In the last issue we described four unusual revenant alleles of
unc-22 (
r750::Tc3)in which Tc3 is still present at the
r750 insertion site, except for minor alterations at the 5' insertional junctions [Mills, Glasner and Collins; WBGv12 #4].In an effort to understand how these changes restore
unc-22 gene function without loss of Tc3 ,we examined the structure of
unc-22 mRNA for
r750 ant the revertants. We amplified first strand cDNA via PCR and determined the nucleotide sequence of the region of interest. The results explain the observed phenotypes, and in the process revealing something interesting about Tc3 itis spliced from the
unc-22 transcript. Splicing of Tc3 is "imprecise", however, leaving a "footprint"- in the mRNA. This footprint alters the
unc-22 reading frame in
r750 ,while preserving it in the revertants.Tc3 is inserted in the middle of a 1.4 kb exon of the
unc-22 gene in
r750 .However, most of the element is spliced from
unc-22 (
r750)mRNA, resulting in a net addition of 29 nucleotides to the message a frameshift. The donor site of this "Tc3 intron" is in the inverted repeat of Tc3 ,40 nucleotides downstream of the 5'-
unc-22 /Tc3junction. Note that this "adds" 40 Tc3 -encodednucleotites to
unc-22 mRNA. The splice acceptor site is in
unc-22 ,13 nucleotides downstream of the 3'-Tc3/unc-22 junction. This results in a net loss of 11
unc-22 -encodednucleotides from the message (the TA target site duplicate flanking the insertion site is not counted). We have cDNA sequence tats for three revertant alleles thus far. For each, Tc3 is spliced from the
unc-22 transcript using the same donor and acceptor sites as in
r750 .Analysis of the fourth revertant allele described previously, and a fifth identified recently, are in progress. At the DNA level, the revertant alleles differ from
r750 by small deletions and/or duplications near the 5' insertional junction, presumably generated by excision of Tc3 ant interruption of the ensuing homolog-dependent double-strand gap repair just short of completion (see Mills, Glasner ant Collins; WBGv12 #4).For the three alleles analyzed so far, the genomic changes affect
unc-22 message, since they are upstream of the splice donor site. As a result, splicing of Tc3 generates footprints in
unc-22 mRNA that are in-frame (
cj206 ,net +18 nucleotides;
cj213 ,+33; 9222, +18). These alleles confer a "partial" reverent phenotype. Incomplete restoration of
unc-22 gene function might result from the additional amino acids encoded by these mRNAs, or might reflect inefficient splicing of Tc3 from the transcript, resulting in reduced levels of
unc-22 message. These results indicate that Tc3 can serve as a "mobile intron", directing splicing of most element-encoded sequences from transcripts in which it has imposed itself. In this case, the intron is bounded by 5'-GU...intron...AG-3', although results of others (Rushforth and Anderson WBGv12 #2and #4; Li and Shaw, personal comm.; Aroian et al. WBGv12 #2;)demonstrate that these sequences, while canonical, are not required for splicing of transposon or non-transposon introns in C. elegans. For splicing of Tc3 from
r750 ,the donor site is in the inverted repeat of Tc3 .The acceptor site is in flanking
unc-22 sequenccs; it is not used in processing of wild-type
unc-22 transcripts. Presumably, Tc3 sequences located upstream in the pre-mRNA are involved in activiation of this "cryptic" acceptor site. It is possible that all sequences required in cis for splicing of Tc3 are contained within the 491 bp inverted repeats of the element. If so, splicing of Tc3 might be independent of orientation in the gene. It is also possible that utilization of these splice sites involves more complex interactions between element- and gene encoded sequences. One possibility we are exploring is that it is influenced by the relative direction of transcription of the affected gene (in this case
unc-22 )and the putative Tc3 -encodedtranscript. For
r750 ,the relative direction is opposite. cDNA sequence analysis is in progress for other
unc-22 ::Tc3aleles, representing both orientations. We would love to examine other alleles of Tc3 ,and other elements, in a variety of genes. No reasonable offer refused! How general is the phenomenon of splicing of transposons from RNA? Others have reported similar observations for Tc1 (Rushforth and Anderson WBG:
v12 #2and #4), and Tc4 /Tc4v (Li et al. Genetics 132:675-689 and prs. comm.) in C. elegans, and Ds and dSpm in maize (reviewed by Wessler, Gene 82:127-133). Such a mechanism has obvious implications for transposon biology and evolution. Accumulating evidence suggests that many (perhaps most) transposon insertions, even in exons, do not cause a mutant phenotype. Splicing of the inserted element from gene transcripts, and the footprints generated in the message may represent important factors that determine the phenotypic consequences if a particular insertion.