Breeding Nutritionally Enhanced Maize: The Tropical Experience

Breeding Nutritionally Enhanced Maize:

The Tropical Experience

K. Pixley, R. Babu, J. Yan, N. Palacios

& colleagues

GEM Cooperator Meeting, 8 December 2010, Chicago

“The Pigs”

QPM also “works” with chickens…

QPM: Nutritionally enhanced maize

Normal Maize

QPM Pellagra: diarrhea, dermatitis; due to

niacin deficiency. Tryptophan is a precursor of niacin

Kwashiorkor: edemas, anorexia, increased susceptibility to infections; due to low quality protein

What is QPM?

• Contains one gene – opaque2 (o2) – that affects protein production in the grain– No change in protein quantity– More of proteins rich in tryptophan and

lysine• o2 was found in maize; QPM is not

transgenic• Looks, cooks and tastes like normal

maize• Must be, and many are,

agronomically competitive

Where is maize an important source of protein?

FAO Stat

WHO, 2007. Protein and amino acid requirements in human nutrition. http://whqlibdoc.who.int/trs/WHO_TRS_935_eng.pdf

Summary: QPM meta-analysis Gunaratna et al.

• 9 studies: 5 countries (Ethiopia, Ghana, India, Mexico, Nicaragua); 48-486 children; 3.5 mo – 5 yr old

• Consuming QPM instead of normal maize resulted in:– 12% (95% CI: 7-18%) increase in weight gain– 9% (95% CI: 6-15) increase in height gain

Gunaratna et al., Food Policy 2010

These results were robust; essentially un-changed by:

Various adjustments/transformations to the data, or methods to calculate CI

Discard studies with most extreme results

Discard any of the studies

Iron deficiency affects >2 billion people• Iron deficiency anemia (IDA)

– Maternal and perinatal mortality– Impaired cognitive skills and physical activity

• Women and children in South Asia and Africa

Zinc deficiency

• Zn deficiency– 800,000 child deaths per year; increased risk

• Diarrhea, pneumonia, malaria– Stunting during early childhood– Equally affects males and females– South Asia and Africa

Photo: N. Palacios

Vitamin A deficiency• Vitamin A deficiency (VAD)

– Night blindness, corneal scarring & blindness

– Weakened immune system: VAD associated with• 20% of measles-• 24% of diarrhea-• 20% of malaria-related mortality

in children; • 20% of maternal mortality

– South Asia and Africa have highest VAD prevalence

– 157 million pre-school children– 30 million pregnant women

* Micronutrient malnutrition affects more than half of the world’s population –

United Nations SCN, 2004.

Dietary sources• Vitamin A

– Meat (esp. liver)– Vegetables (carrot, sweet potato, spinach)

• Iron– Red meat, fish, poultry– Lentils, beans, leafy vegetables

• Zinc– Oysters, animal proteins, – Beans, nuts, whole grains

Animal

Non-staple plants

Staples

Share of expenditures after price rise

Rural Bangladesh

Meat & Fish

Non-staple plants

Meat & Fish

Non-Food

Staples

Non-Food

Share of expendituresbefore price rise

$$$$

$$$

$$

$

Biofortification of staple food crops

• Micronutrients available in staple foods– Sustainable, affordable– Accompanied by dietary/nutrition information– Complemented by supplementation and fortification

• Acute malnutrition• Equal or better agronomic performance of biofortified crops

– Yield, disease resistance, drought tolerance…

Cross high proA x good drought tolerance…

De3, SC55, CI7

BC1S1

BC1S2 BC1S4

XX

X XX

XX

2nd Dose F1’s

BC1 2nd Dose S1’s

XX X

8 promising proA hybrids: 5 sites in Zambia + 2 sites in Zimbabwe

Best hybrid check

Tons

per

hec

tare Hybrid 1: 8.9

Hybrid 2: 7.1Hybrid 3: 6.3Hybrid 4: 6.5Hybrid 5: 7.4Hybrid 6: 5.7Hybrid 7: 6.9Hybrid 8: 5.9

ProA(ug/g)

Msekera

ZamSeed, Lus

GART

MpungweHarare

ART, Harare

SeedCo, Lus

1. Alleles for LCYE identified by: Association mapping Linkage mapping Expression analysis Mutagenesis

LCYE affects the ratio of carotenoids in the biosynthetic pathway

LCYElycopene

δ-carotene

α-caroteneLCYB

HYDb

zeinoxanthin

lutein

HYDE

ABA

HYDb1

LCYB

β-cryptoxanthin

γ-carotene

β-caroteneLCYB

zeaxanthinHYDb

GGPP

Yan et al., Nature Genetics 2010

2. HYDB1 has a large effect on BC

Harjes et al., Science 2008

PSYPDSZ-ISOZDS/CRISTO

De3 (KU1409/DE3/KU1409)S2-18-2-B

Total proA (ug/g) in 9 genotypic classes of 6 crossesLycE HydB Pop 1 Pop 2 Pop 3 Pop 4 Pop 5 Pop 6

4 1 12.03 9.97 6.15 5.74 11.94 9.70 6.12 5.75

4 2 4.27 3.46 3.99 3.90 4.09 3.57 4.08 3.74

4 H 5.60 6.17 4.78 7.42 5.74 4.73 7.51 2 1 7.41 11.11 7.19 6.25 5.10 7.40 10.74 6.92 6.19 5.09 2 2 2.08 3.36 3.73 2.60 4.28 3.31 2.06 3.34 3.71 2.50 4.33 3.29 2 H 4.16 4.07 5.30 3.88 4.42 3.88 4.11 4.07 5.40 3.89 4.43 3.93 H 1 9.45 12.96 9.47 13.20 H 2 3.58 4.14 3.95 3.59 4.21 4.13 H H 4.83 6.82 5.08 5.41 3.64

4.87 6.80 4.98 5.41 3.60 Bank accessions (hets)

KUI carotenoid syn-FS17-3-2-B B104 CML325, CML327, CML460

Seed genotyping pre-planting

Dry chipping using dog nail clippers ≈10,000 seeds will be genotyped pre-planting this season

Steps to develop a hybrid cultivar (w/MAS)

UU x FF -> UFUF x UU -> UU:UF

200UF seeds -> 50UU:100UF:50FF50FF S1 ears -> 40FF S2 ears

40FF x tester -> 15 stage1 hybrids15 best S3 -> 15 best S4

15 S4 (HPLC) x 3 tester -> 15 Stage2 hyb5 best S5 -> 5 best S6

1-2 best S6 x 3 tester -> 15 Stage3 hybS7 -> HPLET

4-5 best x tester -> hybMultilocation on-farm trials

1-2 best -> releaseMultilocation on-farm demo’s

Begin marketing seed

Elim.50%

75%

$25-75/

sample

$10 per

row + time

$5850 150+120+45 nursery

+ (45x6) trial rows

$2250

1

2

4

5

6

7

8

3

year

Photo: N. Palacios

Photo: H. De Groote

What happens to provitamin A during cooking?

25% loss of β-carotene

Shanshan Li et al., 2007

Effect of porridge preparation

Effect of snack preparation

• 36% loss of provitamins A following nixtamalization and snack preparation by deep frying (n=13)

Lozano Alejo et al., 2006

What happens to provitamin A after we eat them? …bioaccessibility

Parker, FASEB J, 1996

Mark FaillaDepartment of Human Nutrition

In vitro assessment of bioaccessibilty of carotenoids from foods

Zn in the action of retinol dehydrogenase (retinol to retinal); essential pigment for vision

Zn is a probable co-factor for b-carotene mono-oxygenase (cleaves proA to vitA)

Zn deficiency depresses synthesis of the carrier protein of vitA => lower plasma retinol concentrations

dehydrogenase Vision retinol retinal Zn monooxygenase Digestion -carotene 2 retinal Zn Protein synthesis Zn retinol binding protein retinol:RBP in blood (RBP)

Pedigree Fe Zn Alppm ppm ppm

KUI carotenoid syn-FS17-3-1-B-B/CML353-B 18.1 33.3 0.6KUI carotenoid syn-FS17-3-1-B-B/(CML-239 x GWIC) -1-7TL-1-1-1-B 17.9 34.8 0.3KUI carotenoid syn-FS17-3-1-B-B/CML356-B 20.5 41.1 0.5KUI carotenoid syn-FS17-3-1-B-B/(CML-356 x GWIB) -1-23TL-1-2-1-B 21.1 37.7 0.5KUI carotenoid syn-FS17-3-2-B-B/CML353-B 19.7 32.0 0.3KUI carotenoid syn-FS17-3-2-B-B/(CML-239 x GWIC) -1-7TL-1-1-1-B 18.7 32.5 0.5KUI carotenoid syn-FS17-3-2-B-B/CML355-B 15.8 29.8 0.2KUI carotenoid syn-FS17-3-2-B-B/CML356-B 22.4 35.2 2.0KUI carotenoid syn-FS17-3-2-B-B/(CML-356 x GWIB) -1-23TL-1-2-1-B 23.1 33.2 2.0KUI carotenoid syn-FS25-3-2-B-B/CML353-B 22.9 33.0 2.0KUI carotenoid syn-FS25-3-2-B-B/(CML-239 x GWIC) -1-7TL-1-1-1-B 22.0 31.9 2.1KUI carotenoid syn-FS25-3-2-B-B/P903 C0 H364-1-8TL-3-2-1-1-B-B-B-B-B -B 19.8 29.3 0.3KUI carotenoid syn-FS25-3-2-B-B/(CML-356 x GWIB) -1-23TL-1-2-1-B 22.6 29.8 0.5Carotenoid Syn3-FS5-1-5-B-B/CML353-B 20.5 28.6 0.6Carotenoid Syn3-FS5-1-5-B-B/CML355-B 32.0 32.9 0.6CML-305-B-B/CML356-B 17.2 23.1 0.4CML-304-B-B/CML353-B 14.5 26.0 0.3Average 20.4 31.6 0.8

Pedigree Fe Zn Al

ppm ppm ppm

KUI carotenoid syn-FS17-3-1-B-B/CML353-B 18.1 33.3 0.6KUI carotenoid syn-FS17-3-1-B-B/(CML-239 x GWIC) -1-7TL-1-1-1-B 17.9 34.8 0.3KUI carotenoid syn-FS17-3-1-B-B/CML356-B 20.5 41.1 0.5KUI carotenoid syn-FS17-3-1-B-B/(CML-356 x GWIB) -1-23TL-1-2-1-B 21.1 37.7 0.5KUI carotenoid syn-FS17-3-2-B-B/CML353-B 19.7 32.0 0.3KUI carotenoid syn-FS17-3-2-B-B/(CML-239 x GWIC) -1-7TL-1-1-1-B 18.7 32.5 0.5KUI carotenoid syn-FS17-3-2-B-B/CML355-B 15.8 29.8 0.2KUI carotenoid syn-FS17-3-2-B-B/CML356-B 22.4 35.2 2.0KUI carotenoid syn-FS17-3-2-B-B/(CML-356 x GWIB) -1-23TL-1-2-1-B 23.1 33.2 2.0KUI carotenoid syn-FS25-3-2-B-B/CML353-B 22.9 33.0 2.0KUI carotenoid syn-FS25-3-2-B-B/(CML-239 x GWIC) -1-7TL-1-1-1-B 22.0 31.9 2.1KUI carotenoid syn-FS25-3-2-B-B/P903 C0 H364-1-8TL-3-2-1-1-B-B-B-B-B -B 19.8 29.3 0.3KUI carotenoid syn-FS25-3-2-B-B/(CML-356 x GWIB) -1-23TL-1-2-1-B 22.6 29.8 0.5Carotenoid Syn3-FS5-1-5-B-B/CML353-B 20.5 28.6 0.6Carotenoid Syn3-FS5-1-5-B-B/CML355-B 32.0 32.9 0.6CML-305-B-B/CML356-B 17.2 23.1 0.4CML-304-B-B/CML353-B 14.5 26.0 0.3Average 20.4 31.6 0.8

2010: Stage 1 High Zn x ProA2011: S2’s High Zn x ProA (HYDB1) to TC

Combining proA and Zn …bioefficacy

Will farmers and consumers grow/ consume biofortified crops?

• ProA sweet potatoes are orange; consumers prefer white

• Will farmers choose to plant the biofortified varieties?

• Will farmers choose to grow orange maize varieties?

• Will seed companies market the new orange varieties?

No complaints from these consumers!

Molecular biology

Plant breeding and agronomy

Agriculture for nutrition and health

+ +

+ ++

Food technologyNutritionSocio-economicsEducation & marketing

Healthy families

AU

-0.02

0.00

0.02

0.04

0.06

0.08

0.10

0.12

Minutes2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00

3.33

2 3.96

14.

485

4.75

05.

473

6.71

7 7.02

47.

217

7.95

7

9.18

3 10.5

7811

.517

13.6

98

16.2

46

18.8

59++

Plant biochemistry

+

+

Seeds of Discovery (SeeD)

A Mexican initiative to contribute to global food security vis-à-vis climate change and resource scarcity by broadening the genetic base of global maize and wheat-breeding programs

P. Wenzl, K. Pixley, G. Atlin, G. Edmeades, M. Banziger & many colleagues

Historical bottleneck

10 – 20 years

SeeD: new genetic variation to raise future

crop production

Genetic resources

Breeding programs

Variety adoption and improvement

Increased agricultural production

Factors limiting the use of GR Factor 1: So many accessions, so little information!

– Challenges to characterize accessions at phenotypic and molecular levels

– Missing or ‘superficial’ passport data Factor 2: Insufficient tools to mine information

– Outdated/user-unfriendly data management tools– Limited query capabilities

Factor 3: How to effectively utilize exotic germplasm?– How to identify beneficial alleles in exotic germplasm?– How to capture novel, useful variation into elite backgrounds

Many of the same challenges and issues of GEM!

Objective 1: To mine maize/wheat genetic resources for novel alleles and beneficial traits combining genotyping and phenotyping methods

Objective 2: To build on-line catalogues that facilitate the identification of beneficial genetic variation, and

Objective 3: To put in place practical delivery pathways that empower maize and wheat programs to broaden their genetic base by incorporating novel variation

Objectives of SeeD

Marker-assisted introgression

pipeline service facility

Doubled haploid service facility?

Develop and release “bridging

lines”

Because of the size and complexity of the initiative, will require strategic alliances with key players.

Key partnerships: 1. Genotyping2. Phenotyping3. High-performance bioinformatics approaches for

genetic analyses4. Cyberinfrastructure for a SeeD web portal

SeeD: A technology intensive project!

Focus on user base targeted by SeeD: maize and wheat breeding programs, especially public and SME breeding programs in developing countries

Design practical delivery paths that enable the adoption of novel and useful genetic diversity in breeding programs.

Concept-development guidelines:

Draft strategy for maize

General points (and working assumptions)

• The CIMMYT maize collection has about 26,000 accessions with no genotypic data, incomplete passport data, and some phenotypic data

• Most accessions are heterogeneous with much more genetic variability among than within accessions.

• Alleles that are rare globally, and at low frequencies in the accessions in which they occur, are unlikely to be very important or detectable.

• Most traits in maize are highly polygenic.

• Haplotypes with small effects likely control most variation

• Main diversity is to be found in the Mexican-Guatemalan germplasm, which has been in long co-existence with teosinte

• Much genetic variability remains in teosinte, but there are few introgression populations available that could allow us access to this variability

• Demand for direct access to landrace or teosinte accessions by breeders will be limited

General points (cont’d)

• Haplotype effect estimates for loci with small effects

• Well-characterized accessions for specific traits to be used as donors for large-effect alleles

Two main products of SeeD for maize

• Because haplotypes are likely to be replicated across many accessions, it is the haplotype rather than the accession itself whose effect we want to estimate, and that is the unit of evaluation or selection

• To sample haplotype frequencies and to begin estimating haplotype effects, one plant per accession will be initially genotyped at >1,000,000-plex, on the assumption that this would detect all but the rarest alleles

Haplotypes as the unit of evaluation

• To estimate haplotype allele effects, the single-plant representatives of core accessions will be crossed to elite, adapted testers (testers as females)

• The testcrosses will be phenotyped in screens of interest.

Test crosses

• Because haplotypes will usually be replicated across accessions, it not necessary to estimate the effect of each testcross with high precision via high levels of replication on individual accessions.

• Based on allelic effects estimated by phenotyping the subset of core accessions, the entire collection will be examined for accessions that have high GEBVs for traits of interest, and that are under-sampled in the existing elite germplasm.

• New pre-breeding populations will be established from these accessions and improved by genomic selection.

GEBV prediction for all accessions

“Details of this approach to delivery of new genetic variation for quantitative traits needs

a lot more thought…”

• The objective is to identify accessions with high frequencies of unusual alleles with large effects on simple or oligogenic traits. There will likely be very few of these.

• The accession is the unit of evaluation

• The core and materials that have a high likelihood of having been selected for the trait (based on passport information) need to be phenotyped at high precision, either per se or in testcrosses.

Large-effect alleles

• Map, and develop gene-based markers

• Introgress the allele of interest into elite inbreds (“bridge inbreds”), which breeding programs will be want to use

• As donors for MAS-based conversion, or• As parents of pedigree starts.

• SeeD probably needs to develop such inbreds as deliverables, to ensure that any genes discovered are truly accessible to smaller public and private breeding programs in the developing world

Large-effect alleles

SeeD and GEM

Many opportunities for complementarity and partnership with GEM

Important to communicate often and learn from each other

Many questions we can work together to answer and enhance the use of genetic resources

[email protected]

[email protected]

[email protected]