Transcript DNA and Gene Expression

Nature, Nurture, and Behaviour
Behavioural Genetics
•
•
•
•
Genetic contribution to behaviour
But also environmental role
Interaction
Could just as well say BG studies the role of
environment on behaviour as that of
genetics…
Polygenic Traits
• Single gene effects (relatively) easy to study
and understand
• Polygenic effects
• Complex behaviour traits
• Environmental interactions for both single
and polygenic effects, but much more
complex for the latter
Heredity
• Obviously important (and easily seen) in
single gene effects
• More difficult to assess heredity’s role in
complex behaviour traits
Synthesis
• Quantitative genetics
– Estimates degree to which differences among
individuals due to genetic and environmental
differences
– Doesn’t identify specific genes or
environmental factors
• Molecular genetics
– Identifies specific genes in behaviours
Humans
• Limited experimental designs
• Ethics
• Twin studies, adoption studies, naturally
existing polymorphisms
Non-human Animal Studies
• Much more empirical control
• Induce polymorphisms, transgenics, etc.
• Even simpler: directed selective breeding
studies and inbred strain studies
Selected Breeding
• Variability in individuals’ behaviour due to genetic
variance
• Variability can be exploited to show role of
genetics on behaviour
• Pick specific trait, cross-breed individuals with
that trait for multiple generations
• Can also study environmental effects this way
– E.g., cross-fostering, different nutrition levels, training
protocols, etc.
Inbred Strain Studies
• Brothers and sisters mated for at least 20
generations
• Result is individuals that are virtual clones
of all other members of strain
• Any differences therefore due to
environmental effects
• Prenatal and postnatal environmental
factors
– Also, genetic factors, of course
Techniques
• Cross different strains to look at changes in behaviour across
subsequent generations
• Diallel design
– Compares several inbred strains and all possible F1 crosses
• Cross fostering
– F1 crosses to mother from one strain or the other
– Assess role of postnatal care
• Environmental manipulation
– Alter specific environmental factors to examine genotypeenvironmental interactions
Scott & Fuller (1965)
• 13 year study of hundreds of crossbreed and
purebreed dogs
– Beagle, Shetland sheep dog, African basenji, wirehaired fox terrier, American cocker spaniel
• Behaviour patterns
• Dogs perhaps the oldest domesticated species
• High chromosomal similarity across breeds, but
high behavioural differences; easily picked out by
selective breeding (Borgaonkar et al., 1968)
Critical Developmental Periods
• Heightened sensitivity to specific
environmental stimuli
• Time critical
• Gene-environmental interaction
Puppy Toileting
• Born blind and deaf, but excellent sense of
smell
• Mother needs to stimulate them to pass
feces and urine
– Mother consumes puppies’ waste; genetic
programming to keep nest clean and free of
smells that would attract predators
– Activated by having litter; otherwise dogs not
normally coprophagic
• After 4 weeks, pups mobile
• Nursing mother trains pups not to toilet near
nest
– Disciplining behaviour can appear harsh
• Pups separated from mother too early don’t
learn proper toileting
– Subsequently have difficulties learning to be
housebroken; critical period missed
First Fear Period
• Fear/hazard avoidance starts in pups at
about 5 weeks, peaks between 8 to 10
weeks
– Typically when people take puppy home
• Easier to condition fear responses at this
time
• Fear of cars, fear of vet clinic, etc.
• (Rawlinson, 2005)
Farm-Fox Experiment
• Domestication
– “Self-domestication” from
colonization of new
ecological niches? (Morey
1994)
– Intentional decisions by
human beings?
• Trut (1999): 40 year
selective breeding study of
45,000 Russian silver foxes
<http://i.pbase.com/o4/29/336929/1/58820261.
T1333SilverFoxFaceOn.jpg>
Patterns of Domestication
• Darwin recognized similar changes in
appearance and behaviour across species
Appearance of dwarf and giant varieties
Piebald coat colour
Wavy or curly hair
Rolled tails
Shortened tails, fewer vertebrae
Floppy ears
Changes in reproductive cycle
All
All
Sheep, poodles, donkeys, horses,
pigs, goats, mice, guinea pigs
Dogs, pigs
Dogs, cats, sheep
Dogs, cats, pigs, horses, sheep,
goats, cattle
All except sheep
Selection
• Dmitry Belyaev
• Believed behaviour, not size or
reproduction, was key factor in
changes from domestication
• Specifically, selection for amenability
(tamability)
• In 1959 started selected breeding of
Vulpes vulpes for tamability
One Change to Many
• Balance of hormones and neurotransmitters
at level of whole organism
• Slight changes to regulatory genes for some
trait (e.g., amenability to domestication) can
have wide-ranging effects on endocrine
system
• Cascading effects on animal development,
physiologically and behaviourally
Belyaev’s Hypothesis
• Two conditions required
• Variations in tamability must be partly
determined by genes
• Domestication places animal under strong
selective pressure
<http://sybilsden.com/images/imgfox
/darla-2006/07-2006/08darla14.jpg>
Cross-Breeding Studies
• Cross-bred foxes of different behaviour
• Cross-fostered newborns
• Transplanted embryos between donor and
host mothers with different tamability levels
• About 35% of variation in foxes’ defense
response to people is genetically determined
(i.e., heritability)
Selective Pressure?
• Attempted domestication of other wild-type
animals
• River otters (Lutra lutra) and gray rats
(Rattus norvegicus)
• Only 14-16% of wild-caught individuals
showing weak defensive reactions produce
offspring that survive to adulthood
• Domestication puts animals under extreme
stress and selective pressure
The Experiment
• 30 male and 100 female foxes, mostly from
commercial fur farms in Estonia
– So, already tamer than wild relatives
• Selected for tameness alone
– Only about 4-5% of male and 20% of female
offspring allowed to breed
• No training of foxes, minimal human
contact
– Ensures tameness due to genetic selection
Behavioural Classes
• Class III
– Flee from experimenters or try to bite when handled
• Class II
– Let themselves be handled but show no emotionally friendly
response
• Class I
– Friendly to experimenters, wagging tails and whining
• Class IE
– Domesticated “elite”
– Eager for human contact, whimper to attract attention, sniff & lick
experimenters like dogs
– Appeared in sixth generation; by 10th gen. 18% of pups; by 20th
gen., 35%; by 35th gen., 75%
Physical Changes
• Domesticated fox pups respond to sounds two days
earlier and open eyes one day earlier than wild type
• Fear response starts at 9 weeks in domesticated foxes
(instead of 6 weeks in wild type)
– More time to become incorporated into human social
environment
• Appearance:
–
–
–
–
From 10th gen., changes in coat colour
From 10th - 15th gen., floppy ears seen
From 15th - 20th gen., tails and legs shortening
These are still unusual traits, but more common than in
wild-type
Traits
Days
11 12 13 14 15 16 17 18 19
4
6
Weeks
8 10
Farmed
Foxes
Domesticated
Foxes
Dogs
Response to sound
Eyes fully open
Window of socialization
Fear of unknown
12
Ontogenetic Changes
• Changes in mechanisms regulating
development
• Effects on normal timing of embryonic and
post-embryonic development
Incidence of Change
Characteristic
Depigmentation (Star)
Animals per 100,000 with trait
Domesticated
Nondomesticated
Population
Population
12,400
710
Increase
in frequency
(percent)
+1,646
Brown mottling
450
86
+423
Gray hairs
500
100
+400
Floppy ears
230
170
+35
Short tail
140
2
+6,900
9,400
830
+1,033
Tail curled up
Inbreeding?
• Maybe changes due to recessive alleles
– Would manifest from long-term inbreeding
• Study designed to prevent this
– Outbreeding with commercial fox farm animals
– Inbreeding coefficients of 0.02-0.07
• Some of new traits are dominant, not recessive
– Couldn’t have been hidden carriers in original
population
Star Gene
• White patch on head
• Affects migration rate of melanoblasts
(embryonic precursors of pigment cells)
– Melanoblasts form in embryonic neural crest,
normally moving to embryo’s epidermis
between 28-31 days after conception
• Foxes with single Star gene
– Melanoblasts’ movement delayed two days
– Leads to death of “tardy” melanoblasts
– Alters pigment pattern, producing star pattern
Corticosteroids
• Changes in plasma levels of corticosteroids,
hormones linked to adaptation to stress
• Adrenal glands producing less adrenalin
– Hence, lowered stress response, and more tamable
• Adrenalin also on biochemical pathway, linked to,
among other things, melanin (pigment)
– Changes in coat colour
•
•
•
•
Different hormone levels have far-reaching effects
Cascade of changes to behaviour and physiology
Polygenic and environmental interaction
Video
Human Behaviour Studies
• Essentially “observational” in nature
• Limited to naturally occurring genetics and
environmental variation
• Adoption designs
• Twin designs
• Combination designs
Adoption Designs
• Traits running in families can be due to
nature and/or nurture
• Adoption studies can separate heredity and
environmental factors
Parents and Offspring
• Genetically related family
– Shared genes and environment
• Adoption produces “genetic” (i.e., birth) and
“environmental” (i.e., adoptive) parents
– Comparing genetic parents and offspring assesses
genetic contribution to parent-offspring similarity
– Comparing environmental parents and offspring
assesses post-natal environmental contribution to
parent-offspring similarity
Siblings
• Similarly, genetic and environmental
siblings
• Again, comparisons made to assess genetic
and post-natal environmental effects
Designs
• Adoptees’ study method
– Examine incidence of trait in adopted away offspring of
parents who possess the trait
• Adoptees’ family method
– Compare adopted offspring who possess trait and lack the
trait of interest (control offspring)
– Assess incidence of trait in the biological and adoptive
families
– Genetic influence if trait greater for biological relatives of
affected adoptees than biological relatives of controls
– Environmental influence if incidence greater for adoptive
relatives of affected adoptees than biological relatives of
controls
Representative?
• Are adopted parents and adopted children
representative of general population?
• Could be biases related to those parents who
adopt and children who are adoptees
• Some studies suggest no bias with respect to
population, others raise questions
Prenatal Environment
• Can be difficult to assess this
• Sometimes simply hard to get access to
information about prenatal conditions of
adopted children post-hoc
• Various environmental factors could be
different between mothers (and fathers) who
keep, as opposed to give up, their offspring
Selective Placement
• Placement of adoptive children may not be
random
– E.g., certainly isn’t based on appearance
• Preference for like characteristics between
adoptive parents and offspring, inflates
apparent correlation between biological
parents and adopted-away children
– Inflates effect of genes on behaviour
Selective Placement (Adjusted)
• Matches biological and adoptive parents
• Genetic influences can increase correlations
b/t adoptive parents and their adopted
children
• Environmental influence can increase
correlations b/t biological parents and their
adopted-away children
Twin Designs
• Monozygotic (MZ)
– From single fertilized egg
– Genotypically identical
– Can be environmental differences, though
• Dizygotic (DZ) twins
– From separately fertilized eggs
– Just siblings born at the same time
– Genotypically different; share 50% genes
• Allows examination of genetic and environmental
variance of a behaviour (a phenotype)
Relations
• First degree relatives
– 50% genetically related, r=0.5 (“coefficient of
relatedness”)
– E.g., parents and children, full non-MZ siblings
• Second degree relatives
– 25% genetically related, r=0.25
– E.g., grandparents and grandchildren, half-siblings
• Third degree relatives
– 12.5% genetically related, r=0.125
– E.g., cousins
Equal Environment Assumption
• Common to twin design methods
• Assumes environmentally caused similarity
about the same for both MZ and DZ twins
• If MZ twins experienced more similar
environments than DZ twins, then
assumption would be violated
• Studies show the assumption probably holds
for most traits
Prenatal Environment
• Interestingly, MZ twins experience greater
prenatal environmental differences than DZ
twins
• Competition for resources hypothesis
– MZ twins identical, so require exactly same
resources
– Limited prenatal resources; competition
produces prenatal stress
Are Twins Representative?
• Twins not representative of population
• Often born prematurely, uterine environment more
adverse
– E.g., twins usually 30% lower in birth weight than
singletons, slower language development, poorer
performance on early verbal and IQ tests
• However, most differences typically disappear by
midchildhood
– Shared and unique postnatal environmental effects
Combination Designs
• Combine family, adoptive, and twin methods
• Twin and family methods
– MZ twins have r=1.0
– Their children (i.e., cousins) will be r=0.25, not typical
r=0.125 of cousins
– Nephew of MZ male twins shares same r-value (0.5)
with his father as his uncle, but there are different
environmental factors in place
• Step-family analysis
– Get half-siblings (r=0.25) and full siblings (r=0.5) in
same family
– Different environmental effects for different children
Heritability
• Proportion of phenotypic variation in a
population attributable to genetic variation
among individuals
• Variation among individuals may be due to
genetic and/or environmental factors
A Warning
• Heritability measures always refer to degree of
variation between individuals in a population
• Can not be applied at level of individual
• E.g., if heritability index of trait is 0.4 it is
incorrect to say that 40% of this trait is from the
parents and 60% is from the environment
– In clones the heritability index for all traits is zero (all
variability between individuals due to environment).
Statistical Significance
• Is similarity between genetically related
individuals significantly greater than
random?
• Depends on sample size
• Note that some significant effects really
have very little “real” effect
– I.e., account for little variance in a trait
Effect Size
• Extent to which individual differences for a trait in
the population can be accounted for by the genetic
difference among individuals
• Effect of heritable genetic disability can be huge
for an individual, but have little effect on
population
– E.g., if trait only exists in 1 in 10,000 individuals, little
overall variation in trait for the entire population
– Many statistically significant environmental effects in
behavioural sciences have small effects in population
Heritability
• The genetic effect size on individuals
• Often accounts for up to 50% of variance for
behaviours
• Heritability estimated from correlations for
relatives
• Consider 1st degree relative
– Correlations reflect half of the effect of genes because
only share 50% genome
– Correlation = 0.5 --> trait heritability = 100%
– Correlation = 0.25 --> trait heritability = 50%
Origins
• Heritability estimates go back to
Francis Galton (1889)
– Data showing relationship between
offspring height as function of parental
height
– Slope of line is 0.57, which approximates
heritability of traits when offspring values
regressed against average traits in parents
– If only one parent’s value is used,
heritability is twice the slope
– (Note: this is the source of the term
“regression” -- offspring values will
always regress to the mean value of the
population; the slope is always less than
one)
<upload.wikimedia.org/wikipedia/en/3/38/
Galton-height-regress.jpg>
Types of Heritability
• Broad-sense
– All sources of genetic variance
– Indicates extent to which genetic factors of any kind are
responsible for trait variation in the population
• Narrow-sense
– Proportion of phenotypic variance explained by
additive genetic effects
– Indicates extent to which a trait will “breed true” (I.e.,
degree of parent-offspring similarity to be expected)
Broad-Sense Heritability
• P = phenotype, G = genotype, E =
environment so that P = G + E
• Considering variance, Var(P) = Var(G)+Var(E)
• H2 = Var(G)/Var(P)
Narrow-Sense Heritability
• Only the portion of the phenotypic variation
that is allelic (“additive”) by nature
• A = allelic contribution, so Var(A) is the
allelic genetic variance
• h2 = Var(A)/Var(P)
Estimates
• Only P can be observed directly
• This makes estimating heritability difficult,
requiring sophisticated statistical methods to
measure genetic and environmental variance
• Not so hard to do in non-human species
– E.g., have same male mate with large number of
females, thereby producing large number of different,
but related, offspring
– Ethics…
• Human heritability studies typically rely on
adoption and twin studies
Correlation of Traits in Twins
• If MZ and DZ twin correlations are the
same, then heritability is estimated as zero.
• If MZ twin correlations are 1.0 and DZ twin
correlations are 0.5, then heritability is
100%.
• Basically, heritability should be about twice
the difference between MZ and DZ twins.
Classic Twin Studies
• Start with assessing variance of a behaviour
in a large group, and attempting to estimate
how much is due to:
• Genetic effects (heritability)
• Shared environment
• Unique environment
ACE Model
•
•
•
•
A (additive, or allelic, genetics)
C (common environment)
E (unique environment)
(More complex models also examine nonadditive genetics effects, D, for dominance)
MZ and ACE
• MZ twins raised in same family share 100%
of their genes and all of the shared
environment
• Differences due to their unique environment
• Correlation observed between MZ twins
provides estimate of A+C
DZ and ACE
• DZ twins have the common shared
environment, share 50% of their genes, and
also have unique environment effects
• The correlation here is an estimate of
0.5A+C
Correlational Statistics
•
•
•
•
•
rMZ = A+C and rDZ = 0.5A+C, then
A = 2(rMZ – rDZ)
C = rMZ – A
E = 1 - rMZ
where rMZ and rDZ are the correlations of the
trait in MZ and DZ twins, respectively
Heritability Estimates
• Heritability for nine psychological traits as
estimated from twin studies.
All sources are twins
raised together (sample
sizes shown inside
bars).
< upload.wikimedia.org/wikipedia/
en/5/54/Heritability-from-twin-correlations1.jpg>
Concordances
• Disorders are diagnosed as either/or
• Therefore, family resemblance is assessed by
concordances, not correlations
• Concordance: an index of risk
• Typically means presence of the same trait in both
members of a pair of twins
• More correctly, the probability that a pair of
individuals will both have a certain characteristic,
given that one of the pair has the characteristic
Concordances
• Twin concordances for seven psychological
traits (sample sizes inside bars)
<upload.wikimedia.org/wikipedia/en/c/cf/Twin-concordances.jpg>
Concordances
• Because MZ twins are genetically identical,
a genetic mutation carried by one will be
carried by the other
• Thus, knowing the concordance rate for a
characteristic should help establish the
extent to which a trait is caused by a genetic
mutation
Three Problems, Though
• A given mutation may not have 100%
penetrance (i.e., may have different
phenotypic consequences in genetically
identical individuals)
• Unique environmental effects may be
present in MZ individuals
• If the trait is polygenic it becomes even
more difficult to evaluate
Penetrance
• Proportion of individuals carrying a
particular variation in a gene (genotype)
that also express a particular trait
(phenotype)
Huntington’s Example
• Known mutations for disease of 95%
penetrance
– 5% with dominant allele for Huntington’s don’t
develop the disease and 95% do
• Trinucleotide repeat of CAG (coding for
amino acid glutamine) forms a
polyglutamine (polyQ) chain
• 36-39 repeats show reduced penetrance
and/or later onset
Age Related Penetrance
• Example: multiple endocrine neoplasia 1
(MEN 1)
– Parathyroid hyperplasia and pancreatic islet-cell
and pituitary adenomas
• Due to mutation in the menin gene on
human chromosome 11q13
• Bessett et al. (1998) found age-related
penetrance of MEN 1 to be 7% at age 10,
but nearly 100% by age 60
Caveats
• Heritability describes what is in a particular
population at a particular time
• Not what could be (or should be)
• Genetic and/or environmental changes will
also alter heritability estimates
Within and Between Groups
• Causes of individual differences in groups have no
implications for causes of average differences
between groups.
• Heritability refers to the genetic contribution to
differences among individuals within a group
– Specific to time and place
• High heritability within a group doesn’t
necessarily mean that average differences between
groups are due to genetic differences
– Environmental factors could be very significant