Three distinct kinds of RNA molecules orchestrate gene expression in all cells. Messenger RNAs (mRNAs) bring the information from the genome to the protein synthesis factory, the ribosome, which itself is composed mostly of RNA (rRNA). Transfer RNAs (tRNAs) serve as the adapter molecules that align the correct amino acids on the mRNA template during the elongation of protein chains.
A fourth class of RNA molecules, called noncoding RNAs, is also crucial for gene expression in higher cells. Small RNAs (having chain lengths of 60–300 nucleotides), which were the first to be characterized, are highly conserved and particularly abundant in the nuclei of mammalian cells. Small nuclear RNAs (snRNAs) exist tightly bound to one or more proteins in particles called small nuclear ribonucleoproteins (snRNPs: pronounced "snurps"). Some snRNPs inhabit the nucleoplasm, which contains the DNA and is devoted to the production of mRNAs for export to the cytoplasm. Others occupy the nucleolus, where ribosomes are assembled before being shunted to the cytoplasm to function in protein synthesis. More recently, a novel family of tiny regulatory RNAs, dubbed microRNAs because they are only about 22 nucleotides long, has been discovered. They too bind proteins (creating microRNPs), but control gene expression at the level of translation and mRNA stability in the cytoplasm. MicroRNAs play important roles in cancer and infectious disease, as well as in normal development.
The most numerous snRNPs of the nucleoplasm (the U1, U2, U5, and U4/U6 particles) participate in pre-mRNA splicing, an early step in gene expression. The DNA and pre-mRNAs in mammalian and other higher cells contain nonsense segments called introns, which must be excised for the mRNA to encode a meaningful protein product. In the nucleus, the pre-mRNA is cut, and the sense segments—exons—are precisely joined back together. Splicing occurs in spliceosomes, which are assembled on each intron and include the above snRNPs and a host of protein factors. A much less abundant minor spliceosome removes 1/300 introns from human pre-mRNAs by recognizing slightly different sequences near the ends of these introns. Our studies of the splicing of the minor introns have revealed the participation of one common snRNP (U5) and four snRNPs (U11, U12, U4atac, and U6atac) that are distinct but structurally and functionally analogous to their counterparts in the major spliceosome. Both kinds of spliceosome deposit on the newly spliced mRNA proteins that later act in the cytoplasm, for instance to survey and elicit decay of faulty messages. We have discovered that one of these proteins, eIF4AIII, also functions in ribosome biogenesis.
Another snRNP related to the spliceosomal snRNPs directs the maturation of histone mRNAs, which are cleaved from longer precursors but not polyadenylated at their 3′ ends (as are the vast majority of vertebrate mRNAs). We have shown that many of the same components contribute to the nuclear maturation of mRNAs both possessing and lacking poly(A) tails, as well as to developmentally controlled poly(A) tail elongation in the cytoplasm, which serves to activate translation. This overlap poses the challenging question of determining how different subsets of common factors are recruited to different molecular machines with different outcomes.
The nucleus contains additional kinds of snRNPs that contribute to the biogenesis of other classes of RNA. In the nucleolus there are about 200 different small nucleolar RNAs (snoRNAs), packaged with specific proteins to form nucleolar snRNPs (snoRNPs). These participate not only in the cutting of the long pre-rRNA precursor into the mature 18S, 5.8S, and 28S rRNAs but also in rRNA modification. About 100 2′-O-methyl groups and about 100 pseudouridines are introduced at highly conserved rRNA positions by base-pairing between a guide snoRNA and the pre-rRNA. Another group of guide RNAs acts on the snRNAs of the spliceosome, which contain multiple 2′-O-methyl and pseudouridine modifications concentrated in regions that engage in RNA-RNA interactions critical for splicing. Some of these guide RNPs reside in nucleoplasmic dots, called Cajal bodies, which are transited by snRNPs during their biogenesis. We have recently identified a protein responsible for localization of these guide RNPs in Cajal bodies.
Deciphering the roles of viral noncoding RNPs found in primate cells infected by gamma-herpesviruses is a challenging problem. This family of viruses is characterized by both lytic (replicative) and latent states. We recently found that viral snRNPs made during latent infection by Herpesvirus saimiri, an agent that causes malignant transformation of the T cells of the immune system of monkeys, are involved in up-regulation of several genes implicated in T cell activation. The molecular mechanism involves the binding and degradation of a specific host microRNA. Epstein-Barr virus (EBV), which causes mononucleosis as well as certain human cancers, encodes two small RNAs called EBERs that are extremely abundant in the nucleus of transformed cells. By identifying changes in the host proteome, we are investigating how the presence of EBERs alters host metabolism to the advantage of the virus.
Investigation of another abundant viral noncoding RNP called PAN (for polyadenylated nuclear), which is present in cells lytically infected by Kaposi sarcoma–associated herpesvirus (KSHV—the cause of the most common cancer afflicting AIDS patients), has revealed that it contains an RNA element (called the ENE) necessary and sufficient for the nuclear accumulation of intronless RNA transcripts. The ENE counteracts rapid nuclear degradation of unspliced RNAs dependent on the presence of a 3′-poly(A) tail. Our data argue that the ENE hybridizes to and sequesters the poly(A) from degradation, the initial step in decay. To probe the mechanism of nuclear RNA surveillance in mammalian cells, we have recently solved the x-ray structure of the ENE bound to oligo(A)9. We observe formation of an elegant triple-stranded structure that clamps the poly(A) tail. These structural insights into the exact interaction between the ENE and poly(A) have allowed identification of ENE-like structures in cellular noncoding RNAs, which are currently being investigated. Concurrent studies have led to the conclusion that the PAN RNA contributes to KSHV lytic infection by avid binding of the cytoplasmic poly(A)-binding protein (PAB), which is relocalized to the nucleus, thereby allowing synthesis of late viral proteins.
Our investigation of the translation of the tumor necrosis factor α (TNFα) mRNA, which contains an AU-rich element (ARE) 3′ of its open reading frame, produced surprising results. Under conditions of cell cycle arrest, which induce a state called quiescence, the efficiency of protein synthesis from this message increases and the proteins Argonaute 2 (AGO2) and fragile X mental retardation–related protein 1 (FXR1) are bound to the ARE. Since these are microRNA-associated proteins previously implicated in the repression of translation, these findings suggested unanticipated links between the microRNP pathway and ARE-mediated translational control. Indeed, the involvement of a particular microRNA in translation up-regulation by the ARE has been documented. Moreover, we have obtained evidence that microRNAs in general can direct up-regulation versus repression of translation dependent on the cell cycle. These observations have been extended to the Xenopus laevis oocyte, which is naturally quiescent, where up-regulation of translation by microRNPs appears to be critical for maintenance of the immature state.
Studies of the biogenesis of microRNAs have revealed that the earliest steps of processing of the primary microRNA transcript occur preferentially before release from the chromatin template. We are investigating the roles of post-translational modifications of Drosha and DGCR8, which are essential for this nuclear processing step.
Research on cellular and viral small RNPs is supported by grants from the National Institutes of Health.
As of May 30, 2012