Alternatives to Fine Mapping Quantitative Trait Loci QTL in Barley Hordeum vulgare

1

• Alternatives to Fine Mapping Quantitative Trait
  Loci (QTL) in Barley (Hordeum vulgare)
• by
• Maqsood Rehman

2
Quantitative trait loci (QTL)

• A quantitative trait locus (QTL) is the location
  of a gene that affects a trait that is measured
  on a quantitative (linear) scale
• These traits are typically affected by more than
  one gene, and also by the environment (such as
  yield)
• Mapping QTL is not as simple as mapping single
  gene that affects a qualitative trait (such as
  flower color)

3
Quantitative trait loci (QTL)

• Mapping QTL has become a reality in the past 10
  years, primarily because of the availability of
  molecular markers. These markers segregate as
  single genes, and they are unaffected by the
  environment.

4
Historical Background

• In the 1920s Fisher and Wright developed the
  concept of quantitative genetics. They considered
  quantitative genetics as a statistical branch of
  genetics based on fundamental Mendelian
  principles
• First example of QTL is usually attributed to
  Sax, K. 1923. The association of size differences
  with seed coat pattern and pigmentation in
  Phaseolus vulgaris. Genetics 8552-560.

5
Tools needed for QTL

• Segregating population (F2, DH, RIL, BC)
• Linkage map for that population constructed of
  genes, molecular markers
• Traits measured on some quantitative scale

6
P1 x P2
F1
F1 x P1 (or P2)
F1
Anther culture
F2
BC1
(Backcross)
DHL
(Doubled haploid line)
F3
RIL
(Recombinant inbred line)
With polymorphic molecular markers and linkage
maps as tools, mapping QTL is simply a matter of
growing and evaluating large populations of
plants, and of applying the appropriate
statistical tools.
7
Mapping QTL

• Up to 1980, the quantitative traits has involved
  statistical techniques based on means variances
  and covariances of relatives, with no actual
  knowledge of the number and location of the genes
  that underlie them
• Statistical tools used for measuring QTL include
  ANOVA, Interval mapping, Composite interval
  mapping

8
Examples

• Polygenes in Drosophila for characters such as
  bristle number and viability by Breese and Mather
  (1957)
• Similar progress was made in wheat using
  aneuploidy as a device to manipulate and fix
  polygenes of interest (Law et al, 1983)

9
Questions

• How many genes are involved in any given trait.
  Which one is major and which one minor, it will
  help in selection?
• What is the nature of dominance, epistasic
  properties of these genes and how do they
  interact with the environment?
• What type of genes are they? Are they structural
  or regulatory and, if regulatory, how wide is
  their sphere of influence?

10
Problems with Mapping QTL

• The methodologies were laborious and based on
  major mutants, either phenotypic or cytological,
  which made it difficult to study larger
  populations
• These impediments were removed by
• a) the discovery of variability at the DNA
  level, which could be used as marker, and b) QTL
  (quantitative trait loci) replaced the heavy
  statistical term polygenes

11
Problems with mapping QTL

• QTL locations obtained from segregating
  populations have very large confidence intervals
  (seldom lt5cM and often gt30cM)
• The reason for these large CI is the lack of
  recombination at meiosis
• Larger populations are needed to get enough
  meiosis to map a particular QTL
• The smaller the heritability of the trait, the
  larger the population required

12
Problems with mapping QTL

• The large confidence intervals have large number
  of potential candidate genes
• Thus, a 10 cM interval will contain, on average
  130 rice genes and over 400 Arabidopsis genes
• It is also difficult to differentiate two QTL
  that are less than 20 cM apart

13
Problems with mapping QTL

• May be misinterpreted as one or a ghost QTL if
  linked in coupling, and possibly no QTL if linked
  in repulsion.
• Statistical problems
• Chromosomes are statistically scanned for a large
  number of positions to find the most likely
  position of QTL

14
Problems with mapping QTL

• The cost of time and money
• Solution for this statistical problem is to
  repeat the experiment (e.g., human genetics and
  maize) but its too costly
• As a result scientists assume what they got was
  correct
• Several biases are also involved with statistical
  methods for example QTL with sufficiently large
  effect will be detected

15
Problems

• Consider a QTL whose true individual size is just
  at the threshold of detection. Because of
  environmental variation, it will be below that
  threshold on 50 of occasions and hence will not
  be detected
• This will give us a false impression of G X E
  interaction on repetition

16
Possible solutions

• Mapping that we have been using could lead us to
  the rough location of QTL. They indicate which
  arm of the chromosome the QTL is on and possibly
  suggest a more precise location within that arm
• To be more precise about the location of these
  QTL, it is necessary to use some form of
  chromosome introgression or substitution lines or
  near-isogenic lines (NILs) OR
• Use Stepped Aligned Recombinant Inbred Strains
  (STAIRS)

17
Possible solutions

• The use of STAIRS will help reduce
• a) the size of CIs around the QTL b) focus
  on a small subset of possible candidate
  genes
• These identifications of potential QTLs will lead
  us to sequencing, expression analysis,
  transformation and gene silencing, etc.

18
Studies

• One similar study by (Bohuon et al., 1998)
• They introgressed sections of donor chromosomes
  from the DH line of calabrese hybrid (Green Duke)
  into the recipient B. oleraceae var alboglabra
• They produced 70 different substitution lines,
  each has an entirely common recipient background
  with just a short region of donor chromosome from
  the calabrese variety

19
Studies

• These part chromosome introgressions varied in
  length
• Compare these introgressed lines with the
  recipient line ( B. oleraceae) to know whether or
  not there are QTLs in that introgressed region
• Comparing different substitution lines
• These lines can be produced by marker assisted
  selection

20
Studies

• They also compared QTL location found by previous
  segregating population with those found in the
  substitution lines
• QTL found in segregating populations were also
  found in substitution lines
• However, there were some new QTLs detected in
  substitution lines because
• Were able to separate QTL linked in coupling
• They also separated pairs in repulsion

21
(No Transcript)
22
Studies (Arabidopsis)

• Koumproglou et al., (2002) described the
• potential of STAIRS in the genetic analysis
  
• of Arabidopsis
• Donor strain- Landsberg (Ler), and recipient
  strain Columbia (Col)
• Major gene and QTLs for Flowering time (ft),
  rosette leaf number (Rln), and plant height (Ht)
• For STAIRS we need to produce
• a) Whole Chromosome Substitution Strains
• b) Stepped aligned recombinant inbred strains

23
(a) The chromosome constitution of the recurrent
parent (recipient) and donor lines, together with
the five, homozygous, whole Chromosome
Substitution Strains (CSS1-5) in Arabidopsis
thaliana, (b) A set of n, single recombinant
lines (SSRL 1-n), making up the Stepped Aligned
Inbred Recombinant Strains (STAIRS), showing the
increasing length of donor segment in a single
chromosome
24
a) Recurrent parent and CSS genotypes and
phenotypes to identify chromosome containing QTL,
(b) Initial coarse mapping to identify
approximate QTL location, (c) Use of more STAIRS
in the marker interval, to allow fine mapping of
QTL, Q. M1-M7 are markers spaced fairly evenly
across the chromosome. STAIRS 3-4 and 4-5 are
SRLs between markers 3 and 4, 4 5, respectively
25
Ideogram of STAIRS for chromo some 3, indicating
Col and Ler regions, line designations and number
of replicate lines in (n). Vertical lines
indicate interstitial marker positions potential
QTL locations are indicated by 'ai' at the base
of the figure. Ler Col
Cross-over region.
26
The reciprocal Col/Ler STAIRS for chromosome 3.
For each parameter in the model, 1 refers to an
allele from Ler or Col, respectively.
Rln30  Rosette leaf number at 30 days,
Ft  flowering time (days from sowing),
Ht35  height at 35 days, Trich  presence/absence
of trichomes
27
Goals

• The goal is to utilize this novel technique
  to identify and analyze QTLs in barley to
  increase the productivity of this important
  cereal crop. In order to achieve this goal we
  propose the following objectives

28
Goals

• Develop whole Chromosome Substitution Strains
  (CSSs) between cultivated Hordeum vulgare ssp. L.
  Harrington, and its wild relative Hordeum
  vulgare ssp. Spontaneum accession 1B-87
• Produce Stepped Aligned Recombinant Inbred Lines
  (STAIRS) to move the study traits valuable for
  cultivar development from the whole chromosome to
  short 1-10cM intervals

29
Goals

• 3. Confirm the location of a Hordeum vulgare
  ssp. Spontaneum QTL conferring resistance to
  Puccinia striiformis f. sp. Hoedei to assess the
  specific ability of STAIRS to fine map QTL of
  interest in barley

30
CSS creation

• Plant material
• H. vulgare ssp. L. variety Harrington is a
  popular malting variety in PNW
• Susceptible to leaf rust, powdery mildew, and
  other foliar pathogens
• H. vulgare ssp. Spontaneum, a wild barley
  collected from Israel and resistant to leaf rust
  and powdery mildew

31
CSS creation
32
CSS creation

• H. vulgare ssp. spontaneum (Donor), and H.
  vulgare Harrington (recipient)
• F1 seeds production
• emasculation of 50 recipient Harrington
• H. spontaneum will be used as a pollen donor
  (target 500-550 seeds)
• BC1 production
• 500 F1 plants will be backcrossed to Harrington
  (expected 20 BC1 seeds/plant)

33
CSS creation

• Marker assisted selection for non-recombinant
  chromosomal lines
• Over 152 polymorphic SSR markers have been
  identified between Harrington and wild barley
  relatives (Matus and Hayes, 2002 Backes et al.
  unpublished)
• 11 polymorphic SSR will be used for screening
  each chromosome

34
CSS creation

• PCR techniques will be utilized to screen
  backcross individuals starting with terminal
  markers for each chromosome
• Select backcross individuals that have all
  non-recombinant chromosomes, either recipient or
  donor chromosomes
• Screening 10,000 individuals (90)

35
CSS creation

• For example for chromosome 1 we will select first
  for terminal markers and discarding any
  individual with recombination
• Then the population will be selected for 9
  remaining interstitial markers (to insure no
  recombination)
• The same process will continue with chromosome 2,
  and so on until we get the whole CSSs for all 7
  chromosomes

36
CSS creation

• Markers
• We will need co-dominant markers near the
  telomeres and at regular intervals
• The more the markers the more expensive
• Markers spaced at 20 cM will be used
• If possible, more markers will be used to confirm
  the integrity of the selected chromosome

37
Chromosome substitution strains

38
Creation of STAIRS

• Cross the CSS for each chromosome with the
  recipient (Harrington)
• Select for the individuals with only one cross
  over
• Self these individuals to get Single Recombinant
  Inbred Lines SRLs by using molecular markers
• Arranging these SRLs in terms of the
  recombination event would produce STAIRS

39
Creation of STAIRS

• Chromosome (100 cM in length), each step will, on
  average be 10 cM above the previous line
  (phenotypic variation)
• Target 10 STAIRS for each chromosome
• Total 70 STAIRS
• If a trait of interest is located between two
  overlapping SRLs, fine mapping of this region
  will be possible by crossing these two lines

40
Creation of STAIRS

• Again 11 well spaced, one at each end and 9
  interstitial will be used to find SRLs
• If the markers are both from the recipient or
  donor, we expect to have 0, two or four
  crossovers and such lines will be discarded
• The approximate location of the recombination can
  be determined by using 9 interstitial markers
• These lines will be self-fertilized
• Seed for each SRL will be kept in groups

41
(No Transcript)
42
The Power of STAIRS

• Methodology
• Score CSSs for leaf rust
• QTL for leaf rust is localized on 2H and 6H
  (Backes et al. 2003)
• A modified leaf segment test for leaf rust
  resistance (Walther, 1991)
• 3-4 leaf segments infected with I-80 (highly
  virulent isolate) that overwhelms all recently
  known major resistance genes except Rph7 (Backes
  et al. 2003)

43
The Power of STAIRS

• If the phenotype shows low score for CSS-2 and
  high for original recipient then gene is on
  chromosome 2
• Screen the groups of SRLs for chromosome 2
• For example, if the first 4 are resistant and the
  last four are susceptible to leaf rust
• QTL is in the vicinity of marker 5(M5) or in
  between M4 and M6

44
How to use for QTL?
(A)
(B)
(C)
45
Summary of STAIRS
Breeding program to produce single whole
Chromosome Substitution Strains (CSSs) (C) and
Stepped Single Recombinant Lines (SSRLs) (E).
Individual Bc1a individuals containing single,
intact whole chromosomes (only chromosome 1 is
illustrated) are selected (B). These are both
selfed to produce the corresponding true breeding
CSS (C) and backcrossed to the recurrent parent
to generate recombinants (D). Individual Bc2
individuals (D) are in turn selfed to generate
the corresponding true breeding SSRLs (E) which
constitute the STAIRS. See text for further
explanation                                       
                 
46
Expected results

• Seven whole chromosome substitution lines will be
  obtained
• STAIRS will be produced for each chromosome
• QTL with confidence interval less than 1 cM can
  be located in three steps
• a) compare CSS with recipient
• b) compare the 10 STAIRS produced for that
• chromosome

47
Potential Pitfalls

• Large population needed to produce CSSs and STAIR
• The possibility of CSSs or SRLs in which two or
  four crossovers events go undetected and may
  affect the results
• More markers needed (expensive, and time
  consuming)

48
Questions