Additive genetic variance and heritability One of the

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 = V A + V D + V I + V E 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 h 2. As opposed to the previous use of the term, this is also called “narrow-sense heritability”. VA/VP = h 2 Experimentally, there are two widely used methods for estimating heritability:

1. In a random mating, non-inbreeding population the 2. correlation between half-sibs (e. g. one male mates with many females, the offspring are half-sibs) should be h 2/4. 3. 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 h 2. 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 h 2.

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 h 2 ?

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… 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 = Sh 2

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 1960 s), 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 m. RNA 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 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”. -----GGATCCGAGAGAGGATCC----------CCTAGGCTCTCTC TCCTAGG----- Variant 1 --GGATCCGAGAGAGAGAGGATCC--CCTAGGCTCTC 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.