Page 1
Background, and why I am here
Insect resistance in Arabidopsis
Epistasis and growth rate
Insect resistance in the field
Approaches to drought tolerance in rice
Thomas Mitchell-OldsInstitute for Genome Sciences & PolicyDepartment of BiologyDuke University
[email protected]
Advances in biotic and abiotic stress tolerance
Page 2
University of Wisconsin: Ph.D. in Botany; Plant Breeding & Plant Genetics Postdoc in Human Genetics
Page 3
Department of Genetics and Evolution
Max Planck Institute of Chemical Ecology
Page 5
Complex trait variation for
Insect resistance
Flowering time
Drought tolerance
NSF NIH
Page 6
146
276
150
Undernourished population in millions
Why am I here?
1) Research to benefit human welfare
2) Rice is similar to Arabidopsis
3) Phenotypes in the field
Page 7
Why are quantitative traits polymorphic?
Page 8
LOD
sco
reInsect resistance
Position on Chromosome V
Correspondence of QTLs for insect resistance & glucosinolate levels
Position on Chromosome V
GlucosinolatesR
S-Glucose
NOSO3-
Insect Resistance in Arabidopsis
Page 9
Myrosinase
Otherproducts
Thiocyanates(RSCN)
Nitriles(RCN + S)
Isothiocyanates(RNCS)
RS-H
NOSO3-
Insect resistance and chemical defense
RS-Glucose
NOSO3-
Arabidopsis glucosinolates:
From amino acid precursor
Influence insect resistance
Genetically variable
Evolve non-neutrally
Breakdown ofglucosinolates leads
to toxic products
Page 10
Cloning QTL for insect resistance and glucosinolatesLO
D s
core
Insect resistance
Position on Chromosome V
Position on Chromosome V
Glucosinolates
Initial QTL mapping
Fine scale mapping of GS QTL
Positional cloning
Verification by transformation
QTL cloning is easier for secondary traits
(like GS) than for insect damage or
yield under drought
Page 11
glucosinolate concentration
MAM2 gene controls
& resistance to generalist herbivores
but not specialist herbivores,
and does not impact growth rate
Kroymann et al. 2003 PNAS
Page 12
Why are complex traits polymorphic?
Data come from cloned QTLs
Hence no information on small effect QTL
Growth rate QTL in Arabidopsis thaliana
Kroymann & Mitchell-Olds, Nature, 2005
Juergen Kroymann
Epistasis and growth rate
Page 13
0
1
2
3
4
5
6
7
8
0 25 50 75 100 125 150 175 200 225 2500
1
2
3
4
5
6
7
8
0 25 50 75 100 125 150 175 200 225 250
F r
atio
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
0 25 50 75 100 125 150 175 200 225 250
position [kb]
Growth rate QTL in the MAM region
MAM2 genedoes not impact
growth rate
Page 14
0
1
2
3
4
5
6
7
8
0 25 50 75 100 125 150 175 200 225 2500
1
2
3
4
5
6
7
8
0 25 50 75 100 125 150 175 200 225 250
F r
atio
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
0 25 50 75 100 125 150 175 200 225 250
position [kb]
Two Growth rate QTL in the MAM region
Right QTL: grLeft QTL: gl
-0.06
-0.03
00.03
0.06
Mas
s (L
er-C
ol)
[mg
]
-0.06
-0.03
00.03
0.06
Mas
s (L
er-C
ol)
[mg
]
-0.06
-0.03
00.03
0.06
Mas
s (L
er-C
ol)
[mg
]
Tightly linked QTL show opposite effects
Ler > ColCol > Ler
Page 15
grLC:
grCL:
grLC:
grCL:
Crosses to identify the QTL on the right
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
RecombinationBreakpoints
Right QTL: gr
Segregatingregion
Reciprocalflanking
backgrounds
Page 16
At5g23140At5g23180
At5g23150 At5g23160 At5g23170 At5g23190 At5g23200 At5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC1grLC2grLC3grLC4grLC5
grCL6grCL7grCL8grCL9
grCL(6-9)
a
b
c
At5g23140At5g23180
At5g23150 At5g23160 At5g23170 At5g23190 At5g23200 At5g23210At5g23150 At5g23160 At5g23170 At5g23190 At5g23200 At5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC1grLC2grLC3grLC4grLC5
grCL6grCL7grCL8grCL9
grCL(6-9)
a
b
c
RecombinationBreakpoints
N > 7000
Page 17
Interval1 2 3 4 5 6 7 8 9
P value n.s. P<0.05 n.s. n.s. n.s. n.s. P<0.05 n.s. n.s.
Col 140 199
Ler 161 189
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
Interval 7 is significant in both reciprocal crosses
Page 18
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
No QTLNo QTL
1 - synonymous SNP1 - intergenic SNP1 - intron indel of 1 bp Epistatic QTL
QTL
serine/threonine protein kinase
Page 19
Interval1 2 3 4 5 6 7 8 9
P value n.s. P<0.05 n.s. n.s. n.s. n.s. P<0.05 n.s. n.s.
Col 140 199
Ler 161 189
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
at5g23140at5g23180
at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210at5g23150 at5g23160at5g23170 at5g23190at5g23200 at5g23210
LJ47C LJ01C CJ10LCJ45L
CJ36LLJ52C
LJ30C CJ40L CJ26L LJ32C LJ35C
222 236230 234232224 228226218 220212 216214206 210208 [kb]222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 222 236230 234232224 228226218 220212 216214206 210208 [kb]
grLC: 1 2 3 4 5
6 7 8 9grCL:
Allelic effect is reversed depending on genetic background
Epistatic growth rate QTL
Page 20
Growth rate QTL within 170 kb
Two tightly linked growth rate QTL
Undetectable with standard QTL mapping
Too small
Opposing effects
Epistasis: QTL change sign depending on background
If typical, then:
Complex traits are massively polygenic
Epistasis is fundamental
Epistasis is fundamental for
drought tolerance(e.g., qtl12.1)
This will complicate association studies
Page 21
Insect resistance in natural environments
Why are quantitative traits polymorphic?
Page 22
Ecological model systems
Clone QTLs
Measure fitnessof allelesin natural populations
Page 23
Genus Boechera (formerly Arabis)
Close relative of Arabidopsis
Genome ~ 250 Mb
Positional cloning
Transformation
Undisturbed natural populations
Diploid apomixis
Page 24
0 20 40 60 80 100
Percent Leaf Damage
0
20
40
60
80
100F
req
ue
ncy
Resistance variation in Boechera stricta
Trichoplusia ni
Page 25
Glucosinolate quantity in Boechera stricta
6 7 8 9 10 11 12 13
Log Total Glucosinolate
0
50
100
Fre
qu
enc
y
About 30% of variation is under genetic control
Page 26
BCAA: branched chain amino acids: Val, Leu, Ile
0.0 0.2 0.4 0.6 0.8 1.0
Proportion of GS from BCAA
0
10
20
30
40
50
60F
requ
ency
Glucosinolate type in Boechera stricta
Met-GS BC-GS
BCMA locus:Branched Chain
Methionine Allocation
T. ni: generalist herbivore BCMA controlsresistance togeneralist herbivore
Met allele is more resistant
Page 27
40N
120W 110W
Montana
Colorado
Page 28
40N
120W 110W
Montana
Colorado
Near Isogenic lines for BCMA:
2304 plants into original populations
Measure herbivore damage & fitness
Today: phenotypic selection
(Still scoring BCMA genotypes)
Page 29
Fitness of alleles in their ancestral environments
MontanaColorado
Page 30
CO MT0
5
10
15
Per
cen
t L
eaf
Dam
age
Susceptible BC allele
Resistant Met allele
High herbivore pressure
Lower herbivore pressure
N = 2,032 plants; 47,109 leaves
Page 31
0 10 20 30 40 50 60Percent Leaf Damage
0.0
0.5
1.0
1.5
Via
bilit
y C
o mpo
n en t
of
Fi tn
ess
ANOVA: significantnatural selectionin the wild, P < 10-9
Colorado: Herbivory reduces survival
Montana: No natural selection on herbivore resistance
Su
rviv
al
993 plants in ColoradoSelection gradient:1% of leaf area removedreduces fitness by 1.1%
How does insect resistance influence
fitness?
How does insect resistance influence
fitness?
How do secondary traits influence yield
under drought?
Page 32
CO MT0
5
10
15
Per
cen
t L
eaf
Dam
age
Susceptible BC allele
Resistant Met allele
High herbivore pressure
Lower herbivore pressure
Strong selection for resistance
No selection for resistance
Page 33
BCMA summary
Positional cloning of BCMA
Influence of trait on fitness in field environment
Page 34
Approaches to drought tolerance in rice
1) Yield in managed stress environments
2) Trait dissection into component traits
Page 35
360 degrees using computer controlled motorized turntable
Benfey labDuke Univ
Nondestructive, automated analysis of root architecture
Page 36
Indicas
Teqing
93-11
IR64Caipo
Jefferson
JaponicasBenfey Lab
Page 37
Topology-based image analysis
High throughput phenotyping
enables cloning of QTLsBenfey Lab
Page 38
Photo by J. Lamo & O. Michael
Approaches to drought tolerance in rice
1) Yield in managed stress environments
2) Trait dissection into component traits
3) Dissecting drought environments
Develop laboratory models of drought environments
Understand their effects on drought tolerance
Quantify axes of environmental variation
Page 39
Ruiz Corral, et al. 2008, Crop Sci 48:1502
Environmental Conditions PC1
Env
ironm
enta
l Con
ditio
ns P
C2
Shapes indicategenetic groups
Different maize races are grown in different environments
Quantify axes of
environmental
variation
Page 40
Hypothetical example: environmental dependence of drought tolerance
0 1 2 3 4 5 6 7Environment PC1
1
2
3
4
5
6
Env
iron m
ent
PC
2
Grow NILs in multiple environments
Plot nurseries based on environments
Measure yield advantage at each site
3D plot: yield advantage vs. environment
Page 41
Hypothetical example: environmental dependence of drought tolerance
What aspects of the
environment control
the yield advantage?
Develop laboratory
models of drought
environments
Page 42
Approaches to drought tolerance in rice
1) Yield in managed stress environments
2) Trait dissection into component traits
3) Dissecting drought environments
4) Integrated pedigrees for QTL & association analysis
- Cannot use association approach for traits that
differ between indica, japonica, or other groups
- Epistasis is a major problem for association studies
Page 43
Bao-Hua SongKasavajhala PrasadAntonio ManzanedaCarrie Olson-Manning
Philip Benfey
Eric SchranzAaron WindsorJuergen Kroymann Maria ClaussKarl SchmidDan KliebensteinJonathan Gershenzon
NSF NIH Duke MPG
Kathy SpringerToyin Aremu-ColeMolly Mitchell-OldsSara Mitchell-OldsJanoo NaqviSlater HurstMichael CameronJ. LutkenhausLaura SaucierElizabeth BallwegCheng-Ruei Lee
Antje Figuth Susi Donnerhacke Kerstin WenigerSylke DietelMarion Kupper Sylke DietelGrit SchubertKatja SchwarzerNancy RichterD. SchnabelrauchTabea BirkKerstin Weniger