We have been analyzing reversion of an
unc-22 ::Tc3allele and an
unc-22 ::Tc5allele to examine two questions: 1) what mutator activities do Tc3 and Tc5 respond to; and 2) what, if any, footprints remain at the "empty sites" generated by excision of each element. To address these questions we started with TR1160 [genotype
unc-22 (
r750::Tc3)]and TR1034 [genotype
unc-22 (
r644::Tc5)].These strains contain the respective
unc-22 insertion alleles in a non-mutator background; neither strain reverts at detectable frequency. To monitor Tc3 and Tc5 activity in response to different mutators, we measured reversion of
r750::Tc3 and
r644::Tc5 following substitution into the
mut-2 and
mut-5 backgrounds. [Thanks to Ron Plasterk for suggesting the use of the tightly linked marker
vab-9 (
e1744)for
mut-5 ,and supplying the
mut-5 ,
vab-9 double mutant used in these studies]. Reversion of both
r750::Tc3 and
r644::Tc5 occurred at a frequency of about 10 (-3) in response to
mut-2 .This was expected, considering that both insertion alleles were originally isolated in a
mut-2 background. Recently we have observed reversion of
r644::Tc5 in the
mut-5 background, providing genetic evidence for activation of Tc5 by
mut-5 .This represents the first evidence for movement of Tc5 in other than a
mut-2 background. Molecular analysis of these revertant alleles is in progress to test whether reversion resulted from excision of Tc5 .We are also screening more populations to determine the frequency of
mut-5 -inducedreversion of
r644::Tc5 .We are interested in the relative levels of Tc5 activity in the
mut-5 vs.
mut-2 backgrounds. Experiments are underway to determine if Tc3 moves in response to
mut-5 . We looked at the footprints generated by excision of each element by amplifying empty sites via PCR and sequencing the PCR products. To date, our analysis has been restricted exclusively to
mut-2 -inducedrevertant alleles. For
r644::Tc5 ,25/25 revertant alleles analyzed had lost Tc5 .These included: 4 "precise" excision events (wild-type
unc-22 sequence restored); 19 "imprecise" events that generated small footprints at the excision site [small (+/- 21bp), in frame insertions or deletions]; and 2 large deletions (size and extent not yet determined). For
r750::Tc3 ,13/17 revertant alleles analyzed had lost Tc3 .These included: 3 "precise" events, 6 small footprints (+/-18 bp in frame insertions or deletions), and one 57 bp deletion. The remaining four are surprising. Tc3 is still there! See the accompanying article by Mills et al. in this issue. For both elements, all the footprints are in frame; presumably this reflects constraints imposed by the selection used to obtain them. What do these excision site sequences tell us? All can be explained by the double-strand gap repair model for transposition/excision proposed by Engels and co-workers (Cell 62:515-525). According to this model, footprints and other rearrangements at the excision site are not the direct result of imprecise excision, but of interrupted repair of the double-strand gap generated by excision. Results of Mori et al. suggest Tc1 moves by such a mechanism, as subsequently demonstrated by Plasterk (EMBO J. 10:1919-1925) and Moermann et al. (Nuc. Acids Res. 19:5669-5672). Our results suggest Tc3 and Tc5 excision can be explained by this model as well. We are currently analyzing reversion of
r750.-:Tc3 and
r644::Tc5 in "heteroallelic" constructs (see Engels paper), in both
mut-2 and
mut-5 backgrounds, to test this idea.