Seeds of Discovery: Why CIMMYT genotyped its entire maize germplasm bank collection. Kevin Pixley

1. Seeds of Discovery (SeeD)
Why CIMMYT genotyped its entire maize
germplasm bank collection (?!)
k.pixley@cgiar.org
13th Asian Regional Maize Conference
Ludhiana, October 2018
Kevin Pixley , Marcela Carvalho, Denise Costich, Kate Dreher, Sarah Hearne,
Huihui Li, Terry Molnar, Cesar Petroli, Fernando Toledo & team

2. Outline
1. Rationale – rising to the challenges
2. Genomics – understanding diversity
3. Pre-breeding – getting from the bank to farmers
4. Capacity Development – unleashing innovation
5. Available products

3. Converging Challenges: we are not on-
track for food security
It’s not only about food security: Increased productivity conserves land for other
uses, e.g. biodiversity reserves
(Byerlee et al., Global Food Sec., 2014)

4. ? With
SeeD
Seeds of Discovery (SeeD)
Vision: Efficient and Equitable
Use of Genetic Diversity
Challenge
to use –
lack of
relevant
information
– need
labels!

5. SeeD – high-density genetic profiles
 28,000 Maize
 100,000 Wheat

6. Finding genetic variation of
breeding value
Genome-wide association studies “GWAS”
Molecular markers associated
with phenotype of interest
Phenotypic data are expensive to generate!
~3500 accessions

7. Passport and GIS information about
accessions

8. Environmental Genome-wide Association Studies “EnvGWAS”
Molecular markers associated
with adaptive variation of
breeding relevance
GIS-derived collection site
data; ~17500 maize
accessions
Finding genetic variation of
breeding value
GIS data is “virtually” free!
500 100 150
Mb
Heatshock factor
Chromosome9
-log10p
0
20
40
60
80
Alberto Romero

9. A candidate gene of interest was
identified via envGWAS from 2700
accessions of the SeeD GWAS panel.
Jeff Ross-Ibarra, Ed Buckler, Sarah Hearne
Provisional gene model, oligopeptide transporter
(glutathione) – known role in drought tolerance
(Hasanuzzaman et al., 2017)
Involvement of the gene in drought
response is supported by:
 expression
data
 phenotypic
data
Environmental GWAS

10. Using genomic and GIS “Big Data” to mine
breeding value from accessions
EigenEnvGWAS

11. 15,384 landraces 1,549 landraces from drought-
prone environments
Euclidean genetic distances
Analyze, visualize, interpret, select
US $150kUS $1.5M

12. 250 Tropical
250 Sub-tropical
100 Sub-tropical
Phenotype and
use best in pre-
breeding
1,549 landraces from drought-
prone environments
Analyze, visualize, interpret, select
US $10k
US $25k
US $25k
US $150k

13. Challenges of Use

14. Maize Lethal Necrosis, MLN = combination
of two viruses:
Maize chlorotic mottle virus (MCMV)
+ Potyvirus, e.g. Sugarcane mosaic virus (SCMV)
• MLN is an occasional and local problem
in US Midwest cornbelt
• MLN has been reported in China
• It is mostly transmitted by insects
• It can be seed borne
• Recently, MLN is a serious problem in
East Africa

15. 1. High density genetic profiles
(28,000 maize accessions)
2. GIS data: identify regions
with likely presence of
MCMV (likely to have
encountered the disease and
developed resistance)
3. Passport data: within each
region, sampled as many
maize races as possible
Selection Process: Geographic,
Phenotypic and Genotypic Criteria

16. Characterization and pre-breeding to
enhance the use of genetic resources
6. Cross with elite lines (BC0
or BC1)
7. Evaluate in testcross trials
with best testers
8. Best lines to breeders for
use in crosses
4. Genomic data: select 1000
accessions with max.
genetic diversity
5. Phenotype in glasshouse:
select MCMV tolerant
accessions
Terry Molnar screening genebank accessions in May 2015

17. Pre-breeding for MLN tolerance:
Naivasha, Kenya (Feb 2018)
Marcela Carvalho

18. Resistance/tolerance to:
 Drought at flowering (>600)**
 Heat at flowering (>200)
 Tar spot complex (>800)
 Maize lethal necrosis (MLN) (>1000)
 Fusarium ear rot
Quality/New markets:
 Pigmented, enhanced anthocyanin
concentration (>200)
Current pre-breeding projects
**(in parentheses, number of landraces or populations evaluated)

19. Evaluate landraces per se
Replicated trials: stressed and ‘normal’ environments
X
F1 X
BC1 X&X Tester
Het. Grp. B
Best
“Dent”
Elite line
Het. Grp. BX
X
BC1 X&X Tester
Het. Grp. A
Best
“Flint”
Elite line
Het. Grp. A
Elite line
Het. Grp. A
Elite line
Het. Grp. B
F1
General pre-breeding strategy

20. Maize pre-breeding networks

21. Pre-breeding: Testcross hybrids of
drought-tolerant lines

22. • Must contain new alleles for the prioritized trait
• Must be better than elite germplasm for the trait
• Or must contribute additive effects to current elite
materials
• Must have yield potential under “unstressed” or
normal conditions
• Otherwise, breeder will not use them
• Must have ‘reasonable’ agronomic performance:
• Plant height, root and stem lodging tolerance, etc…
Pre-breeding  Release of genetic
resource donor lines

23. Elote occidental
CIMMYT (1960s) CUCBA Genebank
University of Guadalajara (2003)
Collaborative Projects
Ears with 8-10 kernel rows; large
grains; floury, purple grains
112 Acc.
50 Acc.

24. Collaborative Projects

25. Product Catalogue http://seedsofdiscovery.org/catalogue/

26. http://seedsofdiscovery.org/es/catalogo/
Product Catalogue

27. Who are the users of
SeeD’s products?
• Breeders: new diversity to accelerate genetic gains
– Impact on national production
– Impact on international commodity prices
• Researchers: stimulate scientific discoveries
• Students: a new generation of agricultural scientists
• Professors: curricula to train the next generation of
scientists
• Genebanks: optimize conservation of genetic resources

28. Vision: A “Sat Nav” for efficient and
equitable use of diversity

29. We thank all
our partners
and funders

30. Thank you
for your
interest!