The Unit of Selection: the level of genetic organization that allows the prediction of the genetic r

1
The Unit of Selection the level of genetic
organization that allows the prediction of the
genetic response to selection
2
Fitnesses in population genetics are assigned to
genotypic classes of individuals rather than
individuals themselves the genotypic classes
can be single locus genotypes, or two locus
genotypes, etc.The unit of selection is the
level of genetic organization to which a fitness
phenotype can be assigned that allows the
response to selection to be accurately
predicted.This means that the unit of selection
must have genetic continuity across the
generations.
3
Meiosis and Sexual Reproduction Break Up
Multilocus Complexes in Outbreeding Species (as a
function of both physical recombination and
assortment, and system of mating), Which Reduces
the Size of the Unit of Selection.Selection
Upon Epistatic Complexes Builds Up Higher Level
Units.
4
The Unit of Selection is a dynamic compromise
between selection building up complexes and
effective recombination breaking them down. The
unit of selection can change as the population
evolves or experiences altered demographic
conditions.
5
Epistasis Is Present, But Recombination Is High
The Unit of Selection Is A Single Locus.
6
Persistence of Fetal Hemoglobin Ameliorates
Impact of Sickle Cell Anemia

• Epistasis Strong, Recombination Weak The Unit
  of Selection Is A Multilocus Complex

7
Various Alleles At Loci in the MHC Complex Are
Predictive of Multiple Sclerosis, an Autoimmune
Disease
Gregersen, J. W., K. R. Kranc, X. Ke, P.
Svendsen, L. S. Madsen, A. R. Thomsen, L. R.
Cardon, J. I. Bell, and L. Fugger. 2006.
Functional epistasis on a common MHC haplotype
associated with multiple sclerosis. Nat.
443574-577.
Extended Haplotype Homozygosity

• Epistasis Strong, Recombination Intermediate
  The Unit of Selection Is A Multilocus Complex,
  But Only Selected Haplotype Shows Extensive D
  Must Be Constantly Built Up By Selection

8
(Templeton et al.,AMJHG 66 69-83, 2000)
LD in the human LPL gene
Recombination is not Uniformly distributed in
the human genome, but rather is Concentrated into
hotspots that Separate regions of low to
no Recombination.
Significant D
Non-significant D
Too Few Observations for any D to be
significant
9
Positive (Diversifying) Selection or Subdivision
Positive (Directional) Selection or Bottleneck
Neutral Genetic Drift, Expanding Population Size
Neutral Genetic Drift, Stable Population Size
Negative Selection
10
Haplotype Network in 5 Region of LPL
Positive (Directional) Selection
11
Haplotype Network in 3 Region of LPL
Positive (Diversifying) Selection
12
Recombinantsand Post-RecombinationalEvolutioni
n LPL
13
12 Recombination Events Occurred Between T-1
Haplotypes With T-2,3, or 4 Haplotypes

• In All 12 Cases, the 5 End Was Of The T-1 Type.
  Under Neutrality, This Has A Probability of
  (1/2)12 0.002.
• Therefore, the 5 End Experienced A Selective
  Sweep Enhanced By Recombination

14
The Unit of Selection For LPL, the unit of
selection is smaller than the gene because of the
recombination hotspot in the middle of this locus.
15
Target of Selectionthe level of biological
organization that displays the phenotype under
selection.
16
Targets Below The Level of The IndividualExample
the t-complex in mice and meiotic drive.
17
The t-complex in mice
20 cM region of chromosome17 of the mouse genome
that constitutes about 1 of the mouse genome.
Inversions suppress most recombination in this
region that contains genes for sperm motility,
capacitation, binding to the zona pellucida of
the oocyte, binding to the oocyte membrane, and
penetration of the oocyte and also notochord
development. Because of strong epistasis and
low recombination, it behaves as a unit of
selection that can be modeled as a single locus.
18
The t-complex in mice
pGtt kGTtGtt1/2GTt-1/2GTtkGTtpGTt(k-1/2)
?pp-pGTt(k-1/2)
19
The t-complex in mice
A single Unit of Selection can have more than one
Target of Selection
At the individual level, the t-alleles affect
viability
20
The t-complex in mice
A single Unit of Selection can have more than one
Target of Selection
Assume random mating then meiotic drive
changes The gamete frequencies to pppq(k-1/2)
After fertilization, selection at the individual
level is governed by
21
The t-complex in mice
The total change in allele frequency is
22
The t-complex in mice
23
Molecular Drive (Dover)

• The nuclear genomes of eukaryotes are subject to
  a continual turnover through unequal exchange,
  gene conversion, and DNA transposition. Both
  stochastic and directional processes of turnover
  occur within nuclear genomes.

24
(No Transcript)
25
Gene Conversion
Holliday Model
Double-Strand Break Model
Unequal gene conversion
Equal gene conversion
26
Gene Conversion Can Be A Major Source of Genetic
Variation in Multigene Families
Holliday Model
Double-Strand Break Model
Takuno, S. et al. Genetics 2008180517-531
Gene conversion increases the number of
haplotypes in multigene systems, particularly
when the tract length is short. If there is
diversifying selection (e.g., MHC, S alleles),
selection often favors these new haplotypes, even
if the source of the converted segment is a
pseudogene.
Unequal gene conversion
Equal gene conversion
27
Walsh (Genetics 105 461-468, 1983)

• 1 Locus, 2 Allele Model (A and a) Such That
• ? the probability of an unequal gene conversion
  event
• ? the conditional probability that a converts to
  A given an unequal conversion occurs
• 1-?the probability of getting a 11 ratio of
  Mendelian Segregation
• ??probability of segregation yielding only A
  alleles
• ?(1-?)probability of segregation yielding only a
  alleles
• Then the segregation ratio Aa in Aa
  heterozygotes is
• 1/2(1- ?) ?? 1/2(1- ?) ?(1-?) or k1-k
  where k 1/2(1- ?)??
• A is fixed if kgt 1/2, and a is fixed if klt 1/2,
  at a rate dependent upon the frequency of
  heterozygotes in the population

28
Extension of Walshs Model To Include Molecular
Drive (g,b), Drift (Nev), and System of
Mating/Population Subdivision (f)

• Probability of Fixation of a Neutral Mutation
  1/(2N)
• Probability of Fixation of a Mutation With Biased
  Gene Conversion
• Thus, the evolutionary impact of gene conversion
  interacts with and is modulated by traditional
  evolutionary forces. For example, as f goes
  down, the probability of fixation of a biased
  gene conversion allele goes up.

Molecular Factors Do Not Override Traditional
Evolutionary Forces Rather, They Strongly
Interact With Them
29
Transposition
30
Transposition Mutator
A flower of Petunia hybrida transposon genotype
derived from the inbred line W138 showing a large
number of white-pink sectors (see Ramulu et al.,
ANL10 19-21, 1998).
31
Transposition Evolutionary Co-option (or
exaptation)
Percentage of TE-derived residues in miRNA
genes. MICRORNAS (miRNAs) are small, 22-nt-long,
noncoding RNAs that regulate gene expression. In
animals, miRNA genes are transcribed into primary
miRNAs (pri-miRNAs) and processed by Drosha to
yield 70- to 90-nt pre-miRNA transcripts that
form hairpin structures. Mature miRNAs are
liberated from these longer hairpin structures by
the RNase III enzyme Dicer. Drosha acts in the
nucleus, cleaving the pri-miRNA near the base of
the hairpin stem to yield the pre-miRNA sequence.
The pre-miRNA is then exported to the cytoplasm
where the stem is cleaved by Dicer to produce a
miRNA duplex. One strand of this duplex is
rapidly degraded and only the mature 22-nt miRNA
sequence remains. The mature miRNA associates
with the RNA-induced silencing complex (RISC),
and together the miRNARISC targets mRNAs for
regulation. miRNAs have been implicated in a
variety of functions, including developmental
timing, apoptosis, and hematopoetic
differentiation
Human miRNA gene sequences
Human mature miRNA sequences
Piriyapongsa, J. et al. Genetics
20071761323-1337
32
Transposition Evolutionary Co-option (or
exaptation)
TEs (brown segments) have been co-opted for both
transcriptional and post-transcriptional
regulation. Can build up regulatory networks in
evolution. Feschotte. 2008. Nat Rev Genet
9397-405.
33
Transposition Direct Phenotypic Effects
34
Transposition Horizontal Vertical Transfer
Anxolabehere et al. Mol. Biol. Evol. 5 252-269,
1988
35
Transposition Horizontal Transfer
Comparison of a) species and b) P-element
phylogenetic histories. Diagonal lines unite
P-element clades with the species from which they
were sampled. From Silva, J.C. and M.G. Kidwell.
Horizontal Transfer and Selection in the
Evolution of P Elements. Mol Biol Evol 17(10)
1542-1557, 2000.
36
Transposition Horizontal Transfer
37
Transposition Horizontal Transfer
38
Unequal Exchange
39
Unequal Exchange Can Also Create New Types of
Genes
40
Unequal Exchange Concerted Evolution
Gene Duplication Without Concerted Evolution
Gene Duplication With Concerted Evolution
41
Unequal Exchange Concerted Evolution
Gentile, K.L., W.D. Burke and T.H. Eickbush.
Multiple Lineages of R1 Retrotransposable
Elements Can Coexist in the rDNA Loci of
Drosophila. Mol Biol Evol 18(2) 235-245, 2001.
42
Unequal Exchange Concerted Evolution
Model of Weir et al. J. Theor. Biol. 116 1-8,
1985 nnumber of repeats in a multigene
family Nideal population size mneutral mutation
rate per repeat per generation 1/(2N)probability
of fixation of new mutant at homologous
sites ?probability of a repeat converting a
paralogous repeat to its state (Molecular drive
exists such that a neutral mutant will eventually
go to fixation at all paralogous sites as well)
1/(2Nn)probability of fixation of a new mutant
at all homologous and paralogous
sites 2Nn?expected number of new mutants per
generation Rate of neutral evolution in multigene
family evolving in concert (2Nn?)1/(2Nn)?
Same neutral rate as if it were a single locus!
43
Unequal Exchange Concerted Evolution
Recall that the time to neutral coalescence of
all homologous copies of a gene to a common
ancestral form 4N The time to neutral
coalescence of all homologous and paralogous
copies in a multi-gene family to a common
ancestral form 2/(1-?) where ? is the Maximum
of one of two forms 1. 1-1/(2N) or
2. 1-? In the first case ?gt1/(2N) that
is molecular drive is more powerful than drift,
then t 2/1-1-1/(2N) 2/1/(2N) 4N the
same rate of coalescence as a single locus and no
effect of ?! In the second case (?lt1/(2N) that
is molecular drive is weak compared to drift), ?
dominates the coalescence process! Therefore,
molecular drive has its biggest evolutionary
impact when it is Weak compared to drift. Under
these conditions, the multigene family will have
much diversity among paralogous copies within a
chromosome.
44
Concerted Evolution With Selection
45
Concerted Evolution With Selection
Mano, S. et al. Genetics 2008180493-505
Considered a similar model, but now assume that
we have a multigene family with n copies and that
the mutant A allele can spread due to molecular
mechanisms leading to concerted evolution. Then,
the probability of fixation of A is given
by Concerted evolution has the effect to
increase the "effective" population size, so that
weak selection works more efficiently in a
multigene family the opposite of Dover!
46
Molecular Drive Does NOT Negate The Importance of
Other Evolutionary Forces.Molecular Drive
INTERACTS With Other Evolutionary Forces In
Determining the Path of Evolution.