Transcript Units&Targets

The Unit of Selection:
the level of genetic organization
that allows the prediction of the
genetic response to selection
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.
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.
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.
Quantitative Genetic Components As a Function
of Allele Frequencies: A. 4 allele at ApoE is
Rare, A2 at LDLR Common; B. Reversed
Genetic Variance
60.00
A.
{
Epistatic Variance
Dominance Variance
Additive Variance
50.00
35.00
B.
30.00
40.00
25.00
30.00
20.00
15.00
20.00
10.00
10.00
5.00
0.00
0.00
ApoE & LDLR
ApoE
LDLR
ApoE & LDLR
ApoE
LDLR
Epistasis Is Present, But Recombination Is High: The Unit of Selection Is A Single Locus.
Persistence of Fetal Hemoglobin Ameliorates
Impact of Sickle Cell Anemia
Epistasis Strong, Recombination Weak; The
Unit of Selection Is A Multilocus Complex
Various Alleles At Loci in the MHC Complex Are
Predictive of Multiple Sclerosis, an Autoimmune Disease
Extended
Haplotype
Homozygosity
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.
443:574-577.
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
Recombination is not
Uniformly distributed in the
human genome, but rather is
Concentrated into “hotspots” that
Separate regions of low to no
Recombination.
Region of Overlap of the
Inferred Intervals Of All 26
Recombination and Gene
Conversion Events Not
Likely to Be Artifacts.
18
Number of Recombination Events
LD in the human LPL gene
16
14
12
10
8
6
Significant |D’|
4
2
Non-significant |D’|
0
0
1000
2000
3000
4000
5000
(Templeton et al.,AMJHG 66: 69-83, 2000)
6000
7000
8000
9000
10000
Too Few Observations
for any |D’| to be
significant
Neutral Genetic
Drift, Stable
Population Size
Neutral Genetic
Drift, Expanding
Population Size
Negative
Selection
Positive
Positive
(Directional) (Diversifying)
Selection or Selection or
Subdivision
Bottleneck
Haplotype Network in 5’ Region of
LPL
84R
4
49N
17
5'-1
13
23J
7
5'-2
8
5'-4
36J
4
12
5'-3
3
5'-5
17
44N
Positive
(Directional)
Selection
5
16
5'-6
6
9
9
2
10
10
17 5'-8
6
4
18
32J
14
15
5'-7
16 14J 8
8J
Haplotype Network in 3’ Region of
LPL
Positive
(Diversifying)
Selection
53
3'-11
75R
59
59
3'-10
53
45J
69
30J
41
16J
36
69
36J
T-1
3'-9
38
39
41
43
44
46
47
60
66
3'-12
46
43J
40J
63
55
37J
40
58
28J
53
60N
59
45
53
67
34J
56
40
24J
39J
56
59
45
59
3'-4
53
41N
38
54
12J
42N
54N
36
38
44N 63 9N
3'7
42
50
53N
56N
46N
62
67N
41
40
48N
42
62
T-2
44
50
68
3'-8
50
T-3
78R
42
8J
32J
49N 53 77R
29J
55
44
14J
41
49
42
44
55
40
53
38J
52
61
50N
64
48
49
3'-6
58
49
81R
46
37
61
45
55
58
56
3'-1 63 3'-2
38
50
44
51N 63 3'-3 42 59N
56
61N
3'-5
T-4
36
55
19J
65
63
57
51
65
61
64J
64
35J
20J
36
41
58
67
26J
T-1 X 11J 9
T-2 X Node n
15
(16-29)
6
(16-19)
12
17
10
17J X T-3
25
Node i X Node k
(17-25)
36J
11
(19-25)
Node f X Node e X T-2
7
14
(19-29)
(16-31)
25J
30
Recombinants
and PostRecombination
al
Evolution
in LPL
32J
58
35
40
29
36
6
41
14
67
15
58
18
19
13
(58-63)
30
80R
30
42
79R X T-2
23
84R
35
20
(33-44)
50N
42
26
46N
30
19
57
58
17
q
50
63N
62
27
23
53
69
55N
38J
Node g X T-2
8
(29-34)
68R 27
X T-2
26
22
(33-35)
6NR
30
50
78R
3
(33-35)
X T-2
X 83R
X 4JN
69R
27
(27-33)
42N
54
1
(27-29)
31
38
53
47N
71R
61
29
28
(19-33)
19
(33-44)
T-2 X Node x
29
(64-68)
62
88R
1
55
26
Node s X Node r X 62N X 45J
21J
49N
5NR
4
40
86R
72R
20
2JNR X T-2
18
27
41 (33-35)
53
35
9
(29-35)
23
36
2
(13-29)
18J 44 48N 67N
62
23
34
77R
16
(29-31)
4
59
Node h X T-4
10
(21-29)
8J
Node m X 73R
21
(5-9)
13J
69
29
39J
26
29
59
Node a X 2JNR
49
35J
75R
26
(23-34)
44N
55
8
44
53
23
(19-33)
55
55
82R
16
14J
3
81R
41
31
58N
49
17
22J
u
15J
T-4 X Node w
38
76R
35
87R X T-2
20J
T-2 X Node t
60
57J
11
Node v X T-3
24
17
15
(18-29)
33J X Node l
21
25
4
(13-29)
28
12
(5-9)
3
Node a X 1JNR
24
14
26J
65N
13
18
61
4
Node m
Gene
Conversion
5
(56-61)
6
6
12J
T-2 X Node p
17
(33-35)
Node j X 28J
28
24J 56
46
Node b X Node c
22
35
36
2
19
7NR
20
3JNR
30
44
31
19
54N
59
59
66N
85R
51N
63
74R
50
33
41N
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
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.
Target of Selection:
the level of biological
organization that displays the
phenotype under selection.
Targets Below The Level of
The Individual:
Example: the t-complex in mice
and meiotic drive.
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.
The t-complex in mice
Adult Population
tt
Tt
TT
Gtt
GTt
GTT
Mechanisms of
1
Producing Gametes
(Violation of Mendel's First Law)
Gene Pool
(Population
of Gametes)
k
(1-k)
1
t
T
p' = Gtt + kGTt
q' = GTT + (1-k)GTt
p’=Gtt+ kGTt=Gtt+1/2GTt-1/2GTt+kGTt=p+GTt(k-1/2)
p=p’-p=GTt
(k-1/
2)
1/
2
Need this 1/2 for t-alleles because
the meiotic drive is expressed
only in males.
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:
TT
1
Tt
1-s
tt
0
The t-complex in mice
A single Unit of Selection can have more than one Target of Selection
TT
1
Tt
1-s
tt
0
Assume random mating; then meiotic drive changes
The gamete frequencies to p’=p+pq(k-1/2):
After fertilization, selection at the individual level is governed by:
at  q(1 s w ) p (w )
Where
w  q2  2 p q(1 s)
The t-complex in mice
The total change in allele frequency is:
p  p  p  p  p  p  p
at
 p   pqk  12
w

Change due
to selection at
the level of
the individual;
always
negative.
Change due to
selection at the
level of the
gamete; always
positive.
The t-complex in mice
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.”
Holliday Model
Double-Strand Break Model
Gene
Conversion
Unequal gene conversion
Equal gene conversion
Double-Strand Break Model
Holliday Model
Gene Conversion Can Be A Major Source of
Genetic Variation in Multigene Families
Takuno, S. et al. Genetics 2008;180:517-531
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
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 1:1 ratio of Mendelian Segregation
=probability of segregation yielding only A alleles
1=probability of segregation yielding only a alleles
Then the segregation ratio A:a in Aa heterozygotes is:
1/ (12
)+ :1/2(1- )+ 1or k:1-k where k= 1/2(1- )+
A is fixed if k> 1/2, and a is fixed if k< 1/2, at a rate dependent upon the
frequency of heterozygotes in the population
Extension of Walsh’s Model To Include
Molecular Drive ,, 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 =
2(2k 1)N ev (1 f )
when 4N ev (1 f )(2k 1) 1
N
(2k 1)N ev (1 f )e 4 N ev (2k1)(1 f )
N
when 4N ev (1 f )(2k 1) 1
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
Transposition
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).
Transposition: Evolutionary
Co-option (or exaptation)
Percentage of TE-derived residues in miRNA genes.
Human miRNA gene sequences
Human mature miRNA
sequences
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 premiRNA 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 miRNA–RISC targets mRNAs for
regulation. miRNAs have been implicated in a
variety of functions, including developmental timing,
apoptosis, and hematopoetic differentiation
Piriyapongsa, J. et al. Genetics 2007;176:1323-1337
Transposition: Evolutionary
Co-option (or exaptation)
TE’s (brown
segments) have
been co-opted for
both transcriptional
and posttranscriptional
regulation. Can
build up regulatory
networks in
evolution.
Feschotte. 2008.
Nat Rev Genet
9:397-405.
Transposition: Direct Phenotypic
Effects
Transposition: Horizontal & Vertical
Transfer
Anxolabehere et al. Mol. Biol. Evol. 5: 252-269, 1988
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.
Transposition: Horizontal Transfer
Transposition: Horizontal Transfer
Unequal Exchange
Unequal Exchange Can Also
Create New Types of Genes
Unequal Exchange: Concerted
Evolution
Gene Duplication Without
Concerted Evolution
Gene Duplication With
Concerted Evolution
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): 235245, 2001.
Unequal Exchange: Concerted Evolution
Model of Weir et al. J. Theor. Biol. 116: 1-8, 1985
n=number of repeats in a multigene family
N=ideal population size
m=neutral 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
2Nnm=expected number of new mutants per generation
Rate of neutral evolution in multigene family evolving in concert =
(2Nnm´1/(2Nn)=m= Same neutral rate as if it were a single locus!
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 [>1/(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 (<1/(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.
Concerted Evolution With Selection
Simple co-dominant model:
start with fixation of a allele, fitness aa is 1
mutation creates A allele, with fitness of
Aa being 1+s, and AA 1+2s
Under Neutrality, Probability of Fixation of A is:
1/(2N)
Kimura showed that fixation probability here is:
Where p is the initial frequency; that is, 1/(2N)
Concerted Evolution With Selection
Mano, S. et al. Genetics 2008;180:493-505
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!
Molecular Drive Does NOT
Negate The Importance of Other
Evolutionary Forces.
Molecular Drive INTERACTS
With Other Evolutionary Forces
In Determining the Path of
Evolution.