Transcript Heritability

Additive genetic variance and heritability
One of the most important questions we can ask to
understand evolutionary change is how the phenotypes of
parents and their offspring are related.
We already know that some phenotypic variance is
environmental in origin. How do we deal with genetic
variation when each generation there is a re-assortment of the
genes into new genotypes? How much of the phenotype of a
trait is heritable via transmission of genes?
A classical approach is to measure what is called breeding
value and normally designated by A.
A is the deviation of the mean phenotype value of an
individual’s progeny from the population mean.
Unrealistically, the measure assumes the individual mates
with a large, random set of individuals to produce these
progeny.
Since it’s a large, random set, we can assume that deviation
from the population mean is due to the genetic characteristics
of the individual.
Also unrealistically, the method only works correctly if the
genes determining the trait have intermediate dominance and
there are no interactions among the genes.
If all those unrealistic assumptions hold, then the genotypic
value, G, depends only on the additive effects of the genes,
and the genetic variance, VG, is comprised totally of additive
genetic variance, VA.
However, in general genetic variance is partitioned into 3
components. Then environmental variance needs to be
included for a complete picture of phenotypic variance:
VP = VA + VD + VI + VE
VA is variance in breeding value (additive genetic variance).
VD is dominance deviation (between alleles at the same
locus)
VI is the interaction term (between alleles at different loci)
The only component responsible for resemblance between
parents and offspring is the additive effects, and these are
what determine responses to selection.
We therefore designate the portion of phenotypic (or genetic)
variance that is additive as the heritability of a trait.
Heritability is usually designated as h2. As opposed to the
previous use of the term, this is also called “narrow-sense
heritability”.
VA/VP = h2
Experimentally, there are two widely used methods for
estimating heritability:
1. In a random mating, non-inbreeding population the
correlation between half-sibs (e.g. one male mates with
many females, the offspring are half-sibs) should be h2/4.
2. The slope of a regression of the ‘value’ (e.g. height) of
offspring phenotypes on the mid-parental value (mean of
the two parents) is a direct estimate of h2.
There are some assumptions you need to be aware of
embedded in these estimates:
a. Individuals occur randomly over environments. There are
no environmentally-driven similarities between relatives;
similarities are entirely genetic.
b. Because the ratio VA/VP involves all components in the
denominator, other factors may determine h2.
What would a high heritability of height indicate
if it were determined by growing a number of
individuals of a single plant line (closely related
or identical genetically) together in the same
greenhouse under identical conditions….?
What if the plants were grown in 2 different
environments? What would you expect the
heritability to be (compared to the results
when grown in a single environment)?
What is the impact of environmental
variability on h2 ?
Thus, heritability is NOT a constant for a given trait in a
given species!
Therefore, comparison of heritability between characters or
species cannot always be carried out directly.
A safe alternative is to measure VA and use it to calculate the
additive genetic coefficient of variation. As with any other
coefficient of variation, it is the standard deviation divided by
the trait mean value, or…
VA

t
Our original goal was to understand the response of a trait to
selection…
We determine the response to selection from the mean trait
values from three groups:
a. the mean for the generation before selection
b. the mean for the parents contributing progeny
and c. the mean of the progeny
Let’s first consider artificial selection – selection using a group
of parents that have high (or low) values for the trait in
question. This is called truncation selection.
The selection differential, S, is the difference between means a
and b. It indicates how much different the selected parents are
from the population.
The difference between means b and c is the response.
If VA is large relative to total phenotypic variance, then there
will be a large response to selection.
Quantitatively, the response to selection is determined by the
strength of selection and the heritability of the trait.
R = Sh2
The figure showed corolla flare (the distance between opposite
tips of the corolla) in Polemonium viscosum.
Heritability was estimated as the slope of a regression between
flare of maternal parents and the flares of their offspring.
In that example, it turned out
that the pollinator exerted
selection on flare. Plants
pollinated by bumblebees had
9% wider flowers than randomly
chosen plants that were hand
pollinated and/or plants
pollinated by other pollinators
at lower elevation in the Rocky
Mountains.
Discrete Genetic Variation
We’ve already used the term marker gene to refer to genes
that are easily scored, e.g. the simple dominant and recessive
genes scored by Mendel.
More useful marker genes are those that show codominance,
so that each genotype can be recognized in the phenotype, i.e.
we can distinguish the phenotypes corresponding to genotypes
AA, Aa, and aa.
Such markers are not very common. With the use of
electrophoretic genetics (developed by Lewontin in the
1960s), it became possible to detect the presence of different
allozymes (alternative alleles for the same enzymatic
function).
Allozymes migrate differently in an electric field on a gel due
to difference in amino acid sequence, resulting in different
charge structure.
An example shows you what you would see if there were two
variants (fast and slow, representing migration rates on the
gel) of both a monomeric enzyme (only one protein subunit in
the enzyme) and a dimeric enzyme (two protein subunits).
If there are multiple (2 or more) alleles evident in a
population, it is described as polymorphic. If there is only
one allele present at a locus, the population is described as
monomorphic.
In plants, some apparent polymorphisms result from different
alleles functioning in different plant organelles. There may be
different alleles active in the cytosol and in plastids. Plastid
and nuclear DNA may even be inherited separately.
Genetic variety in a population is, in part, indicated by the
fraction of loci polymorphic in it. Since genetic variety may
be important in evolutionary adaptability, polymorphism is of
significant interest.
If you check a species that is widely distributed
geographically, the proportion of polymorphic loci across the
whole area may be large – 41% in 16 British populations of
Arabidopsis thaliana. However, the average fraction
polymorphic in a single population was much smaller – 17%.
Geneticists can use electrophoretic analysis of many loci to
assess the genetic distance between populations or species
(using a statistical tool called cluster analysis).
There is more genetic variation than is evident from enzyme
electrophoresis. That variation is revealed using various
modern tools of DNA analysis…
Isozyme variants do not indicate changes in the DNA
sequence that are “silent”, i.e. do not result in a change in the
amino acid sequence.
Other DNA sequence variants occur in introns (parts of
mRNA transcripts that are spliced out before functioning).
Still other sequence variants occur in flanking or intervening
sequences between genes.
Detecting DNA sequence variation does not require complete
sequencing of DNA. Instead, differences in fragment lengths
indicate differences in sequence.
There are a number of methods used in DNA sequence
analysis, with ‘inventive’ acronyms for some.
What they have in common is the use of enzymes (called
restriction enzymes) that cleave the DNA chain when a
specific short sequence is present, e.g. a specific sequence of
4 bases, like GATC.
A probe is used that has the sequence of a gene (or a small
part of a gene) of interest. The probe ‘labels’ the DNA so that
it can be seen after electrophoresis of fragments.
Change of a single base in the sequence identified by the
restriction enzyme changes the places the DNA is cleaved,
and thus the length of the fragments. Electrophoresis moves
different lengths at different rates.
Basically, what I’ve just described is an approach called RFLP,
or Restriction Fragment Length Polymorphism.
PCR (Polymerase Chain Reaction) can amplify the amount
of a given piece of DNA and increase the potential of RFLP
(or related techniques like RAPD or AFLP – see below).
First, primers are used (pieces of DNA that are the sequence
of a gene or a portion of a gene of interest) to amplify the
number of copies of the gene and adjacent non-coding
regions. Then restriction enzymes are used. In most cases the
objective is to expose variation in non-coding regions or in
introns within the gene. We want the sequence variation to be
selectively neutral.
If there is variation with respect to the presence of the
restriction sequence, then different fragment lengths will be
produced.
Allele 1 product
Allele 2 product
RAPD (Random Amplified Polymorphic DNA) uses random
sequences as primers, and produces a variety of bands,
separated by size in electrophoresis.
Genetic difference between individuals is indicated by the
presence of ay least some bands of different size in
comparing the banding patterns from RAPD.
AFLP (Ampified Fragment Length Polymorphism), like
RAPD, typically produces large numbers of bands that
constitute a “genetic fingerprint” of an individual.
There is one more tool used to reveal genetic diversity. It is
the presence and size of microsatellites.
Microsatellites are sequences characterized by small, repeated
DNA motifs. They are codominant and selectively neutral (as
are most molecular markers). They are subject to high
mutation rates in the number of repeats. They are frequently
described as “hypervariable”.
----------GGATCCGAGAGAGAGAGAGGATCC-------------------CCTAGGCTCTCT CTCTC TCCTAGG----------
Variant 1
--GGATCCGAGAGAGAGAGAGAGAGAGAGGATCC--CCTAGGCTCT CTCTC TCTCT CTCTC TCCTAGG-
Variant 2
The approach is to use a restriction enzyme that cleaves just
outside the repeat region.
Variation in size (number of repeats) is easily detectable using
gel electrophoresis. Microsatellites are useful in distinguishing
individuals, e.g. establishing paternity and gene flow. Because
this mutational variation is neutral, it does not contribute to
selection.