Sequence Diversity in Evolution and Crop Improvement

1
Sequence Diversity in Evolution and Crop
Improvement
Teosinte
Maize Landraces
Inbreds/Hybrids

• Sherry Flint-Garcia
• Research Geneticist
• USDA-ARS
• MU Division of Plant Sciences

Photos courtesy J. Doebley
2
Sequence Diversity

• Evolution
• What are the forces that cause evolution?
• Speciation hybridization
• Uncovering evolutionary history
• Crop Improvement
• The teosinte-maize story

3
The Four Forces of Evolution

• Mutation -- spontaneous changes in the DNA of
  gametes. Prerequisite to all other evolution.
• Natural Selection -- genetically-based
  differences in survival or reproduction that
  leads to genetic change in a population.
• Gene flow -- movement of genes between
  populations. In plants this can be accomplished
  by pollen or seed dispersal.
• Genetic drift -- random changes in gene
  frequency. This is very important in small
  populations.

4
Mutation Generation of New Alleles

• Mutations are the result of mistakes in DNA
  replication, exposure to UV or to some chemicals
  (mutagens) and other causes.
• Point mutations
• changing one nucleotide to another
• e.g., C--gtT

5
Sickle Cell Anemia
A single point mutation causes a dramatic change
in phenotype.
6
Other types of mutations

• Indels
• insertions/deletions
• Cause frame-shifts, usually premature stops
• Gene duplication
• May lead to new functions
• Chromosomal mutations
• Inversions, translocations, deletions
• Polyploidy
• Very common in plants
• May lead to new species in one step

7
Most point mutations have no effect or almost no
effect. Why?
Most of the genome seems to be junk -- at least
it doesnt code for proteins. Many mutations
within protein-coding region of genes dont
change the amino acid specified. i.e., there is
redundancy in the genetic code.
For example, 6 different codons specify the
amino acid leucine.
8
The Four Forces of Evolution

• Mutation -- spontaneous changes in the DNA of
  gametes. Prerequisite to all other evolution.
• Natural Selection -- genetically-based
  differences in survival or reproduction that
  leads to genetic change in a population.
• Gene flow -- movement of genes between
  populations. In plants this can be accomplished
  by pollen or seed dispersal.
• Genetic drift -- random changes in gene
  frequency. This is very important in small
  populations.

9
Natural Selection

• Peppered moth (Biston betularia) evolution during
  the industrial revolution in England
• Early 1800s pre-industrial
• Bark of trees were white
• Almost all moths were of typica form
• 1895 Industrial Era
• Bark of trees were covered in black soot
• 98 of moths were of carbonaria form
• Today Clean Air laws enforced
• Prevalence of carbonaria form declining

typica form
carbonaria form
10
(No Transcript)
11
Brassica oleracea
12
The Four Forces of Evolution

• Mutation -- spontaneous changes in the DNA of
  gametes. Prerequisite to all other evolution.
• Natural Selection -- genetically-based
  differences in survival or reproduction that
  leads to genetic change in a population.
• Gene flow -- movement of genes between
  populations. In plants this can be accomplished
  by pollen or seed dispersal.
• Genetic drift -- random changes in gene
  frequency. This is very important in small
  populations.

13
Gene Flow

• Tends to homogenize populations.
• Rates of gene flow depend on the spatial
  arrangement of populations.

Directional movement of alleles
Migration occurs at random among a group of
equivalent populations.
14
Migration along a linear set of populations
Populations are continuous. 
15
(No Transcript)
16
The Four Forces of Evolution

• Mutation -- spontaneous changes in the DNA of
  gametes. Prerequisite to all other evolution.
• Natural Selection -- genetically-based
  differences in survival or reproduction that
  leads to genetic change in a population.
• Gene flow -- movement of genes between
  populations. In plants this can be accomplished
  by pollen or seed dispersal.
• Genetic drift -- random changes in gene
  frequency. This is very important in small
  populations.

17
(No Transcript)
18
Founder effect Gene flow and genetic drift are
responsible for the limited genetic variation on
islands, relative to mainland populations.
19
Speciation and Hybridization

• Speciation how do new species arise?
• What is a species, anyway?
• Most species were originally described by their
  morphology.
• The Problem Convergence
• Similar features in unrelated organisms due to
  evolution of traits that work in similar
  environments

20
Convergent structures in the ocotillo (left) from
the American Southwest, and in the allauidia
(right) from Madagascar.
21
Nectar feeders have converged on this hovering
long-tongued morphology.
22
Speciation and Hybridization

• Biological Species Concept (BSC)
• Based on reproductive compatibility
• Natural spatial, temporal, and morphological
  discontinuities generally correspond to fertility
  barriers
• The Problem In plants, many named species can
  hybridize.

23
Most dandelions are asexual. So the biological
species concept (BSC) doesnt apply. How can you
name species depending on who can mate with whom
when the organisms do not mate at all?!
24
Scarlet and Black oaks can hybridize and inhabit
the same range -- but they have different
microhabitat preferences so hybridization is rare.
25
These pines can also hybridize but they shed
their pollen at different times of the season
26
Speciation by Hybridization
Hybridization often shows how difficult it is
to apply the BSC to plants. The hybrid in this
case is a new species. The rearrangements of its
chromosomes make it infertile with either parent.
hybrid
27
As the climate becomes drier the desert splits
the range of this hypothetical tree species.
This reduces gene flow between the now isolated
populations and sets the stage for speciation.
28
Evolution of species that are geographically
separated. Genetic drift plays a significant
role.
Edge effect where evolution of reproductive
barriers occurs between neighboring populations.
Requires considerable selection pressure.
Establishment of a new population with a
different ecological niche within the same
geographical range of the parental population
29
Uncovering Evolutionary History

• Taxonomy vs. Systematics
• Estimating Phylogeny
• Distance Methods
• Maximum Parsimony Methods
• Maximum Likelihood Methods

30
Taxonomy vs. Systematics

• Taxonomy
• Discovering
• Describing
• Naming
• Classifying
• Systematics
• Figuring out the evolutionary relationships of
  species
• Summarize the evolutionary history of a group

31
Plant Taxonomy

• taxon - any group at any rank
• corn common name
• kingdom Plantae (Viridiplantae)
• division (phylum) Anthophyta
• class Liliopsida
• order Commelinales
• family Poaceae
• genus Zea
• species Zea mays

always capitalized
never capitalized
32
Plant Systematics

• A phylogenetic tree is used to illustrate
  systematicrelationships
• Modern taxonomic groups generally correspond to
  clades on a phylogenetic tree (i.e. cladogram)
• Example phylogenetictree of the grass family

Mathews et al. 2000 American Journal of Botany
33
Angiosperm Phylogeny Group TreeDicots are not
a monophyletic group.
34
Data Types that can be used to Estimate a
Phylogeny

• Cross Compatibility
• Uses the Biological Species Concept
• Morphological
• Continuous traits
• Meristic (countable) traits
• Cytological
• Chromosome number
• Chromosome features
• Pairing in hybrids

• Molecular data
• Secondary chemicals
• Proteins
• DNA
• Allele frequencies at many loci (isozymes, SSR)
• DNA sequences, considered as a whole
• DNA sequences, considered site-by-site

35
Maximum Parsimony (Minimum Evolution) Methods

• The process of attaching preference to the
  pathway that requires the invocation of the
  smallest number of mutational events.
• Most effective when examining sequences with
  strong similarity
• Underlying premises
• Mutations are exceedingly rare events.
• The more unlikely events a model invokes, the
  less likely the model is to be correct.

36
Using only trait 1
Traits must have discrete character states.
Must have same character state in at least 2 taxa.
37
But traits 3 4 disagree with trait 1.
sp2
sp5
Redlt-gtblue
Alt-gtG
sp3
sp1
sp4
38

• Every possible tree is considered individually
  for each informative site (computationally
  intensive).
• After all informative sites have been considered,
  the tree that invokes the smallest total number
  of substitutions is the most parsimonious.

4
2
1
5
3
3
5
2
1
4
Blue
Blue
0
0
G
G
0
Blue
4 substitutions required
5 substitutions required
G
Red
Red
A
A
1
1
39
Distance-based approaches
Compare each taxon to every other taxon to
estimate a distance matrix
Distances are then clustered to estimate a
phylogenetic tree.
d12
d13
d14
d15
d25
d24
d23
d35
d34
d45
40
Distance-based approaches
Compare each taxon to every other taxon to
estimate a distance matrix
Example DNA sequence considered as a whole
10 20 30 40 50Sp1 GTGCTGCACG GCTCAGTAT
A GCATTTACCC TTCCATCTTC AGATCCTGAASp2
ACGCTGCACG GCTCAGTGCG GTGCTTACCC TCCCATCTTC AGATCC
TGAASp3 GTGCTGCACG GCTCGGCGCA GCATTTACCC TCCCATC
TTC AGATCCTATCSp4 GTATCACACG ACTCAGCGCA GCATTTGC
CC TCCCGTCTTC AGATCCTAAASp5 GTATCACATA GCTCAGCGC
A GCATTTGCCC TCCCGTCTTC AGATCTAAAA
9
8
12
15
18
15
11
13
10
5
41
Distance-based approaches
Distances are then clustered to estimate a
phylogenetic tree.
Example UPGMA algorithm Unweighted Pair-Group
Method using Arithmetic means
9
8
12
15
18
15
11
13
10
The smallest distance is identified, the average
of the two combined taxa is calculated, and the
matrix is recalculated. This iteration is
repeated.
5
2.5
2.5
42
Distance-based approaches
Sp1
Sp2
Sp3
4-5
0
9
8
13.5
Sp1
16.5
11
0
Sp2
11.5
0
Sp3
0
4-5
2.5
2.5
4
4
43
Distance-based approaches
Sp2
1-3
4-5
0
10
16.5
Sp2
12.5
0
1-3
0
4-5
2.5
2.5
4
4
5
2
4
5
1
3
44
Distance-based approaches
1-2-3
4-5
0
12.5
1-2-3
0
4-5
6.5
6.5
2.5
2.5
4
4
5
2
4
5
1
3
45
Maximum Likelihood Methods

• Best suited for DNA and protein sequence data
• Requires a model of evolution
• Each nucleotide/amino acid substitution has an
  associated likelihood
• A function is derived to represent the likelihood
  of the data given the tree, branch-lengths and
  additional parameters
• Function is minimized

46
1 A C G C G T T G G G 2 A C G C G T T G G
G 3 A C G C A A T G A A 4 A C A C A G G G A A
L(Tree 1) L0 x L1 x L2 x L3 x L4 x L5 x L6 5
x 10-13
47
1 A C G C G T T G G G 2 A C G C G T T G G
G 3 A C G C A A T G A A 4 A C A C A G G G A A
Repeat for each of node assignment, and each site
in alignment. Probability of that unrooted tree
is the sum of all individual trees. Repeat for
each unrooted tree and choose the tree with the
highest liklihood.
L(Tree 1) L0 x L1 x L2 x L3 x L4 x L5 x L6 5
x 10-13
L(Tree 2) L0 x L1 x L2 x L3 x L4 x L5 x L6 1
x 10-18
48
The Teosinte-Maize Story

• The practical side of sequence diversity
• PLANT BREEDING!
• Sequence Diversity in Teosinte
• Sequence Diversity in Maize
• Selection During Domestication and Improvement

49
Sequence Diversity and Plant Breeding

• Genetic diversity within a crop species is the
  raw material for current plant breeding
• Genetic diversity is the insurance policy to
  enable plant breeders to adapt crops to changing
  environments

50
The Problem

• To what degree is limiting genetic diversity
• inhibiting genetic improvement in corn?

51
Two Views of the Problem

• Most of the corn germplasm in use in the USA
  today is derived from mixtures of only two major
  races out of 300 races total (Wallace and
  Brown, 1956). The simplest means of correcting
  this situation and of increasing the genetic
  diversity of this important crop is to introduce
  unrelated sources of germplasm (Brown and
  Goodman, 1977, Races of Corn, in Corn and Corn
  Improvement)
• From a project comparing sequence diversity in
  21 genes of nine U.S. inbred lines with 16
  diversity maize landraces We found that our
  sample of U.S. inbreds contained a level of
  SNP diversity that was 77 the level of
  diversity in our landrace sample. (Tenaillon et
  al., 2001, PNAS, 989161-9166)

52
Sequence Diversity in Maize

• How has selection shaped sequence diversity in
  maize?
• Survey SNPs from 1800 genes in diverse maize and
  teosinte germplasm
• Screen 4000 candidate genes for evidence of
  selection
• Practical Goal identify genes exhibiting
  selection
• Domestication, agronomic improvement, and local
  adaptation

53
Allele Frequencies
teosinte
Domestication
landraces
Plant Breeding
modern inbreds
Unselected Gene
Domestication Gene
Improvement Gene
54
Can we develop genomic screens to identify genes
that have undergone selection?
1. Invariant SSR approach 2. Direct Sequencing
Approach What proportion of genomic sequences
that have low allelic diversity among inbreds
result from selection for domestication? Contrast
sequence diversity among teosintes, landraces,
and inbreds
55
Screening SSR primers against 12 inbred lines
1,772 total SSRs 1,053 were polymorphic (Class
I) 719 were invariant (Class II)
Invariant SSR primers
56
Invariant SSR Screening

• 470 invariant SSR primer sets
• 321 monomorphic throughout
• 60 polymorphic in both exotics and teosintes
• 14 polymorphic only in exotics
• 75 polymorphic only in teosintes (Class II-E)

Vigouroux et al. 2002. PNAS 99 9650
57
Analysis of Class II-E SSRs

• 31 Class I SSRs and 44 Class II-E SSRs
• 44 teosinte and 45 landrace accessions
• Tested for selection (loss of diversity)
• 0 Class I SSRs showed evidence of selection
• 15 Class II-E SSRs showed evidence of selection
• Extrapolated back to the 1772 total SSRs
• 1.4 genes have been selected

58
Direct Sequencing Approach

• Purpose to develop a SNP resource for the maize
  community
• Result a LOT of data!!!

59
Distribution of SNP Haplotypes (patterns)
470 maize Unigenes in 14 maize lines Mean
haplotype 4.46 gt 80 of unigenes have 2 to
7 haplotypes
For each gene, a few haplotypes account for much
of the diversity
60
Are genes with low inbred diversity enriched for
domestication and improvement candidates?
(Masanori Yamasaki, post-doc in McMullen Lab)
36 genes with no diversity among a 14-inbred
set Sequenced same region in 16 landraces, 16
teosintes, and a Tripsacum dactyloides
sample.   Test for selection on inbreds,
landraces and teosintes compared to four neutral
genes.
61
Selection Tests for 33 (of 36) Genes
5 genes were significant in both the inbreds and
the landraces (evidence for domestication
genes).   7 genes were significant in the inbreds
but not the landraces (evidence for improvement
genes). 1 additional gene was classified as
either domestication or improvement depending on
the test. 13 out of 33 genes 39 !!
Yamasaki et al. submitted
62
Selection on a Genomic Scale

• Sequenced 774 maize unigenes in 14 maize inbreds
  and 16 teosinte accessions
• Tested for selection using coalescent simulations
• Result 2-4 had experienced artificial selection
  
• Assume 59,000 genes in maize
• 59,000 x 2 1200 selected genes

Wright et al. 2005 Science 308 1310
63
Where are we going with this?

• Before genomics, 11 genes had been identified as
  selected by population genetic approaches.
• By sequencing 1000 genes, have 50 novel
  candidates.
• We need
• 1. to completely sequence the maize genome to
  identify ALL genes.
• 2. to resequence all remaining genes in multiple
  maize inbreds and teosinte accessions.

1140 more !
64
Signatures of Selection

• If selected genes were important in the past
  improvement, continued manipulation might
  contribute to future gain.
• If selected genes suffered a loss of diversity
  because of selection, they are prime candidates
  for introgressive breeding from wild relatives.
• Hypothesis manipulation of the expression of
  domestication and improvement genes will alter
  key agronomic traits

65
Selection for Amino Acid Content?

• Four genes that show evidence of selection are
  involved in amino acid biosynthesis

66
Selection for Amino Acid Content?

• Are there more genes in amino acid pathways that
  have been selected?
• Sequenced 16 genes in 28 maize inbreds, 16
  teosinte, and 2 tripsacum.
• Result we found 4 genes that may have been
  selected during domestication/improvement.

67
The Ultimate Selection Project
B73 inbred line
68
Sequence Diversity in Evolution and Crop
Improvement
Teosinte
Maize Landraces
Inbreds/Hybrids

• Sherry Flint-Garcia
• Research Geneticist
• USDA-ARS
• MU Division of Plant Sciences

Photos courtesy J. Doebley
69
Molecular Diversity SNP Single nucleotide
polymorphism InDel Insertion deletion SNPs
and Indels are used markers for genetic analysis