A soil bacterium, Agrobacterium, transfers
the T-DNA containing genes that encode the proteins involved in biosynthesis of
plant growth factors and bacterial nutrients. Because none of the genes carried
by the T-DNA are required for the transfer, foreign DNA can be inserted into
the plant chromosome by putting the sequences between two T-DNA borders and
using the Agrobacterium transfer system. Utilising this naturally occurring
property, efficient transformation protocols have been developed for Arabidopsis
(Bechtold et al., 1993) and rice (Hiei et al., 1994). It is now feasible to
rapidly generate the thousands of transformants necessary for investigating
genome-wide mutagenesis (Clogh and Bent 1998; Jeon et al., 2000).
The T-DNA integration patterns mediated by
Agrobacterium usually are much simpler than those observed after
transformations that use direct gene transfer procedures, e.g., particle
bombardment, PEG-mediated or electroporation. This integrated T-DNA consists of
one to a few copies, rendering easy analysis and reducing the risks of
expression instabilities that could be caused by further recombination or
gene-silencing events. T-DNA does not require additional steps to stabilise the
insert.
Several research groups have independently
created large collections of T-DNA insertional mutants of Arabidopsis and rice.
The Arabidopsis knockout facility at the University of Wisconsin recently
established a population of 60 480 T-DNA-tagged lines (Krysan et al., 1999).
The T-DNA tagging strategy has also been employed in rice. Jeon et al. (2000),
Jeon and An (2001) and Sallaud et al. (2002) have been generating a large
number of fertile rice lines, and expect to tag more than 100 000 using several
different types of T-DNAs. These are significant steps towards the production
of genome-wide coverage of mutations in plant species.
It is assumed that T-DNA insertion events
occur randomly, because no hot spots or preferences for a certain gene have
been observed. This is advanta-geous when developing a large population of
well-distributed mutations. In species with a large genome, the number of
tagging lines necessary for mutating nearly all the genes is apparently smaller
than the estimated value. Barakat et al. 1997, 2000 have found that T-DNA insertions
occur preferentially in gene-rich regions. They isolated nuclear DNAs from a
pool of independent transformants of both Arabidopsis and rice, and
fractionated the DNAs by preparative centrifugation in Cs2SO4 density gradients
in the presence of 3,6-bis(acetatomercurimethyl)-1,4-dioxane (BAMD).
Hybridisation of T-DNA to the Cs2SO4/BAMD fractions showed that the signals
were localised in fractions, which corresponds to gene-rich regions in both
species. Because genes occupy most of the Arabidopsis genome, T-DNA appears to
be integrated essentially everywhere in the genome. In contrast, integration of
T-DNA into rice was detected mainly in gene-rich regions, i.e., the
transcriptionally active regions that correspond to 100 Mb or only 25% of the
entire rice genome. That, however, is approximately the size of the entire
Arabidopsis genome, suggesting that saturation of transcribed regions by T-DNA
insertion in rice may only require approximately the same number of insertions
as in Arabidopsis.
Generated mutant populations can be
maintained and distributed primarily as seeds. The Arabidopsis Biological
Resource Center (Columbus, OH) has provided mutant lines that are essential for
its functional genomics. In some other species, however, it is difficult to obtain
enough seeds from the first generation after transformation. In an experiment
with 10 743 primary transgenic rice plants, 39% bore fewer than 50 seeds and
only 50% yielded more than 100 (Jeon et al., 2000). Yin and Wang (2000) had
similar results from their analysis of 2633 transgenic rice plants. Although
the majority of the transgenic plants became fully fertile in the next
generation, 22% of the lines showed less than 50% fertility. Reduced fertility
is probably due, in part, to the lethal effects of T-DNA insertion.
However, a considerable number of lethal
mutations likely are induced during the T-DNA transformation process.
Chromosome rearrangement or modification might also occur during
transformation. Nevertheless, those lines that show low fertility must be
individually amplified before being used for further analysis.
T-DNA-tagged lines are useful resources
when studying gene function. A number of genes have been isolated from those
lines through several methods (Azpiroz-Leehan and Feldmann, 1997). Insertion
lines can be used to identify mutants whose functions are altered by T-DNA
insertion. Because most mutations are recessive, their phenotypes can be
detected in a segregating population. After segregation analysis of a large
number of transformants, an average number of independent inserts per diploid
genome has now been estimated. In Arabidopsis, genetic characterisation of a
subset of the transformants indicates that they contain an average of 1.4
inserts each (Feldmann, 1991). This is consistent with an average of 1.5
inserts found in a transgenic rice population (Jeon et al., 2000).
Approximately 8000 lines of T-DNA-inserted transgenic Arabidopsis have been
screened under a variety of growth conditions for visible alterations in
phenotype (Feldmann, 1991). These mutants fall into several classes, such as
seedling-lethals (3–5%); size variants that do not include dwarfs (3–5%);
pigment defect (2–3%); embryo-defective (2.5–3.5%); reduced-fertility (1–2%);
morphological mutants (2.5–3.0%) in flowers, roots, root hairs, trichomes and
dwarfs; and physiological mutants (>1%) involving flowering time,
eceriferum, and high florescence.
In total, 15–26% of transgenic Arabidopsis
plants show phenotype abnormalities. However, in species such as rice, whose
plants are relatively tall and wide, it is more difficult to find mutants that
show phenotypic alterations in the adult stages. Therefore, such a forward
screening of phenotypic mutants can be employed only for limited lines and
times. Analysis of 1600 T-DNA-tagged rice lines in the T2 generation resulted
in identification of various mutants at the vegetative stages (Jeon and An,
2001). The most common characteristics were dwarf phenotype (7.0%) and
leaf-pigment mutations (9.5%), such as albino, pale-green, chlorina, striped or
zebra (transverse green and chlorotic bands). Spotted leaves (1.0%) and
leaf-morphology mutations (1.2%) were also found. Seedling mortality was 1.1%.
The T-DNA-tagged rice population also displayed mutant phenotypes in the reproductive
organs, such as depressed paleae, filamentous flowers, extra glumes and long
sterile glumes. About 1.8% of the lines were completely sterile. Flowering-time
mutations included early-flowering (0.2%) and late-flowering (0.1%) phenotypes.
Co-segregation analysis of mutant
phenotypes with the T-DNA insertion is needed to determine whether these
phenotypes are due to that particular insertion. DNA blot analysis can be
employed. Alternatively, PCR (polymerase chain reaction)-based methods can be
used when the sequence that flanks the T-DNA is determined. Several techniques
are available for identifying this flanking sequence, such as thermal
asymmetric interlaced (TAIL) PCR (Liu and Whittier, 1995), adapter-ligated PCR
(Balzergue et al., 2001), inverse PCR (Triglia et al., 1988), a universal
biotinylated adapter amplification procedure (Hanley et al., 2000), panhandle
PCR (Walbot, 2000) or a plasmid rescue system (Weigel et al., 2000) (see
section on ‘isolation of a gene tag from insertion lines’ in this chapter).
Many phenotypes may be manifested as only
subtle changes in growth or development, therefore demanding a sensitive
methodology for their detection. For Arabidopsis, Boyes et al. (2001) have
established an analysis procedure based on a series of defined growth stages
that serve as developmental landmarks and triggers for the collection of
morphological data involving growth and development over the entire life of the
plant. Their data collection process is divided into two complementary
platforms: characterisation of early seedling growth on vertical plates for two
weeks, plus an extensive examination of plants grown in soil for two months.
Using this method, they have demonstrated that each of the tested mutants that
had no reported morphological phenotypes actually had some developmental
alteration compared with wild-type plants. Establishing such a high throughput
process of phenotypic analysis should facilitate the functional discovery of
genes that are essential for growth and development in other plant species.
As a reverse genetics approach, the T-DNA
insertional line can be identified in a given gene via PCR-based screening
(Krysan et al., 1999; Sato et al., 1999; see section on ‘reverse genetics’ in
this chapter). Using a gene-specific primer and a primer located near the end
of the T-DNA, a DNA fragment flanking the inserted T-DNA can be amplified and
its sequence then determined. DNA pools of a large number of lines are commonly
used in this approach. Ultimately, establishment of a sequence database of
T-DNA-flanking regions can excel identification of insertional mutants.
T-DNA insertional lines also contain genetic
mutations not linked with T-DNA. Such mutations may be attributed to endogenous
transposons. In rice, Tos17 is one such retrotransposon that is activated by
tissue culture stress (Hirochika et al., 1996b; see section on ‘retrotransposon
tagging’ in this chapter). The T-DNA-tagged lines in rice possess an average of
four new copies of Tos17 (Jeon and An, 2001). Hence, the T-DNA tagging lines
can be used for screening insertional mutants caused by such an identified
endogenous transposon.
0 komentar:
Post a Comment