Marker-assisted backcrossing_CTHash_ICRISAT.ppt

marker

1. Marker-assisted breeding,
with a focus on backcrossing
C Tom Hash, ICRISAT-Patancheru

2. Two closely related activities:
•Genetic mapping
–Select parental lines
–Produce population
–Phenotyping
–Genotyping
–Linkage map
construction
–QTL analysis
•Marker-assisted breeding
–Donor and elite parents
–Marker polymorphism
–Breeding population
–DNA sampling
–Genotyping at flanking
markers
–Marker-aided selection
–Rapid generation advance
(and alternatives)

3. Marker-assisted breeding
•Choose donor and elite parents
•Marker polymorphism assessment
•Breeding population production
•DNA sampling
•Genotyping at flanking markers
•Marker-aided selection
•Rapid generation advance

4. Choose donor and elite parents
•Donor parent for mapped QTL(s)
–Genetically distant from elite parent, at least in
genomic regions carrying putative QTLs of interest
•Mapping population parental line
•Mapping population progeny
•Elite parent for pedigree or backcross
–Economically important crop genotype, but weak for
economically important target trait with mapped
QTLs
–Ideally the susceptible parent of mapping population
–Marker and trait polymorphism relative to donor

5. Marker polymorphism assessment
•Characterize extent of marker polymorphism
between trait donor parent and candidate elite
parents
–In region of target QTL(s) polymorphism is essential
–In non-target regions of the genome, marker
polymorphism will permit speedy recovery of
recurrent parent genotype by marker-assisted
background selection, but is likely to be associated
with undesirable traits that may require additional
crosses or backcrosses to eliminate from product
lines

6. Breeding population production
•Breeding population size (F2 or BCnF1) must
be large enough to ensure reasonable
probability of obtaining desirable segregants
given the number of target QTLs
–Tables and prediction equations are available,
but a conservative rule of thumb:
•3X the minimum perfect population size will
give <5% failure rate,
•5X the minimum perfect population size will
give <1% failure rate

7. How many plants and progenies?
One target
gene
Minimum
perfect
population
Plants
5
per_

7
family
11
Generation
F
1 1
BC
1F
1 2 5% 1% <1%
F2 or BC
2F
1 4 10% 2% <1%

8. How many plants and progenies?
Two
target
genes
Minimum
perfect
population
Geno
11

typed
17
plants

23
per
46
family
94
Generation
F
1 1
BC
1F
1 4 5% 2% <1%<<1%<<<1%
F2 or
BC
2F
1
16 __%__% __% 5% <1%

9. DNA sampling
•Small sample of “reasonable” quality
•Every individual considered for selection for generation
advance: tracking samples!!!
–Tissue sampling
–Tissue disruption:
•Manual: mortar & pestle with dry ice or liquid nitrogen
•Semi-automated: “paint-shakers” and other devices
–DNA extraction
•Kits
•Single-tube CTAB prep
–Quality assessment by electrophoresis and spectroscopy
–Slow and expensive (up to 384 samples per day at $1 to $2
each)

10. DNA sampling: The new way
•Very small sample of “reasonable” quality
•Every individual considered for selection for generation
advance
–tracking samples is still the biggest headache
•Tissue sampling: leaf disc punched from labeled plant
(or from part of a seed) and placed in micro-tube or
plate well
–Tissue disruption: not required
–DNA extraction
•Use kit: add extraction buffer and heat to 95C for 15 min to produce
self-stable DNA for >50 PCR reactions
–Maybe so little DNA that you cannot see it on conventional
agarose get, but cheap (<US$0.25 per sample), quick
(>1000 per day easy) and good enough for SSRs (and
probably SNPs or DArTs)

11. Genotyping at flanking markers
•Three-five marker loci per target QTL must be
genotyped in progeny considered for selection
–One-two flanking loci “above” target QTL
–Locus co-segregating with target QTL
–One-two flanking loci“below” target QTL
•Selective genotyping can reduce supplies
costs, but may be operationally difficult
B
D
C
E
A

12. Marker-aided selection
•Foreground: Select for advancement at least
one of the genotyped individuals having donor
alleles (heterozygous or homozygous,
depending on breeding scheme and generation)
at all three genotyped loci in vicinity of each
target QTL
–Pre-flowering selection can reduce number of
crosses to be made in a backcrossing program
•Background: Select for recombination near
target QTL, and then for recovery of elite
genotype elsewhere

13. Rapid generation advance
•When marker genotyping was relatively slow
compared to the crop life cycle, it was often
expedient to advance two generations between
stages with genotyping (maintaining population
size needed to retain desired segregants), and just
use additional BC generations to recover recurrent
parent background:
–e.g., genotyping in BC2F1, BC4F1, BC6F1, and BC6F3
•Push forward generation advance as quickly as
possible to speed delivery of finished product
for testing
•Generally no longer needed, instead, use more
markers and reduce number of BC generations
to recover eliteness

14. Marker-assisted breeding
•Choose donor and elite parents
•Marker polymorphism assessment
•Breeding population production
•DNA sampling
•Genotyping at flanking markers
•Marker-aided selection
•Rapid generation advance and/or reduce
numbers of generations to recover elite
products for testing

15. Full-scale conventional M-A BC
•Starts only when certain of trait, marker, and QTL
polymorphism, so long lag phase but low risk
•Starting point is typically a BC
1
F
1
or a genotyped
mapping population progeny homozygous for donor
parent marker alleles flanking target QTL(s)
•Marker genotyping (optional in non-target regions
to hasten recurrent parent allele recovery)
•Most cost effective when have PCR-compatible co-
dominant markers, and/or a long-lived crop with
large individuals and a long juvenile phase

16. Minimal M-A BC
•Similar starting points, so long lag & low risk
•Genotype only 2-3 marker loci per target QTL
•Use backcross generation advance to recover
recurrent parent alleles in non-target regions
•Each generation use appropriate numbers of plants &
progenies to minimize probability of failure to
advance target QTLs (individually)
•Allows cost-effective slow-marker applications in
otherwise conventional breeding programs
•No longer necessary for maize, rice, sorghum, …

17. Jump-started M-A BC
•Assumes lag time delays are more expensive than
probability of failure
•Begin backcross transfer before markers flanking
target QTLs are identified
•Proceed as for minimal M-A BC
•Most efficient when you have a high degree of
confidence in donor parent, high-value target
trait(s), little time, and ‘slow’ markers or not enough
cheap high-throughput markers for effective
background selection
•Again, no longer really necessary for major crops

18. Reduce generations needed
•Now marker genotyping quick & less
expensive, so use more to reduce number
of BC generations needed to recover
eliteness:
–Foreground genotype (SSRs?) a large BC1F1
generation
•identify several individuals with recombination on one side of target,
•backcross all of these to recurrent parent
–Background genotype selected BC1F1 individuals
(DArT?)
•identify one with smallest proportion of donor parent genome
•advance backcross of this selected individual,

19. Reduce generations needed
–Foreground genotype large BC2F1 generation from
selected BC1F1 plant
•identify several individuals with recombination on the
other flank of foreground target
•self all of these
–Background genotype selected BC2F1 individuals
•identify one with largest portion of (homozygous)
recurrent parent genome
•advance selfed progeny of this selected BC2F1 individual
–Foreground genotype small BC2F2 generation
•identify introgression homozygotes, and self these
•resulting BC2F3 families are finished products
ready for testing

20. Contiguous segment substitution
lines (contig lines)
•Their development is a logical extension of
minimal M-A BC
•Expensive in terms of marker data demands
and population size requirements
•Permits ‘bin-level’ mapping of many QTLs
of small effect by phenotyping smaller
numbers of progenies with greater accuracy
•Facilitates fine mapping for multiple traits

21. How to make marker-assisted breeding
more practical and cost-effective?
•Cheaper and easier DNA isolation
•Fast becoming a reality
•Reliable marker data that takes less
time to obtain
•Service labs!!!
•Better data tracking systems
•The weak link at present

22. Thanks for your
attention!
Time for some examples?