Research Group - IPK Gatersleben

Abteilung Genbank/Department of Genebank
In the context of the map-based isolation of the GA-insensitive
semi-dwarfing gene sdw3, the target interval,
comprising 0.04 cM on a high resolution genetic map of
barley, has been further analysed by comparative mapping
in rice and Brachypodium sylvaticum. As to the latter,
a contiguous sequence of 365 kb was obtained from
screening a BAC library and sequence analysis of an overlapping
set of four BACs. The target interval in rice encompasses
180 kb and comprises 4 genes of which 15 are
not present in Brachypodium. The corresponding region
in Brachypodium sylvaticum measures about 0 kb with
16 annotated genes of which seven are not present in
rice. The remaining nine genes are conserved between
both species and show perfect collinearity. Screening of
a barley BAC library by using eight genes from the target
interval led to the identification of seven BAC contigs. Of
these, only two contigs identified by two genes spaced
by 800 bp in Brachypodium showed overlaps. Based on
survey sequencing of one representative BAC per barley
contig, 11 complete and two fragmented genes were
identified. Seven genes were found in a colinear position
in rice while four genes are unique to barley. The results
confirm previous observations on varying gene repertoires
of orthologous regions in rice, Brachypodium and
Triticeae and provide further proof of the necessity to rely
on several genomes when exploiting synteny with model
genomes for map-based cloning. Based on the sequence
information, three candidate genes for GA-signalling are
further investigated by re-sequencing sdw3 mutant and
wild type genotypes in barley and by analysing induced
mutants identified from a TILLING population (V.T.H.
Giang).
To assist the functional verification of candidate genes
a TILLING (targeted induced local lesions in genomes)
population is being developed. The establishment of an
M mutant population consisting of 10,49 independent
plants was completed. For systematic screening of this
population, DNA of 8,500 M mutants has been arranged
in 8-fold D pools. During the reference period, screening
of a subset comprising 7,348 lines was completed for
five genes. The averaged mutation frequency amounts to
1/500kb. 35 % of the mutations revealed missense alleles
and 4 % resulted in truncations. 35 % of the silent mutations
were present in noncoding regions (S. Gottwald).
As a step towards the development of genome-based
strategies for the utilisation of genetic diversity a study
on linkage disequilibrium (LD) mapping in barley has
been completed. To this end, a world-wide collection of
5 spring barley cultivars was phenotyped in multi-location
trials for five agronomic traits (heading date, plant
height, thousand grain weight, starch content, protein
content). Single Nucleotide Polymorphisms (SNPs) in nine
candidate genes were correlated with trait data in order
to identify associations. Using a mixed model that includes
information on population structure, row-type and origin
of the accessions, significant marker trait associations
34
were identified for five of the candidate genes explaining
up to 5 % of the phenotypic variation, respectively (see
Fig. 11). Interestingly, heading date in the present population
was influenced only by allelic variation in the gene
PpdH1, while allelic variation in other candidate genes of
the flowering time signalling pathway (Hv-CO1, Hv-FT1)
did not yield significant effects. While the results are very
encouraging regarding the potential of LD-based association
mapping approaches in barley, they also demonstrate
that population structure is a major limitation for
the statistic and genetic power of the approach. Hence,
Fig. 11 Association mapping in spring barley. A) Geographic origin of
4 accessions used for LD mapping of candidate genes. B) Frequency
distribution and broad sense heritability (h²) for Raw Protein Content
(RPC) for the subpopulations of -rowed (n=96) and 6-rowed barleys
(n=1 8). C) Association analysis of SNPs detected in the candidate
gene Blz1, a bZIP transcription factor expressed in early endosperm
development. SNPs with minor allele frequencies > 5 % are indicated
on top by their respective position in the annotated gene. SNPs printed
in bold are associated (p< 0.05) with RPC. Haplotypes of that gene
explain 10.5 % of the phenotypic variation. Pairwise LD values were
calculated as r². The r² scale is binned by a colour code. Complete LD
(1.0) is indicated by red whereas white refers to alleles at Hardy Weinberg
equilibrium (r²=0) (G. Haseneyer et al.).