Subversion of a protein–microRNA signaling pathway by hepatitis C virus

Hepatitis C virus (HCV), which is a member of the Flaviviridae family, causes human liver disease with an estimated 170 million people being infected worldwide (1). Chronic HCV infections eventually lead to liver cirrhosis and hepatocellular carcinoma (1, 2). The discovery of sofosbuvir, a nucleotide analog inhibitor of the viral RNA-dependent RNA polymerase NS5B, introduced a very potent compound with a high barrier to resistance (3). Unfortunately, no vaccines are available against HCV. The viral genome is approximately 9,600 nucleotides in length and comprises a single open reading frame (ORF) which encodes a ~3,000 amino acid polyprotein that is subsequently cleaved by viral and cellular proteinases to yield structural and nonstructural viral proteins. The viral ORF is flanked by structured 5′ and 3′ untranslated regions (UTRs), which are essential for translation, replication, and stability of the viral RNA.

Because HCV is dependent on host factors for its gene amplification, the search for viral susceptibility genes with high barrier to resistance is an active area of research (4, 5). One essential factor for HCV gene amplification is liver-specific microRNA-122 (miR-122) that associates in tandem at the very 5′ end of the HCV RNA genome (6, 7) and modulates translation (8), replication, and stability (6) of the viral RNA. Clinical trials in which miR-122 was sequestered with antisense oligonucleotides greatly diminished HCV titers in patients (9, 10). While sustained virological responses were observed (9), resistance-associated mutations were identified in miR-122 binding sites in the viral RNA (10).

“Seo et al. report that a protein–miR-122 complex is modulated by a signaling cascade to enhance HCV gene amplification.”

In PNAS, Seo et al. (11) reported that a protein–miR-122 complex is modulated by a signaling cascade to enhance HCV gene amplification. This finding is very exciting because it opens the possibility to target upstream signaling factors of miR-122 than directly miR-122. Specifically, the authors discovered that RNA-binding protein ELAVL1/HuR (embryonic lethal-abnormal vision like 1/human antigen R) (12) is strongly associated with the 3′ end of miR-122 in cultured liver cells. This interaction was dependent on the presence of a 3′ terminal guanosine residue in miR-122, rather than the seed sequence in miR-122, suggesting that ELAVL1/HuR protects miR-122 from the action of 3′ riboexonucleases. ELAVL1/HuR was further shown to facilitate the expression of mature miR-122, while pri-miR-122 abundances were not affected, indicating that ELAVL1/HuR plays a key role in regulating miR-122 biogenesis at a posttranscriptional step. ELAVL1/HuR, which is known to be essential for HCV RNA replication (4), has been shown to relocalize from the nucleus to the cytoplasm in HCV-infected cells, where it can associate with the viral RNA to enhance viral replication (13). Thus, the authors examined whether ELAVL1/HuR-miR-122 complexes affect HCV RNA abundances. Indeed, ELAVL1/HuR depletion led to the concomitant decrease in miR-122 and HCV RNA abundances (Fig. 1). The abundance of a G28A mutant viral replicon, which is less dependent on miR-122, was less affected when ELAVL1/HuR was depleted, further suggesting that ELAVL1/HuR is an essential host factor for HCV RNA expression through regulating miR-122 expression. However, rescue of HCV RNA abundances in ELAVL1/HuR-depleted cells by miR-122 mimetics did not restore HCV RNA abundances to wild-type levels. ELAVL1/HuR is a ubiquitously expressed RNA-binding protein that is known to bind to certain miRs and to stabilize 3′ UTRs of target mRNAs (14). Thus, effects of additional ELAVL1/HuR-RNA complexes may contribute to the observed phenotypes. To examine whether ELAVL1/HuR-miR122 complexes associate with the viral RNA, the authors performed RNP immunoprecipitation analyses using anti-ELAVL1/HuR antibodies. While the antibodies precipitated miR-122 and viral RNA, it is not known as of yet whether ELAVL1/HuR alone or in complex with miR-122 bound to site 1 or site 2 at the viral 5′ end (Fig. 1). Selective or combinatorial binding of ELAVL1/HuR-miR-122 and ELAVL1/HuR to the viral end sequences may explain why Shwetha et al. (13) observed effects of ELAVL1/HuR depletion on replication and Seo et al. (11) noted effects on translation of the viral RNA. Of course, effects of ELAVL/HuR-miR-122 depletion on translation may be a consequence of stabilization of the viral RNA by this protein–miR complex against the attack of 5′ RNA triphosphatase DUSP11 (15).

Fig. 1.

Proposed PLK1-ELAVL1/HuR-miR-122 signaling cascade that enhances HCV RNA abundances. The top shows a diagram of the HCV RNA genome with RNA-binding sites 1 and 2 for miR-122 at the 5′ end of the viral genome. The potential interactions of miR-122 and PLK1-ELAVL1/HuR-miR-122 with the viral RNA are indicated. The enhancement of miR-122 abundances by ELAVL1/HuR is shown. Inhibition of the PLK1-ELAVL1/HuR-miR-122 signaling axis and the PI3K/Akt pathway by rigosertib is diagramed. Red (+) signs represent positive regulation of the downstream factor, while red (−) signs indicate inhibition. Potential binding of the ELAVL1/HuR-miR-122 complex to HCV 5′UTR is highlighted by the gray dotted rectangle. Gray dashed lines represent modulation of additional factors by PLK1 or rigosertib. ELAVL1/HuR: embryonic lethal-abnormal vision like 1/human antigen R. miR-122, microRNA-122; PLK1, polo-like kinase 1; PI3K, phosphoinositide 3-kinases.

To identify regulatory components of the ELAVL1/HuR-miR-122 complex, the authors used a cleverly designed kinase inhibitory screen that examined effects of compounds on the decrease of HCV replicon RNA abundances and the increased activity of an miR-122 sensor mRNA. Using this approach, chemical inhibitors of polo-like kinase 1 (PLK1), including rigosertib, were identified as potential anti-HCV drug candidates that regulate ELAVL1/HuR-miR-122 (Fig. 1). PLK1 is a known regulator of cell cycle progression and has been shown to upregulate HCV RNA replication in Huh7 liver cells without affecting cell growth (16). Depletion of PLK1 also displayed an miR-122-dependent phenotype on HCV, and addition of exogenous miR-122 mimetics or overexpression of catalytically active PLK1 largely rescued viral proliferation in PLK-1-depleted cells. Furthermore, association of tagged-PLK1 with ELAVL1/HuR and upregulation of PLK1 by HCV replicon RNAs suggest that a PLK1-ELAVL1/HuR-miR-122 signaling axis modulates HCV proliferation.

How do PLK1 inhibitors modulate the PLK1-ELAVL1/HuR/miR-122 signaling pathway? Rigosertib, a styryl benzyl sulfone compound, is currently being evaluated in clinical trials as a potential cancer therapeutics (17). It preferentially induces mitotic arrest and apoptosis in tumor cells, with little or no effect on normal cells. In the PNAS paper (11), rigosertib not only exhibited a strong anti-HCV activity but also decreased the expression of ELAVL1/HuR and mature miR-122. These inhibitory effects were further demonstrated in mice xenografted with Huh7 cells harboring HCV replicon RNAs. Rigosertib is also known to inhibit the phosphoinositide 3-kinase PI3K/AKT signaling pathway (18). Interestingly, the PI3K/AKT pathway is known to upregulate HCV RNA translation through the activation of sterol regulatory element-binding proteins, which are downstream effectors of the PI3K/AKT signaling pathway (19). Thus, rigosertib may affect HCV gene expression by inhibition of both PLK1 and PI3K/AKT pathways (Fig. 1). Because of these properties, is rigosertib a novel anti-HCV therapeutic compound? The authors note that the inhibitory efficacy of rigosertib is like that of sofosbuvir in cellular assays. Clinical trials have shown that rigosertib-treated patients with high-risk myelodysplastic syndromes did not show improved survival rates compared to those treated with best supportive care (20), questioning whether the compound is active in patients. However, it is not known whether patients who have been infected with HCV or other viruses respond to rigosertib.

The study by Seo et al. (11) presents compelling evidence that the host protein, ELAVL1/HuR, is essential in HCV gene expression by contributing to the stability of mature miR-122. It is known that the binding of miR-122 to the 5′ end of the viral RNA induces several structural changes in the viral RNA that modulates translation, stability, and replication of both wild-type and resistance-associated variant RNAs (21, 22). Thus, it is possible that this regulatory PLK1-ELAVL1/HuR-miR-122 axis induces structural changes in the viral RNA that could be explored in antiviral therapeutic approaches.