"Frontmatter". In: Plant Genomics and Proteomics

bidopsis can be accounted for by the relative amounts of methylated repeated
DNA. Thus the level of gene methylation may be about the same in the two
species. However, using the methylation status will render some parts of the
genome that are of interest inaccessible. Thus, for example, the methylation
status of genes, or portions of genes, can differ markedly between tissues
and/or developmental stages. Furthermore, the association between methy-
lation and expression is certainly not absolute. Therefore, the choice of tissue
from which the library is constructed is important (Yuan et al., 2002). 
The removal of methylated regions can be achieved at two different
points in the cloning procedure. The first approach is with the use of a
methylation-sensitive restriction enzyme, that is, one where the enzyme does
not cleave the DNA if a cytosine in its recognition site is methylated. This
approach has been used successfully to isolate low-copy-number sequences
for molecular mapping probes, especially when using the enzyme PstI. The
second method takes advantage of the E. coli pathway that digests incom-
ing sequences if they are methylated. Mammalian and plant DNA can be
cloned more efficiently in bacterial strains that are mutant at the mcrA and
mcrBC loci because these strains do not eliminate recombinants that contain
methylated cytosines. Most of the time the cloning strategy is to get a high
representation of all the sequences present. By using E. coli that still have a
functional mcr system, all those recombinants containing methylated DNA
will be eliminated, leaving a library highly enriched in nonmethylated
sequences. Therefore, with the known distribution of much of the methyla-
tion in higher plants, the mcr system can be used to filter out methylated
regions, resulting in an enrichment of genes at the expense of repeated
sequences in such libraries. 
T
RANSPOSON
T
AGS
For some well-characterized genomes such as that of maize, certain families
of transposable elements are known to have a preference to transpose into
M
E T H O D S O F
F
R A C T I O N AT I N G T H E
G
E N O M E
5 7

---

genes. Therefore, any method of isolating the regions adjacent to these ele-
ments will also isolate sequences enriched in genes. However, it is unlikely
that this could be extended to use as a general method for sequencing a gene
set for most higher plants because the distribution and characterization of
the transposon content has not been done. 
S
ELECTING
BAC C
ONTIGS
E
NRICHED FOR
E
XPRESSED
G
ENES
Rather than sequencing all the BACs that have been identified and placed
on the physical map, those BAC contigs enriched for genes can be identified
by using hybridization to overgo oligonucleotide probes (Han et al., 2000).
Overgo probes are paired 24-mer oligonucleotides that contain an 8-bp 
complementary overlap. By annealing the oligonucleotides to each other 
and performing a [
32
P]dCTP fill-in reaction, a labeled 40-mer is created 
that can be used for highly specific filter hybridization. This method 
has proven very efficient to construct sequence-ready BAC contigs for the
human genome-sequencing project, where up to 40 distinct overgos have
been used for simultaneous hybridization. The probes are designed from
nonredundant cDNA or genomic sequences, but they still must be masked
to eliminate repetitive elements. Therefore, some knowledge of the major
families of repetitive elements for the species is necessary. The overgo probes
are then designed from the masked sequence according to melting temper-
ature, GC content, and potential for internal base pairing. The overgo probes
are then hybridized to the BACs, and the labeled BACs are placed in the
sequencing pipeline. Just sequencing the BACs containing expressed
sequences will reduce the absolute amount of sequencing necessary. A pos-
sible second benefit to this strategy would arise if there were gene-rich
islands, because the BACs selected in this way would have a high density
of genes on them.