Harnessing interdisciplinary approaches for germplasm development

International Institute of Tropical Agriculture – Institut international d’agriculture tropicale – www.iita.org

“IITA - the appliance of science”

Harnessing interdisciplinary approaches for

germplasm development

Sarah Hearne


Development of varieties that are:

High yield potential

High yielding under farmer

conditions

Pest and disease resistant

Resistant to abiotic stress

Effective use of nutrients

Market preferred characteristics

Nutritious

Desired duration

The bottom line


Breeding

Agronomy

IPM

Pathology

Weed science

Entomology

Biometrics

Bioinformatics

Nematology

Physiology

Biotechnology


Tools to

accelerate /

better target /

improve

efficiency of

germplasm

improvement

Molecular tools -markers

Statistical methodologies

Data management, analysis and

decision support tools

Phenotyping

Methodologies

Benchmark sites

Breeding and selection

methodologies


How these tools

support

breeders?

Rapid targeted selection of segregants

Molecular breeding

Rapid integrated data analysis

Stress interactions / host-pest interactions

Target breeding for specific stress

complexes / pest genotypes

Selection of parents

Heterotic grouping

Diversity assessment – allelic / general

Predictive breeding

Variety production and seed systems

DUS (distinct, uniform, stable)& VCU (value for cultivation and use)

Seed purity

Impact assessment


Three years-

outputs and

outcomes

Maize molecular breeding

projects, methodologies

applications

Plans for the future

Maize

Cowpea

Cassava

Musa

Striga

Some highlights

Maize genetic diversity

global diversity

Striga

distribution, diversity

pathogenicity

Resource development and use

cowpea, cassava and musa


DTMA

Interdisciplinary team with many talents working

across centers!


DTMA

Polygenic with high degree of epistasis

GxE issues

Phenotyping complexity – precision issues

Can’t apply markers across populations for MAS

Drought

Multiple interconnected themes within the initiative

Genomics

Breeding

Seed systems

Impact assessment


Phenotyping Increase throughput

Enhance precision

Benchmark sites with precision irrigation

-Nigeria, Kenya, Zimbabwe, Zambia, Mexico


Phenotyping

New tools

implemented

Collaboration with

breeders and

physiologists

Biomass/ senescence

Spectroradiometry – NDVI – rapid biomass

assessment

TranspirationLeaf / canopy temperature

NIRS & ash contentWater use

efficiency /

effectiveness


Molecular

breeding

DTMA- largest public sector molecular breeding

project for maize in the world

Monitor and coordinate phenotyping, genotyping,

data analysis and breeding turnaround across

projects

Work with physiologists on phenotyping methods

and site characterization

Work with molecular team on genotyping

technologies and bottlenecks

Work with biometricians on issues of pop size

and marker number and on simulations of new

approaches using real data

Work with bioinformaticians on issues of data

handling and processing


Molecular

breeding

What strategy to use for molecular

breeding?

What markers to use?

How to do the genotyping?

What traits to look at using markers?:

Make adapted materials more

drought tolerant

Make drought tolerant materials

more disease tolerant


Molecular

breeding

What strategy to use for molecular

breeding?

What markers to use?

How to do the genotyping?

What traits to look at using markers?:

Classic QTL

ABQTL

MARS

GWS

For breeding - SNP

Both drought and disease

-list of diseases and populations


Molecular

breeding

Molecular breeding training course 2008

DTMA scientists

NARS and private sector partners

Side meetings

Breeding strategy?

Evaluated our own QTL data and looked at

necessary population sizes and marker

numbers – NO classic QTL

Private sector present advising

ABQTL and MARS breeding strategies

Defined a template for population planning

and molecular breeding implementation

Worked with breeders to define possible

populations – 35 potentials


Select

parents

Make

populationsPhenotype test cross

populations

in 3 or more

multilocation trialsGenotype parents

Using ~1500 SNP markers

Find marker trait

associations for

drought tolerance

Recombine the best materials

using marker data to stack or

pyramid favorable markers

for two or more cycles

(no phenotyping)

Isolate new lines

and phenotype

under drought

Use lines to

create new

drought tolerant

varieties

Marker assisted recurrent selection: MARS

-new lines

Genotype populations

Using 200-300 polymorphic

SNP markers

Seed based DNA extraction


Y1

Y2

Y3

mid p

Sel C0

Breeder lines

MARS C2S1 lines

If you have 20 regions under selection, freq of optimum

genotype goes from 1 per trillion in cycle 0 to 1 in 5 in

cycle 3

Molecular

breeding

Power of MARS


Select

parents

Make

populationsPhenotype test cross

populations

in 3 or more

multilocation trialsGenotype parents

Using ~1500 SNP markers

Find marker trait

associations for

drought tolerance

Recombine the best materials

using marker data to stack or

pyramid favorable markers

for two or more cycles

(no phenotyping)

Isolate new lines

and phenotype

under drought

Use lines to

create new

drought tolerant

varieties

Marker assisted recurrent selection: MARS

-new lines

Genotype populations

Using 200-300 polymorphic

SNP markers

Seed based DNA extraction


Molecular

breeding

Annual meeting 2008

Scheduling of phenotyping and genotyping

and cost of genotyping

Some seasons ~6500 individuals to

genotype

3 weeks to extract DNA and genotype

Genotyping option

Illumina golden gate – 1536 SNP markers


Molecular

breeding

Illumina golden gate – 1536 SNP markers


Molecular

breeding

Workshop Feb 2009

SNPlex no longer an option

BeadXpress

Illumina system 384 markers

Single plex assays – KBiosciences

DTMA – buy BeadXpress

Analysis of throughput and cost

Discussion with private sector

Visit to KBioscience

Presented data at 2009 DTMA annual

meeting in Zim - use KBiosciences


Molecular

breeding

DTMA – 20 molecular breeding projects. 18

MARS/GWS (explained in next slide), 2 ABQTL (line

conversion) – WA had 5 MARS/GWS populations

Two genotyping platforms-

Illumina golden gate-

One diverse 1536 illumina OPA, 1330 good

SNP

KASPar – single plex primer extension based

assays designed for 1111 SNP

Will be using new 60k infinium assay in 2010

Genotyping by sequencing under evaluation


Building on MARS

populations

Some of the hit list:-

MSV

Striga hermonthica

GLS

Nematodes

We have high

breeding value

segregating

populations

We have

genotype data

What other high

value traits we

can phenotype

for…


Molecular

breedingTolerance to MSV

Breeder, Virologist, Biometrician

and Mol Geneticist

Abebe + – populations, field phenotyping

Lava – Screenhouse phenotyping and indexing,

field phenotyping

Sarah – Candidate marker identification,

population genotyping, tracking, data curation

Sarah and Jose – Data analysis and marker

identification

Contribute knowledge and markers for simple

traits to incorporate into breeding work

Guy – Data handling and manipulation


Molecular

breedingResistance and tolerance to Striga

Breeders, Physiologist, Biometrician

and Mol Geneticist

Abebe and Baffour – populations, field

phenotyping

Sarah – Screenhouse phenotyping

Sarah – Candidate marker identification,

population genotyping, tracking, data

curation

Sarah and Jose – Data analysis and

marker identification

Contribute knowledge and markers for

identified traits to incorporate into breeding

work


GLS Field phenotyping issues - Reliability

Not a simple trait – or a single species (Lava)

Development of improved phenotyping

e.g. HTP detached leaf assay for GLS

evaluations with pathologist - Ranajit

Enable controlled infestation with different

genotypes of pathogen (pathotypes / species

with agroecological niches)


CML 488 maize inbred line

Un-infected Meloidogyne infested

Black

Lesions

Reduced root

mass

“Root galling”

Nematodes

Pratylenchus

Meloidogyne


Nematodes

Interaction

between drought

and nematode –

flowering date

and ASI

Nematode and drought interaction

Breeders, Nematologist, Physiologist,

Biometrician and Mol Geneticist

Abebe and Baffour – lines – parents of MARS

populations

Danny – Screenhouse phenotyping-

nematodes

Sarah and Jorge – Data analysis

Contribute knowledge and markers for

identified traits to incorporate into breeding

work

Sarah – Screenhouse phenotyping-

drought


What other

applications to

support

breeders?

Selection of parents

Heterotic grouping

Diversity assessment – allelic / general

Predictive breeding

Variety production and seed systems

DUS (distinct, uniform, stable)& VCU (value for cultivation and use)

Seed purity

Impact assessment

Genotyping of potential parents

Testing the quality of outgrower produced

hybrid seed in Zimbabwe

Genotyping IITA released maize lines

using 60k SNP chip to facilitate tracking of

IITA germplasm for impact assessment


Numbers,

numbers

everywhere…

Molecular breeding is a balancing

act

$How do we balance competing

demands for funds and optimize

genetic gain?


Analysis of pop

size and marker

density on QTL

detection

0

10

20

30

40

50

60

70

80

90

100

40

14

02

40

34

04

40

54

04

01

40

24

03

40

44

05

40

40

14

02

40

34

04

40

54

04

01

40

24

03

40

44

05

40

40

14

02

40

34

04

40

54

04

01

40

24

03

40

44

05

40

40

14

02

40

34

04

40

54

04

01

40

24

03

40

44

05

40

Po

we

r (%

)

Population size

MD=5 cM

MD=10 cM

MD=20 cM

MD=40 cM

PVE=1% PVE=2% PVE=3% PVE=4% PVE=5% PVE=10% PVE=20% PVE=30%

A B

C D

E F

G H

0.100.150.200.250.300.350.400.450.500.550.600.650.70

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Esti

mat

ed

ge

ne

tic

eff

ect

Population size

MD= 5 cM

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Stan

dar

d e

rro

r o

f e

stim

ate

d e

ffe

ct

Population size

PVE=1%

PVE=2%

PVE=3%

PVE=4%

PVE=5%

PVE=10%

PVE=20%

PVE=30%

MD= 5 cM

0.100.150.200.250.300.350.400.450.500.550.600.650.70

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Esti

mat

ed

ge

ne

tic

eff

ect

Population size

MD= 10 cM

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Stan

dar

d e

rro

r o

f e

stim

ate

d e

ffe

ct

Population size

PVE=1%

PVE=2%

PVE=3%

PVE=4%

PVE=5%

PVE=10%

PVE=20%

PVE=30%

MD= 10 cM

0.100.150.200.250.300.350.400.450.500.550.600.650.70

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Esti

mat

ed

ge

ne

tic

eff

ect

Population size

MD= 20 cM

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Stan

dar

d e

rro

r o

f e

stim

ate

d e

ffe

ct

Population size

PVE=1%

PVE=2%

PVE=3%

PVE=4%

PVE=5%

PVE=10%

PVE=20%

PVE=30%

MD= 20 cM

0.100.150.200.250.300.350.400.450.500.550.600.650.70

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Esti

mat

ed

ge

ne

tic

eff

ect

Population size

MD= 40 cM

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

20

40

60

80

10

01

20

14

01

60

18

02

00

22

02

40

26

02

80

30

03

20

34

03

60

38

04

00

42

04

40

46

04

80

50

05

20

54

05

60

58

06

00

Stan

dar

d e

rro

r o

f e

stim

ate

d e

ffe

ct

Population size

PVE=1%

PVE=2%

PVE=3%

PVE=4%

PVE=5%

PVE=10%

PVE=20%

PVE=30%

MD= 40 cM

Huihui Li, Sarah Hearne, Yunbi Xu, Marianne Bänziger, Zhonglai Li, & Jiankang Wang, Heredity in press


MAS versus

GWSMAS

Only track significant markers

GWS

All markers have value

Which is optimal?

How many cycles of selection?

Best estimate?

BLUE (GxE), BLUP (no GxE but

marker by marker)

Selection index?


MAS versus

GWSSimulations

Cerón-Rojas, J. J., Crossa, J., Alvarado, G., Burgueño, J., Wang, J., Atlin, G., Bänziger, M., Hearne, S. J. & Davenport, G., F., Sub Crop Sci

Fig.1

-5

0

5

10

15

20

25

30

325

330

335

340

345

350

355

360

0 1 2 3 4 5 6Cycle of selection

YB 400MM BLUE YB 40MM BLUE

YA 400MM BLUE YB 400MM BLUP

YB 40MM BLUP YA 400MM BLUP

EPP

GY

ASI

GY

(gra

ms

pe

rp

lot)

EP

P (

nu

mb

er)

AS

I (d

ays)

Fig. 1a


MAS versus

GWSGWS better than MAS in all situations

400 markers better than 40

BLUP better than BLUE when looking at

well watered environments

BLUE better than BLUP when looking at

drought stressed environments

Genetic gain for grain yield up to cycle six

Fig.1

-5

0

5

10

15

20

25

30

325

330

335

340

345

350

355

360

0 1 2 3 4 5 6Cycle of selection

YB 400MM BLUE YB 40MM BLUE

YA 400MM BLUE YB 400MM BLUP

YB 40MM BLUP YA 400MM BLUP

EPP

GY

ASI

GY

(gra

ms

pe

rp

lot)

EP

P (

nu

mb

er)

AS

I (d

ays)

Fig. 1a


Three years-

outputs and

outcomes



applications


Maize

Cowpea

Cassava

Musa

Striga

Some highlights


global diversity

Striga


pathogenicity




Maize diversity Maize from Americas, Africa, Asia, Europe

and teosintes

Group of collaborators from ten different

institutes

Population geneticists, breeders, genebank

Curators, molecular geneticists, GIS

specialists

How has maize migrated across the

globe?

What is the pattern of global diversity?


Maize diversity Maize from Americas, Africa, Asia, Europe

and teosintes

Group of collaborators from ten different

institutes

Population geneticists, breeders, genebank

Curators, molecular geneticists, GIS

specialists

How has maize migrated across the

globe?

What is the pattern of global diversity?


Maize diversity

Africa

Globally landraces are in general tropical

in origin

Twelve main landrace clusters exist in the

world


Maize diversity

Africa

Western Sub-Saharan Africa -tropical

maize

Diversity data strongly supports hypothesis

of introduction via Portuguese slave

West Africa very distinct - Sao Tome and

Cape Verde – groups originating from both

can be defined:- rapid differentiation of

original gene pools- environment and

human uses

East Africa – data suggests direct diffusion

of US maize after World War II – potential

replacement of tropical landraces


Mir, C., Warburton, M.L., Taba, S., Bedoya, C., Franco, J.,Zhang, S., Xie, C., Prasanna, B.M., Hearne, S., Muthamia, Z., Yunus, M., Cuong,

B.M., and Charcosse. In Prep


Three years-

outputs and

outcomes



applications


Maize

Cowpea

Cassava

Musa

Striga

Some highlights


global diversity

Striga


pathogenicity




Striga

-from the parasite

perspective

Striga is one of the most significant biotic

constraints to maize production in SSA

Hearne, Pest Mgmt Sci, 65, 2009

Striga sp., while often lumped together have

distinct and different breeding systems. S.

asiatica is inbreeding species while S.

hermonthica is self incompatible and is highly

outbreeding in nature.

These differences have implications on diversity

within and between parasite populations per se

and also has implications control technology

efficacy from region to region – be that host

resistance and tolerance to parasite or herbicide

tolerant germplasm.


Striga

-from the parasite

perspective

Surveying Striga endemic areas in Nigeria, DRC

and Kenya. In 2011 survey Tanzania.

Taking basic farmer perception data and

collecting Striga leaf and seed samples from

individuals and populations to assess parasite

diversity.

Using location data to prepare and atlas of Striga

distribution working with GIS


Striga

-from the parasite

perspective

Pilot project completed assessing diversity of S.

hermonthica in Kenya using SSR markers

Within population variation contributed to some

95.89% of the total variance with only 4.11%

being due to among populations. High even for

an allogamous species.

The total heterozygosity observed across all

markers and populations was very high at

0.72054, within population heterozygosities

ranged from 0.4829 to 0.73988.


Striga

-from the parasite

perspective

The level of heterozygosity may reflect the

obligate out-breeding nature of the species and

the level of population diversity needed to ensure

fitness

Marker availability for Striga is limited – 8

informative SSR markers

Link with evolutionary biologist, Claude de

Pamphilis at Penn State on NSF initiative to

sequence parasitic plant ESTs

Striga tissue collected from diverse sources in

Nigeria and sent to US for RNA isolation and

sequencing using 454 Titanium sequencing


Striga

-from the parasite

perspective

Assemblies being constructed – mine for SNP

and polymorphic SSR to conduct further studies

Claude using data to complete the Striga

chloroplast genome and construct a mitochondrial

genome – understand evolution of parasitism

across parasitic plants – Striga, Alectra,

Orobanche

Use new markers to evaluate within and between

population S. hermonthica diversity. SSR in

Ibadan and SNP at Kbiosciences.

Data analysis, Sarah, Jorge and GIS


Striga Pathogenicity work – implication of

diversity?

NSF proposal to sequence Striga genome – great

resource to start to understand parasitism and

pathogenicity at the genomic level

First need a system for maintaining the discrete

Striga populations collected in the field

Working with Mike Timko, U . Virginia to look at

methodologies for inter population mating to

maintain population diversity. Working on

Nigerian isolates of Striga


Striga-host

interactions

Field screening $, uses only one or two populations of

Striga and does not provide an understanding of the

mechanisms of tolerance / resistance seen

Implemented medium and high throughput lab and

screenhouse based phenotyping protocols to assess

the Striga-host interaction at biochemical, physiological

and morphological levels.

Assess resistance at germination, haustorial initiation,

attachment and post attachment stages

Enable improved selection of favorable recombinants

and facilitate studies of gene action

Integrate these assays with current molecular breeding

work to identify molecular markers for key traits of

interest


Three years-

outputs and

outcomes



applications


Maize

Cowpea

Cassava

Musa

Striga

Some highlights


global diversity

Striga


pathogenicity




Marker

development

JCVI, ILRI, UC

Riverside

Program of EST sequencing projects – cowpea,

cassava and musa

Generated:-

Cowpea-

41949 sequences

3367 putative SNP

1805 putative SSR, 916 di- and trinucleotide repeats

Cassava

5046 sequences (41173)

2699 putative SSR

2486 putative SNP

Musa

5494 sequences (52907)

1937 putative SSR

28815 putative SNP


Marker development

Consolidation and

utilization of EST

HarvEST cowpea database - 17 libraries -

Two IITA libraries, 12 UCR libraries.

http://harvest.ucr.edu/

HarvEST cassava database - 17 libraries -

Two IITA libraries, 12 UCR libraries.

HarvEST musa database - 17 libraries - Two

IITA libraries, 12 UCR libraries.

















Marker

development and

use

Cowpea

Developed Illumina Goldengate SNP assay

IITA Reference collection, breeders lines, bi-

parental populations

IITA reference collection encompassed genetic

diversity assessed using SNP

Ref set and additional germplasm extensively

phenotyped – tool for allele mining

Development of consensus genetic map of

cowpea, and synteny with soybean and Medicago

Provides genomic framework for identification of

marker trait association, map-based cloning,

selection of markers for assessment of genetic

diversity, association mapping

Muchero, W., Diop, N., Bhat, P., Fenton, R., Pottorff, M., Hearne, S., Ndiaga, C., Fatokun, C., Ehlers, J.,

Roberts, P., Close,T. 2009. PNAS


Marker

development

Consolidation and

utilization of EST

410 EST derived SSR - Xu et al 2009, Molecular

breeding

Mike Timko’s group compared EST with GSS

data

Determine the gene discovery rate Timko et al

2008 BMC Genomics

Predict open reading frames for the genes (oligos)

used to create a cowpea microarray – 385k.

Microarray used to study cowpea-Striga

interaction- Li et al 2009 Pest Mgmt Sci


Three years-

outputs and

outcomes



applications


Maize

Cowpea

Cassava

Musa

Striga

Some highlights


global diversity

Striga


pathogenicity




Building on good

foundations

Breeding

DTMA DTMA DTMA

TLI TLII

DTMA

Cassava

Musa Yam

Adapt and apply know how to achieve

rapid advances in other IITA crops

Genome sequence

ESTs


Be brave and

look at

complexity

In farmers fields plant are subject to

multiple stresses

Start looking at the basis of stress

tolerance resistance. Then look at

complexes of commonly occurring

stresses


We have a bright future

Thank you

Harnessing interdisciplinary approaches for germplasm development

Technology

breeding turnaround

breeding snp

meetings breeding strategy

mars breeding strategies

breeding dtma scientists

qtl data

molecular team

issues of data handling