Worm Breeder's Gazette 11(5): 66
These abstracts should not be cited in bibliographies. Material contained herein should be treated as personal communication and should be cited as such only with the consent of the author.
Transposon tagging remains an efficient way to clone mutationally
defined genes in C. elegans, whether by cloning out the flanking DNA
or by looking for allele-specific rearrangements with wild-type probes.
Often, multiple alleles of the gene of interest are obtained from
mutator backgrounds, but the effort required to finely map or clone
out more than just a type specimen does not usually warrant the
exercise. It is useful to know the location of insertion sites
relative to the established structure of the gene, because insertions
in exons indicate probable null alleles, while insertions upstream or
downstream of the transcript can identify potential cfs-acting
regulatory sequences. Here I describe a novel strategy (
'echolocation') to obtain quite precise (+ or - 10bp) estimates of the
site of insertion by PCR (recently published in NAR 18: 6741 qv for
more details).
The ends of most transposons are defined by short sequences that are
inverted with respect to each other. Thus a single oligonucleotide,
oriented such that it would prime DNA synthesis away from the
transposon, can do so from both ends of the transposon simultaneously (
see also Hill and Sternberg CSH '89 Abstracts p.124). The strategy
involves pairwise combination of such an oligonucleotide with each of
several oligonucleotides of either orientation derived from the
genomic sequence throughout the region of suspected transposon
insertion. Whenever the transposon primer has sufficient proximity to
an appropriately oriented genomic primer, a product ('echo') will be
formed, whose length can be determined by high-resolution agarose gel
electrophoresis. For example, the x508 allele of the lev-1 gene of C.
elegans (which encodes a muscle acetylcholine receptor sub unit)
arose in TR679, and contains the transposon Tc1 (Barnes and Lewis, WBG
11(2):47). Nine different primers were derived from the wild-type
genomic sequence of the lev-1 gene of C. elegans (Mike Squire, pers.
comm), (Fig. A). The transposon primer was a 20mer derived from bp
1582 to 1601 of the Tc1 sequence (1), designed such that it avoided
the exact end of Tc1 which is somewhat similar between different
transposons (2). Each of the genomic primers was then paired with the
Tc1 primer in a PCR reaction, using unbackcrossed lev-1(x508) genomic
DNA as template. The products were resolved on a 2% agarose gel,
allowing the size of the smallest fragment to be determined with high
accuracy. Only four primers produced products, and from their sizes
they indicate that the site of insertion lies near an intron/exon
boundary (Fig. B). The two favorably oriented primers that failed to
produce products (primers 1 and MS2) are probably too distant from the
Tc1 primer to work with a 2 minute extension time. This method has
also been applied to the two other Tc1 alleles, x504 and x562, which
has located them near a different intron/exon boundary, and in an exon,
respectively. The precise estimate can be converted into an exact
guess by dint of the site preference for Tc1 (4). Each of the defined
regions had just two TpA sequences, and one was clearly a better match
to the consensus than the other in each case. Thus x504 is predicted
to have occurred in a splice donor, and x508 in an exon.
The reliability of the strategy is demonstrated by the fact that
different genomic primers, from either side of the transposon, produce
fragments which are consistent with the same single site (Fig. B).
The technique is also fairly robust: the DNA used has ~600 priming
sites for the Tc1 primer per genome (3), yet there is little problem
with spurious bands; when evident, these are also produced by TR679
DNA. Finally, to determine the site of insertion exactly, the PCR
products can be easily cloned and sequenced.
[See Figure 1]