University Press

Consider new alleles brought into a population by mutation:

1. The majority of new mutations will be deleterious to fitness

2. Most of new mutations will be quickly eliminated from the 
population by natural selection

3. A small minority may be advantageous under some or all 
environments

4. Some mutations will have little or no effect on fitness and 
are effectively neutral

The probability of fixation of a neutral allele is 1/(2Ne)

Therefore, a steady substitution of one allele for another is 
expected to occur due to genetic drift alone


A new mutation that eventually becomes fixed takes 4Ne 
generations to do so

Thus fixation time is longer in large populations than 
in small ones

CONCEPT OF POPULATION "SUBSTITUTION"

Important to distinguish between 
"Mutation" and "Substitution" 
with respect to individuals and populations

MUTATION: alteration of a single copy of a gene during DNA replication

SUBSTITUTION: replacement of one DNA or protein sequence by another in a 
population or species

Most new mutations are lost, and therefore very few of them will eventually result 
in population substitutions

The continuing debate about the neutral theory concerns how the majority of the 
standing variation in natural populations is maintained

The neutral school maintains that the vast majority of molecular variation in 
natural populations can be explained by the presence of alleles that are neutral

The selectionist school maintains that a substantial variation is maintained by 
positive selection

The neutralist position has shifted over time to focus on alleles that are "nearly 
neutral" rather than strictly neutral.

Tomoko Ohta, a colleague of Kimura's has been responsible for much of the work, in 
the 70s and 80s, leading to this shift.





RATE OF GENE SUBSTITUTION:
 Defined as: the number of mutants reaching fixation per unit time.

1. Neutral mutations

If neutral mutations occur at a rate of u per gene per generation, 
then the number of mutants arising in a diploid population of size N 
is 2Nu per generation. 

 Since the probability of fixation for each of these mutations is 1/(2N), we obtain 
K, the rate of substitution of neutral alleles, 
by multiplying the total number of mutations by the probability of their fixation:
					
			          K =  2Nu/2N
                            Therefore, K = u                                

Thus, for neutral mutations, 

THE RATE OF SUBSTITUTION IS EQUAL TO THE RATE OF MUTATION 

a remarkably simple result (Kimura 1968a).  

This result can be intuitively understood by noting that, in a large population, the 
number of mutations arising every generation is high but the fixation probability of 
each mutation is low.  

In comparison, in a small population, the number of mutations arising every 
generation is low, but the fixation probability of each mutation is high.  

As a consequence, 

THE RATE OF SUBSTITUTION FOR NEUTRAL MUTATIONS IS 
INDEPENDENT OF POPULATION SIZE.

2. For advantageous mutations
	the rate of substitution is obtained by multiplying the rate of mutations by the 
	probability of fixation for such alleles 
 
	For genic selection with s > 0, we obtain
	K = 4Nsu                                  
	The rate of substitution for the case of genic selection depends on the population 
	size (N) and the selective advantage (s), as well as on the rate of mutation (u).




MOLECULAR DATA IN 
PHYLOGENETIC ANALYSIS

BACKGROUND

In recent years molecular data has increasingly been used both in 
place of, and together with, more traditional traits such as those 
relating to morphology.

In particular, DNA and amino acid sequence data are now widely 
used as they can often provide more information than other kinds of 
traits.

Different  types of sequences evolve at widely different rates. 
Therefore it is important to choose a sequence appropriate to the 
time span of interest in your study.

Illustrations here -














MOLECULAR CLOCKS

Controversial topic

It was early suggested that contrary to the situation with 
morphological and other non molecular traits, 
macromolecules may change at a roughly constant rate over long 
periods of evolutionary time.

However, DIFFERENT macromolecules clearly evolve 
at DIFFERENT rates

E.g., cytochrome C evolves very slowly,
fibrinogen evolves very fast

Graph here



 


















The MOLECULAR CLOCK HYPOTHESIS postulates that:

 For any given macro-molecule (a protein or DNA sequence) 
the rate of evolution is approximately constant over time 
in all evolutionary lineages 
(Zuckerkandl and Pauling 1965).  

This hypothesis has stimulated much interest in the use of 
macromolecular evolution studies, for two reasons:

1.  If macromolecules evolve at constant rates, they can be used to 
date species divergence times and other types of evolutionary events, 
similar to the dating of geological times using radioactive elements.  
	
Moreover, phylogenetic reconstruction is much simpler under 
constant rates than under nonconstant rates. 

2. The degree of rate variation among lineages may provide much 
insight into the mechanisms of molecular evolution (e.g., see Kimura 
1983; Gillespie 1991).

Under neutral mutation the rate of evolution is equal to the rate of 
mutation (see above).  

Therefore, if the neutral mutation hypothesis is true 
and if the rate of neutral (or nearly neutral mutation)
 in a protein has not changed with time, 
the rate of evolution in that protein should be nearly constant.  

Thus, a large change in substitution rate may indicate a large change 
in evolutionary factors. 

e.g., a large increase in the rate of evolution in a protein, in a particular lineage 
may indicate:
	a. adaptive evolution, 
	b. relaxation of functional constraints (or loss of function), 
or 	c. a large reduction in effective population size in that lineage. 
Although the concept of a molecular clock has had a strong impact 
on the study of evolution, it has always been controversial..  

There has been a wide range of views on this issue.  

In one extreme, Ochman and Wilson (1987) suggested the 
existence of a universal clock of synonymous substitution that is 
applicable to all organisms.  

In the other extreme, Goodman(1976,1981) and associates denied 
even the existence of approximate constancy.  

Also, there has been a strong controversy on whether generation 
time can have a significant effect on the rate of molecular 
evolution.  

Classical genetic studies indicated that mutation rates are more 
comparable among organisms when measured in terms of 
generation than in terms of absolute time.

For this reason, the molecular clock should run faster in organisms 
with a short generation time, for they will go through more 
generations per unit time than organisms with a long generation 
time.

The generation-time effect hypothesis.


RELATIVE RATE TESTS

Tests the molecular clock hypothesis

Does not require knowledge of divergence times

Effect of generation time
(Insert illustration here)

A LOCAL CLOCK IN MICE, RATS, AND HAMSTERS

Since the molecular clock hypothesis is controversial, the first 
question to ask is, "Does there exist a molecular clock in any 
group of organisms?" 

(Such a clock is known as a local clock.) 

The best organisms to look for the existence of a local clock are 
a group of organisms with similar physiology and life histories 
such as generation time.  

The muroid rodents (e.g., mice and rats) and their relatives 
would be such a group for which there is abundance of DNA 
sequence data.  

There have been two lines of evidence suggesting that the clock-
behavior is approximately maintained in this group of rodents.  

First, the DNA-DNA hybridization studies of Brownell (1983) 
and Catzeflis et al. (1987) revealed a constant rate among lineages 
in the murine (mouse and rat) family and the microtine (hamster) 
family.  

Second, in an analysis of nucleotide sequences using the 
relative-rate test with human sequences as references, Li et al. 
(1987a) found nearly equal rates in the mouse and rat lineages.  

We review below the study of O'hUigin and Li (1992), who used 
extensive sequence data from mice, rats, and hamsters. 

(Insert TABLE 8.1)



First, let compare the substitution rates in the mouse and rat lineages, using the 
hamster lineage as a reference.  

The number of substitutions per synonymous site (Ks) is 30.3% between mouse 
and hamster and 31.1% between rat and hamster (Table 8.1). 

The difference (0.8%) is not statistically significant because it is smaller than the 
standard error of Ks (1.0%). 

So, the synonymous rates in the mouse and rat lineages are nearly equal.  

The difference in the non-synonymous rate (KA), i.e., d = 2.9% - 2.7% = 0.2%, is 
equal to two times the approximate standard error of KA and may be considered 
statistically significant.  

Thus, the nonsynonymous rate seems to be slightly faster in the mouse lineage than 
in the rat lineage.

Second, we compare the substitution rates in the mouse (or rat) and hamster lin-
eages, using the human lineage as a reference.  

The Ks value is 53.4% between mouse and human and 52.3% between hamster and 
human, so the difference 1.1% is smaller than the approximate standard error 1.5% 
and is not statistically significant.  

Similarly, the difference between the two lineages in the rate of nonsynonymous 
substi-tution is also not significant.  

The same conclusion can be drawn when the mouse lineage is replaced by the rat 
lineage (Table 8.1). 

Thus, the mouse, rat and hamster lineages have evolved at nearly equal rates in 
terms of nucleotide substitution.

In conclusion, there appears to be an approximate molecular clock in these rodents, 
at least in terms of synonymous substitution.  
This clock may be used to date divergence times among these rodents.  

For example, since the Ks value is 18.0% between mouse and rat and is about 
31.0% between mouse-rat and hamster, the hamster lineage is estimated to have 
branched off 0.31/0.18 = 1.7 times earlier than the mouse-rat divergence.


LOWER RATES IN HUMANS THAN IN MONKEYS

There has been a long-standing controversy over the hominoid rate-slowdown 
hypothesis, which postulates that the rate of molecular evolution has become slower in 
hominoids (humans and apes) after their separation from the Old World (OW) 
monkeys.  

This hypothesis, proposed by Goodman (1961) and Goodman et al. (1971), was based 
on rates estimated from immunological distance and protein sequence data.  Sarich and 
Wilson (1967) and Wilson et al. (1977) contended that the slowdown was an artifact, 
owing to the use of an erroneous paleonto-logical estimate of the ape-human divergence 
time.  

CAUSES OF RATE VARIATION AMONG LINEAGES 
(Li, Molecular Evolution 1997)

1. The efficiency of DNA repair mechanisms may vary among 
lineages.
This hypothesis was proposed by Britten (1986) to explain the difference in 
substitution rates between the primate and rodent lineages. 

Evidence from cultured cells supports this hypothesis, 
but no in vivo data are available.

2. The generation-time effect hypothesis,

Postulates a higher rate of evolution in organisms with a short generation time than in 
organisms with a  long generation time.

Considerable evidence exists to support this hypothesis.

3. The metabolic rate hypothesis 

Postulates a higher rate of evolution for organisms with a high metabolic rate than for 
organisms with a low metabolic rate.

This hypothesis may hold for mtDNA, but whether it holds for nuclear DNA remains 
to be seen.

These three hypotheses are not mutually exclusive.