Small RNAs in HnRNP fibrils and their possible function in splicing.

Several arguments are in favor of a function of snRNA in the processing of premessenger RNA. A large fraction of snRNA is localized in hnRNP which are assumed to be the site of processing. The different snRNA species are not bound to hnRNP in a unique manner but are associated with both proteins and hnRNA which suggests the possibility of metabolic exchanges in the course of processing. There is approximately 1-2 molecules of snRNA per individual hnRNP. We reexamined the possibility that U1A RNA might serve for the alignment of the extremities of the intron sequences of premessenger RNA insuring correct condition for cutting and splicing. We found that only a UCCA (3' leads to 5') sequence at position 8-11 of U1A RNA was complementary to an AG-GU (5' leads to 3') around a putative splice point for 69 different introns sequenced so far. On the basis of secondary structure of U1A RNA, the UCCA sequence would be available for hybridization. The UCCA sequence is also present in U2 RNA and 4.5 S RNAI. It might associate with AG-GU in a manner similar to that of codon-anticodon, the stability of the complex being insured by the configuration of hnRNP. The possible formation of larger hybrids stable by themselves is unlikely upon examination of the nucleotide sequence of various introns adjacent to the splice point. As there is no direct experimental evidence for the function of snRNA in splicing, there considerations are speculative at the present time. The possibility that adenovirus encoded VA RNA would play a role in splicing was also examined. Various arguments suggest that this possibility is rather remote.