By searching Genbank, we have identified C. elegans cDNAs encoding proteins known in mammals to be needed for RNA processing: 3' end formation and splicing. Full length antisense RNAs were generated by in vitro transcription from plasmids (kindly supplied by Y. Kohara) and injected into the gonads of adult worms in order to mimic the knockout phenotype of these genes. Injection of antisense RNA to the subunits of the 3' end formation protein, CstF, which binds RNA downstream of the cleavage site and is thought to be involved in cleavage site selection, resulted in an embryonic lethal phenotype. This is perhaps not surprising as null alleles of the Drosophila homolog of the 77K subunit of CstF are also embryonic lethal (1,2). To investigate the phenotypes of knockouts of genes involved in splicing, antisense RNAs were injected into
smg-2 animals to prevent degradation of any resulting incompletely processed RNAs. Injection of antisense RNA to the integral U1-associated protein, U1C, resulted in a developmental delay phenotype. In contrast, injection of RNA antisense to another integral U1 protein, U170K, caused embryonic lethality. The difference in phenotype between knockouts of these two genes may simply reflect injection of different concentrations of antisense RNA or different maternal protein contributions. We have used RT-PCR to look at processing of the
myo-3 gene in these delayed or dead embryos. The embryos from antisense-injected mothers accumulated large amounts of incompletely cis-spliced and unprocessed pre-mRNAs suggesting that the phenotype is indeed due to perturbation of splicing. Interestingly, only the mRNAs from which both introns had been removed were found to be trans-spliced, suggesting that either cis-splicing must occur in order for trans-splicing to occur in vivo or trans-splicing is even more sensitive to lack of U1 snRNP than is cis-splicing. The latter explanation would be quite surprising, since trans-splicing in Ascaris has been shown to be U1 snRNP-independent in vitro (3) (although the possibility of these proteins serving as components of the SL snRNP has not been investigated). We have recently identified two U1A protein homologs arranged in an operon. The two encoded proteins appear to be quite different from one another, particularly at key residues within the second RNA binding domain. Nevertheless, the antisense injection experiments indicate they are functionally redundant: neither antisense RNA on its own gave a detectable phenotype, but when both were injected together approximately half the embryos died and the remainder exhibited delayed development, hatching 48 hours after laying instead of the usual 12 hours. We have not yet tested for cis- or trans-splicing defects in these embryos. Overall, our results demonstrate that genes involved in mRNA processing are vital to embryonic development and also confirm the power of antisense technology in predicting the knockout phenotypes of genes. (1)Mitchelson, A., M. Simonelig, C. Williams and K. O'Hare (1993) Homology with Saccharomyces cerevisiae RNA14 suggests that phenotypic suppression in Drosophila melanogaster by suppressor of forked occurs at the level of RNA stability. Genes Dev 7: 241-9 (2)Takagaki, Y. and J. L. Manley (1994) A polyadenylation factor subunit is the human homologue of the Drosophila suppressor of forked protein. Nature 372: 471-474 (3) Hannon, G. J., P. A. Maroney and T. W. Nilsen (1991) U small nuclear ribonucleoprotein requirements for nematode cis- and trans-splicing in vitro. J Biol Chem 266: 22792-5