In C. elegans some transcripts receive a 22 base spliced leader (SL) from the SL snRNP by trans-splicing, while others do not. We are interested in the question of what property of the recipient transcripts marks them for trans-splicing. We are testing the simple idea that the presence of a splice acceptor and branch site, with no donor site upstream, is the signal for trans-splicing. The recent discovery that some RNAs receive SL1 while others receive SL2, however, suggests that this simple hypothesis cannot be the whole story. There must also be a specificity determinant. In order to test the idea, we created a modified vit gene (normally not trans-spliced). We engineered an almost-complete intron from vit- 5 into our standard
vit-2/vit-6 fusion gene (J. Spieth et. al., 1988, Dev. Biol. 130, 285 ). The 11 nucleotide stretch of 5' untranslated RNA normally found in this gene was converted to a 95 nucleotide region containing a
vit-5 intron lacking only the 5' splice- donor site. Two stable transgenic strains containing this gene, both of which express the fusion protein, were selected. The transgene is transcribed exclusively in the adult hermaphrodite intestine but in low amounts. Both strains are viable only as heterozygotes for the fusion gene and produce very small broods. We have successfully performed two types of experiments with the small amount of RNA we have been able to obtain from these two strains. The results of both experiments support the idea that the fusion message is trans-spliced to SL1. In the first experiment, we performed primer extension with an oligonucleotide capable of hybridizing to either the endogenous
vit-2 mRNA or the fusion mRNA. Primer extension products were found corresponding to the expected lengths of the former (a very intense band) and of the predicted trans- spliced product (a very weak band). The latter band was generated only from RNA preparations made from the two transgenic strains carrying the modified vit fusion gene. In the second experiment we amplified reverse transcription products of the fusion mRNA using PCR. A downstream oligonucleotide, complementary to
vit-2 mRNA, was the primer extended in the reverse transcription step. Three different upstream oligonucleotides were used: oligo A corresponds to the 5' end of the initial transcript of the modified fusion gene; oligo SL1 corresponds to the SL1 sequence; and oligo SL2 corresponds to the SL2 sequence. The PCR products were probed with an oligonucleotide complementary to a region of
vit-2 between the primers. With oligo A, the PCR product was not dependent on the reverse transcription step and so was probably due to DNA contamination. With oligo SL1, an intense band of the size predicted for the trans-spliced mRNA was seen only with RNA isolated from transgenic lines containing the modified fusion gene. A much weaker band was seen with oligo SL2, which may be due to a small amount of trans-splicing to SL2, but could also result entirely from cross-hybridization of the oligo SL2 with the SL1 sequence. Vit-2 mRNA is not normally trans-spliced. The fact that addition of most of an intron into its 5' untranslated region results in trans- splicing indicates that this DNA contains the information to specify trans-splicing to SL1. The simplest explanation is that SL1 splicing represents a default mode for splice-acceptor sites which have no upstream splice-donor, and that SL2 splicing may require a more specific signal. However, we have not eliminated the possibility that the
vit-5 intron sequences inserted in the modified fusion gene coincidentally harbor the information for an SL1-specific splice.