Structure‐function relationships in the vertebrate U4‐U6 snRNP have been analysed by assaying the ability of mutant RNAs to form U4‐U6 snRNPs and to function in splicing complementation in Xenopus oocytes. The mutants define three categories of domain within the RNAs. First, domains which are not essential for splicing. These include regions of U6 which have previously been implicated in the capping and transport to the nucleus of U6 RNA as well as, less surprisingly, regions of U4 and U6 which have been poorly conserved in evolution. Second, domains whose mutation reduces U4‐U6 snRNP assembly or stability. This group includes mutations in both the proposed U4‐U6 interaction domain, and also, in the case of U6, in a highly conserve sequence flanking stem I of the interaction domain. These mutants are all defective in splicing. Third, regions not required for U4‐U6 assembly, but required for splicing complementation. This category defines domains which are likely to be required for specific contacts with other components of the splicing machinery. Combinations of mutants in the U4 and U6 interaction domain are used to show that there are not only requirements for base complementarity but also for specific sequences in these regions.

We have characterized the U2 snRNA gene family in the higher plant Arabidopsis thaliana. It consists of 10‐15 genes which do not appear to be closely clustered. Six of the U2 genes were sequenced and the structure of the Arabidopsis U2 RNA termini was determined in order to define the coding regions. Each of the genes codes for a distinct RNA differing from the others by 2‐13 point mutations, localized in the 3′ part of the 196 nt‐long RNA. The upstream non‐coding regions of all genes show strong sequence similarity in positions −81 to −1 and contain three highly conserved sequence elements: GTCCCACATCG (positions −78 to −68; 100% conservation), GTAGTATAAATA (−37 to −26) and CAANTC (−6 to −1). The coding regions are followed by the sequence CAN7‐9AGTNNAA, a putative termination signal. The expression of three of the genes was studied in electroporated Orychophragmus violaceus and Nicotiana tabacum protoplasts. The genes, one of which contains a T → C change in the Sm antigen binding site, were actively transcribed and processed into U2 RNAs of the expected size and containing trimethylguanosine caps. Deletion analysis indicates that sequences upstream of the conserved −80 to −1 region are not important for transcription in protoplasts. The 5′‐terminal parts of U2 RNAs from several monocot and dicot plants were sequenced. This region, containing the sequence implicated in base‐pairing with the branch point in pre‐mRNA introns, is identical in all U2 RNAs examined.

The U2 and U5 snRNA genes of Arabidopsis thaliana contain in their promoter regions two elements with conserved sequence and position. To test the significance of this conservation we have made a construction in which the promoter of the U2 RNA gene is replaced by the synthetic 98 bp long sequence containing the two conserved elements: an upstream sequence element, GTCCCACATCG (USE, pos. ‐78 to ‐68), and a TATA‐like sequence TATAAATA (‐33 to ‐26), positioned approximately three helical turns apart, as in the wild‐type promoter. This synthetic promoter efficiently drove transcription of the U2 gene in transfected protoplasts of Nicotiana plumbaginifolia. The importance of the individual elements and of their position within the promoter was investigated. Deletion of the USE, change of its orientation, and some single point mutations all decreased transcription 10‐ to 20‐fold, and replacement of the TATA‐like element by an unrelated sequence inactivated the promoter. Mutants in which the spacing between the USE and TATAAATA was changed were less active but no correlation was observed between promoter activity and insertion of either odd or even numbers of half helical turns. Insertion of a spacer between TATAAATA and the cap site resulted in accumulation of U2 RNA with an extended 5′ end, indicating that the TATAAATA element is responsible for selection of the initiation site. The data indicate that the promoters of RNA polymerase II‐specific U‐snRNA genes in higher plants differ from their animal counter‐parts and also from plant mRNA gene promoters. They contain two essential elements, an USE, an element found only in U‐snRNA genes, and a TATA element which is indistinguishable from the TATA boxes of mRNA‐coding genes.

A series of U4 and U6 snRNA mutants was analysed in Xenopus oocytes to determine whether they block splicing complex assembly or splicing itself. All the U4 and U6 mutants found to be inactive in splicing complementation resulted in defects in assembly of either U4/U6 snRNP or of splicing complexes. No mutants were found to separate the entry of U5 and U6 snRNAs into splicing complexes and neither of these RNAs was able to associate with the pre-mRNA in the absence of U4. In the absence of U6 snRNA, however, U4 entered a complex containing pre-mRNA as well as the Ul and U2 snRNAs. U6 nucleotides whose mutation resulted in specific blockage of the second step of splicing in Saccharomyces cerevisiae are shown not to be essential for splicing in the oocyte assay. The results are discussed in terms of the roles of U4 and U6 in the assembly and catalytic steps of the splicing process.

We have previously characterized the U2 small nuclear (sn) RNA gene family of Arabidopsis thaliana. To find out the structural features of upstream and downstream non-coding regions that are shared by different U-RNA genes in higher plants we have isolated the gene encoding a 125 nt-long U5 snRNA of Arabidopsis. Activity of the cloned gene was demonstrated in stably transformed tobacco calli and by transient expression in transfected protoplasts of Nicotiana plumbaginifolia. Southern analysis indicated that the Arabidopsis genome contains 8-9 copies of the U5 gene. Alignment of upstream non-coding regions revealed two elements conserved between all plant U-RNA genes characterized so far: the sequence RTCCCACATCG (-70/-80 region, 100% conservation) and the TATA homology around position -30. The coding regions in all genes are followed by the sequence CAN4-9AGTN (A/T)AA which may correspond to a termination and/or processing signal.