Research ArticleANTIVIRAL IMMUNITY

The host microRNA miR-301a blocks the IRF1-mediated neuronal innate immune response to Japanese encephalitis virus infection

+ See all authors and affiliations

Sci. Signal.  14 Feb 2017:
Vol. 10, Issue 466, pp. eaaf5185
DOI: 10.1126/scisignal.aaf5185

Fighting flavivirus infection

Japanese encephalitis virus (JEV), which is related to the Zika, West Nile, and dengue viruses, targets the central nervous system, causing viral encephalitis. Because of the high mortality rate in JEV-infected individuals and the neurological damage that afflicts survivors, effective antiviral therapies are urgently needed. Hazra et al. found that JEV infection of mouse and human neuronal cells increased the abundance of the microRNA miR-301a, which inhibited the cells from producing type I interferons (IFNs), cytokines that are critical to the antiviral immune response. One of the targets of miR-301a is the transcription factor IRF1, which is required for the expression of genes encoding type I IFNs. Treating JEV-infected mice with an inhibitor of miR-301a restored IRF1 abundance and function and led to increased type I IFN production, decreased viral replication, and increased survival. Together, these data suggest that targeting miR-301a may be an effective therapy against JEV infection.

Abstract

Effective recognition of viral components and the subsequent stimulation of the production of type I interferons (IFNs) is crucial for the induction of host antiviral immunity. The failure of the host to efficiently produce type I IFNs in response to infection by the Japanese encephalitis virus (JEV) is linked with an increased probability for the disease to become lethal. JEV is a neurotropic virus of the Flaviviridae family that causes encephalitis in humans. JEV infection is regulated by several host factors, including microRNAs, which are conserved noncoding RNAs that participate in various physiological and pathological processes. We showed that the JEV-induced expression of miR-301a led to inhibition of the production of type I IFN by reducing the abundances of the transcription factor IFN regulatory factor 1 (IRF1) and the signaling protein suppressor of cytokine signaling 5 (SOCS5). Mechanistically, induction of miR-301a expression during JEV infection required the transcription factor nuclear factor κB. In mouse neurons, neutralization of miR-301a restored the host innate immune response by enabling IFN-β production, thereby restricting viral propagation. Inhibition of miR-301a in mouse brain rescued the production of IRF1 and SOCS5, increased the generation of IFN-β, and reduced the extent of JEV replication, thus improving mouse survival. Thus, our study suggests that the JEV-induced expression of miR-301a assists viral pathogenesis by suppressing IFN production, which might be targeted by antiviral therapies.

INTRODUCTION

Japanese encephalitis (JE) is a disease of public health importance because of its epidemic potential and high fatality rate. JE virus (JEV) is the leading cause of viral encephalitis in the Asia-Pacific region since its initial outbreak in 1871 in Japan (1). JEV is a single-stranded, positive-sense, mosquito-borne flavivirus that belongs to the same genus as the dengue virus (DENV), yellow fever virus, Zika virus, and West Nile virus (WNV). After entering the body, JEV invades the central nervous system (CNS) and clinically manifests as fever, headache, and vomiting. About one-third of patients die, and half of the survivors suffer permanent neuropsychiatric sequelae (2). According to the World Health Organization, most of the countries in Southeast Asia and the Western Pacific region have endemic JEV transmission, exposing more than 3 billion people to the risks of infection (3). With rapid globalization and climate shift, JEV has started to emerge in areas where the threat was previously unknown and has become the cause of worldwide pandemics (4). Although multiple vaccines with varying degrees of effectiveness exist to control JE, there is no cure for the disease (5), and treatment is mainly palliative.

Type I interferons (IFNs) are best known for their ability to induce an antiviral state by stimulating the transcription of IFN-stimulated genes (ISGs), the products of which affect multiple stages of the viral replication cycle (6). In addition, type I IFNs activate key components of the innate and adaptive immune systems, including the maturation of antigen-presenting cells and the production of cytokines involved in the activation of T cells, B cells, and natural killer cells (7). Pattern recognition receptors (PRRs) have become universally important in inducing the production of type I IFNs during viral infections by recognizing various viral components (8). In the case of JEV, we previously showed that the recognition of viral products by PRRs, including Toll-like receptor 7 (TLR7) and retinoic acid–inducible gene I, serves to elicit immunological responses involving type I IFNs in neurons (911). Despite this, different strategies are adopted by JEV to overpower the antiviral activities of type I IFNs (12, 13). Failure of the host to produce IFNs competently against the virus is linked with an increased probability of the disease to become lethal (14). However, evasion of the immune system by JEV and its consequences might be mitigated if the type I IFN response could be augmented during viral infection, a proposal that is supported by studies reporting inhibition of the replication of different flaviviruses by IFN-based therapy (1517).

MicroRNAs (miRNAs) have emerged as key posttranscriptional regulators of gene expression, and they affect nearly every cellular process (18). Because individual miRNAs can regulate the expression of multiple mRNAs, it is difficult to conclude whether an increase in the abundance of a specific miRNA represents a part of the host innate immune response to restrict viral infection or whether it was specifically induced by the virus to promote a cellular environment more favorable to viral replication (19). Increasing evidence indicates that cellular miRNAs can exert positive or negative influences on flaviviral infection. WNV infection induces the production of miR-154, which expedites virus-mediated apoptosis by targeting the mRNAs corresponding to cellular antiapoptotic factors (20). As a part of the host immune response, miR-532-5p exhibits antiviral activity against WNV through its suppression of the expression of the host genes SEC14 and spectrin domain containing 1 (SESTD1) and transforming growth factor–β–activated kinase 1/MAP3K7 binding protein 3 (TAB3) (21). The miRNA let-7c inhibits DENV infection by targeting the DENV genome (22). Another host miRNA, miR-30e*, substantially suppresses DENV replication by promoting IFN-β production, whereas miR-223 inhibits DENV by negatively regulating the microtubule-destabilizing protein stathmin 1 (STMN1) (23, 24). Another flavivirus, JEV, reduces the abundance of miR-33a-5p in the host to stabilize the components of the JEV replicase complex and thus enhance JEV replication (25). Two other miRNAs, miR-155 and miR-15b, are induced during JEV infection and modulate JEV-induced innate immunity and inflammatory responses, respectively (26, 27). In our previous studies, we found that the expression of many miRNAs was induced abruptly during JEV infection (28); among these, the regulation of two host miRNAs, miR-29b and miR-155, in JEV-induced microglial inflammation has already been documented (28, 29).

Here, we investigated the role of miR-301 family (miR-301a and miR-301b) and initially investigated whether the increase in miR-301 abundance during JEV infection exclusively occurred in microglia. We found that the abundances of both these miRNAs were increased in microglia together with that of miR-301a in neurons early during JEV infection. The altered amounts of miRNAs early during infection were generally coupled with pathogen detection and the host immune response. Because a proinflammatory function of miR-301a is well established in different models (3033) but its role in the immune response is not well characterized (34), its prompt induction in neurons motivated us to investigate host immune regulation of miR-301a during JEV infection. Here, we postulate that the JEV-induced increase in host miR-301a abundance may inhibit type I IFN signaling and enable JEV to take advantage of a weakened immune system.

RESULTS

miR-301a expression is enhanced early during JEV infection of neurons

We previously used miRNA polymerase chain reaction (PCR)–based arrays to analyze the miRNA expression profile in JEV-infected microglia (BV2 cells) and found that the abundance of members of the miR-301 family (miR-301a and miR-301b) was markedly increased compared to that in uninfected cells (28). Here, we performed quantitative reverse transcription PCR (qRT-PCR) analysis of the abundance of miR-301 family members in BV2 cells and in HT22 cells, a mouse hippocampal neuronal cell line. As shown previously, BV2 cells exhibited increased abundance of miR-301a and miR-301b 12 hours after infection with JEV, whereas a substantial early increase in miR-301a abundance (up to 6 hours) was also observed in JEV-infected HT22 cells (fig. S1, A and B). Furthermore, HT22 cells infected with varying concentrations of JEV for 6 hours had a higher abundance of miR-301a than mock-infected (MI) cells, suggesting that infection with JEV at early times induces the expression of miR-301a in HT22 cells (Fig. 1A). Similar results were obtained from the analysis of the time and dose dependency of miR-301a expression in JEV-infected primary cortical neurons (Fig. 1, B and C). To detect the early induction of miR-301a expression in mouse brain neurons, we performed both in situ hybridization (ISH) analysis for miR-301a or U6 small nuclear RNA (snRNA) (as a positive control) and immunohistochemical (IHC) analysis of neurons in uninfected and JEV-infected (2 days after infection) mouse brain sections. Although both sections showed similar amounts of U6 snRNA, the increased abundance of miR-301a was observed only in the neurons of JEV-infected mice (Fig. 1D). Next, we validated the expression of miR-301a in JEV-infected human neuroblastoma (SH-SY5Y) cells. Our qRT-PCR analysis showed both time- and dose-dependent increase in the abundance of miR-301a early during JEV infection (Fig. 1, E and F). Together, these results suggest that miR-301a expression in neurons is enhanced early during JEV infection.

Fig. 1 miR-301a expression is induced early in JEV-infected neuronal cells.

(A) HT22 cells were left uninfected (MI) or were infected with JEV at the indicated multiplicity of infections (MOIs) for 6 hours before the abundance of miR-301a was measured by qRT-PCR analysis and normalized to that of SNORD68 snRNA. *P < 0.05, **P < 0.01 compared to uninfected cells. (B and C) Primary neuronal cells were isolated from postnatal day 0 (P0) to P2 BALB/c mouse pups, cultured for 7 days, and exposed to JEV for the indicated times (B) or were infected with JEV at the indicated MOIs for 6 hours (C). In both cases, miR-301a abundance was quantified by qRT-PCR analysis, and the results are expressed as the fold change compared to that in uninfected cells. *P < 0.05, **P < 0.01. (D) ISH of miR-301a (purple chromogen) in neuronal cells (brown chromogen) from mouse brain. Brain samples from P10 mice were collected from MI mice or from mice 2 days postinfection (d.p.i.) with JEV, and sections were hybridized with the miRCURY LNA miR-301a probe or the LNA U6 snRNA probe, which was followed by IHC analysis of neurons with DAB (3,3’-diaminobenzidine). Scale bars, 20 μm; magnification, ×40. The ubiquitously expressed U6 snRNA was used as a positive control for miRNA ISH. Data are representative of four mice per group. (E) SH-SY5Y cells were left uninfected (MI) or were exposed to JEV for the indicated times, and the abundance of miR-301a was evaluated by qRT-PCR analysis. *P < 0.05, **P < 0.01 compared to uninfected cells. (F) SH-SY5Y cells were left uninfected (MI) or were infected with JEV at the indicated MOIs for 6 hours. Relative miR-301a abundance was then determined by qRT-PCR analysis. *P < 0.05, **P < 0.01 compared to uninfected cells. h.p.i., hours postinfection. All data in bar graphs are means ± SD of three biological replicates. P values are calculated by analysis of variance (ANOVA) followed by Bonferroni’s post hoc test.

miR-301a inhibits the antiviral IFN response and promotes JEV replication

Induction of miRNAs early during infection usually regulates the antiviral IFN response and viral replication. To explore whether the silencing of miR-301a could augment the production of type I IFNs, we transfected HT22 and SH-SY6Y cells with either an miR-301a inhibitor (anti–miR-301a) or a negative control (anti–miR-Con) 24 hours before the cells were subjected to JEV infection. We observed a substantial decrease in miR-301a abundance in the cells transfected with anti–miR-301a compared to that in cells transfected with anti–miR-Con (Fig. 2A and fig. S2A). Inhibition of miR-301a led to a substantial increase in IFN-β mRNA abundance, whereas IFN-α mRNA abundance remained unchanged up to 6 hours after infection of HT22 cells (Fig. 2, B and C). We observed a similar trend in IFN-α (fig. S2B) and IFN-β (fig. S2C) mRNA abundances in JEV-infected SH-SY5Y cells. The amount of IFN-β secreted by infected HT22 cells transfected with anti–miR-301a was also increased compared to that secreted by control cells as determined by enzyme-linked immunosorbent assay (ELISA) (Fig. 2D). Inhibition of miR-301a led to the increased expression of ISGs in both HT22 (Fig. 2E) and SH-SY5Y cells (fig. S2D) up to 6 hours after JEV infection. Because IFNs play a pivotal role in the host antiviral defense, we further assessed whether anti–miR-301a affected viral replication, and we found that there was a substantial reduction in viral RNA abundances in both HT22 (Fig. 2F) and SH-SY5Y cells (fig. S2E) that were transfected with anti–miR-301a, as analyzed by qRT-PCR. We also checked the long-term effect of the miRNA inhibitor on JEV infection and detected a marked decrease in the viral titer (number of viral particles) in the culture medium of HT22 cells 48 hours after infection (Fig. 2G).

Fig. 2 miR-301a inhibits the antiviral IFN response and enhances viral replication.

(A) HT22 cells transfected with an miR-301a inhibitor (anti–miR-301a) or a negative control (anti–miR-Con) were left uninfected (MI) or were infected with JEV for the indicated times. Relative miR-301a abundance was then determined by qRT-PCR analysis. *P < 0.05, **P < 0.01, ***P < 0.001 compared to the inhibitor control. (B and C) HT22 cells were transfected with the miR-301a inhibitor or the nonspecific inhibitor control. Twenty-four hours later, the cells were infected with JEV for the indicated times before the cells were subjected to qRT-PCR analysis of the relative abundances of IFN-α (B) and IFN-β (C) mRNAs. *P < 0.05, **P < 0.01, ***P < 0.001 compared to the inhibitor control. (D) Cell culture medium from the cells shown in (B) and (C) was analyzed by ELISA to determine the amount of secreted IFN-β protein. **P < 0.01, ***P < 0.001 compared to the inhibitor control. (E) HT22 cells were transfected with the miR-301a inhibitor or the nonspecific inhibitor control. Twenty-four hours later, the cells were left uninfected (MI) or were infected with JEV for the indicated times before the cells were subjected to qRT-PCR analysis of the relative abundances of IFIT1, ISG15, and OAS1 mRNAs. *P < 0.05, **P < 0.01, ***P < 0.001 compared to the inhibitor control. (F) HT22 cells were transfected with the miR-301a inhibitor or the nonspecific inhibitor control. Twenty-four hours later, the cells were infected with JEV for the indicated times before the cells were subjected to qRT-PCR analysis of the relative abundances of viral RNA. *P < 0.05, ***P < 0.001 compared to the inhibitor control. (G) HT22 cells were transfected with the miR-301a inhibitor or the nonspecific inhibitor control. Twenty-four hours later, the cells were infected with JEV for the indicated times before the cell culture medium was collected for plaque assays, which were performed as described in Materials and Methods. ***P < 0.001 compared to the inhibitor control. All data are means ± SD of three biological replicates. P values were calculated by two-way ANOVA followed by the Holm-Sidak method.

Virus-induced miR-301a targets a number of host genes

To gain insight into the mechanism underlying miR-301a function, we analyzed miR-301a target genes that might influence the IFN response in our model of JE. We found more than 500 common target genes of mouse miR-301a by analyzing four widely used miRNA target prediction databases: Miranda (35), TargetScan (36), PicTar (37), and miRDB (38). Among these targets, 100 common targets that had high binding scores were identified (table S1). To identify prospective candidate target genes whose products might control JEV replication, we chose to investigate suppressor of cytokine signaling 5 (SOCS5) and IFN regulatory factor 1 (IRF1) because they are potential regulators of innate immunity (39, 40). The targeting of the genes encoding SOCS5 and IRF1 by miR-301a could have a broad effect on the innate response against JEV. All computational analyses indicated that there was a conserved miR-301a binding site within the 3′ untranslated regions (3′UTRs) of SOCS5 (Fig. 3A) and IRF1 (Fig. 3B).

Fig. 3 SOCS5 and IRF1 are targets of miR-301a.

(A and B) Predicted miR-301a binding sites in the 3′UTRs of SOCS5 mRNA (A) and IRF1 mRNA (B). Perfect matches in the seed regions are indicated in orange. (C and D) Top: Diagrams of constructs containing the 3′UTR of SOCS5 (C) or IRF1 (D) downstream of a luciferase reporter. The WT 3′UTR (WT UTR) contains an intrinsic miR-301a binding site, whereas the mutant 3′UTR (Mut UTR) contains mutations that eliminated the seed match with miR-301a. Mutations (magenta) in the 3′UTRs of both SOCS5 and IRF1 were generated for reporter gene assays. Bottom: Dual luciferase assays of HT22 cells cotransfected with firefly luciferase constructs containing the WT or mutant 3′UTRs of SOCS5 (C) or IRF1 (D) and either the miR-301a mimic (Mimic) or the control mimic (Mimic Con) were performed. Firefly luciferase activity was normalized to Renilla luciferase activity. Data are shown as the relative luciferase activity of cells transfected with the miR-301a mimic compared to that of cells transfected with the control mimic. Data are means ± SD of nine experiments from three independent transfections. **P < 0.01, ***P < 0.001, by Student’s t test. ns, not significant. (E) HT22 cells were subjected to mock transfection [mock-transfected (MT)] or were transfected with either the miR-301a miRNA mimic or the negative control mimic. Top: Twenty-four hours later, the cells were subjected to Western blotting analysis of the abundances of SOCS5 and IRF1 proteins. Bottom: Densitometric analysis of the Western blots from three experiments was performed to determine the fold changes in SOCS5 and IRF1 protein abundance (normalized to that of β-actin) relative to the MT cells. The relative abundance of miR-301a as determined by qRT-PCR analysis of each set of cells is shown above the blots to confirm effective transfection. Data are means ± SD of three independent experiments. ***P < 0.001, by one-way ANOVA followed by Bonferroni’s post hoc test. (F) HT22 cells were left untransfected (MI) or were transfected with either the miR-301a mimic or the negative control mimic. Twenty-four hours later, the cells were subjected to qRT-PCR analysis of the relative abundances of SOCS5 and IRF1 mRNAs. Data are means ± SD of three independent experiments. Statistical analysis of the data by one-way ANOVA with Bonferroni’s multiple comparisons showed that there were no statistically significant differences.

To verify these database predictions, we cloned the 3′UTRs of mouse SOCS5 and IRF1 into a firefly luciferase reporter vector. We then generated six- and seven-base mutations in the predicted seed matching site in the 3′UTRs of SOCS5 (Fig. 3C) and IRF1 (Fig. 3D) to test the miRNA-target interactions. HT22 cells were then transfected with individual reporters containing wild-type (WT) or mutant (Mut) UTRs together with an miR-301a mimic or a mimic control. The miR-301a mimic effectively reduced the luciferase activity of the WT UTR reporter compared to that in cells transfected with the mimic control. In contrast, the miR-301a–dependent reduction in luciferase activity was disrupted by mutating the 3′UTR binding sites in SOCS5 (Fig. 3C) or IRF1 (Fig. 3D). We further investigated whether the amounts of SOCS5 and IRF1 mRNAs and proteins were affected by the miR-301a mimic. Transfection of HT22 or SH-SY5Y cells with the miR-301a mimic substantially increased the abundance of miR-301a and subsequently attenuated the production of both SOCS5 and IRF1 proteins (Fig. 3E and fig. S3, A and B) but did not substantially affect their mRNA abundances (Fig. 3F). These results suggest that miR-301a suppressed the production of SOCS5 and IRF1 proteins by inhibiting translation rather than inducing mRNA degradation.

miR-301a inhibits the production of SOCS5 and IRF1 proteins during JEV infection

We next examined whether miR-301a targeted SOCS5 and IRF1 during JEV infection. We cotransfected HT22 cells with either the miR-301a inhibitor or the inhibitor control together with either WT or mutant SOCS5 or IRF1 3′UTR reporter constructs, which was followed by infection of the cells with JEV for 6 hours. We detected a substantial increase in luciferase activity in the cells transfected with the WT UTR construct and the miR-301a inhibitor (Fig. 4, A and B). In contrast, mutating the SOCS5 and IRF1 3′UTRs blocked the anti–miR-301a–mediated increase in luciferase activity in these cells (Fig. 4, A and B). To present direct evidence that the miR-301a induced by JEV infection suppressed the production of SOCS5 and IRF1 proteins, we examined the abundances of SOCS5 and IRF1 in HT22 cells infected with JEV for different times as well as in HT22 cells infected for a fixed time with different viral concentrations, in which miR-301a was inhibited. We found that inhibition of miR-301a reconstituted SOCS5 and IRF1 protein production in JEV-infected cells compared to that in control cells (Fig. 4, C and D). Similarly, silencing of miR-301a in SH-SY5Y cells led to the restoration of SOCS5 and IRF1 protein production compared to that in control cells infected with JEV (fig. S4, A and B).

Fig. 4 JEV-induced miR-301a suppresses SOCS5 and IRF1 protein production.

(A and B) HT22 cells were cotransfected with either the miR-301a inhibitor (anti–miR-301a) or the inhibitor control (anti–miR-Con) together with a firefly luciferase reporter plasmid encoding the WT or mutant 3′UTRs of SOCS5 (A) or IRF1 (B). Twenty-four hours later, the cells were infected with JEV for 6 hours before luciferase activities were measured with a dual luciferase assay kit and normalized to that of Renilla luciferase. Data are expressed as the relative luciferase activity of the anti–miR-301a–transfected cells compared to that of the anti–miR-Con–transfected cells. Data are means ± SD of nine experiments from three independent transfections. **P < 0.01, by Student’s t test. (C and D) HT22 cells were transfected with either the miR-301a inhibitor or the inhibitor control and then were either infected with JEV at an MOI of 5 for the indicated times (C) or infected for 6 hours with JEV at the indicated MOIs (D). Top: The cells were analyzed by Western blotting to determine the relative abundances of SOCS5 and IRF1 proteins. β-Actin was used as a loading control. Western blots are representative of three independent experiments. The relative abundance of miR-301a in each set of cells was determined by qRT-PCR analysis and is shown below the blots to confirm effective transfection. Bottom: Densitometric analysis of the Western blots was performed to determine the fold changes in the indicated protein abundances (normalized to that of β-actin) in infected cells compared to those in uninfected (MI) cells. Data are means ± SD of three individual experiments. *P < 0.05, **P < 0.01, ***P < 0.001, by one-way ANOVA followed by Bonferroni’s post hoc test.

miR-301a hampers the JEV-induced production of type I IFN by suppressing SOCS5 and IRF1 mRNA translation

To validate the roles of SOCS5 and IRF1 in the anti-IFN action of miR-301a, we performed experiments in which SOCS5 or IRF1 were knocked down or overexpressed. HT22 cells were cotransfected with the miR-301a inhibitor and with Con-esiRNA [control (Con) endoribonuclease-prepared small interfering RNA (siRNA) (esiRNA)] or esiRNAs specific for SOCS5 and IRF1 alone or in combination before being infected 24 hours later with JEV for 6 hours. The induction in expression of IFN-β and the ISGs IFIT1, ISG15, and OAS1 by the miR-301a inhibitor observed in Con-esiRNA–transfected cells was disrupted by the knockdown of SOCS5, IRF1, or both (Fig. 5, A to D). Consequently, the decrease in viral RNA caused by miR-301a inhibition was restored by knockdown of both SOCS5 and IRF1 (Fig. 5E). The viral titer in the cell culture medium of HT22 cells transfected with anti–miR-301a was increased by the concomitant knockdown of either SOCS5 or IRF1 (Fig. 5F). In contrast, transfection of miR-301a inhibitor–treated HT22 cells with plasmids encoding SOCS5 or IRF1 alone or in combination followed by infection with JEV enhanced the expression of IFN-β and the ISGs compared to cells cotransfected with the control vector (Con-vector) (Fig. 5, G to J). Accordingly, the impairment of viral replication caused by miR-301a inhibition was augmented by overexpression of SOCS5 or IRF1 (Fig. 5K). The viral titer in the culture medium of cells overexpressing SOCS5, IRF1, or both was also reduced compared to that for cells transfected with Con-vector (Fig. 5L). Thus, the miR-301a induced upon JEV infection exhibited its viral-promoting function mainly through the suppression of SOCS5 and IRF1 expression and the subsequent reduced IFN response.

Fig. 5 miR-301a inhibits the IFN response by repressing SOCS5 and IRF1 protein production.

(A to E) HT22 cells were left untransfected (MT) or were transfected with anti–miR-301a together with Con-esiRNA or esiRNAs specific for SOCS5 or IRF1 alone or in combination before being infected with JEV at an MOI of 5. Six hours after infection, the relative abundances of IFN-β (A), IFIT1 (B), ISG15 (C), OAS1 (D) mRNAs, and viral RNA (E) were quantified by qRT-PCR analysis. (A) Cells were also subjected to Western blotting (WB) analysis of SOCS5 and IRF1 abundances to confirm effective transfection of the cells. Bar graphs show densitometric analysis of the fold changes in the abundance of the indicated proteins (normalized to that of β-actin) compared to that in the MT, JEV-infected cells. ##P < 0.01, ###P < 0.001 compared to MT cells; *P < 0.05, **P < 0.01, ***P < 0.001 compared to cells transfected with anti–miR-301a and Con-esiRNA. (F) HT22 cells transfected as described in (A) to (E) were infected with JEV for 48 hours before the culture medium was collected for plaque assays. Data are means ± SD of three independent experiments. #P < 0.05 compared to MT cells; *P < 0.05 compared to cells transfected with anti–miR-301a and Con-esiRNA. (G to K) HT22 cells were left untransfected (MT) or were transfected with the indicated combinations of plasmids and anti–miR-301a. Twenty-four hours later, the cells were infected with JEV at an MOI of 5. At 6 hours after infection, the relative abundances of IFN-β (G), IFIT1 (H), ISG15 (I), and OAS1 (J) mRNAs, and of viral RNA (K) were measured by qRT-PCR analysis. (G) Cells were also subjected to Western blotting analysis of SOCS5 and IRF1 abundances to confirm effective transfection of the cells. Bar graphs show densitometric analysis of the fold changes in the abundance of the indicated proteins (normalized to that of β-actin) compared to that in the MT, JEV-infected cells. #P < 0.05, ##P < 0.01, ###P < 0.001 compared to untransfected cells (MT); *P < 0.05, **P < 0.01, ***P < 0.001 compared to cells transfected with anti–miR-301a and Con-vector. (L) HT22 cells that were transfected as described in (G) to (K) were infected with JEV for 48 hours before the culture medium was collected for plaque assays. Data are means ± SD of three independent experiments. ##P < 0.01 compared to untransfected cells (MT); *P < 0.05, **P < 0.01, ***P < 0.001 compared to cells transfected with anti–miR-301a and Con-vector. P values were calculated by one-way ANOVA followed by Bonferroni’s post hoc test.

Reduction in SOCS5 leads to epidermal growth factor receptor activation and inhibits IRF1

IRF1 is a potential regulator of the antiviral IFN response (41), but little is known about the role of SOCS5 in antiviral innate immunity. Therefore, the function of SOCS5 in the miR-301a–mediated suppression of IFN-β production in JEV-infected HT22 cells was investigated. First, we found that SOCS5 protein abundance in JEV-infected cells was substantially reduced up to 24 hours after infection (Fig. 6A). We also confirmed a decrease in SOCS5 protein abundance in JEV-infected mouse brain and primary neurons by immunofluorescence analysis (fig. S5, A and B). Because SOCS proteins are potential regulators of IRF signaling (42), we evaluated whether suppression of SOCS5 during JEV infection might have any effect on IRF1 expression. We found that transfection of cells with SOCS5-specific esiRNA reduced IRF1 expression compared to JEV-infected HT22 cells transfected with control esiRNA (Fig. 6B). Conversely, overexpression of exogenous SOCS5 substantially rescued IRF1 expression in cells after infection with JEV (Fig. 6C).

Fig. 6 Loss of SOCS5 facilitates the suppression of IRF1 production through EGFR activation.

(A) HT22 cells were left uninfected (MI) or were infected with JEV at an MOI of 5 for 24 hours. Top: Cells were analyzed by Western blotting with antibodies against the indicated proteins. Bottom: Densitometric analysis was performed to show the fold change in SOCS5 protein abundance (normalized to that of β-actin) in infected cells compared to that in uninfected cells (MI). *P < 0.05, **P < 0.01, ***P < 0.001 compared to uninfected cells. (B and C) HT22 cells were left untransfected or were transfected with either SOCS5-specific or Con-esiRNA (B) or with either the SOCS5 expression plasmid or the control plasmid (C). Top: Twenty-four hours later, the cells were infected with JEV for 6 hours before the relative abundance of IRF1 mRNA was quantified by qRT-PCR analysis. ***P < 0.001 compared to untransfected cells. Bottom: Transfection efficiency was determined by Western blotting analysis of SOCS5 abundance. Histograms show the densitometric analysis of the fold change in SOCS5 protein abundance (normalized to that of β-actin) in transfected cells compared to that in untransfected cells. **P < 0.01 compared to untransfected cells. (D) HT22 cells were left untransfected or were transfected with the indicated esiRNAs or expression plasmids before being infected with JEV for 6 hours. Top: Samples were then analyzed by Western blotting with antibodies against the indicated proteins. Equivalent gel loading was confirmed by analysis with an antibody against β-actin. Bottom: Densitometric quantification of the abundance of p-EGFR normalized to that of β-actin. **P < 0.01, ***P < 0.001. (E and F) HT22 cells were treated with 10 μM AG 1478 or vehicle before being left uninfected of infected with JEV for 6 hours. The relative abundance of IRF1 mRNA and viral RNA were analyzed by qRT-PCR (E). **P < 0.01, ***P < 0.001 compared to untreated control cells; ##P < 0.01 compared to JEV-infected cells. In addition, the amount of IFN-β secreted by the indicated cells was measured by ELISA (F). **P < 0.01 comparison to untreated control cells; ##P < 0.01 compared to JEV-infected cells. (G) HT22 cells treated with or without 10 μM AG 1478 were infected with JEV for the indicated times before the culture medium was collected for plaque assays. All data are means ± SD of three biological replicates. P values were obtained by one-way ANOVA followed by Bonferroni’s multiple comparisons.

SOCS5 inhibits epidermal growth factor receptor (EGFR) signaling (43), and active, phosphorylated EGFR (p-EGFR) inhibits the IRF1-dependent antiviral IFN response (44). Therefore, to investigate the mechanism underlying the SOCS5-mediated regulation of IRF1, we investigated the effect of SOCS5 on p-EGFR abundance in JEV-infected HT22 cells. EGFR activation in JEV-infected cells was further enhanced when SOCS5 was knocked down, whereas SOCS5 overexpression markedly reduced the abundance of p-EGFR (Fig. 6D), which suggests that JEV activates EGFR by suppressing SOCS5 in neuronal cells. Furthermore, to assess whether EGFR activation suppressed IRF1 expression, we treated JEV-infected HT22 cells with a selective EGFR tyrosine kinase inhibitor, AG 1478. We found that IRF1 mRNA abundance in JEV-infected cells was increased to a greater extent in the presence of AG 1478 than in cells infected with JEV alone; however, AG 1478 substantially decreased the abundance of viral RNA (Fig. 6E). In addition, AG 1478 increased the amount of IFN-β secreted by JEV-infected cells (Fig. 6F). The viral titer in the culture medium of AG 1478–treated, JEV-infected cells was similar to that in the culture medium of cells infected with JEV alone (Fig. 6G). Together, these data suggest that the miR-301a–mediated reduction in SOCS5 abundance during JEV infection leads to EGFR activation, which then represses the IRF1-mediated production of IFN-β.

miR-301a expression is induced through an nuclear factor κB–dependent mechanism

We next investigated the underlying mechanism by which miR-301a expression is enhanced during JEV infection. The human miR-301a promoter has a binding site for RelA, and nuclear factor κB (NF-κB) activation promotes miR-301a expression (33). Here, we examined the time-dependent activation of NF-κB and found that JEV infection increased the amounts of the phosphorylated forms of p65, IκBα, and IKKα/β in HT22 (Fig. 7, A and B) and SH-SY5Y (fig. S6, A and B) cells up to 6 hours after infection. Hence, we used bioinformatics to investigate the binding sites for RelA in the promoter region of mouse miR-301a. The miR-301a gene in the mouse resides on chromosome 11: 86920763 to 86924763 (+), within the first intron of a ~15-kb mRNA transcript of spindle and kinetochore-associated protein 2 (Ska2). Using the PROMO software (45), a search engine that identifies transcription factor binding sites in DNA sequences from a given species, we analyzed the regions 1.5 kb upstream and 0.3 kb downstream of the transcription start site (TSS) of Ska2. Binding sites for more than 15 transcription factors were predicted (table S2). Among these, we identified a putative RelA binding site within a CpG island upstream of the TSS (1169 to 1178) (Fig. 7C).

Fig. 7 JEV induces miR-301a expression through the activation of NF-κB.

(A) HT22 cells were left uninfected (MI) or were infected with JEV at an MOI of 5 for the indicated times. Cells were then analyzed by Western blotting with antibodies specific for the indicated proteins. β-Actin served as a loading control. Western blots are representative of three independent experiments. (B) Densitometric analysis of the indicated bands in the Western blots shown in (A). Fold changes in the abundance of the indicated proteins (normalized to that of β-actin) are shown relative to those in uninfected cells. Data are means ± SD of three separate experiments. *P < 0.05, **P < 0.01, ***P < 0.001 compared to uninfected cells. Data were analyzed by one-way ANOVA followed by Bonferroni’s post hoc test. (C) Schematic representation of the predicted RelA binding site in the miR-301a promoter region within the host gene Ska2. The WT sequence GGTTCTTCCC in the miR-301a promoter was replaced by GGTTCTTAAA to generate the mutant construct. (D) Top: HT22 cells were transfected with either the WT or mutant miR-301a promoter construct alone or together with plasmid encoding Flag-tagged RelA (RelA cFlag pcDNA3). Twenty-four hours later, promoter activity was analyzed by dual luciferase assay. Firefly luciferase activity was normalized to Renilla luciferase activity, and the fold change in luciferase activity compared to that in cells transfected with the empty plasmid pGL3 was determined. Data are means ± SD of nine experiments from three independent transfections. **P < 0.01 compared to plasmid control, by one-way ANOVA with Bonferroni’s corrections. Bottom: Transfection efficiency was determined by Western blotting analysis of Flag-p65 abundance. (E) HT22 cells were transfected with the WT or mutant miR-301a promoter reporter. Six to 8 hours later, the cells were left uninfected or were infected with JEV at an MOI of 5 for the indicated times before luciferase assays were performed. The relative luciferase activity of the mutant reporter compared to that of the WT reporter after normalization to Renilla luciferase was determined. Data are means ± SD nine experiments from three independent transfections. ***P < 0.001 by two-way ANOVA followed by the Holm-Sidak method. (F) HT22 cells pretreated with or without 50 μM PDTC were left uninfected or were infected with JEV for 6 hours. Nuclear and cytoplasmic fractions of the cells were then isolated and analyzed by Western blotting with antibodies specific for the indicated proteins. PCNA and β-actin served as loading controls for the nuclear and cytoplasmic extracts, respectively. Values below the blots show miR-301a abundance in the corresponding samples as determined by qRT-PCR analysis. (G) Densitometric analysis of the indicated bands in the Western blots shown in (F). The fold changes in the abundance of the indicated proteins (normalized to that of the appropriate loading control) compared to those in uninfected control cells were determined. Data are means ± SD of three biological replicates. *P < 0.05, **P < 0.01, ***P < 0.001 compared to JEV-infected, untreated cells, as analyzed by one-way ANOVA followed by Bonferroni’s post hoc test. (H to J) HT22 cells were left uninfected or were infected with JEV at an MOI of 5 for 6 hours before being subjected to chromatin immunoprecipitation (ChIP) assay with an anti–p-p65 antibody to measure recruitment to the miR-301a promoter. The nuclear extract before treatment with antibody served as input. (I) PCR analysis of the DNA isolated by ChIP compared to input DNA with primers flanking the RelA binding site in the predicted miR-301a promoter region. (J) The relative abundance of the miR-301a promoter fragment in the indicated samples, normalized to that in the input, was determined by qRT-PCR analysis. ***P < 0.001 compared to the uninfected cells, as determined by Student’s t test. Data are means ± SD of three independent experiments, each with triplicate measurements.

To confirm whether this DNA fragment was a functional promoter and that the binding site was genuine, we cloned the 1.8-kb DNA sequence into a luciferase reporter plasmid and further prepared a construct containing a mutation in the RelA binding site within this promoter region. HT22 cells were transfected separately with plasmids encoding the WT or mutant reporter constructs, and we observed a minor difference in luciferase activity between cells expressing the WT or mutant reporters (Fig. 7D). However, ectopic expression of p65 in cells expressing the WT reporter construct substantially enhanced the transcriptional activity of the miR-301a promoter and induced luciferase expression, suggesting that this fragment is an active promoter with an authentic RelA binding site (Fig. 7D). Furthermore, we transfected cells with the WT or mutant miR-301a promoter constructs before infecting the cells with JEV to analyze whether NF-κB activation stimulated promoter activity as a response to JEV infection. Although there was a substantial increase in luciferase activity in cells expressing the WT construct, mutation of the RelA binding site substantially reduced promoter activity during the first 6 hours of JEV infection (Fig. 7E). This result suggests that the JEV-induced activation of NF-κB increased miR-301a expression.

Next, we investigated whether the activation of NF-κB suppressed SOCS5 and IRF1 during JEV infection. To address this issue, we introduced ammonium pyrrolidinedithiocarbamate (PDTC) into HT22 cells before viral infection to block the activation of NF-κB and its translocation to the nucleus, where it could bind to the miR-301a promoter and induce expression. The abundance of p-p65 in the nucleus of JEV-infected cells was reduced in response to PDTC; however, PDTC had no effect on the abundance of p-p65 in the cytoplasm (Fig. 7, F and G). PDTC inhibited NF-κB translocation, and miR-301a abundance was partially reduced during JEV infection (Fig. 7F). Furthermore, the abundances of SOCS5 and IRF1 proteins were restored in JEV-infected HT22 cells treated with PDTC (Fig. 7, F and G). Similar results were observed in PDTC-treated SH-SY5Y cells in response to JEV infection (fig. S6, C and D). Next, we performed a ChIP assay with lysates of uninfected or JEV-infected HT22 cells to further verify the binding of endogenous RelA to the putative binding site in the miR-301a promoter. We found that JEV infection resulted in the recruitment of NF-κB to the putative miR-301a promoter, which did not occur in uninfected cells, as determined by the increased abundance of miR-301a by PCR and qRT-PCR analysis (Fig. 7, H to J). Together, these results suggest that JEV infection leads to the activation of NF-κB, which then binds to the miR-301a promoter to facilitate its expression.

Mice treated with a miR-301a Vivo-Morpholino are protected from JEV infection

Morpholino oligomers are a proven antisense reagent used to block mRNA translation or interfere with RNA processing, including splicing and miRNA maturation. To evaluate the effect of miR-301a inhibition on the production of type I IFN in vivo, we designed a Vivo-Morpholino to specifically target the seed sequence of miR-301a (miR-301a-VM) and an appropriate negative control (VM-NC) in which five nucleotides of the seed sequence were mutated (Fig. 8A). The Vivo-Morpholino system provides effective delivery of morpholino antisense oligomers into a wide range of tissues in live mice (46). Because the efficiency of Vivo-Morpholino delivery into the brain through the intravenous or intraperitoneal route is quite low, mice were initially injected intracranially with sequentially increasing amounts of miR-301a-VM for dose standardization. We observed a substantial reduction in endogenous miR-301a expression in mice treated with miR-301a-VM doses of ≥16 mg/kg body weight (fig. S7A). However, at doses >18 mg/kg, the mice exhibited decreased survival and lost weight (fig. S7, B and C), indicating that when administered above this dose, miR-301a-VM resulted in toxicity. Next, we administered miR-301a-VM (intracranially) into WT mice at a dose of 18 mg/kg 12, 24, or 48 hours after the mice were infected intraperitoneally with JEV and monitored the mice for clinical symptoms daily. Treatment with miR-301a-VM at 12 or 24 hours after infection improved the survival percentage (60 and 20%, respectively) of the mice as compared with the survival of the mice treated with VM-NC (Fig. 8B). Furthermore, the percentage of paralyzed mice and the degree of body weight loss were substantially reduced in the mice treated with miR-301a-VM 12 or 24 hours after infection (fig. S7, D and E) compared to those of mice treated with VM-NC. Thus, miR-301a-VM not only improved survival but also lessened disease severity in the CNS.

Fig. 8 Inhibition of miR-301a in vivo restores the IFN response and improves survival in JEV-infected mice.

(A) Stem-loop sequence of pre–miR-301a. Bottom: Sequences of the miR-301a Vivo-Morpholino (miR-301a-VM), which targets mature miR-301a (blue), and a scrambled Vivo-Morpholino that was designed as a negative control (VM-NC). (B) BALB/c mice were intracranially treated with miR-301a-VM or VM-NC (18 mg/kg) at the indicated times after they were infected with JEV (3 × 105 PFU) and then were observed daily to determine their survival rate. The asterisk represents comparison between the JEV/VM-NC group versus the JEV/miR-301a-VM 12 h.p.i. and JEV/miR-301a-VM 24 h.p.i. groups. P values were determined by one-way ANOVA followed by Bonferroni’s post hoc test. (C) Brain samples from the mice shown in (B) were collected on day 3 (72 h.p.i.) of infection, and their abundance of miR-301a was quantified by qRT-PCR analysis. Data are represented as the fold change in miR-301a abundance (normalized to that of SNORD68) in the infected samples relative to that in the uninfected samples. Data are means ± SD of four mice from each group. **P < 0.01, ***P < 0.001 compared to the JEV + VM-NC group, as analyzed by one-way ANOVA followed by Bonferroni’s multiple comparisons. (D) Viral loads in the brain tissues of the mice shown in (B) were evaluated by plaque assay in porcine kidney cells. Plaque assays were performed twice, and the data are means ± SD of three mice from each group. *P < 0.05 when compared to the JEV + VM-NC group, as determined by one-way ANOVA with Bonferroni’s correction. (E to G) Mice were left uninfected or were infected with JEV in the presence of VM-NC or were infected with JEV and treated with miR-301a-VM at the indicated times. Three days after infection, brains were collected from the mice and analyzed by qRT-PCR to determine the abundances of viral RNA (E), IFN-β mRNA (F), and IFIT1, ISG15, and OAS1 mRNAs (G). Data are means ± SD of four mice from each group. *P < 0.05, **P < 0.01, ***P < 0.001 when compared to the VM-NC group. Data were analyzed by one-way ANOVA followed by Bonferroni’s post hoc test. (H) Brain samples from the mice described in (E) to (G) were analyzed by Western blotting with antibodies specific for SOCS5 and IRF1. β-Actin served as loading control. Blots are representative of four mice from each group. (I and J) Densitometric analysis of the blots shown in (H). The histograms show the fold changes in SOCS5 (I) and IRF1 (J) protein abundance (normalized to that of β-actin) in infected samples compared to those in uninfected samples. Data are means ± SD of four mice from each group. ##P < 0.01, ###P < 0.001, **P <0.01, ***P < 0.001, as calculated by one-way ANOVA followed by Bonferroni’s multiple comparisons.

In another set of experiments, the mice were left uninfected or were infected with JEV and treated with VM-NC (18 mg/kg) or miR-301a-VM at 12, 24, and 48 hours after infection and then were left for another 24 hours, at which point brain samples were collected. Although mice treated with VM-NC showed a substantial increase in miR-301a abundance in response to infection compared to uninfected mice, this enhancement was markedly reduced in the mice treated with miR-301a-VM (Fig. 8C). Because of the increased survival of miR-301a-VM–treated mice compared to VM-NC–treated mice, we measured the titers of JEV in the mouse brains. Both plaque and qRT-PCR assays showed that there was a substantial reduction in viral load in the miR-301a-VM–treated mice, which suggested that viremia was restricted because of the inhibition of miR-301a in the JEV-infected mice (Fig. 8, D and E). To determine whether the reduction in viremia was caused by the IFN response, we evaluated the expression of IFN-β and various ISGs. We found that all of these mRNAs were substantially increased in abundance in the mice treated with miR-301a-VM 12 and 24 hours after infection compared to those in the VM-NC–treated mice (Fig. 8, F and G). As expected, the suppression of SOCS5 and IRF1 protein production in the VM-NC–treated, JEV-infected mice was partially restored in the JEV-infected mice treated with miR-301a-VM (Fig. 8, H to J). Thus, these data suggest that the inhibition of miR-301a expression in vivo improved the survival of the JEV-infected mice by inducing the type I IFN response to inhibit viral propagation.

JEV infection of peripheral organs has no effect on miR-301a expression

Although JEV is neurotropic, it also infects some peripheral tissues before invading the CNS. To explore whether miR-301a might regulate JEV infection in peripheral tissues, we infected liver (HepG2) and kidney [human embryonic kidney–293 (HEK 293)] cell lines with JEV and then analyzed the cells for NF-κB activation and the expression of miR-301a, IFN-β, IFIT1, and ISG15, as well as the generation of SOCS5 and IRF1 proteins, at various times. We found that JEV infection did not induce substantial increases in miR-301a abundance in either cell type (fig. S8A); however, NF-κB was activated up to 6 hours after infection (fig. S8, B and C). Consequently, no changes in the abundances of SOCS5 and IRF1 proteins were observed in JEV-infected cells transfected with anti–miR-301a compared to those in JEV-infected cells transfected with the negative control (anti–miR-Con) (fig. S8, D and E). Furthermore, the effect of the inhibitor on the JEV-induced expression of IFN-β and ISGs in HepG2 and HEK 293 cells was examined. The abundances of IFN-β and ISG mRNAs were equivalent in inhibitor- and control inhibitor–treated cells (fig. S8, F and G); thus, JEV replication was not inhibited in either cell type, as evident from the measurement of viral RNA and titers (fig. S8, H and I). Consistent with these data, liver and kidney samples taken from JEV-infected mice had similar amounts of miR-301a to those from uninfected mice (fig. S9A). For further confirmation, we injected JEV-infected mice with miR-301a-VM or VM-NC intraperitoneally at various times and then collected liver and kidney samples after 72 hours. The amounts of IFN-β mRNA and viral RNA in the miR-301a-VM–treated, JEV-infected mice were similar to those in the VM-NC–treated infected mice (fig. S9, B and C). These results suggest that the NF-κB–mediated induction of miR-301a expression and the suppression of SOCS5 and IRF1 production are not ubiquitous mechanisms in the livers and kidneys of JEV-infected mice.

DISCUSSION

Upon viral infection, host innate immunity is the first line of antiviral defense, which is tasked with the recognition of viral components and the production of proinflammatory cytokines and type I IFNs (4749). Neuronal innate immunity in flaviviral infection, including that by JEV, is critical because of the neurotropic nature of these viruses. After the initial infection of peripheral tissues by flaviviruses, they invade the CNS, where neurons are the main target (50). Because the CNS lacks secondary lymphoid tissues and cannot induce adaptive immune responses, it relies typically on intrinsic neuronal innate immunity to limit the extent of neuroviral infection. IFN-β, an early type I IFN, which is an inherent part of the neuronal innate immune response, is vital for the efficient restriction of neurotropic viral propagation (51).

miRNAs have emerged as important posttranscriptional regulators of many biological systems, including mammalian innate immunity (52, 53), and they sometimes play a decisive role when the innate immune system is challenged by viral infection. Viruses may exploit cellular miRNAs, and in some cases, the increased abundance of cellular miRNAs may reshape cellular gene expression to the benefit of the virus. Here, we observed the substantial early enhancement of neuronal miR-301a expression after JEV infection, which led us to investigate whether miR-301a affected the antiviral IFN response. Inhibition of miR-301a in JEV-infected neuronal cells resulted in the considerable restoration of IFN-β production and the restriction of viral replication, suggesting that miR-301a abrogates the JEV-induced IFN response in neurons. Consistent with a previous report that showed that IFN-β elicits its antiviral actions by inducing the expression of a wide array of ISGs through its binding to IFN-stimulated response elements in the promoters of these genes (54), we found that the expression of ISG15, IFIT1, and OAS1 was induced under conditions in which miR-301a expression was restricted. Under basal conditions, the expression of these ISGs is usually undetectable, but their expression is stimulated upon viral infection to mediate the antiviral effector functions of IFN-β (55).

To determine the biological role of the enhanced expression of miR-301a in the host response to JEV infection, we analyzed specific target genes and identified IRF1 and SOCS5 as miR-301a targets involved in the IFN-β response in neuronal cells. Delivery of miR-301a into neuronal cells resulted in decreased IRF1 and SOCS5 protein abundances but did not affect the amounts of IRF1 and SOCS5 mRNAs, which suggests that miR-301a reduced the amounts of these proteins by inhibiting mRNA translation without affecting mRNA abundance. Moreover, knockdown of miR-301a in JEV-infected neuronal cells substantially rescued the generation of IRF1 and SOCS5 proteins, demonstrating that JEV-induced miR-301a suppresses IRF1 and SOCS5.

Studies previously showed that IRF1 is a potent inhibitor of a broad spectrum of viruses, including flaviviruses (56, 57). IRF1 shapes both innate and adaptive immune responses to protect the host from lethal WNV infection (41). Furthermore, IRF1 expression in neurons prevents the replication of fatal neurotropic vesicular stomatitis virus (VSV) (58). The transcriptional profile of antiviral genes induced by type I IFNs is also stimulated by IRF1 (56). Thus, both IRF1 and IFN-β appear to promote an active and critical antiviral program. Although IRF1 is implicated as a regulator of the expression of genes encoding IFN-α and IFN-β, it specifically binds to the upstream regulatory region of the IFN-β gene and induces IFN-β expression in response to viral infection (40, 59). Because miR-301a directly targets IRF1, we observed the substantial stimulation of the expression of IFN-β, but not IFN-α, when miR-301a was inhibited in JEV-infected neuronal cells. Knockdown of IRF1 in JEV-infected cells prevented the induction of IFN-β expression that occurred in miR-301a–inhibited cells, whereas ectopic expression of IRF1 conversely enhanced the expression of IFN-β in cells in which miR-301a was inhibited. Together, these observations suggest that miR-301a contributes to the suppression of IFN-β production through its effects on IRF1.

The other specific target of miR-301a that we found was SOCS5, which is not associated directly with the regulation of IFN-β production. SOCS5 is preferentially found in T helper 1 (TH1) cells, and it inhibits the differentiation of naïve CD4+ T cells into TH2 cells, signifying a key role in regulating the balance between these two cell subsets (60). Members of the SOCS family can be manipulated for viral benefit, and the expression of SOCS-encoding genes is usually induced upon different viral infections (6163). However, the role of SOCS5 in innate immunity was unknown until a report by Watanabe et al. (39) described how the overexpression of SOCS5 enhances the innate immune response in murine septic peritonitis and improves survival. Consistent with this previous study, we found that knockdown or overexpression of SOCS5 in JEV-infected cells respectively attenuated or amplified IFN-β expression in neuronal cells in which miR-301a was inhibited.

The activation of EGFR during viral infection suppresses the IRF1-mediated antiviral response (44). Previous studies showed that SOCS5 interacts with the EGFR in a ligand-independent manner to inhibit EGFR signaling (43, 64). Thus, we hypothesized that the miR-301a–mediated suppression of SOCS5 during JEV infection might initiate EGFR activation and thereby suppress IRF1 expression. Accordingly, we observed that EGFR was activated in JEV-infected neuronal cells and that manipulation of endogenous or exogenous SOCS5 abundance potentially regulated the abundance of p-EGFR. Furthermore, the decreased expression of IRF1 caused by knockdown of SOCS5 and the enhanced expression of IRF1 in cells overexpressing SOCS5 suggest that SOCS5 stimulates IRF1 expression in JEV-infected neuronal cells. The amplification of the IRF1-mediated IFN-β response in cells treated with the EGFR inhibitor AG 1478 further suggested the participation of EGFR in the SOCS5-dependent generation of IRF1 protein. The substantial reduction in JEV RNA abundance compared to the modest increase in IRF1 expression in AG 1478–treated cells suggested that, in addition to enabling induction of IRF1 expression, AG 1478 might directly inhibit JEV replication. Several reports showed that multiple viruses use EGFR for cell entry and that inhibitors of the kinase activity of EGFR inhibit viral infection (6567). We assume that AG 1478, in conjunction with enabling induction of IRF1 expression, might also hamper the entry of JEV into cells by inhibiting EGFR activation. The observation of the almost comparable viral titers in the culture medium of cells infected with JEV in the presence or absence of AG 1478 at late times during infection also supported our assumption. IRF1 and SOCS5 are involved in the miR-301a–mediated inhibition of JEV infection in neurons; however, they are not the only contributory factors. There are many other genes that are predicted to be potential targets of miR-301a (table S1) and thus may also be involved in the miR-301a–mediated regulation of host-virus interactions.

Numerous studies showed that virus-induced host transcription factors induce miRNA expression by binding to their promoter regions (68, 69). We analyzed the miR-301a promoter with a software-based method and detected the binding sites of several transcription factors, including NF-κB. We focused on NF-κB because NF-κB activation stimulates the expression of miR-301a in a human pancreatic cancer cell line (33). However, other transcription factors may regulate miR-301a expression through binding to its promoter. Several miRNAs that are critical to innate and adaptive immunity regulate NF-κB signaling and influence the mammalian response to microbial infection (70, 71). Here, we found that the promoter for mouse miR-301a contains an authentic RelA binding site and that miR-301a expression in neurons after JEV infection was dependent on NF-κB activation. The extent of NF-κB activation in JEV-infected neurons was directly proportional to the degree of miR-301a expression. Overexpressing NF-κB subunits in HT22 cells increased the transcription of miR-301a, as confirmed by mutagenesis studies in which the NF-κB binding site was disrupted. Furthermore, the NF-κB–driven transcription of miR-301a in JEV-infected cells was confirmed by luciferase reporter assays. Next, we treated the cells with PDTC. PDTC is an established antioxidant and NF-κB inhibitor, and it blocks NF-κB activation by inhibiting the ubiquitylation of IκB (inhibitor of nuclear factor κB) independently of its antioxidative functions (72). In neurons, PDTC prevents the nuclear translocation of NF-κB and its binding to DNA (73). Consistently, when we inhibited NF-κB translocation in JEV-infected HT22 cells with PDTC, we attenuated the expression of miR-301a. This resulted in the restored generation of the products of the target genes of miR-301a, thus suggesting that NF-κB promotes the expression of miR-301a during JEV infection. The binding of NF-κB to the miR-301a promoter element was further confirmed by ChIP analysis.

We validated the in vivo effect of miR-301a in a JEV-infected mouse model with miR-301a-VM. Several studies have already shown that Vivo-Morpholino actively decreases viral titers in animal models by blocking the translation of viral mRNA (7476). Vivo-Morpholino has been used in vivo for the effective inhibition of specific miRNAs (77); therapeutics involving Vivo-Morpholino have also fruitfully reached human clinical trials (78, 79). Previously, we showed that a Vivo-Morpholino against the JEV genome serves a neuroprotective role and blocks viral replication in mice to an extent similar to that by which a Vivo-Morpholino against TLR7 inhibits type I IFN production and promotes viral infection in neurons (9, 80). Here, we injected miR-301a-VM into mice to demonstrate its potential application against JEV-induced immune evasion in a mouse model. Therapeutic treatment during the early stage of JEV infection efficiently relieved clinical signs, including paralysis and loss of body weight, as well as improved survival. For further clarification, we checked the interactions of miR-301a with its targets and the subsequent effects on IFN-β production and JEV propagation in vivo. Inhibition of miR-301a restored the abundances of IRF1 and SOCS5 proteins and further promoted the expression of IFN-β and ISGs, effectively resulting in a substantial reduction in JEV replication. Thus, Vivo-Morpholino–based antisense therapy targeting miR-301a could be a promising therapeutic approach against JEV infections. JEV pathogenicity is fundamentally similar to that of other encephalitic flaviviruses, where it is believed that initial virus replication occurs in peripheral organs including the liver, kidney, heart, and lung, and then invades the CNS to develop neuropathogenesis (81). Because miR-301a is an abundant miRNA in the brain and there is no evidence of its signaling in liver and kidney, it did not affect IFN production in peripheral organs after infection with JEV.

Most miRNAs, together with cellular mRNAs, are transcribed by polymerase II using many of the same transcription factors (19). Thus, cellular transcription factors, whose abundances are increased by viral infection, play a vital role in changing miRNA expression patterns. Viral infection activates NF-κB, which is crucial for initiating the production of antiviral IFNs; however, sometimes this immune response, if it is too strong and persists for too long, can be harmful to the host. Thus, cells must have negative feedback mechanisms to tightly regulate the response (82). There are several reports indicating that miRNA is part of a negative feedback mechanism, which inhibits inflammatory cytokine production in response to microbial stimuli (83, 84). A study that used bacterial lipopolysaccharide revealed that the NF-κB–dependent induction of miR-146a and miR-146b expression inhibits the innate immune response by targeting IL-1 receptor–associated kinase 1 (IRAK1) and tumor necrosis factor receptor–associated factor 6 (TRAF6) (70). This regulatory mechanism is exploited by many viruses, including Epstein-Barr virus (EBV), Kaposi’s sarcoma–associated herpesvirus (KSHV), Enterovirus 71 (EV71), Sendai virus (SeV), VSV, and hepatitis C virus (HCV), to evade the host innate immune system. EBV stimulates the NF-κB–dependent expression of cellular miR-155, which in turn targets intermediates of innate immune signaling pathways (71). The miRNA miR-132 inhibits innate antiviral immunity during KSHV infection by inhibiting expression of the transcriptional coactivator p300 (85). EV71 induces the expression of miR-146a, which targets IRAK1 and TRAF6, which are involved in TLR signaling and type I IFN production (86). The expression of IFNA1 is repressed by miR-466I in cells infected with SeV and VSV (87). The increased expression of miR-21 promotes HCV replication by inhibiting IFN-α signaling through the targeting of MyD88 and IRAK1 (88). HCV also induces the expression of a liver-specific miRNA, miR-122, to enhance viral genome stabilization and protein translation (89). Here, we found that JEV infection increases the abundance of the host miR-301a to inhibit IFN-β production and escape the host immune response.

In summary, we have identified a previously uncharacterized, neuron-specific host-virus interaction, in which the NF-κB–mediated inducible expression of miR-301a inhibits the antiviral IFN-β response by suppressing IRF1 and SOCS5 production. Thus, the virus evades the host innate immune system. Deactivation of JEV-induced miR-301a expression conversely impaired this molecular pathway, thereby restoring the antiviral immune response and reducing mortality in JEV-infected mice (fig. S10). Thus, miR-301a might be a potential therapeutic target in JEV infection of humans. These findings further illuminate the influence of host miRNA on the antiviral response and may help to understand the molecular pathogenesis driven by flavivirus-host interactions.

MATERIALS AND METHODS

Mice

P10 BALB/c mice of either sex were housed together with their respective mothers under a 12-hour dark/12-hour light cycle at a constant temperature and humidity. All experiments were performed after obtaining approval from the Institutional Animal Ethics Committee of the National Brain Research Centre (NBRC) (approval no. NBRC/IAEC/2014/96). The animals were handled in strict accordance with good animal practice as per the guidelines of the Committee for the Purpose of Control and Supervision of Experiments on Animals, Ministry of Environment and Forestry, Government of India.

Cell culture

The mouse hippocampal neuronal cell line HT22, a gift from S. K. Sharma (NBRC); the mouse microglial cell line BV2 and the human neuroblastoma cell line SH-SY5Y, a gift from S. Levison (Rutgers University–New Jersey Medical School, NJ); the human hepatocellular liver carcinoma cell line HepG2, a gift from E. Sen (NBRC); the HEK 293 cell line, provided by A. Krishnan (Institute of Molecular Medicine, New Delhi, India); and porcine stable kidney cells (PS cells), a gift from G. R. Medigeshi (Translational Health Science and Technology Institute, Faridabad, India), were cultured at 37°C in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin (100 units/ml), and streptomycin (100 μg/ml). All cell culture reagents were obtained from Sigma-Aldrich, unless otherwise specified. HT22 cells were used for our experiments with previous permission from D. Schubert of the Salk Institute from whom these cells were originally obtained.

Primary cortical neuronal cultures

Cortical neurons were cultured according to a previously described protocol (9). Briefly, cortices of P2 BALB/c mouse pups were dissected aseptically in calcium- and magnesium-free (CMF) Tyrode’s solution after decapitation. After the meninges were removed, the tissue was chopped into smaller pieces and collected in the CMF-Tyrode’s solution. The samples were treated with trypsin-deoxyribonuclease and then dissociated in the same solution by trituration to make a single-cell suspension. The suspension was centrifuged, and the pellet was resuspended in neurobasal medium, supplemented with 2 mM l-glutamine, 1% glucose, 5% FBS, 10% horse serum, and penicillin-streptomycin. Cells were plated at a density of 5 × 103/cm2 onto poly-d-lysine–coated Lab-Tek chamber slides for immunofluorescence staining, whereas for RNA preparation, 3 × 105 cells were seeded in 60-mm plates. After 48 hours of incubation at 37°C, the serum-containing medium was replaced with serum-free medium, and the cells were incubated for 4 hours with antibiotics alone. For experimental treatments, the resting medium was exchanged for DMEM with N2 and B27 supplements and antibiotics. Arabinoside (20 μM) was used to inhibit astrocyte proliferation.

Viral infection of cells and mice

All cells were seeded at the desired density in culture plates as per the requirements for different experiments. After the cells reached 90% confluence, they were further incubated for 3 to 4 hours in serum-free medium and infected with JEV (strain GP78) at an MOI of 5. Cells were harvested at different times for the time course study. For the dose-dependent study, the cells were infected for 6 hours separately with JEV at MOIs of 1, 5, or 10. The infection of primary neuronal cells was performed according to a similar procedure. Mock infection (MI) consisted of adding the same amount of medium as that containing the JEV inoculum but without virus. P10 BALB/c mouse pups were divided into two groups, each having four pups; the gender of the mice in each group was random unless otherwise indicated. Members of the JEV group of mice were injected with 3 × 105 plaque-forming units (PFU) of virus, whereas members of the uninfected group received phosphate-buffered saline (PBS) through the intraperitoneal route. After 2 days of infection, the animals were sacrificed, and the brains were excised after repeated transcardial perfusion with ice-cold PBS. For ISH and immunofluorescence analysis, tissues were fixed with 4% paraformaldehyde (PFA) for 48 hours at 4°C and then dipped in 30% sucrose until they were completely immersed in the solution. The tissues were then processed for cryosectioning. For miRNA, RNA, and protein isolation, the brains were stored at −80°C until needed for use.

Combined ISH and IHC analysis

Cryosections, prepared with a Leica CM3050S cryostat, were processed for ISH with the miRCURY LNA microRNA ISH Optimization Kit (Exiqon), as described previously (29). Briefly, thawed sections were fixed in fresh 30% formalin overnight. After washes with PBS, the sections were treated with proteinase K (7 μg/ml) at 37°C for 15 min. The sections were then thoroughly washed with PBS, dehydrated with ethanol, and incubated with hybridization buffer containing 60 nM double digoxigenin (DIG)–labeled LNA miR-301a probe (5′-GCTTTGACAATACTATTGCACTG-3′; Exiqon) or 5 nM 5′-DIG–labeled LNA U6 snRNA probe (5′-CACGAATTTGCGTGTCATCCTT-3′; Exiqon) for 1 hour at 55°C in a humidified chamber. After hybridization, the sections were sequentially washed with 5×, 1×, and 0.2× saline-sodium citrate buffer and incubated with blocking solution for 15 min at room temperature. Sections were then incubated with sheep anti–DIG–alkaline phosphatase (AP) antibody (1:400 dilution; Roche Life Science) for 60 min at room temperature in a humidified chamber and then were washed twice with 1× PBS-Tween 20 (PBST). Freshly prepared AP substrate (Roche Life Science) was then added to the sections and incubated for 2 hours at 30°C in the humidifying chamber under dark conditions. The reaction was terminated by the addition of KTBT buffer (an AP stop solution composed of 50 mM tris-HCl, 150 mM NaCl, and 10 mM KCl) for 5 min. After stringent washing with water, sections were incubated in blocking solution (5% bovine serum albumin in PBS) for 20 min followed by overnight incubation with anti-MAP2 antibody (1:200, Millipore) at 4°C. After extensive washing with PBS, the sample slides were incubated with biotinylated anti-mouse immunoglobulin G (IgG; 1:200, Vector Laboratories) for 30 min at room temperature followed by a 20-min incubation with horseradish peroxidase (HRP)–conjugated streptavidin (1:250, Vector Laboratories). After the samples were rinsed with PBS, a brown reaction product was developed with a DAB substrate kit (SK-4100, Vector Laboratories) according to the manufacturer’s instructions. After a 2-min reaction, the slides were dehydrated in a series of solutions containing different percentages of alcohol and mounted with DPX (Qualigens). The slides were observed with a Leica DMRXA2 microscope, and images were taken under the appropriate magnification for each sample.

Transfection of cells with miRNA mimics and inhibitors

To overexpress or inhibit miR-301a, we transfected cells with mimics of mouse or human miR-301a (double-stranded RNAs that mimic mature endogenous miR-301a) or with miR-301a inhibitors (single-stranded, modified RNAs that specifically inhibit endogenous miR-301a) (Qiagen), respectively, with the HiPerFect Transfection Reagent (Qiagen) according to the manufacturer’s instructions. Twenty-four hours later, the cells were harvested or infected with JEV for specific times and then were analyzed to determine the abundances of the miRNAs, mRNAs, and proteins of interest. Negative controls of the mimic or inhibitor (Ambion) were used in the transfections as the matched controls. MT cells received an equal volume of HiPerFect reagent that did not contain any nucleic acid.

Enzyme-linked immunosorbent assay

HT22 cells were seeded in six-well plates (0.3 × 106), incubated overnight, and transfected with the miR-301a inhibitor or inhibitor control. Twenty-four hours later, the cells were infected with JEV for the appropriate times. In another set of experiments, the cells were infected with JEV for 6 hours in the presence of Tyrphostin AG 1478, as indicated in the figure legends. In both sets of experiments, the concentration of IFN-β in the cell culture medium was determined by ELISA using the antibody pair for the detection IFN-β and the IFN-β standard (BioLegend) according to the manufacturer’s recommendations.

Plasmid construction

The 354–base pair (bp) segment of cDNA encoding the 3′UTR of mouse SOCS5 and a 429-bp cDNA segment encoding the 3′UTR of IRF1 containing the putative miR-301a binding site were amplified by PCR from mouse brain cDNA with the SOCS5 Luc and IRF1 Luc primers (forward and reverse), respectively (table S3). The DNA fragments were subcloned into the Hind III and Spe I sites downstream of the firefly luciferase gene in the pMIR-REPORT plasmid. Site-directed mutagenesis performed with the SOCS5 Luc Mut and IRF1 Luc Mut primers (forward and reverse; table S3) generated mutants with the appropriate mutations. A PCR-based method involving Phusion High-Fidelity DNA Polymerase and Dpn I (New England Biolabs) was performed for mutagenesis. The SOCS cds and IRF1 cds primers (forward and reverse; table S3) were respectively used to amplify the coding regions for SOCS5 and IRF1 from mouse cDNA. These PCR products were digested with Hind III and Xho I or Hind III and Bam HI, respectively, and then cloned into the pcDNA 3.1(+) plasmid (which was provided by D. Chattopadhyay, Amity University, Kolkata, India). The ~1.8-kb promoter element of miR-301a was amplified from mouse cDNA with the miR-301a Pro forward and reverse primers (table S3), digested with Mlu I and Xho I, and inserted into the pGL3-Basic vector (a gift from P. Chattopadhyay, All India Institute of Medical Sciences, New Delhi, India) immediately upstream of the firefly luciferase gene. The mutation in the RelA binding site in the promoter segment was achieved with the miR-301a Pro Mut forward and reverse primer (table S3) according to the site-directed mutagenesis procedure described earlier. Hence, all of the constructs were commercially sequenced at Invitrogen BioServices India Pvt. Ltd., Gurgaon, India.

Transfection of cells with esiRNA and plasmids

esiRNA specific for mouse SOCS5 (EMU021991) and IRF1 (EMU052481), as well as negative Con-esiRNA (sense, 5′-GTGAGCAAGGGCGAGGAGCTGTTCACCGGGGTGGTGCCCATCCTGGTCGAG CTG GA-3′) were purchased from Sigma-Aldrich. HT22 cells were transfected with either esiRNAs (10 pM) or plasmids encoding SOCS5 or IRF1 with Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. Briefly, after the cells reached 70 to 80% confluence, they were cotransfected with the miR-301a inhibitor and either the esiRNAs or expression plasmids with commercial medium (Opti-MEM, Invitrogen) and Lipofectamine 2000. Twenty-four hours later, the cells were infected with JEV (MOI, 5) for 6 hours, and the cells were then subjected to mRNA analysis. Transfection efficiency was assessed by measuring the amounts of the proteins of interest. MT cells received an equivalent volume of Opti-MEM and Lipofectamine mix in the absence of any nucleic acid.

Luciferase reporter assays

HT22 cells (2 × 104) were seeded in a 24-well plate for 16 to 18 hours and then were transiently transfected with firefly luciferase reporter constructs together with either an miR-301a mimic or inhibitor and their respective controls with Lipofectamine 2000. The cells were also cotransfected with a Renilla luciferase vector (pRL-TK, a gift from E. Sen, NBRC) for normalization of transfection efficiency. Twenty-four hours later, the cells were harvested at the times indicated in the figure legends and were lysed in Passive Lysis Buffer (Promega). The luciferase activity of each sample was measured with the Dual-Luciferase Reporter Assay System (Promega) with a GloMax 96 Microplate Luminometer according to the manufacturer’s protocol. For promoter analysis, HT22 cells were cotransfected with the luciferase reporter constructs, and the mouse RelA cFlag pcDNA3 clone (Addgene plasmid repository) with Lipofectamine 2000. The pGL3-Basic vector was used as a negative control. Luciferase activity was determined 24 hours later. In another set of experiments, the cells were transiently transfected with the promoter constructs. Twenty-four hours later, the cells were infected with JEV, and luminescence was measured at the times indicated in the figure legends.

ChIP assays

ChIP assays were performed with the Dynabeads Protein G Immunoprecipitation Kit (Life Technologies) according to the manufacturer’s instructions. Briefly, uninfected or JEV-infected HT22 cells were first treated with 1% formaldehyde for 10 min, which was stopped by the addition of 0.25 M glycine (pH 7.0). After centrifugation, the cell pellet was lysed in radioimmunoprecipitation assay buffer containing protease inhibitors and then was sonicated to shear the chromatin. The lysate was then mixed with Dynabeads Protein G (Invitrogen) (50/250 μl of lysate) in a HulaMixer (Invitrogen) for 1 hour at 4°C. The precleared supernatant, referred to as input sample, was subjected to immunoprecipitation with 10 μl of anti-RelA (p-p65) antibody conjugated to Dynabeads (50 μl). The precipitated complexes were then eluted and subjected to reverse cross-linking by the addition of 0.2 M NaCl, which was followed by incubation at 65°C for 6 hours. The coimmunoprecipitated DNA was purified with a QIAquick PCR purification kit (Qiagen). Occupancy of the promoter was detected by PCR and qRT-PCR with primers flanking the p50-RelA binding site in the mouse miR-301a promoter (forward primer, 5′-AGTCGTGAAGCATTGCTATC-3′; reverse primer, 5′-AAGACTAAGGACCTGTGGTC-3′). Input sample (40 μl) was subjected to reverse cross-linking and resolved by SDS–polyacrylamide gel electrophoresis (SDS-PAGE). After ChIP was performed, the Dynabeads were boiled with loading buffer for 10 min and analyzed by Western blotting to assess the immunoprecipitation efficiency.

Quantitative RT-PCR

For quantitative determination of mature mRNA and miRNA abundances, qRT-PCR analysis was performed. After the isolation of total RNA from treated cells with Tri Reagent (Sigma-Aldrich), cDNA was prepared with the Advantage RT-for-PCR Kit (Clontech Laboratories). qRT-PCR analysis of mouse and human IFNs, ISGs, and viral RNA were performed with the Power SYBR Green PCR Master Mix (Applied Biosystems) with gene-specific primers (table S1). The relative abundance of an mRNA of interest was determined by normalization to that of GAPDH mRNA through the 2−ΔΔCt method (Ct refers to the threshold value). miRNA was isolated with a miRNeasy kit (Qiagen), and cDNA was prepared with a miScript II RT Kit with miScript HiSpec buffer (Qiagen) according to the manufacturer’s instructions. The forward primers for amplification of different miRNAs were as follows: mouse miR-301a, 5′-CAGUGCAAUAGUAUUGUCAAAGC-3′; human miR-301a, 5′-CAGUGCAAUAGUAUUGUCAAAGC-3′; and mouse miR-301b, 5′-CAGUGCAAUGGUAUUGUCAAAGC-3′. qRT-PCR was performed with specific forward primers and the miScript SYBR Green PCR Kit containing the miScript Universal Reverse Primer (Qiagen). The snRNA SNORD68 was used as a normalization control. The thermal cycler ViiA 7 Real-Time PCR (Applied Biosystems) was used for qRT-PCR, and the data were analyzed with the iCycler Thermal Cycler software (Applied Biosystems).

Western blotting

Cells and mouse tissues were lysed in complete lysis buffer containing 1% Triton X-100, 10 mM tris-HCl (pH 8.0), 150 mM NaCl, 0.5% NP-40, 1 mM EDTA, 0.2% EGTA, 0.2% sodium orthovanadate, and a protease inhibitor cocktail (Sigma-Aldrich); and the protein concentration in each sample was measured with the BCA reagent (Sigma-Aldrich). Equal amounts of the proteins were resolved by SDS-PAGE and transferred onto a nitrocellulose membrane, which was blocked with 5% skim milk in 1× PBST and incubated with primary antibodies specific for SOCS5 (1:1000, Abcam), IRF1 (1:1000, Cell Signaling Technology), NF-κB (1:1000, Santa Cruz Biotechnology), pNF-κB (1:1000, Cell Signaling Technology), pIκBα (1:1000, Cell Signaling Technology), IκBα (1:1000, Cell Signaling Technology), pIKKα/β (1:1000, Cell Signaling Technology), p-EGFR (1:1000, Abcam), the Flag tag (1:1000, Sigma-Aldrich), proliferating cell nuclear antigen (PCNA) (1:2000, Cell Signaling Technology), or β-actin (1:10,000; Sigma-Aldrich). β-actin was used as internal control except for samples containing nuclear proteins for which PCNA acted as the internal control. The secondary antibodies used for detection were HRP-conjugated goat anti-mouse and goat anti-rabbit IgG (1:5000, Vector Laboratories). The blots were developed by exposure in a ChemiGenius Bio-Imaging System (Syngene) with GeneSnap software. Representative Western blots are shown in the figures. Samples of the Western blots used to generate the densitometric data shown in Figs. 3 to 8 can be found in fig. S11, whereas those used to generate the densitometric data shown in figs. S3, S4, S6, and S8 can be found in fig. S12.

Immunofluorescence

Brain sections were permeabilized with 0.1% Triton X-100 in PBS and then incubated with a blocking buffer for 1 hour at room temperature, which was followed by overnight incubation with anti-NeuN (1:250, Millipore) and anti-SOCS5 (1:250, Abcam) antibodies at 4°C. After extensive washing, the sections were incubated with Alexa Fluor 488– or Alexa Fluor 594–conjugated secondary antibodies (1:1500, Molecular Probes) for 1 hour. Finally, the sections were mounted with 4′,6-diamidino-2-phenylindole (Vector Laboratories Inc.) and observed with a Zeiss Apotome microscope at the specified magnification. Primary cortical neuronal cells were infected with JEV for 6 hours, fixed with 4% PFA, and then subjected to the same immunofluorescence analysis.

Inhibition of EGFR activation

HT22 cells were seeded into six-well plates and infected with JEV in the presence or absence of the selective EGFR tyrosine kinase inhibitor, Tyrphostin AG 1478 (10 μM) (Sigma-Aldrich). Six hours after infection, the cells were harvested to assess IRF1 mRNA and viral RNA abundances by qRT-PCR analysis. In addition, cell culture medium was collected to measure IFN-β secretion and JEV titers at the times indicated in the figure legends.

Inhibition of NF-κB activation

HT22 cells were treated with 50 μM PDTC (an inhibitor of NF-κB activation) (Sigma-Aldrich) for 2 hours before being infected with JEV for 6 hours. The abundances of cytosolic and nuclear proteins of interest were analyzed by Western blotting, whereas the amounts of miRNAs were analyzed by qRT-PCR analysis.

JEV infection and administration of Vivo-Morpholino to mice

The GP78 strain of JEV was propagated in suckling BALB/c mice (P3 or P4) of either sex until the symptoms of sickness (limb paralysis, poor pain response, and whole-body tremors) appeared. At this point, viral suspensions were prepared as described previously (29). The virus was titrated by plaque formation assay with the PS cell line as reported previously (29). For in vivo experiments, P10 mice of either sex were randomly assigned to eight groups. Among them, group 1 was the MI group and received only PBS, whereas mice from the other seven groups were injected intraperitoneally with JEV (3 × 105 PFU). In addition to receiving virus, mice in groups 3, 4, and 5 were treated with the Vivo-Morpholino (Gene Tools LLC) negative control (VM-NC; 18 mg/kg) intracranially at the times indicated in the figure legends, whereas those in groups 6, 7, and 8 received miR-301a Vivo-Morpholino (miR-301a-VM; 18 mg/kg). After 72 hours, the mice were euthanized, and brain samples were collected for qRT-PCR, plaque assay, and Western blotting analyses. For peripheral infections, WT mice were injected intraperitoneally with miR-301a-VM (12 mg/kg) or VM-NC (12 mg/kg), according to the manufacturer’s instructions, after 12, 24, and 48 hours of virus infection (3 × 105 PFU) as indicated. Three days later, liver and kidney tissues were collected to evaluate miRNA and mRNA expression.

Plaque assays

Viral titers in cell culture medium and brain samples were assessed by plaque assay. Plaque formation was performed on monolayers of PS cells. Cells were seeded in six-well plates at a density to give semiconfluent monolayers in about 18 hours. Monolayers were inoculated with serial dilutions of culture medium or virus sample (obtained from brain) in MEM containing 1% fetal calf serum and incubated for 1 hour at 37°C with occasional shaking. The inoculum was removed by aspiration, and the monolayers were overlaid with MEM containing 5% FBS, 1% low–melting point agarose, and a cocktail of penicillin-streptomycin (0.5%). Plates were incubated at 37°C for 3 to 4 days until plaques were visible. To enable counting of the plaques, we stained the cell monolayer with crystal violet after the cells were fixed with 10% PFA (80).

Statistical analysis

All experiments were performed in triplicate unless otherwise indicated. To analyze statistical difference between the two groups, we used a Student’s two-tailed unpaired t test. Comparisons involving multiple groups were assessed by one-way ANOVA followed by Bonferroni’s post hoc test, whereas differences between multiple groups influenced by two factors were evaluated by two-way ANOVA followed by the Holm-Sidak method. P < 0.05 was considered statistically significant. The results are expressed as means ± SD, and graphs were prepared with KyPlot (version 2.0 beta 13) and SigmaPlot 11.0.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/10/466/eaaf5185/DC1

Fig. S1. Expression profile of miR-301 family members during JEV infection.

Fig. S2. miR-301a inhibits the antiviral type I IFN response and increases viral replication.

Fig. S3. SOCS5 and IRF1 are targets of miR-301a.

Fig. S4. Suppression of SOCS5 and IRF1 protein production is reversed by inhibition of miR-301a.

Fig. S5. Loss of SOCS5 protein early during JEV infection.

Fig. S6. JEV induces miR-301a expression through the activation of NF-κB.

Fig. S7. In vivo dose standardization and treatment with miR-301a-VM in JEV-infected mice.

Fig. S8. Aberrant regulation of miR-301a expression during JEV infection in mouse liver and kidney.

Fig. S9. The regulation of miR-301a expression is not ubiquitous in the peripheral organs of JEV-infected mice.

Fig. S10. The role of miR-301a early in JEV infection.

Fig. S11. Examples of Western blots used for densitometric analysis.

Fig. S12. Further examples of Western blots used for densitometric analysis.

Table S1. Potential targets of mouse miR-301a.

Table S2. Prediction of transcription factor binding sites within the miR-301a promoter by PROMO software.

Table S3. Primer sequences.

REFERENCES AND NOTES

Acknowledgments: We are grateful to P. K. Roy [National Brain Research Centre (NBRC)] for helping with statistical analysis of the data. We are obliged to S. K. Sharma (NBRC), E. Sen (NBRC), D. Schubert (Salk Institute), S. Levison (Rutgers University), A. Krishnan (Institute of Molecular Medicine), G. R. Medigeshi (Translational Health Science and Technology Institute), D. Chattopadhyay (Amity University), and P. Chattopadhyay (All India Institute of Medical Science) for providing cell lines and plasmids. We acknowledge K. Dutta (University of Laval, Quebec, Canada) for critically reading the manuscript and giving insightful comments. We thank A. Nazmi (University of Gothenburg, Sweden) for his valuable suggestions. We are highly thankful to M. Dogra for his technical assistance. Funding: The study was supported by the NBRC Core Fund and a Tata Innovation Fellowship (BT/HRD/35/01/02/2014) to A.B. from the Department of Biotechnology, Ministry of Science and Technology, Government of India. B.H. is supported by a postdoctoral fellowship from the Indian Council of Medical Research, Government of India (80/901/2014-ECD-I). Author contributions: A.B. and B.H. designed the study, generated the hypothesis, analyzed the data, and wrote the manuscript. B.H. performed all of the experiments, interpreted the results, and executed statistical analysis. K.L.K. helped in handling and performing all animal experiments. All of the authors read and approved the final manuscript. Competing interests: The authors declare that they have no competing interests.
View Abstract

Related Content

Navigate This Article