Download Keywords Lectures 7 and 8

Bio1B
Evolution section (Thomson)
Oct 9, 2003
Fall 2003
lectures 7 and 8: Genetic variation, genetic drift, and selection (ch. 21: 402-412)
• Extent of genetic variation in natural populations
• Examine the extent of genetic variation in natural populations, and understand the
concepts of the neutral, balancing selection and evolutionary lag schools to explain this
variation
• Deviations from Hardy Weinberg (HW)
• Explain the consequences of violating each of the assumptions of the HW law especially
non-random mating, genetic drift, and selection
• Genetic drift
• Understand the short term effects of genetic drift on the genetic structure of populations
and the consequences of founder effects and bottlenecks
• Selection
• Distinguish between single gene traits and quantitative traits
• Contrast the short and long term effects of balancing versus directional selection on the
genetic structure of a population for Mendelian and quantitative traits
• Describe examples of heterozygote advantage, disruptive selection, etc.
• Understand the concept and describe examples of sexual selection
• Contrast Mullerian and Batesian mimicry
• Relate observations on artificial selection to natural selection
• Describe the concept of inclusive fitness and how it relates to Darwinian selection
How much genetic variation is there in natural populations: before 1966 there were two disparate views
on the extent of overall genetic variation in natural populations.
The classical theory assumed that at nearly every locus every individual is homozygous for a wild-type
allele. In addition, each individual is heterozygous for rare deleterious alleles, and occasionally
heterozygous for a selected allele maintained in the population by balancing selection.
The balance theory in its extreme form on the other hand assumed that there was a lot of genetic
variation in populations so that most individuals will be heterozygous for alternative alleles at very
many of their loci. This genetic variation was believed to be maintained by some form of balancing
selection.
The year 1966 is important in population genetics, as it marks the use of an objective test to measure the
extent of genetic variation in populations—gel electrophoresis. The initial, and later, studies showed
that more than approximately 30% of loci (and this is an underestimate) exhibit variation in natural
populations.
So, we now know, and more recent DNA based technologies have confirmed this, that a great deal of
variation does exist in natural populations. In humans approximately 1/1,000 DNA base pairs is
1
polymorphic (referred to as a SNP—single nucleotide polymorphism). In contrast, humans differ from
chimpanzees approximately every 1/100 base pairs.
From these observations, it would seem that the balance school wins out. However, the classical theory
has been retained in terms of the so-called neutral (or neo-classical) theory.
Also to consider is that some, or much, of the variation in natural populations may represent a transient
polymorphism—the evolutionary lag school. The argument is that there will ultimately be changes in a
species ecosystem (via environmental changes or evolutionary advances by other species) and
consequently if a species is to survive it must evolve continually and rapidly to catch up to the latest
changes in its ecosystem.
neutral school: much of the genetic variation in populations is evolutionary noise, and the allelic variants
are selectively equivalent.
balance school: most variation has adaptive significance and is maintained by some form of balancing
selection.
evolutionary lag school: much of the variation in a population is transient variation, as advantageous
alleles replace other alleles. Even if an allele is selected it will take a long time to become established in
the population unless the selection is extremely strong (for example, with selection of 1% it takes 2,000
generations to fix an allele in a population, which equates to about 45,000 years for humans).
Which school is right?: we will discuss the theoretical backgrounds of these three schools in the next
lectures. There is controversy as to which is the predominant factor creating the high level of genetic
variation seen in most natural populations. While selection certainly operates, nevertheless, much
genetic variation is probably neutral. All three factors probably play an important role.
apportionment of genetic variation: initial studies of the degree of genetic differentiation of human
populations and ethnic groups using allozyme data from gel electrophoresis studies showed that most
genetic variation in humans is found within populations (85%), with the remaining variation equally
divided (7.5% each) between populations within ethnic groups, and between ethnic groups. Similar
results have been found with RFLPs, microsatellites, and HLA data.
deviations from Hardy Weinberg assumptions: the strength of the Hardy Weinberg (HW) law is that one
can deviate from the assumptions quite a bit and the data will still approximate Hardy Weinberg
proportions (HWP). The weakness of the HW test is that the deviation from the assumptions has to be
very strong in order to detect the effect of this evolutionary force, e.g., selection. Deviations from HW
assumptions involve:
(1) non-random mating, e.g., inbreeding,
(2) effects of mutation are negligible as mutation rates are low—but mutation is an important force in
creating new variation,
(3) migration is important if the migration rate is high and the two population are very distinct genetically,
(4) genetic drift due to small population size (chance effects)—genetic drift effects are important in both
small and large (but finite) populations in terms of long term effects of changes in allele frequencies
over generations due solely to drift effects (this will be discussed in a later lecture on molecular
evolution), however the finite size of a sample taken from a population is taken into account in the
2
statistical tests for HWP and finite population size itself does not cause significantly detectable
deviations from HWP,
(5) selection has to be strong to cause deviations from HWP, e.g., it can be detected with sickle cell anemia,
see below.
mutation: in and of itself does not change allele frequencies to a noticeable extent as mutation rates are
low. However, mutations are the raw material of evolution, the ultimate source of genetic variation.
Although the frequencies of mutants are initially rare, and most are lost from the population,
nevertheless some increase in frequency due to genetic drift effects and also selection (see below and
later lecture on molecular evolution).
migration: is the movement of individuals from one population into another, which can alter allele
frequencies, and if there are large genetic differences cause a statistically significant deficiency of
heterozygotes from Hardy-Weinberg expectations.
non-random mating: individuals with certain genotypes sometimes mate with one another more
commonly than would be expected on a random basis.
When like mates more often with like we term this positive assortative mating, e.g., height, IQ. Positive
assortative mating increases the proportion of homozygous individuals but does not alter the allele
frequencies.
With self-fertilizing plants the level of heterozygosity is reduced by 1/2 each generation.
Self-fertilizing plants have more homozygotes than expected under Hardy-Weinberg and often show
significant deviations from HWP.
inbreeding: mating with close relatives is another form of non-random mating.
Relatives are more likely to carry the same recessive allele for a rare recessive trait—inbreeding
increases the number of affected individuals with rare recessive traits. Marriages between first cousins
have about twice the rate of birth defects as random matings.
genetic drift: (chance effects) random change in the frequency of alleles at a locus.
short term genetic drift effects: cause changes in allele frequencies, both in small and large populations.
The change in allele frequency due to genetic drift in a small population appears larger, statistical
testing can determine whether changes are larger than expected by chance.
As an example, an allele frequency change in a population of size 50 from p = 0.5 to 0.56 in 1
generation is within the range expected by drift, whereas in a population of size 5,000 such a change
would be much too large to be due solely to drift effects.
Think in terms of tossing a coin, if you tossed a coin 50 times you would expect 25 heads and 25 tails,
but due to the finite number of tosses would not be surprised to observe 28 heads and 22 tails (the
change given for the allele frequency above, 56% heads). In fact, 95% of the results of tossing a coin 50
times would fall within the range from 30 heads (20 tails) to 20 heads (30 tails).
If you toss a coin 5,000 times 95% of the results would fall within the range from 2,550 heads (2,450
tails) to 2,450 heads (2,550 tails), and an observed outcome of 2,800 heads (2,200 tails) (56% heads
when 50% expected) is well outside the range expected by chance (this outcome would occur by chance
less than 1 in a million times).
3
founder effect: the change in allele frequencies when a new colony is formed by a very small number of
founding individuals from a larger population.
Alleles found in the general population may be absent from the founder population, e.g., of an island
population, or a religious isolate, such as the Amish and Hutterite populations in the United States.
On the other hand, by chance rare alleles from the general population may be more frequent in the
founder population, e.g., a form of dwarfism in the Amish.
bottleneck: drastic reduction in population size due, e.g., to over fishing or a natural disaster such as a
hurricane, will lead to changes in allele frequencies.
Selection
Why are there so few examples of selection?: selection has to be very strong to detect it via deviations
from HW, or direct measurement of differential fitness or fertility. On the other hand, even relatively
weak selection, e.g., 1%, can have a major effect on long term evolution.
malaria and selection: one strong selective pressure, which has led to polymorphism of a number of genes
in humans is malaria. Some examples related to Plasmodium falciparum are:
1. Variants at the hemoglobin β gene, notably Hb-S (homozygotes have sickle cell anemia, also see
below), and also there is another variant Hb-C where homozygotes have a milder anemia, in both cases
there is heterozygote advantage in malarial environments. Both alleles have relatively high frequencies
in parts of Africa and the Mediterranean.
2. Both α and β thallasemias are found where there is a reduced amount or no production of either the α
or β chains of hemoglobin (the hemoglobin molecule is a tetramer of 2 α and 2 β chains). Again there is
heterozygote advantage in malarial environments, affected individuals have mild to severe anemia, and
these polymorphisms are relatively frequent in the Mediterranean and Asia.
3. An X-linked recessive trait, G6PD (glucose-6 phosphate-dehydrogenase) deficiency also provides
protection against malaria; life threatening anemias can occur under certain conditions, e.g., eating fava
beans or taking certain drugs. Different mutational origins led to the polymorphisms seen in the
Mediterranean and Asia.
4. HLA associations: in the Gambia, the HLA alleles B53 and DR13 are relatively frequent and have
been shown to provide protection from malaria.
AIDS and selection: two major examples of AIDS causing selection on common polymorphisms in
humans are as follows (in these cases AIDS has not led to these polymorphisms, as in the malaria
selection examples above, since it is much too recent a phenomenon):
1. Individuals homozygous for the CCR5 ∆32 mutation are protected from AIDS; this polymorphism is
largely confined to individuals with European ancestry. One possible explanation for its relatively high
frequency is that it may have provided protection from smallpox; the myxoma virus, which is a member
of the poxvirus family, also exploits CCR5 for infection.
4
2. HLA: (a) specific HLA alleles provide protection, and (b) people heterozygous for a number of HLA
genes (there are a total of 6 classical HLA genes) live longer after infection (presumably these
individuals have a wider immune response).
other possible examples of heterozygote advantage: there is some evidence with the following diseases
of an advantage to heterozygotes:
(a) With phenylketonuria (PKU) excess of the amino acid phenylalanine in carriers (and those with the
disease) inactivates ochratoxin A a fungal toxin that causes miscarriages. PKU carriers have a lower
than average incidence of miscarriages. PKU is most frequent in Scotland and Ireland, and this may
relate to times of famine when people were forced to eat moldy grains to survive.
(b) There is some evidence from World War II that healthy relatives of children who had Tay Sachs
(hence the relatives are more likely to be carriers) did not contract tuberculosis as often as others in
crowded ghetto conditions.
(c) In the case of cystic fibrosis, there may be protection from diarrheal diseases such as cholera and
typhus; carriers have less chloride channels in intestinal cells, blocking cholera toxin entry, or
Salmonella typhi, which uses the CFTR protein as a receptor.
relative fitness: the average number of offspring produced by individuals with a certain genotype, relative
to the number produced by individuals with other genotypes.
heterozygote advantage: the heterozygote has a higher relative fitness than both homozygotes, this leads
to a balanced polymorphism (see sickle cell anemia example below).
sickle cell anemia: Individuals with sickle cell trait (AS heterozygotes) are more resistant to malaria than
individuals who are homozygous AA; SS individuals are also more resistant to malaria than AA
individuals, but they have sickle cell anemia which drastically reduces their fitness. The relative fitness
values in a malarial environment have been estimated as:
genotype
relative fitness
AA
0.88
AS
1.0
SS
0.14
The selection in this case is very strong and causes detectable deviations from HWP.
Genotypes
AA
AS
SS
Observed no. among adults
9,365
2,993
29
Total: 12,387
Observed freq. among adults
0.756
0.242
0.002
Expected no. among adults (HWP)
9,523
2,676
188
Total: 12,387
Expected freq. among adults (HWP)
0.769
0.216
0.015
frequency dependent selection: is another cause leading to balanced polymorphism. The reproductive
success of a phenotype depends on its frequency, with higher fitness associated with lower frequency of
the phenotype in the population, e.g., 'right-mouthed' and 'left-mouthed' species of cichlid fish.
5
balanced polymorphism: 2 or more alleles are maintained in a population due to selection. Both
heterozygote advantage and frequency dependent selection are examples of balancing selection, they
both lead to a stable polymorphic equilibrium state.
directional selection: leads to fixation of an allele. An example would be relative fitnesses
genotype
AA
AB
BB
relative fitness
1.0
0.9
0.8
which would lead to fixation of the A allele.
Industrial melanism: is a term used to describe the evolutionary process by which initially light
colored organisms become dark as a result of natural selection in an industrial environment.
The process, which is common among moths that rest on tree trunks, takes place because the
dark organisms are better concealed from their predators in habitats that have been darkened
by soot and other forms of industrial pollution.
lead tolerance: plants able to grow on lead-rich soil are found in association with mines less than
a century old. Populations of the grass Agrostis tenuis are clearly able to evolve through
changing their genotypic frequencies quickly when the environment demands it. Similar rapid
changes have now been documented for other populations of organisms.
quantitative trait: determined by a large number of genes each of small effect and environmental factors,
e.g., height and weight.
stabilizing selection: selection acts to eliminate both extremes from an array of phenotypes (this is the
quantitative equivalent of balancing selection for a single gene trait).
directional selection: selection acts to eliminate one extreme from an array of phenotypes.
disruptive selection: (also called diversifying selection) selection acts to eliminate the intermediate type,
favoring both extremes (this is the quantitative equivalent of heterozygote disadvantage for a single
gene trait).
artificial selection: selective breeding, carried out by humans, to alter a population. Artificial
selection in domesticated animals and plants demonstrates the ability to change the gene pool
in a small amount of time if the selection is strong. Some examples are:
(1) Selection of cattle for meat quantity and quality.
(2) Selection of cows for milk yield.
(3) selection of sheep for meat or for specific types of fleece to make into wool.
(4) Selection of horses for speed, strength, or racing ability.
(5) Selection of specific features of plants. For example broccoli, cauliflower, cabbages, kale,
and Brussels sprouts all have a common ancestor in one species of wild mustard. By selecting
different parts of the plant to accentuate, breeders have obtained these divergent results.
(6) Selection for increased yield and conformity of size and shape for mechanical harvesting
and durability for travel to market, e.g., tomato.
6
(7) Selection of all the different breeds of dogs has taken place over approximately 5,000 years
with specializations for different features such as hunting ability, etc.
sexual selection: differential reproduction owing to variation in the ability to obtain mates, i.e.,
selection on mating behavior, either through choice by members of one sex (usually females)
of certain members of the other sex, e.g., the showy plumage of many male birds; or through
competition among members of one sex (usually males) for access to members of the other
sex, e.g., male competition can take the form of direct fighting.
Batesian mimicry: a palatable species masquerades as an unpalatable one.
Mullerian mimicry: several unpalatable species converge in appearance, each species gaining
protection from its similarity to the other one.
sociobiology: systematic study of the biological basis of all social behavior.
kin selection: a phenomenon of inclusive fitness, used to explain altruistic behavior between
related individuals.
inclusive fitness: the fitness of a gene or genotype measured by its effect on survival and
reproduction both of the organism bearing it, and of the genes, identical by descent, borne by
the individual's relatives.
altruism: the aiding of another at one's own risk or expense.
7
Problems
1. In a population of self fertilizing plants of size 100, the genotypes at a codominant locus are 20 AA, 60
AB, and 20 BB individuals. What will the genotype counts be in the next generation (assume a
population size of 100 again)? What would they be if there was random mating?
2. In Africa in an area where malaria is prevalent, 1,000 newborns and 1,000 adults are typed for the
sickle-cell polymorphism:
Genotypes
AA
AS
SS
Newborns (O)
780
204
16
Total: 1,000
Adults (O)
769
229
2
Total: 1,000
For both samples, determine the allele frequencies of the A and S alleles, and test fit to HWP. Comment
on any differences, or not, between the allele frequencies and the test of HWP in the two samples.
3. For the following selection schemes, where the relative fitness values for each genotype are given, state
whether one or the other allele will be fixed (and if so which allele) or whether a balanced
polymorphism will be maintained:
AA
AB
BB
(a)
0.8
0.9
1.0
(b)
0.2
1.0
0.6
(c)
1.2
1.1
1.0
8
Answers to Problems
1. With selfing, the frequency of heterozygotes is reduced by 1/2 each generation, with the remaining 1/2
divided equally between the two homozygous classes. The genotype counts in the next generation will
be 35 AA, 30 AB, and 35 BB. If there had been random mating, the genotype counts would be 25 AA,
50 AB, and 25 BB.
2. Among the newborns, f(A) = p = 0.8820, f(S) = q = 0.1220,
and in the adults f(A) = 0.8835, f(S) = 0.1165.
Newborns (O) 780
Newborns (E) 777.9
204
208.2
16
13.9
Total: 1,000
Total: 1,000
Adults (O)
Adults (E)
229
205.9
2
13.5
Total: 1,000
Total: 1,000
769
780.6
The adult sample shows deviation from HWP, while the newborn sample does not. This reflects the fact
that at birth (pre-selection) the genotype frequencies are expected to be in HWP, whereas after selection
has acted (malaria and sickle-cell anemia), the adult population shows significant deviation from HWP.
The fact that the allele frequencies are very similar in newborns and adults may indicate that the
population is close to equilibrium.
3. (a) fixation of the B allele, loss of the A allele
(b) stable polymorphism
(c) fixation of the A allele, loss of the B allele
9