Worm Breeder's Gazette 11(3): 25
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.
In the last WBG some of us (Conrad, Thomas, Spieth and Blumenthal) reported that a conventional gene (vit-2) can be converted into a trans-spliced gene by insertion of all of an intron except the splice donor into its 5' untranslated region. We concluded that the primary signal for trans-splicing is the presence of a splice acceptor and branch site without a donor site upstream. In order to learn more about the sequence requirements for a trans-splice acceptor, we are making mutants of the rol-6 acceptor and testing them in vivo. The experimental system involves injection of a plasmid containing the 2.2 kb HindIII fragment which carries the dominant roller allele of rol-6. We altered several nucleotides in the coding region which result in no change in the protein but which allow us to test specifically for the RNA product of the transgene. When this plasmid is injected, rollers are obtained and the RNA is found to be trans-spliced exclusively at the normal site immediately preceding the AUG at which translation initiates. Most transcripts contain SL1, but surprisingly, some begin with SL2. If the UUCCAG which serves as the splice acceptor is converted to UUCCAA, rollers are nevertheless obtained. Analysis of the RNA from these strains revealed that SL1 was spliced onto a different site ( CUUUAG), 20 nucleotides upstream of the normal SL1 acceptor, a site which is not used at all when the wild-type acceptor is present. In this mutant, trans-splicing is relatively inefficient (Unspliced transcripts are also seen.) and no SL2-spliced products have been observed. We have also tested a few less drastic alterations in the splice acceptor site. One (AUUCAG) still splices solely at the normal acceptor, one (UAACAG) splices predominantly at the normal acceptor, and one (AAAAAG) splices solely at the upstream acceptor. All three splice at reduced efficiency and only to SL1. We hypothesize that when splicing of the RNA products from the tandem arrays is very efficient, which so far is true only with the wild-type splice acceptor, the localized SL1 snRNP concentration in the expressing cells is reduced, thereby allowing the normally less-favored SL2 snRNP to donate its 5' end. To further test the idea that an acceptor site without a donor site upstream is the primary signal for trans-splicing, we inserted a consensus donor site (GTAAG) 131 nucleotides upstream of the trans- splice acceptor. RNA from transformants containing this construct was mostly cis-spliced from this novel donor site to the normal trans- splice acceptor, but a small amount of SL1 trans-splicing still occurred at this site. Our working hypothesis is that, if the right signals were present, cis-splicing, trans-splicing with SL1 and trans- splicing with SL2 could all compete. We don't yet know all the factors that may mitigate towards cis- or trans-splicing, but the quality of the U1 snRNP binding site (the donor site) is certainly important. We are currently trying to locate a signal which biases towards an SL1- or SL2-specific splice.