microRNAs

MicroRNAs

In 1998, the Fireand Mello labs described a new technology based on specific gene silencing by double-stranded RNA, termed RNA interference (RNAi) (Fire et al., 1998). They showed that in C. elegans, the presence of just a few molecules of double-stranded RNA (dsRNA) was sufficient to essentially abolish the expression of a gene homologous to the dsRNA. This discovery was soon to illuminate the posttranscriptional gene silencing phenomenon in plants, connecting previous observations of silencing activity to mechanisms relating to dsRNA.

Specifically, an important observation in plants suggested that the mechanism of dsRNA silencing might occur through the action of small ~25 nt double-stranded RNAs (Hamilton and Baulcombe, 1999). Following this key observation, studies were initiated to dissect the biochemistry and molecular mechanisms of the RNAi pathway, using extracts prepared from Drosophila embryos (Bernstein et al., 2001; Hammond et al., 2000; Hammond et al., 2001; Tuschl et al., 1999; Zamore et al., 2000). These studies, which were supported by genetic analysis of RNAi in C. elegans (Grishok et al., 2001; Grishok et al., 2000; Tabara et al., 1999), suggested a two-step model, in which ~21 nt short interfering RNAs (siRNAs) were processed from long dsRNA and used to specifically degrade mRNA. It didn’t take long for the structure of the siRNA to be defined, and investigators rapidly embraced RNAi biology by using chemically synthesized siRNAs to perform gene function studies in mammalian cells (Elbashir et al., 2001). While the RNAi pathway was being dissected through biochemical and genetic means, it became clear that the pathway served a broad and important role in generating small RNAs to silence endogenous gene expression.

The lin-4 Gene

The story actually begins almost 15 years ago, when Victor Ambros and colleagues discovered that lin-4, a gene important in the regulation of C. elegans developmental timing, did not encode for a protein but instead encoded for a small RNA of approximately 22 nt in length (Lee et al., 1993). The lin-4 RNA acts as a negative regulator of the developmentally important lin-14 and lin-28 protein-coding genes (Lee et al., 1993; Moss et al., 1997; Olsen and Ambros, 1999; Wightman et al., 1993). The 22-nucleotide lin-4 RNA contains imperfect homology to specific regions of the 3’ untranslated regions (UTRs) of lin-14 and lin-28, and likely other developmentally important genes. Genetic data suggested a potential mechanism of action: deletion of the lin-4 target sequences in the UTR causes an unregulated gain-of-function phenotype. In support of this mechanism, reporter genes containing fusions of the lin-4 UTR renders it susceptible to developmental regulation (Lee et al., 1993; Wightman et al., 1993). These studies suggested that the lin-4 acts to downregulate gene expression by binding to the regions of lin-4 homology in the 3-UTR. However, it is generally not believed that the down-regulation is mechanistically akin to siRNA action. Northern blot analysis of the target mRNAs indicated that the message remains stably associated with polysomes, and the current models suggest that the action of lin-4 acts via translational repression (Olsen and Ambros, 1999).

Additional microRNAs

When lin-4 was originally described, it seemed that this small RNA was a feature peculiar to worms. But then a second small RNA was found in worms, encoded by the heterochronic let-7 gene, which was found to be present in other animals (Pasquinelli et al., 2000). The conservation of let-7 gene is striking: it is conserved among bilaterally symmetrical animals, from sea urchin to human. Suspecting that even more small RNAs might exist among other organisms, three different groups independently developed clever methods for cloning such small RNAs from worms, flies, and mammals. These studies surprised the scientific community when it was reported that literally hundreds of different small RNAs exist within cells (Lagos-Quintana et al., 2001; Lau et al., 2001; Lee and Ambros, 2001). Because of their small size and developmentally-temporal manner of expression, these RNAs were originally named small temporal RNAs, but are now generally referred to as microRNAs, or miRNAs (Lagos-Quintana et al., 2001; Lau et al., 2001; Lee and Ambros, 2001).

Many of the microRNA sequences are conserved throughout evolution, and we now know that RNA silencing based on microRNAs is ancient in origin and conserved in plants, animals, and fungi. Each of these organisms produces double-stranded RNA (dsRNA) as part of the normal pathways of gene regulation (for review, see (Baulcombe, 2004; Lippman and Martienssen, 2004; Meister and Tuschl, 2004; Mello and Conte, 2004)). The overall mechanisms by which a small RNA may target an mRNA for degradation or translational repression have been assimilated into a general two-step model, in which the small RNA is used to direct the silencing of genes (Figure 1). This review will not cover the details of small RNA mechanisms of action, but instead highlight general aspects of small RNAs that may relate to immune function. These aspects include the role of the RNAi pathways in intracellular immunity against invading nucleic acids (such as transposable elements and viruses) and the role of the microRNA pathway in cellular differentiation and cancer.

Even though microRNAs were only recently discovered, much has been learned in only a short time (for a comprehensive review in microRNA biology/biogenesis, see (Bartel, 2004; He and Hannon, 2004)). Most microRNAs are present as solitary elements in the genome, residing in intergenic regions that were previously thought to be barren of genes and perhaps not even transcribed (Lagos-Quintana et al., 2001; Lau et al., 2001; Lee and Ambros, 2001). However, some can be found in clusters reminiscent of polycistrons, containing the same or different sequences of microRNAs. In some cases, the microRNAs may be present within the introns of protein-encoding genes, perhaps hitchhiking on the transcription and regulation of the upstream promoter of the protein-encoding gene (Lagos-Quintana et al., 2001; Lau et al., 2001; Lee and Ambros, 2001). It may be speculated that the presence of such microRNA clusters may be an evolutionary result of the need to swiftly and simultaneously co-regulate several different targets.

How many microRNAs are there? Several independent lines of evidence predict that within the human genome there are approximately 200-255 microRNA genes; Drosophila, 96-124 microRNA genes; and C. elegans, 103-120 microRNA genes (Lim et al., 2003a; Lim et al., 2003b). Thus, it can be estimated that microRNAs may typically constitute approximately 1% of an organisms known genes. These estimates are based on a combination of computational and experimental approaches, and although a few additional microRNAs may be found, we suspect that the majority of microRNAs have been identified in mammalian lineages.

Some microRNAs are highly expressed within a given cell, and others are expressed at levels that are difficult to detect. For the most highly expressed microRNAs in C. elegans, the number of transcripts has been estimated to be more than 50,000 molecules per cell– a greater abundance than that of the U6 snRNA of the spliceosome (Lim et al., 2003b). Other microRNAs appear to be expressed at levels that are barely detectable by Northern blot analysis. However, when estimating the expression level of a given microRNA from a given tissue, it is important to realize that the microRNA may be expressed in only a small subset of cells within a given tissue. Thus, it may be prudent to use methodologies that may probe the expression patterns of microRNAs within a given tissue, such as the use of microRNA microarrays on selected cell types (Calin et al., 2004a; Esau et al., 2004; Miska et al., 2004) or in situ approaches similar to what has been recently reported for use in the mouse (Mansfield et al., 2004).

Little is known about the regulation of microRNA expression, but the emerging evidence suggests that most, if not all microRNAs are transcribed by Polymerase II (Cai et al., 2004; Lee et al., 2004). Although Polymerase II is the enzyme generally known to transcribe protein-coding genes, it is not unusual that it transcribes noncoding RNAs. In fact, Polymerase II transcribes some noncoding RNAs, including the small nucleolar RNAs and four of the small nuclear RNAs of the spliceosome. Like other Polymerase II transcripts, the primary microRNA transcripts (pri-miRNAs) contain cap structures, undergo splicing, and contain poly (A) tails (Figure 2) (Cai et al., 2004; Lee et al., 2004). In support of this, many microRNA primary transcripts can be found in est libraries, and can be quite long (3 kb or longer). In addition, the microRNA promoter can be used to express protein encoding reporter genes, and there is at least one documented case in which a MYC gene was expressed from a microRNA promoter in an aggressive B-cell leukemia (Gauwerky et al., 1989). These data are important, since they suggest that the regulation of microRNA expression may be as rich and as complicated as protein-encoding genes, sharing many of the same transcription factors.

How do microRNAs work? Current data argue that microRNAs can exhibit their silencing effects by two separate mechanisms: translational repression and mRNA degradation. The mechanisms for how the translational repression occurs are not understood, but the interaction is characterized by 1) imperfect pairing between the microRNA and the target and 2) a correspondence between the degree of repression and the number of target sites. This is illustrated by the interaction of let-7 with the lin-41 untranslated region (Reinhart et al., 2000) (Figure 3A). Those microRNAs that promote mRNA degradation do so by acting as siRNAs. Here, only one interaction of perfect/near-perfect hybridization is needed to suppress the expression because the microRNA guides the cleavage of the mRNA, which occurs between the tenth and eleventh nucleotide, starting from the small RNA 5’-end. In plants, this second mechanism is principally how microRNAs act (Rhoades et al., 2002); however, in animals, only one microRNA so far has been shown to act as a siRNA. This is the microRNA termed miR-196, which appears to target the HoxB8 mRNA for degradation (Figure 3B) (Mansfield et al., 2004; Yekta et al., 2004). At present, it is unknown whether other animal microRNAs may act by targeted degradation of the mRNA.