Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
Improving water use efficiency and water capture in
Solanum and Brassica, transgenic and QTL approaches
VeGIN meeting 05 November 2009
Growth in food production has exceeded population growth
(so far!), but production must double by 2050
Andrew J. Thompson
Scope for further growth in food production by increasing
irrigation is severely limited
What is drought resistance?
Crossover point
Ideal drought
resistant cultivar
drought-resistant
cultivar
lti
Yield gain
under stress
(drought
resistance)
Crop yield
Loss of
yield
i ld potential
t ti l
Water stress
Crop yield is closely related to seasonal water use
Yield in water limited environment can
be improved by:
ƒ Increasing the slope of the line increased water use efficiency
(ton ha-1 mm-1)
ƒ Increased
I
d use off available
il bl water
t increased water capture, e.g. deeper
roots (but may reduce river flows/aquifer
recharge)
Seasonal evapotranspiration (mm)
Topics
1. Improving water use efficiency
by increasing the production of
the plant hormone abscisic
acid (tomato, B. napus).
GM and natural variation
adaptation to
water deficit
dormancy
2. Quantitative trait loci (QTL) for
water use efficiency
(vegetable brassicas)
3. QTL for water capture (tomato)
Grain yield as a function of seasonal
water use, oilseed crops, Eastern India
From: Kar et al 2007 Ag. Wat. Manag. 87:73-82
1
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
VeGIN meeting 05 November 2009
Abscisic acid synthesis
1. Abscisic acid
?
• Demonstrate that we can engineer plants with higher ABA content
• Describe the impact of long-term high ABA on plant growth and
development, and water use
• To explore how we should optimise ABA biosynthesis to maximise crop
productivity when water is limiting
Key regulatory step:
9-cis-epoxycarotenoid
dioxygenase (NCED)
“High ABA” tomato plants created: sp::NCED lines
Thompson et al 2000, Plant J
WT - well watered
WT - drought
sp12
12 - wellll watered
t d
sp12 - drought
Xylem sap ABA (ng ml-1)
Transgenic tomato lines: sp::NCED plants have
higher abscisic acid in well-watered conditions
Closure of stomata in sp::NCED “high ABA” tomato plants has
a bigger impact on water loss than on photosynthesis, so
improves intrinsic water use efficiency of leaves
50
40
30
• Stomatal conductance
(gs) is “capped” in sp
lines
20
10
9
8
7
6
5
4
1
2
3
4
• Reducing stomatal
conductance from 600
to 300 has little
impact on rate of
photosynthesis (A)
5
Days
• sp::NCED plants become equivalent to WT plants once stress is imposed
• Physiology under well-watered conditions is different to WT
“High ABA” sp::NCED plants only close stomata in dry air when
transpiration is high (maximising photosynthesis when the risk
of water loss is low)
sp::NCED “high ABA” lines have up to 80% increase
in water use efficiency at the whole plant level
Genotype
WT
sp12
sp5
Initial biomass (g DW)
3.31
3.50NS
3.29NS
Biomass gain over 25 days (g DW)
37.96
31.37NS 34.91NS
Transpiration (kg H20 plant-1)
8.5
5.5*
4.4*
WUEp (g DW kg-1 H2O)
4.5
5.7**
8.0**
Thompson et al 2007, Plant Physiol.
2
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
VeGIN meeting 05 November 2009
In sp::NCED plants germination is delayed or prevented this is reversed by norflurazon
Water + 5 mg/l norflurazon
Water
Growth of “high
g ABA” p
plants
6 days from imbibition
Healthy seedlings can be produced using low
concentrations of the herbicide norflurazon
sp::NCED plants have delayed early growth
Nz (mg l-1)
• Shoot dry weight at 10 DAE was
greatly reduced
(sp12 -28%, sp5 - 54%)
2.0
• By 40 DAE,
DAE dry weight
differences were non-significant
(sp12 +4%, sp5 -10%)
1.0
0.5
n = 10
sp::NCED plants have delayed early growth
Cotyledon expansion and early growth may be inhibited by ABA signalling pathways:
e.g. post-germination growth arrest (Arabidopsis)
40-50% reduction in cotyledon size:
“…ABA…can also reversibly block growth
during a narrow developmental time interval
gg
germination and before the onset of
following
vegetative growth”.
Luis Lopez-Molina et al 2001 PNAS 98:4782-4787
n = 10
3
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
Long-term biomass accumulation of sp::NCED plants is normal,
leaf area is increased (glasshouse-grown plants in pots)
Increased leaf expansion in sp::NCED plants
n=9
Petiole length and epinasty
25
3000
20
2500
15
10
5
25
AC
p < 0.001
1500
1000
15
10
5
500
0
0
0
WT
SP5
SP5
20
2000
Leaf numbe
er
2
Leaf area (c
cm )
Plant DW (g)
n=9
VeGIN meeting 05 November 2009
50
WT
100
150
200
250
300
350
Leaf length (m m )
SP5
WT
SP5
Sp5 has higher specific leaf area (cm2 g-1)
Conclusions from sp::NCED construct
Moderate increases in ABA:
Positive effects
• dramatically increase water use efficiency by limiting gs at high VPDs
• little or no impact on long-term biomass (well watered CE/glass)
• enhances
h
expansion
i off lleaves ((turgor/ethylene?)
/ h l
?)
• enhances biomass under water limited conditions (drought resistant)
• increases root hydraulic conductivity
Further optimization of ABA biosynthesis for
productivity – can we separate the good from the
bad?
Negative effects
• inhibition of germination and seedling establishment (increased soil evap.)
NCED over-expression in tomato using a light-regulated gene
promoter (rbcS::NCED) to avoid germination problems
Tung et al 2008, PCE
rbcS::NCED lines: too much ABA severely inhibits growth
High ABA
Wild-type
4
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
Optimising ABA content with
different gene promoters:
promoter X
PX::NCED1
(het)
SP::NCED
(SP5)
VeGIN meeting 05 November 2009
Avoiding GM: exploiting natural variation in NCED activity
• NCED alleles from ten wild species are being introgressed into tomato c.v.
Ailsa Craig
WT
S. lycopersicum S. pimpinellifolium S. galapagense
(LA1673)
(LA2646)
(LA317)
S. peruvianum
Northern
population
(LA2561)
S. chilense
(LA2884)
S. habrochaites
(LA1315)
S. neorickii
(LA247)
S. pennellii
(LA1376)
S. chmielewskii
(LA1306)
S. lycopersicoides
(LA2772)
Spooner et al., 2005 TAXON 54 (1)
Avoiding GM: exploiting natural variation in NCED activity
(preliminary data)
- comparing the effect of three NCED alleles on stomatal conductance in nearisogenic lines
A
B
A
“High ABA” Arabidopsis thaliana transgenics: how do
they match up to natural variation in WUE?
C
Allele
Genetic variation in water use efficiency in 96 Arabidopsis ecotypes
ABA content of rosette leaves
4.8
4.4
4.0
3.6
32
3.2
2.8
2.4
2.0
614
649
645
619
617
633
632
631
613
618
648
567
593
628
603
611
630
641
606
622
588
610
650
612
576
627
580
625
646
598
644
642
601
629
590
596
564
653
624
604
623
626
568
605
656
658
659
608
638
636
643
654
635
594
602
587
615
582
600
584
637
595
647
577
599
566
616
571
651
607
634
589
575
592
652
570
565
569
585
639
591
620
655
572
621
574
581
578
586
573
Gravimetric WU
UEp (mg DW/g water)
“High ABA” transgenic Arabidopsis thaliana
Very significant genotype effect on WUE, but also on anatomical traits (Chlorophyll
content, Leaf Thickness, water content, etc)
5
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
“High ABA” transgenic Arabidopsis thaliana always have higher water
use efficiency
VeGIN meeting 05 November 2009
“High ABA” Arabidopsis plants grow better under terminal drought
15-26% penalty in
yield potential
Crossover point
around 50%
reduction in dry
weight
30-50% increase in WUE
under all conditions
“High ABA” Arabidopsis transgenic plants have much greater
WUE than is found in nature
2. Quantitative trait loci for water use efficiency
Brassica oleracea
(vegetable brassicas)
Arabidopsis
• Ecotypes may have been under
selective pressure to use soil water
rather than leave it in the soil for
their competitors to use
10 cm
1 cm
Field crop
Model plant
Comparative genetics/genomics
QTL analysis in two-parent mapping populations
A
Parents
X
A12 x GD33 DH population – QTL trials
(A12 = Chinese kale, GD33 = broccoli/calabrese)
B
F1
2 x sites
2 x treatments (well-watered & drought)
2 x transplantings
2 x replicates per transplanting
8 plants per plot
(heterozygous)
Automatic rain shelters – Gleadthorpe (sandy soil)
microspore
culture, conversion
to doubled haploid
(DH) lines
Spanish tunnels – Kirton (heavy soil)
Selfing for recombinant
inbred lines (F7)
Mapping population
6
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
VeGIN meeting 05 November 2009
A12 x GD33 DH population: δ13C QTL
Trait assays to asses both WUE and indicators of yield potential
•
•
C1
C2
C3
C4
C5
C6
C7
C9
C8
0
WUE is a ratio, selection on this alone could drive gs down too far, leading to
loss of yield (very low gs will limit A)
Other trait data is needed to assess components of the ratio
50
leaf chl (SPAD), leaf thickness (MLA)
biomass
WUEIntrinsic =
A
gs
≈
A
T
x VPD
100
δ18O Gleadthorpe
δ18O Kirton
δ13C Gleadthorpe
δ13C Kirton
SPAD Gleadthorpe
SPAD Kirton
Specific leaf area Gleadthorpe
Specific leaf area Kirton
Biomass response Gleadthorpe
Biomass response Kirton
A = CO2 assimilation rate
δ18O
δ13C
gs = stomatal conductance
T = transpiration
cM
150
VPD = vapour pressure deficit
A12 x GD33 DH population: δ13C and δ18O QTL
C1
C2
C3
C4
C5
C6
C7
Mean
Drought
Control
δ13C, δ18O, SPAD, SLA & Biomass response QTL
C1
C9
C8
0
C2
C3
C4
C5
C6
C7
C9
C8
0
H2
M
50
H1
50
H2
100
cM
100
δ18O Gleadthorpe
δ18O Kirton
δ13C Gleadthorpe
δ13C Kirton
SPAD Gleadthorpe
SPAD Kirton
Specific leaf area Gleadthorpe
Specific leaf area Kirton
Biomass response Gleadthorpe
Biomass response Kirton
150
Mean
Drought
cM
δ18O Gleadthorpe
δ18O Kirton
δ13C Gleadthorpe
δ13C Kirton
SPAD Gleadthorpe
SPAD Kirton
Specific leaf area Gleadthorpe
Specific leaf area Kirton
Biomass response Gleadthorpe
Biomass response Kirton
150
Control
Mean
Drought
Control
Investigation of top of C7 with substitution line SL118
Co-localisation of traits on C7
C7
C1 C2
δ13C N X G Field
0
C3
C4
C5
C6
C7
C8
C9
0
• A12 parent provides:
• high WUEi
• low transpiration (high δ18O).
50
δ13C N X G Glass
• no effect on leaf traits or
biomass detected
50
• supported by QTL in a different
population
100
cM
150
100
δ18O Gleadthorpe
δ18O Kirton
δ13C Gleadthorpe
δ13C Kirton
SPAD Gleadthorpe
SPAD Kirton
Specific leaf area Gleadthorpe
Specific leaf area Kirton
Biomass response Gleadthorpe
Biomass response Kirton
Mean
Drought
cM
150
SL118: introgressions of GD33 DNA (red) in
an A12 background (green)
Control
7
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
VeGIN meeting 05 November 2009
Line SL118 (with a substitution on chr. 7):
Line SL118 (with a substitution on chr. 7):
• A12 alleles causes 20-25% lower stomatal conductance
relative to GD33
• A12 alleles
causes 15-20%
higher WUE
relative to GD33
LSD (5%), n = 10
Genetic variation in intrinsic water use efficiency in vegetable brassicas
Association mapping in B. oleracea?
WUE was assessed with carbon isotope composition (δ13C) in two field transplantings at Kirton
(107 accessions selected from the vegetable gene bank at Wellesbourne)
Each point represents a single genotype:
there is consistency between the two
transplantings, and the range of value
indicates differences in actual water use
efficiency of > 50%
BoDFS: Brassica oleracea diversity foundation set
• Collect trait data from BoDFS and genotype within regions of interest
• Search for association between trait values and specific haplotypes
Comparison of two-parent QTL mapping and allelic association mapping on C9
3. Traits for water capture
C9
0
50
A12 provides:
• High WUE
• High SLA
• Low transpiration (Kirton)
• Allelic association hits
correspond to A x G QTL
•
Root depth and root length
density at depth
•
Ability to penetrate sub-soil
•
Ability to conduct water to
the shoot
AxG QTL Key
100
cM
150
δ18O Gleadthorpe
δ18O Kirton
δ13C Gleadthorpe
δ13C Kirton
SPAD Gleadthorpe
SPAD Kirton
Specific leaf area Gleadthorpe
Specific leaf area Kirton
Biomass response Kirton
AxG QTL
Association
Mapping
Drought
Control
Image from Larcher 4th ed
8
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
VeGIN meeting 05 November 2009
50 introgressed lines screened in 64 rhizotrons
6 reps per line, 2 chromosomes per run
S. pennellii x S. lycopersicum (M82) introgressed line (IL) population
Dani Zamir, The Hebrew University of Jerusalem
S. pennellii growing on dry
rocky slopes in central Peru
S. lycopersicum (LA3475, cv M82)
Determinate (bush)
cultivated field tomato
Images courtesy of the TGRC, UC Davis
QTL for root traits
Image analysis of root images
using a custom-designed MATLAB based program
12 chromosomes of Solanum lycopersicum
(Nick Parsons, University of Warwick)
Master flash
unit
Slave flash
unit
Flash light path
Digital camera
Up arrow indicates
increasing effect of
S. pennellii introgression
Root length (2 m tube hydroponics)
Root penetration (geotextiles)
Rhizotron
Root distribution (soil rhizotrons)
Side elevation
Photographic hood
Vertical profile level
A QTL for the vertical distribution of
the tomato root system
Data from imaging of soil-grown plants
in rhizotrons
0
2
4
6
8
10
12
0
2
4
6
8
10
12
0
2
4
6
8
10
12
0
2
4
6
8
10
12
Shoot biomass
Root mass fraction (deep trough hydroponics)
7 days
Greater root penetration of introgression 4-1
M82
IL4-1
14 days
Identification of QTL for
penetration of
geotextile membranes
200
150
100
50
0
M82
IL4-1
IL7-1
IL9-1
21 days
28 days
0
A
Root mass fraction (soil rhizotrons)
Number of penetrating roots
Captured root image
Processed root image (pixel
density at 12 depths)
5
10
15
20
25
Root pixel density
9
Improving water use efficiency and water capture Solanum
and Brassica, transgenic and QTL approaches.
Dr Andrew J Thompson
Fine mapping of root penetration QTL (NORP1) on chr 4
VeGIN meeting 05 November 2009
• Root hydraulic conductivity is important for water capture and for regulating
shoot water status under variable transpiration rates
WSS1011
WSS1090
IL4-1-1
IL4-1
• Root hydraulic conductivity is strongly regulated by abscisic acid
SSR72 (0)
High ABA maize lines recover leaf water
potential faster when rehydrated (Parent et
al 2009)
ABA deficient roots are worse at
extracting water from drying soil in 3way grafts (Holbrook et al 2002)
NORP1 (0)
At5g41270 (0)
Hero (6.7)
T0529 (8)
SSR43 (9.4)
CT175 (40)
TG182 (46)
14-fold range of root hydraulic conductivity (Lpr) from ABA deficient or
ABA over-producing tomato lines
Conclusions
400
Lp
pr (% WT)
300
200
•
Much of agriculture uses water in an unsustainable way, breeding for WUE
is one approach to address this issue
•
Transgenics: over-expression of NCED doubles WUE
– a strong effect of a single gene! Beyond natural variation.
– but
b t need
d tto understand
d t d th
the pleiotropic
l i t i effects
ff t and
d iimpactt on productivity
d ti it
– more subtle engineering to enhance benefits and limit negative effects
100
•
sp
5
sp
12
co
nt
ro
l
no
ta
bi
lis
0
Marker assisted selection for WUE will be difficult, but rewards are large
– there is a wealth of natural genetic variation in WUE
– many genes with small effects
– different genes will be important in each environment – which trait to measure?
Data from:
Tal and Nevo (1973) (notabilis)
Thompson et al (2007) (NCED overexpression)
Acknowledgements
ABA transgenics:
Liz Harrison
John Andrews
Howard Hilton
Alison Jackson
Ian Taylor
Alan Burbidge
Brassica Quantitative genetics:
Dave Pink, Guy Barker
Graham Teakle
Philip White (SCRI)
Jean-Charles Deswarte
Carol Ryder, Howard Hilton
Chemical genetics:
Martin Sergeant
Tim Bugg (Warwick Chemistry)
Stable isotopes:
Graham Farquhar (ANU)
Warwick/Nottingham
joint PhD students
Matthew Jones
Sajjad Awan
Swee-Ang Tung
Rachel Smeeton
Charlotte White
Root trait analysis:
John Andrews
10