Hepatitis C Virus Regulates its Replication by Maturing miR-122 Through Akt-Dependent Phosphorylation of KSRP

The liver-specific micro-RNA-122 (miR-122) is required for the replication of hepatitis C virus (HCV). The direct interaction between miR-122 and the 5’ untranslated region of the HCV genome promotes viral replication and protects HCV RNA from degradation. Because HCV RNA is its own substrate for replication, infected cells are submitted to the sequestration of increasing levels of miR-122 and to global de-repression of host miR-122 mRNA targets. Whether and how HCV regulates miR-122 maturation to create an environment favorable to its replication remains unexplored. We discovered that Akt-dependent phosphorylation of KSRP host protein at Serine residue 193 is essential for miR-122 maturation in hepatocytes. Moreover, we showed the existence of a reciprocal regulation loop where HCV replication can modulate the proviral effect mediated by KSRP-dependent maturation of miR-122. These data support a mechanism by which HCV regulates the expression of miR-122 by hijacking KSRP, thereby fueling its own replication.


KSRP exerts a proviral effect on HCV replication mediated by miR-122 maturation
Because the ubiquitous cellular protein KSRP is involved in the processing of miRNAs in HeLa cells (Trabucchi et al., 2009), we assessed whether maturation of miR-122, the most abundant liver-specific miRNA and a key actor of the HCV lifecycle, is KSRP-dependent. miR-23b, another less abundant miRNA present in hepatocytes but not involved in the HCV lifecycle, was used as a control. KSRP was silenced by means of siRNA transfection in non-infected Huh7.5 cells and total RNA and proteins were analyzed at different time points. Transient KSRP silencing led to a significant decrease of the intracellular amount of both miR-122 and miR-23b at 48 and 72 hours post-transfection ( Figure 1A). We confirmed by western blot analysis the decrease of KSRP protein quantity as compared to the negative silencing control and the lack of cytotoxicity of KSRP silencing at the selected time points (Supplementary Figure S1).
The role of KSRP on HCV infection was then studied. For this, Huh7.5 cells treated with siKSRP or the negative silencing control were cotransfected with exogenous mimic miR-122, mimic miR-23b (control miRNA) or a negative mimic control. Twenty-four hours post-transfection, the cells were infected with the HCV Jad strain at an MOI of 0.1 and HCV RNA was quantified 24 h post-infection. As shown in Figure 1B, HCV replication was significantly reduced 24 h post-infection in siKSRP-treated Huh7.5 cells as compared to the negative control. The exogenous addition of mimic miR-122 restored HCV replication in KSRP-deficient Huh7.5 cells, whereas mimic miR-23b or the negative mimic control did not ( Figure 1B To confirm these results, we used Hep3B cells, hepatoma-derived cells with undetectable levels of miR-122, in which HCV replication and the production of viral particles are fully dependent on mimic miR-122 supplementation (Thibault, Huys, Dhillon, & Wilson, 2013;Varnholt et al., 2008). Hep3B cells were co-transfected with siKSRP or the negative silencing control on the one hand, mimic miR-122 or the mimic control on the other hand, 24 h before HCV infection with the Jad strain at an MOI of 0.1 (the experimental design is shown in the Supplementary Figure S3A).
The respective amounts of miR-122 and KSRP protein obtained after co-transfection in Hep3B cells are shown in the Supplementary Figure S3B and S3C. As shown in Figure 1C, HCV replication was restored in Hep3B cells by mimic miR-122 transfection and it was not altered when KSRP was silenced in miR-122-cotransfected cells.
Together, these results suggest that KSRP plays a proviral role in HCVinfected hepatocytes by promoting the maturation of miR-122, a key actor of the HCV lifecycle.

The proviral role of KSRP in HCV infection requires its phosphorylation at serine residue 193
Akt-dependent phosphorylation of KSRP at serine residue 193 has been reported to promote miRNA maturation by facilitating pri-miRNA to pre-miRNA processing (Briata et al., 2013;Briata et al., 2012). Therefore, we investigated the effect of non-phosphorylable KSRP-S193A mutagenesis on miR-122 quantities and HCV replication. Huh7.5 cells were transfected with a pool of 3 siRNAs targeting the 3'-UTR region of KSRP mRNA (siKSRP e ) to block the production of endogenous KSRP.
Then, plasmids expressing exogenous HA-tagged wild-type KSRP or KSRP-S193A were transfected for 48 h before HCV infection of Huh7.5 cells at an MOI of 0.1. HCV replication and miR-122 amounts were quantified by means of RT-qPCR at 24 h and 48 h post-infection (see experimental design and the quantities of KSRP and KSRP-HA proteins obtained 48 h post-infection in the Supplementary Figure S4). As shown in Figure 1D, the reductions of miR-122 and HCV RNA quantities induced by KSRP silencing at 24 and 48 h were reversed by transfection of pKSRP-HA but not with non-phosphorylable pKSRP-S193A-HA or the plasmid control. These results suggest that the phosphorylation of KSRP at serine residue 193 is essential for miR-122 maturation and HCV replication.

KSRP phosphorylation at serine residue 193 is increased in the presence of HCV replication
We then aimed to study the Akt-dependent phosphorylation of S193-KSRP in Huh7.5 cells 24 h post-infection by the genotype 2a JFH1-derived HCV strain Jad at an MOI of 0.1. Because no specific antibodies against the phosphorylated forms of KSRP are currently available, we quantified and analyzed the subcellular localization of S193-phosphorylated KSRP by means of in situ proximity ligation assay (PLA) using anti-KSRP mouse and anti-RXXS*/T* primary antibodies. As shown in Figure 2A, in situ PLA showed markedly increased amounts of both cytoplasmic and nuclear phospho-S193-KSRP in Jad-infected Huh7.5 cells as compared to uninfected cells.
Control experiment without primary antibodies is shown in the Supplementary Figure   S5.

Phosphorylation of Akt at serine 473 is induced by HCV infection, is required for the phosphorylation of KSRP at serine 193 and for HCV replication
To assess whether the increased quantity of phospho-S193-KSRP observed in presence of HCV could result from a viral-induced activation of the kinase Akt Then, we tested the effect of Akt inhibition by Akt inhibitor VIII on S193-KSRP phosphorylation by means of in situ PLA experiments using anti-KSRP mouse and anti-RXXS*/T* primary antibodies coupled with an immunofluorescence assay using specific antibodies to detect phospho-S473-Akt, in Huh7.5 cells harboring the HCV genotype 1b subgenomic replicon. As shown in Figures 3C and 3D, the amounts of both phospho-S193-KSRP and phospho-S473-Akt were significantly reduced by Akt inhibitor VIII treatment in HCV replicon-harboring cells, as compared to untreated cells. These results confirm the dependency of S193-KSRP phosphorylation for Akt activity.
We further analyzed the effects of the inhibition of Akt phosphorylation at serine 473 by Akt inhibitor VIII on HCV replication and viral protein expression.
Inhibition of Akt activity was associated with a decrease of the amount of HCV replicon RNA and NS5A protein in replicon-harboring cells, whereas the total amount of KSRP protein remained unchanged ( Figures 3E and 3F).
These findings demonstrate a viral-induced activation of Akt, as measured by the amount of phosphor-S473-Akt and show a correlation between Akt activity and S193-KSRP phosphorylation. Finally, our results emphasize the requirement of Akt activity for HCV replication and protein expression.

Akt-dependent phosphorylation of S193-KSRP induces miR-122 maturation and enhances HCV replication
As shown in Figure 4A, Akt inhibitor VIII treatment of Huh7.5 cells harboring the HCV genotype 1b subgenomic replicon was also associated with significant dose-dependent increase of the amount of pri-miR-122 and decrease of the amount of mature miR-122. This result is in keeping with the requirement of phospho-S193-KSRP for miR-122 maturation. In light of these findings, we hypothesized that the diminution of HCV replication in Akt inhibitor VIII treated-cells ( Figure 3E and 3F) was linked to the reduced phosphorylation of S193-KSRP ( Figure 3C and 3D) and thus, to the decrease of miR-122 maturation ( Figure 4A).
To evaluate this hypothesis, Huh7.5 cells were pre-incubated with DMSO or 10 !M Akt inhibitor VIII for 8 h, transfected with mimic Ctrl or mimic miR-122 and infected 24 h post-treatment with the HCV Jad strain at an MOI of 0.1. Akt inhibitor VIII treatment or DMSO was maintained for the total duration of the experiment, i.e. until 48 h post-infection. As shown in Figure 4B, there was a 2.5-fold decrease of miR-122 and a 8.3-fold decrease of HCV RNA quantities in cells treated with Akt inhibitor VIII in the absence of mimic-miR-122, as compared to the DMSO-treated control. The amount of miR-122 was increased by the administration of exogenous mimic miR-122 in both DMSO-treated and Akt inhibitor VIII-treated cells. An exogenous mimic miR-122 transfection to Akt inhibitor VIII-treated cells restored a level of HCV replication close to that observed in DMSO-and mimic control-treated cells, but slightly lower than that in DMSO-and mimic miR-122-treated cells ( Figure   4B). These results suggest that the effect of Akt inhibition on HCV replication is at least partly related to a reduction of the amount of hepatocyte miR-122 as a result of a defect in its maturation.
We then assessed whether the antiviral effect of Akt inhibitor VIII could be reversed by KSRP transfection. Huh7.5 cells were pre-treated for 8 h with 10 µM of Akt inhibitor VIII or DMSO, transfected with KSRP-HA plasmids for 16 h and infected with the Jad strain of HCV. Again, Akt inhibitor VIII or DMSO treatment was maintained for the total duration of the experiment, i.e. until 48 h post-infection. As shown in Figure 4C, Akt inhibitor VIII reduced the amounts of miR-122 and HCV RNA in control cells and in cells transfected with non-phosphorylable KSRP-S193A-HA. In contrast, transfection of a plasmid expressing phosphomimetic KSRP-S193D-HA rescued both miR-122 quantities and HCV replication in the presence of Akt inhibitor VIII ( Figure 4C).
These results suggest that Akt-dependent phosphorylation of KSRP at serine 193 induces the maturation of miR-122, thereby favoring HCV replication.
These findings suggest that Akt-dependent phosphorylation of KSRP at serine 193 is required for the interaction of pri-miR-122 with DROSHA that leads to pri-miR-122 maturation.  Figure 6A). Akt inhibitor VIII treatment was also associated with an approximately 2-fold decrease of HCV RNA and mature miR-122 amounts and no change in KSRP mRNA expression ( Figure 6B). KSRP silencing, performed 24 h before JFH1 infection of PHHs at a MOI of 2, was associated with a marked reduction of both miR-122 expression and HCV replication as compared to nonsilenced PHHs ( Figures 6C and 6D).

Akt activation and KSRP expression are required to maturate miR-122 and promote HCV replication in primary human hepatocytes
These results confirm the essential role of Akt activity and KSRP in miR-122 maturation and HCV replication in HCV-infected PHHs.

Discussion
miR-122 is the most abundant tissue-specific miRNA in hepatocytes and is required to achieve a full HCV lifecycle. This requirement partly explains the hepatic tropism of HCV. Indeed, the ectopic expression of miR-122 and apolipoprotein E (ApoE) has been shown to be sufficient to render a mouse liver-derived cell line permissive for HCV replication and for the release of infectious particles (Frentzen et al., 2014). Additionally, the expression of exogenous miR-122 in non-hepatic or hepatic cell lines facilitates efficient HCV replication. The susceptibility of Huh7.5 cells to the propagation of infectious HCV strains in cell culture is attributable to their high level of miR-122 expression (Fukuhara et al., 2012).
Binding of miR-122 at two tandem sites, S1 and S2, located within the 5' untranslated region of the viral genome enhances RNA stability and promotes viral replication (Jopling, Schutz, & Sarnow, 2008;Li, Masaki, & Lemon, 2013). Until now, whether and how HCV can regulate the maturation of miR-122 to sustain its own lifecycle remained unexplored. Our results show that the host protein KSRP is essential for miR-122 maturation in hepatocytes ( Figure 1A) and therefore is a proviral host factor for HCV ( Figures 1B). We also demonstrate that Akt phosphorylation of S193-KSRP is pivotal for this function (Figures 1D Previous studies indicated that miR-122 levels are decreased in cells and in sera from patients chronically infected with HCV (Bala, Marcos, & Szabo, 2009;Marquez et al., 2010). These observations could be linked to the decreased stability of miR-122 resulting from the interaction between HCV core protein and GLD-2, leading to the inhibition of this non-canonical cytoplasmic poly(A) polymerase (Kim et al., 2016). In addition, miR-122 "sponging" by HCV RNA has been reported during HCV infection. Indeed, HCV RNA acts as a competitive inhibitor of miR-122 activity by establishing the stable interactions required for viral genome stability and replication. Because HCV RNA is a template for its replication, a positive feedback loop sequesters additional miR-122. This "sponging" phenomenon has the potential to result in a global de-repression of cellular miR-122 target genes (Luna et al., 2015).
Therefore, compensatory mechanisms are required to maintain the cellular levels of miR-122 while sustaining both HCV RNA replication and post-translational regulation of miR-122 cellular target genes. Thus, the regulation of miR-122 maturation by KSRP described here could be of importance for acute HCV infection and the longterm homeostasis of infected hepatocytes in the context of chronic HCV infection.
In summary, we describe here a so far unknown mechanism by which HCV "hijacks" KSRP, a host cell protein implicated in RNA metabolism, to regulate the hepatocyte expression of miR-122 and therefore fuel its own replication. Whether this "hijacking" is specific for HCV infection or could represent a general mechanism used by different viruses to establish and sustain their chronic infection remains to be investigated. miR-122 has been shown to be an attractive target for HCV inhibition. Indeed, a single injection of an antagomir drastically reduced viral replication for several weeks, leading to a cure of infection in some cases (van der Ree et al., 2017). These results validated the concept that antagonists or agonists of miRNAs with proviral or antiviral effects, respectively, could be used in the treatment of acute or chronic viral infections. Our results suggest that KSRP may constitute a valuable tool to identify new targets of antiviral intervention.

RNA extraction and quantification
Total RNA was prepared using the RNeasy mini kit (Qiagen

Plasmid transfections and co-transfections with siRNAs
A human KSRP cDNA sequence (isoform 1), cloned into a pCMV expression plasmid and fused with an HA-tag (pKSRP-HA) was used to obtain KSRP mutants.
The Quick Change XL Site-directed mutagenesis kit (Stratagene, San Diego, California, USA) was used to generate the phosphomimetic KSRP-S193D-HA and non-phosphorylable KSRP-S193A-HA proteins. For transfection, cells were seeded in 12-well plates at 180,000 cells/well and maintained in culture medium for 24 h.

In situ proximity ligation assay
In situ proximity ligation assay (PLA) was used to visualize protein-protein interactions and post-translational protein modifications in fixed cultured cells using secondary antibodies with attached oligonucleotides. KSRP post-translational modifications were monitored in Huh7.5-SGR and Huh7.5 cells grown on Nunc Lab-Tek Chamber Slide™ system 8 wells (Permanox, Sigma-Aldrich), 10,000 cells/well.
After incubation and infection with Jad, Huh7.5 cells were fixed with 2% paraformaldehyde in PBS for 10 min at room temperature and permeabilized with cold methanol for 10 min at -20°C. Slides were blocked for 1 h with 1% BSA (fraction MRc 3 CCD sensor and 40x or 100x objectives with filters for DAPI, FITC and Cy3.

Primary human hepatocytes
Primary human hepatocytes (PHH, Biopredic, Rennes, France) were isolated from normal-appearing liver tissues obtained from adult patients undergoing partial hepatectomy for the treatment of metastases who were seronegative for HCV, hepatitis B virus and human immunodeficiency virus. Freshly isolated PHHs were seeded at high density on collagen-coated plates and maintained in primary culture as described previously (Podevin et al., 2010 HCV replication and miR-122 maturation were measured by means of RT-qPCR as previously described (Carriere et al., 2007). For KSRP silencing, siRNAs (ON-TARGETplus smartpool, Dharmacon, GE Healthcare) were diluted into PBS to a final concentration of 2 nM and mixed with 2 !l of RiboCellin siRNA Delivery Reagent (BioCellChallenge), as recommended by the manufacturer. Data are presented as means of 3 replicates in cell culture.

Statistical analyses
The results are expressed as mean±SEM of three replicate experiments.
Comparisons were made by means of the Student's t test wherever appropriate. A p value <0.05 was considered significant. * differs from no treatment or zero time, p<0.05, ** differs from no treatment or zero time, p<0.01, *** differs from no treatment or zero time, p=<0.001.