Transcript Lecture PPT - Carol Lee Lab

Constraints on Natural Selection
Natural Selection does not always
proceed in the expected direction,
because several forces could interfere
with the action of Natural Selection
Constraints on Natural Selection
• Genetic Variation: Selection can only act on existing genetic
variation (we talked about this last lecture)
• Phylogenetic Inertia (Historical Constraints): can only build on
what is there (hard to make wings without appendages)
• Pleiotropy: one gene might affect more than one trait. So if you
alter a gene, it could have multiple effects. So you might not be
able to alter that gene
• Linkage: alleles close together on a chromosome could share a
fate just do to physical proximity and undergo selection and be
inherited as a unit
• Genetic Drift will interfere with the action of Natural Selection in
small populations
Constraints on Natural Selection
• Phylogenetic Inertia (Historical Constraints): can
only build on what is there (hard to make wings
without appendages)
• Physical Constraints (anatomical constraints): a
structural or anatomical feature could prevent
a modification of a trait (brain might not be
able to grow if body cannot grow)
Physical Constraint
Developmental constraint
Constraint in Body Plan
If body size increases, brain size has to increase
If a larger eye evolves, need a bigger socket (the
socket itself is not the target of selection)
Analogy: the Spandrels of San Marco
Gould & Lewontin:
The metaphor:
The spandrels of San Marco
San Marco Cathedral, Venice
Gould & Lewontin on
Physical Constraint:
The spandrels of San Marco
might not have been created
for a reason, but might
simply be a by product due
to the creation of arches
San Marco Cathedral, Venice
Physical Constraints (examples)
•
•
•
•
Limits to brain size due to head size
Limits to eye size due to space on face
Limits to finger number due to hand size
Limits to body size on land
Constraints on Natural Selection
• Genetic Variation: Selection can only act on existing genetic
variation (we talked about this last lecture)
• Phylogenetic Inertia (Historical Constraints): can only build on
what is there (hard to make wings without appendages)
• Pleiotropy: one gene might affect more than one trait. So if
you alter a gene, it could have multiple effects. So you might not
be able to alter that gene
• Linkage: alleles close together on a chromosome could share a
fate just do to physical proximity and undergo selection and be
inherited as a unit
• Genetic Drift will interfere with the action of Natural Selection in
small populations
Pleiotropy:
when a gene
affects many traits or functions
• Selection might not be able to
act on trait if the gene that
encodes the trait is
Pleiotropic, and also affects
other traits. So, changing the
gene could negatively affect
the other traits
Gene Network
• Conversely, a seemingly unbeneficial trait might get
selected for because the gene that codes for it also
enhances fitness
• Pleiotropy could sometimes lead to evolutionary tradeoffs
(you can have evolutionary tradeoffs that are not
pleiotropic– between traits encoded by different genes)
Pleiotropy:
when a gene
affects many traits or functions
• Difficult to select on a gene that affects
multiple traits: Because, changing the
gene could negatively affect the other traits
• Conversely, a seemingly unbeneficial trait might get
selected for because the gene that codes for it also
enhances fitness, due to beneficial impacts on another trait
• Antagonistic Pleiotropy could lead to evolutionary
tradeoffs
• Antagonistic Pleiotropy: when one gene
controls for more than one trait where at
least one of these traits is beneficial to
the organism's fitness and at least one
is detrimental to the organism's fitness.
Hard to select on a gene that affects multiple
traits (pleiotropy)
Antagonist pleiotropy can lead to trade offs
• Water retention might be good for desiccation
resistance, but also cause hypertension
• High estrogen could increase fertility, but also increase
cancer (estrogen has many targets in the body, and
many consequences)
• Some genes (“Thrifty genes”) are helpful in famines
but also lead to diabetes and obesity
(https://docs.google.com/viewer?a=v&q=cache:pgRlIhEJligJ:www2.uah.es/salud-yenfermedad/pdf/The%2520double%2520puzzle%2520of%2520diabetes.pdf+&hl=en&gl=us&p
id=bl&srcid=ADGEESgMnHzClSUTNI0ZGz690SR4DqSzAVrwRfKILLhBESRcADGmLWJZAUg0IKftVhXpxQumQ0Ox8zixtrylo9xmuZLG30zXRue2MXl3GB7XYwxxS0PQZjFlJaoZ6rAOqF0KSuEC5I&sig=AHIEtbT8VrWAgqBRyM46mdeCd3Ayce
mudQ)Ta
Antagonist pleiotropy can lead to
evolutionary trade offs
(we are interested in fitness trade offs)
• Mutations at single genes that cause high estrogen levels could
increase fertility (trait 1), but also cause cell proliferation in
breast and ovarian tissue (trait 2) leading to cancer (estrogen
has many targets in the body, and many consequences)
• In HIV, slow and careful reverse transcriptase confers AZT
resistance (trait 1), but slow growth rate (trait 2)
• Sickle Cell Anemia - heterozygote for the Hb locus (HgbS) leads
to resistance to malaria (removal of parasite from blood, trait 1)
but lower oxygen carrying capacity (trait 2)
Not all Evolutionary Tradeoffs are
due to Antagonistic Pleiotropy
• Evolutionary tradeoffs could arise
between traits encoded by multiple genes
• With limited resources, there would be
evolutionary tradeoffs between growth
and reproduction
• There could be tradeoffs between
lifespan and reproduction
Constraints on Natural Selection
• Genetic Variation: Selection can only act on existing genetic
variation (we talked about this last lecture)
• Phylogenetic Inertia (Historical Constraints): can only build on
what is there (hard to make wings without appendages)
• Pleiotropy: one gene might affect more than one trait. So if you
alter a gene, it could have multiple effects. So you might not be
able to alter that gene
• Linkage: alleles close together on a chromosome could share a
fate just do to physical proximity and undergo selection and be
inherited as a unit
• Genetic Drift will interfere with the action of Natural Selection in
small populations
Linkage
Definition: The tendency
for certain alleles to be
inherited together due to
their physical proximity on
the chromosome
Human linkage map
• Phenotypic evolution could arise due to linkage
(≠adaptation): Genes might experience an
evolutionary shift because another gene closely
linked on the chromosome is under selection
(selective sweep, genetic hitchhiking)
• This is a genetic mechanism of evolutionary change
that is NOT adaptive
Linkage
Definition: The tendency
for certain alleles to be
inherited together due to
their physical proximity on
the chromosome
Human linkage map
• Consequence: Selection at a locus (gene)
might cause selection at many other genes
closely linked on a chromosome, even if there
is no reason for those other genes to evolve
Constraints on Natural Selection
• Genetic Variation: Selection can only act on existing genetic
variation (we talked about this last lecture)
• Phylogenetic Inertia (Historical Constraints): can only build on
what is there (hard to make wings without appendages)
• Pleiotropy: one gene might affect more than one trait. So if you
alter a gene, it could have multiple effects. So you might not be
able to alter that gene
• Linkage: alleles close together on a chromosome could share a
fate just do to physical proximity and undergo selection and be
inherited as a unit
• Genetic Drift will interfere with the action of Natural Selection in
small populations
Genetic Drift and Natural Selection
• Because of the randomness introduced
by Genetic Drift, Natural Selection is
less efficient when there is genetic drift
• Thus, Natural Selection is more
efficient in larger populations, and less
effective in smaller populations
Selection acts only on genes
that are expressed
• Remember that selection acts
on the phenotype… the traits
that are expressed, given the
genotype
Adaptation
vs
Plasticity
QuickTime™ and a
Photo - JPEG decompressor
are needed to see this picture.
Conceptual Confusions
Trait variation is often assumed to be
due to Adaptation, when the differences
might be due to Phenotypic Plasticity or
nonadaptive genetic causes
The Problem:
• People often wish to jump to the conclusion
that a trait change they see is the result of
adaptation (Natural Selection)
• However, that is not always the case. There
are other mechanisms that could cause
phenotypic variation
• This is what Stephen Jay Gould called the
“Adaptationist Paradigm”
The Problem:
• Adaptations are ubiquitous,
but demonstrating that a
particular trait is an adaptation
is not always easy
The Problem
• Need to distinguish genetically-based
differences from plastic (environmentallyinduced) differences
• Need to distinguish adaptive geneticallybased differences (due to Natural
Selection) from nonadaptive geneticallybased differences (due to other forces)
Example in which adaptation
and plasticity were poorly
distinguished:
• The Bell Curve, 1994 by Murray and Hernstein
• Murray and Hernstein used IQ score differences to
make the argument that African-Americans are
genetically (and evolutionarily) less intelligent than
European and Asian-Americans
• Scientifically, there are key flaws in their argument
• Their argument has key flaws:
(1) Not account for environmental effects on IQ
intelligence is known to be a plastic trait, affected by
pollution (lead), nutrition, cognitive stimulation, etc.
(2) Correlation is associative, not causative
If IQ and “race” are correlated, it does not mean that
race is responsible for IQ
A third factor could be correlated with both race and IQ
that might actually be the underlying cause
• Fischer reanalyzed Murray and
Hernstein’s data and found that socioeconomic status was a stronger
determinant of IQ scores than race, even
in their examples (Inequality by Design:
Cracking The `Bell Curve' Myth)
• Rigorous tests are required to determine
whether or not IQ is genetically based in
human populations (later in the lecture)
Components of Phenotypic Variation
• Quantitative traits are controlled by many loci, many of
which with small effects. For quantitative traits, we depict
the sources of variation as follows:
• Phenotypic variation (VP) is a result of variation that is
due to both genetic effects (VG), and variation due to
environmental factors (VE) and their interaction (VGxE)
VP = VG + VE + VGxE
• So, some of the variation could be due to genetic
causes, but some might be induced by the environment
(as a result of gene expression).
In other words:
• VP = VG + VE + V(GxE)
• VG:
Genetically based (and heritable)
differences, resulting from natural selection
(adaptation) or other genetic factors (genetic
drift, etc)
• VE:
Differences that are inducible by the
environment by a given genotype (acclimation,
phenotypic plasticity)
AND Not All Heritable Variation is due
to Adaptation (Natural Selection)!!!
Phenotypic change and variation could have other
causes:
– Plasticity: Changes that are not due to genetic
changes, but due to changes in gene expression:
Phenotypic Plasticity (including nonheritable epigenetic
modifications)
–
Changes that are Genetic, but NOT adaptive:
•
Genetic Drift: random chance
•
Linkage and Genetic Hitchhiking: Genetic changes that occur
because the gene was right next to another gene on a chromosome that
was under selection
• Physical or structural Constraints (like the Spandrels)
Critique of the
“Adaptationist Programme”
Gould & Lewontin 1979. The Spandrels of San Marco and the Panglossian
Paradigm: A Critique of the Adaptationist Programme.
• One of the most important papers in Evolutionary
Biology
• They critique the “Adaptationist” and “Panglossian
Programme” that assumes that a phenotypic change is
the result of adaptation
• Gould & Lewontin point out that not all phenotypic
variation or phenotypic evolution is the result of
adaptation
Gould & Lewontin:
The spandrels of San Marco
San Marco Cathedral, Venice
Gould & Lewontin on
Physical Constraint:
The spandrels of San Marco
might not have been created
for a reason, but might
simply be a by product due
to the creation of arches
San Marco Cathedral, Venice
Adaptation
Requires Natural Selection
Requires polymorphism in a population
MUST have an effect on Fitness
Is a frequency (%) change in a population
There must be a Selective Force
How can you tell if a trait evolved as a
result of adaptation?
(1) The trait must be heritable
(2) The differences between populations are
genetically based differences rather than inducible
differences (plasticity)
(3) The trait has fitness consequences (promotes
survival, performance, and number of offspring)
(If a trait evolved due to genetic drift, linkage or
pleiotropy, the change is genetic, but may confer
no fitness advantage)
Phenotypic Plasticity
Definition:
• Differences in phenotype that a genotype
exhibits across a range of environments
• Some traits with a plastic component:
intelligence, height, temperature tolerance,
salinity tolerance, muscle mass…
Acclimation (≠ Adaptation)
1) Result of Phenotypic Plasticity
2) Not heritable
3) Short term or developmental response within
a single generation
4) Arises through differential gene expression
or other regulatory mechanism rather than
natural selection
Nature (genetics) vs
Nurture (environment)
• Both environment and genetics affect many
traits, but need to experimentally or statistically
separate these factors
• How?
• Example: Common-garden experiment
• Having appropriate controls
• Statistically assessing the effect of environment
This is a general problem
• This type of problem is a factor in all studies that
attempt to associate a gene with a trait
• You need to account for the effects of
environment
• For example, problems arise when different labs
attempt to associate a gene with a disease
using laboratory mice that have been reared
under different conditions
Types of Plasticity
• Short-term reversible
• Development acclimation:
generally irreversible
Genotype --> Development --> Phenotype
• Within normal tolerance range
• In response to Stress
Plasticity can be depicted
graphically as a Reaction Norm
Response
Environment
Reaction Norm: the function which describes the plastic response
Response
Environment
• In the case of plasticity, the different phenotypes in
different environments are NOT the result of Adaptation…
• The Genotype(s) in the environments are NOT changing
• The differences between them are due to differences in
response (such as gene expression) in different
environments
Dodson, SI. 1989. Predator induced reaction norms. BioScience 39:447–452
Predator induced
formation of helmets
in Daphnia
Hebert and Grewe, 1985
The next 4 slides are optional because of lack of time
Genotype x Environment
Interaction
• VP = VG + VE + V(GxE)  this last
interaction term
• Changes in rank or level of performance among
genotypes when tested in different environments
• Reveals genetic variation for plasticity
• Could reflect tradeoffs between fitness of
different genotypes in different environments
When lines cross, the implication is that different
environments will select for different phenotypes
Response
Environment
Trade-offs in different environments
Select for this reaction
norm in cold
environments
big
Size
Select for this reaction
norm in hot
environments
small
cold
hot
Temperature
Could get selection for different reaction norms
(different plasticity) in different environments
Genetic variation for plasticity can be determined by
examining the significance of the interaction term from an
Analysis of Variance (ANOVA)
Genetic Variation for Plasticity
No Genetic Variation for
Plasticity
Response
Environment
Environment
Example: Human IQ Data
• Data: Many studies use survey data on human
populations in the US (not a common-garden
experiment, where environment is held
constant)
• Did not statistically account for differences in
environment
• A correlation is associative, and not necessarily
causative
Problems with data and
statistical analysis:
• Several reanalyses have found that
socio-economic status (and historical
factors) was a stronger determinant of
IQ scores than “race.”
• Socio-economic status could reflect
nutrition, access to education, etc.
Impact of environment must
be accounted for:
• There is an IQ gap between blacks and whites in America,
Japanese and Koreans in Japan, Ashkenazi and
Sephardic Jews in Israel, and Protestants and Catholics in
Northern Ireland. As economic conditions improve for the
subordinated groups, the gaps are reduced
• A common-garden experiment has never been
performed on humans with respect to IQ scores to
determine actual genetic differences with environmental
effects removed
• Do not know of any study of IQ that has effectively
controlled for socio-economic differences
Most Importantly,
• Must distinguish (1) genetically based
differences from phenotypic plasticity AND (2)
genetic differences due to adaptations (natural
selection) vs other causes of genetic differences
(drift, just different mutations arising in different
populations, etc.
• Many experiments fail to do this
• Examples: drug response, hormone
replacement therapy
How to distinguish between genetically based
traits vs. phenotypic plasticity?
• Animal Model Analyses: Determine how much of a
trait is due to additive, dominance, genetic variance
etc (quantitative genetic methods) – not cover here
• Common-garden experiment: rear different
populations in a common environment to remove
the effects of environmental plasticity, and
determine how much variation is remaining (and due
to genetic effects).
• Molecular Genetic Approaches: transgenic or gene
knockout studies, to determine the impacts of
particular genes on a trait
If the phenotypic change is genetically based,
How to detect if the change was due to selection?
• Selection in the Wild: Look at selection response in
nature (R= h2S, breeder’s equation)
• Selection Experiments (Experimental Evolution):
Impose selection on a population, then examine
evolutionary shift
• Genetic Signatures of Selection: Look for genetic
signatures of natural selection in the DNA sequences
Is the trait change
genetically based?
What is a common-garden
Experiment?
• An Experiment in which individuals from
different populations or species are reared
under identical conditions (can be over a
range of conditions)
• Remove differences due to environmental
plasticity
Example:
Different Populations
A saltwater population and a freshwater population of
a small crustacean (copepod) show differences in
salinity tolerance.
Are the differences due to simply being reared at
different salinities, or are the differences due to
genetically based differences?
Common Garden
Experiment
Different Populations
Rear under common conditions
To determine the differences
when the environment is held
constant
Common Garden
Experiment
If the populations still differ
under common-garden
conditions, the differences are
genetically based.
But are these genetic
differences the result of
adaptation? (or some other
genetic cause)
Different Populations
Molecular genetic
approaches to determine
the genetic basis of a trait
• Is that gene causing the trait?
• Transgenics, gene knockout studies,
CRISPR gene editing, RNAi, etc.
• Generally use model systems, such as
mice, fruit flies, C. elegans, etc.
Detecting Selection
Laboratory Selection Experiments
• But is salinity really the factor causing the
evolutionary physiological change?
• Perform selection experiments to test whether the
evolutionary change happens in response to salinity
alone.
Control
saline ancestors
Selection for several generations
ancestor
5
freshwater selected lines
5
acclimate at same salinity
5
Selection in
the Lab
Take the saline population
and then imposed selection
for freshwater tolerance
Compare the populations
before and after selection
Do the selection lines show
the same evolutionary shift
to fresh water as the wild
population?
0
5
15
Common garden experiment at the end of the selection
Detecting signatures of
selection in DNA sequences



So, how does one detect Natural Selection in
a population?
Many types of tests: use signatures of
Linkage Disequlibrium, reduced variation
(Selective Sweep), genetic deviation (PhiST)
The challenge is distinguishing natural
selection from signatures of Genetic Drift
RNA Codons



In the case of amino
acids
Mutations in Position 1,
2 lead to Amino Acid
change
Mutations in Position 3
often don’t matter
(1)
Simplest Test: Ka/Ks Test
Nonsynonymous substitution rate
Synonymous substitution rate

Ka
Ks
>1

Need coding sequence (sequence that codes proteins)

Ks is used here as the “control”, proxy for neutral evolution


A greater rate of nonsynonymous substitutions (Ka) than
synonymous (Ks) is used as an indication of selection
(Ka/Ks >1)
Substitution rate: out of all the possible number of
mutations, how many fixed
(1)
Ka/Ks Test
Nonsynonymous substitution rate
Synonymous substitution rate




Ka
Ks
>1
In the absence of selection, you’d expect Ks = Ka, and for
Ka/Ks = 1
A greater rate of nonsynonymous substitutions (Ka) than
synonymous (Ks) is used as an indication of positive
selection (Ka/Ks >1)
If Ks > Ka, or Ka/Ks < 1, that would suggest purifying
selection, that selection might be preserving amino acid
composition
Examples: Adaptation
or not?
• A plant grows taller to obtain more
sunlight
• Weeds in cornfields (corn is tall) are on
average taller than weeds of the same
species in soybean fields in order to
obtain more sunlight
Examples: Adaptation
or not?
• A plant grows taller to obtain more
sunlight
Plasticity
• Weeds in cornfields (corn is tall) are on
average taller than weeds of the same
species in soybean fields in order to
obtain more sunlight
Not enough information
Examples: Adaptation
or not?
• Weeds in a cornfield have been found to
grow taller than those in soybean fields
when both populations are reared in
common-garden conditions
• Taller weeds in the cornfields survive and
a greater rate and leave more offspring
Examples: Adaptation
or not?
• Weeds in a cornfield have been found to grow
taller than those in soybean fields when both
populations are reared in common-garden
conditions
They are genetically different, but not know for
sure if it is adaptation (could be linkage, genetic
drift)
• Taller weeds in the cornfields survive and a
greater rate and leave more offspring
Very like to be Adaptation
1. Which of the following is an example of a pleiotropic
effect that would likely lead to an evolutionary tradeoff?
(a) A gene that increases reproduction also shortens life span
(b) A gene that encodes for high sprint speed is negatively correlated
with another gene that encodes for low muscle mass
(c) A gene that causes deafness also causes blindness
(d) A gene that encodes for blue eyes also increases the risk for
cataracts
(e) A gene that increases reproduction also reduces susceptibility to
disease
16. Which of the following cases provides the best evidence for
adaptation (due to natural selection)?
(a) When two populations of mice, one from the North and one from the
South, are reared in the same conditions in the lab, the population from
the North has larger body size than the Southern population
(b) Two populations of rabbits differ in coat color. When they are reared
under the same conditions, the differences in coat color remain. The
genes underlying coat color show signatures of positive selection.
(c) A plant that is reared on more nutritious soil grows taller than other
plants that are grown on nutrient deficient soils
(d) A population of frogs living on green foliage has green skin, whereas a
population of frogs living on brown foliage has brown skin
(e) Two populations of birds differ in beak size. The differences remain
when they are reared under common-garden conditions. The alleles
frequencies at the genes for beak size differ between the two
populations.
• 1-a
• 2-b