The Genes and Transcription of the Major Small Nuclear RNAs

Summary

Several unusual features distinguish snRNA genes and make snRNA synthesis an important and interesting subject for study. Although the snRNA genes are very efficiently and accurately transcribed by transcription complexes which use RNA polymerase II (or III, in the case of U6 genes), these genes contain transcription signals that differ from those normally recognized by RNA polymerase II.

The genes for the major snRNAs constitute a collection of multigene families, each of which, in mammals, has 5–30 members (fewer in chickens or insects) (Table 1). Each member of an snRNA gene family contains a significant amount of DNA that is dispensable for gene expression, but which is nonetheless conserved between members of the same family. Apparently, each of these gene families arose by amplification of an ancient progenitor DNA sequence containing only one or a few snRNA genes. The amounts of genomic DNA that were co-amplified varies considerably from gene to gene and species to species.

Amplification of relatively short DNA segments have yielded snRNA genes arranged in short, perfect tandemly repeated units (e.g., human U2 genes) whereas duplication of large segments has resulted in snRNA genes that are loosely clustered but appear to be independent units with considerable flanking region homology (e.g., human U1 genes). In certain species (e.g., Xenopus) the numbers of particular types of snRNA genes have been expanded to more than 500 per haploid equivalent, thereby allowing extremely high rates of snRNA synthesis at specific stages of early development.

In addition to the multiple true genes, many genomes, the mammalian ones in particular, contain a large number of untranscribed snRNA pseudogenes (Table 2). The presence of a high level of retroposon-like pseudogenes, which evidently arose by RNA-mediated events, indicates that within the germline cells enzymes such as reverse transcriptase have access to snRNAs. In contrast, the lack of pseudogenes in genomes of such species as Xenopus and Drosophila may be due to differences between oogenesis in these organisms and the mammals (as discussed by Weiner et al. 1986). The other class of pseudogenes, which is generated entirely by DNA-mediated events and which bears strong resemblance to snRNA true genes, apparently arose as a result of the continued evolution of snRNA gene families via mutation and alternating cycles of transposition and amplification.

Several characteristics set transcription of snRNA genes apart from that of other genes. First, the major snRNA transcription signals are unique to snRNA genes (Table 3). For example, snRNA enhancers apparently are unable to activate mRNA promoters, and conversely, although mRNA enhancers can activate snRNA transcription, they do so nonspecifically. Moreover, formation of the 3’ ends of snRNAs is coupled to transcription from snRNA promoters. Secondly, the snRNA promoters are among the most powerful in the cell, being capable of initiating transcription precisely at position +1 once every 2–4 s. Thus, in spite of the relatively small number of snRNA transcription units (totalling perhaps 200–300 per HeLa cell), the number of initiation events at snRNA promoters needed to synthesize ∼ 2-3 x 10⁶ snRNA transcripts per cell per generation must account for a substantial fraction of all the initiations performed by RNA polymerase II. Presumably, this efficiency and accuracy of snRNA gene expression is due to unique properties of the snRNA-specific transcription complexes.

One of the most striking features of this system is the immediate export of the newly synthesized snRNAs to the cytoplasm. Once in the cytoplasm the snRNAs undergo maturation and assembly into snRNPs prior to their reentry in-to the nucleus. This series of events, is the opposite of that of other RNA polymerase II (and III) transcripts, which remain in the nucleus until they are fully processed. In all cases, however, a common feature is the sequestration of the precursor RNAs in cell compartments that are different from the ones where the RNAs ultimately function; transport into the latter compartments (the nucleus for snRNAs and the cytoplasm for both mRNAs and tRNAs) occurs once the RNAs are mature. Presumably the rapid export of newly synthesized snRNA precursors is required to prevent their inactivation by polyadenylation or other nuclear processing events, or by association with hnRNP binding proteins.

Perhaps the unusual structure of snRNA promoters and 3’ end signals are responsible for the coupling of snRNA synthesis to the immediate export of the RNA product. In this regard it is interesting that the 5’ flanking regions of both U1–U5 genes (transcribed by RNA polymerase II) and U6 genes (transcribed by RNA polymerase III) contain a similar sequence (snRNA-TATA-box, Table 5) which is essential for transcription. That sequence may target both types of snRNA genes to a nuclear compartment at or close to the inner nuclear membrane from which the snRNAs could be rapidly exported.

SnRNA gene transcription in invertebrates probably obeys many of the same general rules as those described for the vertebrate genes; however, the snRNA- specific transcription signals differ in sequence from those of vertebrate genes and their functions have yet to be tested experimentally. Thus, short sequences shared between the 5’ flanking regions of Drosophila U1-U5 and U6 genes (Table 4) may have functions comparable to those of similarly located sequences shared between vertebrate U1-U5 and U6 genes.

It is unclear why attempts to make in vitro transcription systems for vertebrate U1–U5 genes have been unsuccessful. Perhaps the coupling of synthesis and immediate export of snRNA from the nucleus makes such systems sensitive to the disruption or loss of a nuclear structure during extract preparation. Moreover, the standard RNA polymerase II in vitro systems may lack other factors which directly or indirectly activate transcription of snRNA genes.

In recent years our knowledge of snRNA gene structure and transcription has increased greatly. However, as we discuss throughout this chapter, many important questions remain to be answered. We expect that an even better understanding of the mechanisms and control of snRNA gene expression will emerge in coming years, fostered (hopefully) by the development of an accurate and efficient system for transcription of vertebrate snRNA genes in vitro.