A novel function for a micro RNA: Negative regulators can do positive for the hepatitis C virus


  • Potential conflict of interest: Nothing to report.


MicroRNAs are small RNA molecules that regulate messenger RNA (mRNA) expression. MicroRNA 122 (miR-122) is specifically expressed and highly abundant in the human liver. We show that the sequestration of miR-122 in liver cells results in marked loss of autonomously replicating hepatitis C viral RNAs. A genetic interaction between miR-122 and the 5′ noncoding region of the viral genome was revealed by mutational analyses of the predicted microRNA binding site and ectopic expression of miR-122 molecules containing compensatory mutations. Studies with replication-defective RNAs suggested that miR-122 did not detectably affect mRNA translation or RNA stability. Therefore, miR-122 is likely to facilitate replication of the viral RNA, suggesting that miR-122 may present a target for antiviral intervention.

Jopling, CL, Yi M, Lancaster AM, Lemon SM, Sarnow P. Modulation of hepatitis C virus RNA abundance by a liver-specific MicroRNA. Science 2005;309:1577-1581. (Reprinted with permission.)


MicroRNAs (miRNAs) are a group of about 22 nucleotides (nts) short nucleic acids that are involved in regulating a multitude of different biological processes.1, 2 Based on computational analysis up to 1,000 miRNAs appear to be encoded in the human genome that potentially regulate more than 5,000 genes.3, 4 MiRNAs are initially transcribed as part of a much larger precursor termed pri-miRNA that is processed in the nucleus by a ribonuclease complex containing the endonuclease III Drosha.5–7 After recognition and binding of the pri-miRNA Drosha cleaves at both sites of a 60 - 80 nts stem loop and liberates an approximately 60 nts hairpin termed pre-miRNA.8 It is exported into the cytoplasm where a second processing occurs. This cleavage is mediated by the cytoplasmic endonuclease III Dicer and results in the formation of an approximately 22 nts long RNA duplex bearing 2-nt overhangs at both 3′ ends. After unwinding one strand of the mature miRNA is integrated into the RNA-induced silencing (RISC) complex. It is guided via complementary base pairing to a homologous target sequence in the mRNA resulting in either nucleolytic degradation,9, 10 or an arrest of mRNA translation,11, 12 depending on the degree of complementarity. Current evidence suggests that binding of miRNAs to their target sequence dramatically influences mRNA stability.13, 14 It was found that miRNA-mRNA duplexes are transferred to “P-bodies”, which are the sites of mRNA degradation.15, 16 In addition, P bodies do not contain ribosomal components and therefore mRNAs sequestered into P-bodies are excluded from translation.

Although miRNAs have a length of about 22 nts, a mere 6 or 7 nts within the miRNA confers most of the binding specificity with the target sequence. This “seed sequence” usually resides between nts 2-7 of the miRNA, counting from the 5′ end. It is thought that the seed sequence initiates the binding to the target site and facilitates further base pairing -if possible- between the target and the 3′ terminal miRNA sequence.

All small regulatory RNAs described thus far (small interfering RNAs, miRNAs or repeat associated small interfering RNAs) negatively regulate gene expression. This paradigm has now been broken by a recent study showing for the first time that a human liver-specific miRNA acts as a positive regulator.17 The object of this study was not a cellular gene but rather a virus, the hepatitis C virus (HCV). Infections with this pathogen frequently lead to chronic hepatitis and may result in cirrhosis and hepatocellular carcinoma. The HCV genome is a single stranded RNA molecule of positive polarity and composed of a 5′ non-translated region (NTR), a single open reading frame (ORF), and a 3′ NTR.18 The ORF encodes a polyprotein of about 3,000 amino acids that is cleaved co- and post-translationally into 10 different products: core, E1, E2 p7, NS2, NS3, NS4A, NS4B, NS5A, and NS5B. Translation of the polyprotein is mediated by an internal ribosome entry site (IRES) comprising domains II, III and IV in the 5′ NTR (Fig. 1). Domain II, together with domain I is also required for RNA replication.

Figure 1.

(A) Schematic representation of the HCV 5′ NTR according to27. The region required for HCV RNA replication is highlighted by the grey box. Complementary nucleotide sequences negatively regulating HCV IRES activity are indicated by hatched boxes and the double headed arrow. (B) Primary sequence of the first 40 nts of the HCV genome containing domain one (dI). The seed sequence of miR-122 is highlighted with stars, the region involved in base pairing with the core sequence (nt 428-442) is marked with the line. (C) Watson–Crick base pairing between miR-122 and the complementary region in the HCV 5′ NTR according to17.

Studies of HCV replication became possible with the advent of replicons comprising at least the NTRs and the NS3 to NS5B coding region.18 Replication of these RNAs was restricted for a long time to the human hepatoma cell line Huh-7, but could recently be established in other cell lines including other human liver cell lines (e.g., HepG2),19, 20 human cervical carcinoma-derived cells (HeLa),21, 22 and murine liver cells (Hepa1-6).21 Thus, HCV RNA replication is neither liver specific nor restricted to human cells.

Building on the observation that the miRNA miR-122 is expressed to high levels specifically in the liver,23 Jopling and coworkers first monitored miR-122 abundance in several cell lines.17 They observed well-detectable expression in Huh-7 and Hepa1-6 cells, but not in HepG2 or HeLa cells corroborating earlier findings.23 Assuming that miR-122 might contribute to the efficient replication of HCV in Huh-7 cells, the viral genome was scanned for complementarity to the miR-122 seed sequence. Two such positions were identified: one residing in a conserved stretch in the variable region of the 3′ NTR and one just downstream of the first stem-loop in the 5′ NTR (Fig. 1). Taking advantage of a Huh-7 cell line carrying a selectable genomic HCV replicon of genotype 1b, Jopling and colleagues found that sequestration of miR-122 by transfection of a chemically modified antisense-oligonucleotide reduced the level of HCV RNA up to 5-fold. This result was confirmed by two alternative approaches: first, in a transient assay in which replication of a cell culture-adapted full length HCV genome of genotype 1a (H77c) transfected into Huh-7 cells turned out to be equally sensitive to miRNA-122 depletion; second, by ectopic expression of miR-122 in HCV transfected Huh-7 cells resulting in an about 3-fold increase of viral RNA levels. Site-directed mutagenesis was then used to demonstrate that the miR-122 target sequence in the 5′ NTR is responsible for these effects. These results support the notion that miR-122 levels modulate HCV RNA amounts in Huh-7 cells.

In a series of elegant experiments Jopling and coworkers were able to demonstrate that miR-122 acts via direct binding to the HCV target sequence.17 Single nucleotide substitutions introduced into the miR-122 seed sequence resulted in an up to 5-fold reduction of HCV RNA replication as determined in a transient replication assay using the full length H77c genome. This defect was rescued by ectopic expression of a mutated miRNA in which the seed sequence perfectly matched the altered seed sequence in the HCV genome, but it was not rescued by wild-type miR-122. These results provide firm proof for a direct interaction between miR-122 and HCV RNA and they exclude off-target effects.

Three mechanisms may account for the increase of HCV RNA abundance by miR-122. First, an increase of viral RNA translation; second, an increase of HCV RNA stability; third, a direct enhancement of viral RNA replication. To address the first two possibilities, Jopling and co-workers compared RNA translation upon transfection of Huh-7 cells with either the full length H77c HCV genome or a mutant that carried a single nucleotide substitution in the seed sequence of the 5′ NTR. No obvious difference in the amount of core protein expressed from the two transfected genomes was found arguing that miR-122 does not affect HCV RNA translation and RNA stability. In a complementary approach Jopling and colleagues used a subgenomic reporter replicon in which HCV RNA amounts can be deduced indirectly from the level of expressed reporter.24 It was found that reporter gene expression with a replicon carrying a single point mutation in the miR-122 seed sequence was as efficient as reporter gene expression achieved with the analogous replicon with a wild-type seed sequence. This result supports the notion that neither RNA half life nor translation from the HCV IRES is affected by miR-122. Based on these exclusion criteria it is assumed that miR-122 directly or indirectly enhances HCV RNA replication.

What can we learn from these studies and what is the possible relevance for HCV replication in vivo? The study by Jopling and colleagues clearly shows that HCV RNA replication can be modulated by the abundance of miR-122 in Huh-7 cells. However, miR-122 is not expressed in several other cell lines that also support HCV RNA replication such as HeLa,21, 22 HepG2,19, 20 or 293T cells22, 25 arguing that miR-122 is not a major determinant of host cell permissiveness and that miR-122-HCV RNA interaction per se is not required for viral RNA replication. However, it is possible that in HeLa or HepG2 cells carrying a stably replicating HCV RNA miR-122 levels are upregulated and in this way host cell permissiveness is enhanced. Moreover, it will be important to determine whether ectopic expression of miR-122 in non-permissive or poorly permissive cells positively affects HCV RNA replication. Finally, studies in primary human hepatocytes expressing high amounts of miR-12217, 23 will ultimately be required to get an idea about the role of miR-122 in vivo.

Thus far, it is also unclear by which mechanism miR-122 enhances HCV RNA abundance. The studies performed by Jopling and coworkers suggest that miR-122 neither affects RNA stability nor RNA translation. However, this can not be ruled out rigorously at this stage. Although not very likely, mutations introduced into the seed sequence may enhance RNA translation and at the same time reduce RNA stability. This possibility could be addressed by careful comparisons of RNA half lives based on direct measurements of transfected HCV RNAs.

Interestingly, in 2003 Kim and coworkers reported a long range RNA-RNA interaction between 24-38 of the HCV 5′ NTR and nts 428-442 of the core coding sequence (Fig. 1A).26 This interaction was shown to impair translation efficiency of the HCV IRES in rabbit reticulocyte lysate and in HepG2 cells. It would be interesting to determine whether ectopic expression of miR-122 in these cells that do not contain this miRNA affects HCV RNA replication and whether mutations affecting the long range RNA-RNA interaction also affect miR-122 mediated elevation of viral RNA abundance.

In summary, the study by Jopling and coworkers describes an unexpected role of a miRNA. In the light of their work the paradigm that miRNAs are negative regulators of gene expression obviously is no longer valid illustrating that miRNAs can have multiple regulatory functions. As shown here for miR-122, it can act as a positive regulator for a viral RNA, but presumably as a negative regulator for its natural target assumed to be the mRNA of the cationic amino acid transporter gene.23 Given the large number of miRNAs encoded in the human genome, one may expect that not only miR-122 but perhaps also other miRNAs may act as negative regulators on one target but as a positive regulator on another target. Moreover, because viruses are obligate intracellular parasites that exploit host cell functions in multiple ways it is conceivable that apart from HCV other viruses use miRNAs as a way to regulate individual steps in their life cycle not only in a negative way, but eventually also in a positive one.