Agriculture Reference
In-Depth Information
et al., 2005). Lowly expressed genes and highly
specialized genes that are transcribed only during
specifi c developmental stages, in specifi c tissues,
or under specifi c environmental conditions will
generally not be represented in EST databases.
These genes might be obtained using gene-
enrichment techniques that target gene-contain-
ing genomic DNA fragments. These methods are
based either on the different kinetics with which
low-copy sequences and repeats undergo reasso-
ciation following denaturation (high C
0
t
; Yuan et
al., 2003) or the differential methylation of genes
and repeats. Transcribed genes tend to be
hypomethylated while repeats generally carry
5-methyl groups at CG and CNG locations.
Preferential digestion of genic regions with meth-
ylation-sensitive restriction enzymes (Emberton
et al., 2005) and selective propagation of hypo-
methylated fragments in methylation-restrictive
bacterial hosts (Rabinowicz et al., 1999) are two
methods that can be employed to generate librar-
ies of cloned genomic fragments enriched for
genes. Pilot studies of these methodologies in
maize showed that high C
0
t
and methyl-fi ltered
reads tagged >95% of genes (Springer et al.,
2004) and reduced the effective size of the genome
to be sequenced at least fi vefold (Barbazuk et al.,
2005).
Application of the high C
0
t
technology to
hexaploid wheat genomic DNA resulted in a
13.7-fold enrichment in gene sequences, a 5.8-
fold enrichment in unknown low-copy sequences,
and a 3-fold reduction in repetitive DNA
(Lamoureux et al., 2005). The gene enrichment
obtained in wheat was slightly higher than in
maize (4.2%), which is probably a refl ection of
the lower relative gene content and higher repeat
content in wheat compared with the maize
genome. When methyl fi ltration was applied to
wheat, the outcome was rather surprising. In
maize, the gene-enrichment factor (GEF),
which is the percentage of genes present in a
methyl-fi ltered library relative to the percentage
of genes present in a WGS library, was 13.15.
One would expect the GEF to increase with
increasing genome size (Rabinowicz et al., 2005),
but in hexaploid wheat, the GEF was only 4.7.
A low GEF was also observed in diploid wheat.
Barley, on the other hand, had an expected GEF
of 18.7. In diploid wheat, the low GEF appeared
to be due mainly to the high percentage of
unmethylated repeats. Methyl fi ltration is there-
fore unlikely to be useful as an enrichment strat-
egy in diploid wheat (Li et al., 2004; Rabinowicz
et al., 2005). In hexaploid wheat, the proportion
of repeats in the methyl-fi ltered library was
similar to that in other grasses, but the low GEF
was due to the high number of genes identifi ed
in the WGS library. Many of these genes
appeared to be methylated and are likely pseu-
dogenes (Rabinowicz et al., 2005). It should be
noted that the gene number of 98,000 identifi ed
by Rabinowicz et al. (2005) in the WGS library
was considerably higher than the 36,000 genes
estimated from 3B BAC-end sequences (Paux
et al., 2006). If the GEF in wheat is indeed
artifi cially defl ated due to the presence of large
numbers of methylated pseudogenes in the WGS
(control) libraries, then methyl fi ltration does
provide actual enrichment of active genes in
hexaploid wheat.
One problem with gene-enrichment tech-
niques is that the gene sequences are not placed
in a genomic context. To obtain information on
the location of the genes in the genome, the
cDNA or genomic fragments need to be super-
imposed on a low-redundancy BAC draft
sequence. Another technology that could aid in
connecting the sequences generated by gene-
enrichment techniques is methylation spanning
linker libraries. These are large-insert clones
generated with methylation-sensitive restriction
enzymes. These clones have low-copy sequences
at their ends, and the remaining sequence
consists of methylated repeats. Hence they span
the distance between neighboring genes (Yuan
et al., 2002). This technique would work for
connecting genes present in the more distal
chromosome regions of wheat where intergenic
distances are likely to fall in the range of a
BAC cloning vector, but it might not be techni-
cally feasible to span the large intergenic dis-
tances that are likely to exist in the proximal
chromosome regions.
