key: cord-0907026-a29zl9pp
authors: Lin, Xian; Yu, Shiman; Ren, Peilei; Sun, Xiaomei; Jin, Meilin
title: Human microRNA‐30 inhibits influenza virus infection by suppressing the expression of SOCS1, SOCS3, and NEDD4
date: 2019-12-26
journal: Cell Microbiol
DOI: 10.1111/cmi.13150
sha: 3ad35815b1c3ec5d6b964d6d6b9465ebd555a0da
doc_id: 907026
cord_uid: a29zl9pp

Influenza A virus (IAV) has evolved multiple mechanisms to compromise type I interferon (IFN) responses. The antiviral function of IFN is mainly exerted by activating the JAK/STAT signalling and subsequently inducing IFN‐stimulated gene (ISG) production. However, the mechanism by which IAV combat the type I IFN signalling pathway is not fully elucidated. In this study, we explored the roles of human microRNAs modulated by IAV infection in type I IFN responses. We demonstrated that microRNA‐30 (miR‐30) family members were downregulated by IAV infection. Our data showed that the forced expression of miR‐30 family members inhibited IAV proliferation, while miR‐30 family member inhibitors promoted IAV proliferation. Mechanistically, we found that miR‐30 family members targeted and reduced SOCS1 and SOCS3 expression, and thus relieved their inhibiting effects on IFN/JAK/STAT signalling pathway. In addition, miR‐30 family members inhibited the expression of NEDD4, a negative regulator of IFITM3, which is important for host defence against influenza viruses. Our findings suggest that IAV utilises a novel strategy to restrain host type I IFN‐mediated antiviral immune responses by decreasing the expression of miR‐30 family members, and add a new way to understand the mechanism of immune escape caused by influenza viruses.

antagonistic factor of host IFN-β production via inhibition of the RIG-I-MAVS signalling pathway (Krug, 2015; Pachler & Vlasak, 2011; Rajsbaum et al., 2012) , and the PB1-F2 protein was identified to block IFN-β induction by targeting MAVS (Varga, Grant, Manicassamy, & Palese, 2012; Varga et al., 2011) . Influenza virus can also modulate the antiviral response at the JAK/STAT pathway level. The HA protein was shown to antagonise the IFN response by inducing the degradation of type I IFN receptor 1 (IFNAR1) (Xia et al., 2015) . The NS1 F I G U R E 1 Influenza virus infection downregulates miR-30 expression. A549 cells were infected with 0.1 MOI of H5N1 influenza virus, and the expression levels of miR-30a/d/e (a), miR-30b (b), and miR-30c (c) were detected at 12, 24, 36, and 48 hpi by qRT-PCR. (d) A549 cells were infected with heat-inactivated H5N1, and after 36 hr, miR-30 expression was detected by qRT-PCR. A549 cells were transfected with 200 ng poly(I:C), IFN-β and MX1 (e), and miR-30 (f) were tested at the indicated time points post-transfection by qRT-PCR. A549 cells were infected by VSV with an MOI of 0.1, IFN-β and MX1 (g), and miR-30 (h) were detected at the indicated time points post-infection by qRT-PCR. The values are shown as the mean and SD and are representative of three independent experiments. The data in (a-c) were analysed using two-way ANOVA; data in (d) were analysed using Student's t test. The values are shown as the mean and SD and are representative of three independent experiments. #: non-significant, ***p < .001; **p < .01; *p < .05 protein was demonstrated to negatively regulate JAK/STAT signalling by elevating SOCS1 and SOCS3 expression (Jia et al., 2010) . In addition, influenza virus can restrict JAK/STAT pathway activation via the NF-κB-dependent induction of SOCS3 expression (Pauli et al., 2008) .

MicroRNAs (miRNAs) are a class of noncoding small RNA molecules, which can regulate gene expressions by binding to the 3 0 -untranslated region (3 0 UTR) of targeted mRNA to induce degradation or suppress translation (Ambros, 2004) . Growing evidence has shown that miRNAs are involved in regulating development, apoptosis, host immunity, and viral infection (Bartel, 2004; Cullen, 2013; Wienholds & Plasterk, 2005) .

MiRNAs can regulate viral infection by modulating the antiviral immune response at multiple levels. MiR-146 was shown to inhibit type I IFN production by regulating expressions of IRAK1, IRAK2, and TRAF6 (Hou et al., 2009) , and miR-3570 was found to modulate MAVS expression following rhabdovirus infection (Xu, Chu, Cui, & Bi, 2018) . Many studies have demonstrated the significant effect of miRNAs on modulating influenza virus infection. For examples, miR-33a was suggest to disrupt influenza A virus (IAV) replication by targeting APCN1 (Hu et al., 2016) ; miR-34a contributed to influenza virus-mediated apoptosis by binding to BAX (Fan & Wang, 2016) ; and miR-let-7c targeted and inhibited M1 expression of the H1N1 influenza virus (Ma et al., 2012) . Although several studies linked miRNAs to type I IFN production mediated by RIG-I in influenza virus infection (miR-144, miR-485, and miR-136) (Ingle et al., 2015; Rosenberger et al., 2017; Zhao et al., 2015) , the underlying regulation of the type I IFN-mediated antiviral response by miRNAs during influenza virus infection remains unclear.

In this study, we found that microRNA-30 (miR-30) family members were downregulated upon IAV infection. Further analysis demonstrated that miR-30 overexpression suppressed SOCS1, SOCS3, and NEDD4 and thus positively regulated the type I IFN signalling pathway and IFITM3 expression, thereby inhibiting influenza virus infection. For the first time, we identified miR-30 as a positive regulator of the antiviral response, and our data revealed that influenza virus could disrupt the host antiviral immune response by downregulating miR-30 expression, which may be a new mechanism of influenza virus inhibition of the host antiviral immune response.

F I G U R E 2 MiR-30 suppresses influenza virus replication. A549 cells were transfected with 80 nm miR-30a/b/c mimics (a-c) or 100 nm inhibitors (d-f); 24 hr later, cells were infected with 0.2 MOI of H5N1 influenza virus (HM/H5N1). At the indicated time post-infection, the RNA, total protein, and the supernatant were collected for NP mRNA (a and d), NP protein (b and e), and viral titre (c and f) detection by qRT-PCR with GAPDH as housekeeping gene, western blot, and plaque assay analyses, respectively. NC, negative control; IN, inhibitor. The values are shown as the mean and SD and are representative of three independent experiments. Data were analysed using two-way ANOVA; ***p < .001; **p < .01; *p < .05 2 | RESULTS

Many studies have investigated microRNAs expression upon IAV infection. Previous studies suggested that miR-30 family members (miR-30) expression could be modulated by IAV challenge (Y. Y. Li et al., 2011; Tambyah et al., 2013) . We hypothesised that miR-30 may exert some important effects on influenza virus infection.

Human miR-30 family contains five members (miR-30a-e), which are broadly conserved. We first assessed miR-30 expression in A549 cells infected with H5N1 influenza virus in a time-course assay. Sequence analysis indicated that only one or two nucleotides differed among miR-30a, miR-30d, and miR-30e, which have large differences in their 3 0 terminals compared with miR-30b and miR-30c, while miR-30b and miR-30c have significant differences in their 3 0 terminal sequences ( Figure S1 ). Therefore, each member of miR-30 family could be analysed using three specific primers: miR-30a (which cross-reacts with miR-30d and miR-30e), miR-30b, and miR-30c. Our results showed that miR-30a/d/e, miR-30b, and miR-30c were remarkably decreased in A549 cells in a time-dependent manner following H5N1 infection (Figure 1a -c). However, their expression was not substantially changed at 36 hr post-infection (hpi) when A549 cells were treated with heat-inactivated viruses (Figure 1d ). MiR-30 expression was also decreased in H1N1 and H9N2 subtype influenza virusinfected A549 cells compared with uninfected cells ( Figure S2 ). Poly(I: C), a double-stranded RNA analogue that can activate type I IFN response, was used as a stimulus to investigate miR-30 expression.

Although poly(I:C) significantly increased IFN-β and MX1 expression 

To investigate the biological effect of miR-30 on viral infection in host cells, we inspected the role of miR-30a/b/c in H5N1 influenza virus replication in A549 cells. As shown in Figure 2a (Figure 2f ). Given the high sequence similarity of miR-30a, miR-30d, and miR-30e, we did not determine the effect of miR-30d and miR-30e on influenza virus infection, but we hypothesised that they have the same role as miR-30a/b/c in influenza virus infection. These data suggested that miR-30 could suppress influenza virus replication, while miR-30 inhibition facilitated influenza virus infection.

F I G U R E 3 Influenza virus genomic RNA is not the target of miR-30. (a) Predicted interactions between miR-30b/c and the influenza virus M gene using RNA22 V2. (b) Effects of miR-30a/b/c/d/e mimics on the expression of the firefly luciferase gene from the pmirGLO reporter constructs containing the putative miR-30b/c binding sites from the M gene (pmirGLO-M) and 3 0 UTR of human p53 (pmirGLO-p53). First, 293T cells were co-transfected with reporter constructs together with miR-30a/b/c/d/e mimics or NC. After 24 hr, cells were lysed, and luciferase activities were measured. The luciferase activity was normalised to the Renilla luciferase activity, and the data are expressed relative to that of the NC. The values are shown as the mean and SD and are representative of three independent experiments. Data were analysed using Student's t test. ***p < .001; **p < .01; *p < .05 2.3 | MiR-30 does not target putative influenza virus genomic RNA Host miRNAs can affect RNA virus replication either by modulating host gene expression or by directly targeting viral RNA for degradation. To determine whether miR-30 inhibits influenza virus replication by directly targeting viral RNA, we performed computational analysis of the potential binding sites in viral RNA for miR-30. Analysis using RNA22 V2 miRNA detection software identified a potential binding site of miR-30b and miR-30c at position 952-975 of the M gene sequence from HM/H5N1 in this study (Figure 3a ). The prediction was then experimentally tested by a luciferase reporter assay.

The nucleotide sequence from the M gene of HM/H5N1 was inserted into the pmirGLO vector. Then, 293T cells were transfected F I G U R E 4 Legend on next page. F I G U R E 4 MiR-30 targets the 3 0 UTRs of SOCS1 and SOCS3. (a) Predicted target sites of miR-30a/b/c in the 3 0 UTRs of SOCS1 and SOCS3. Nucleotides in blue are seed regions of miR-30a/b/c; nucleotides in red are complementary to the miR-30a/b/c seed region. (b and c) Effects of miR-30a/b/c mimics on expression of the firefly luciferase gene from reporter constructs containing the SOCS1 and SOCS3 3 0 UTRs or SOCS1 and SOCS3 mutated 3 0 UTRs. First, 293T cells were transfected with NC or miR-30a/b/c, together with the SOCS1 or SOCS3 3 0 UTR (3 0 UTR WT) or 3 0 UTR mutant (3 0 UTR Mut), and 24 hr after transfection, cells were harvested to assess luciferase activity. (d and e) MiR-30a/b/c inhibit the SOCS1 and SOCS3 3 0 UTRs in a dose-dependent manner. First, 293T cells were transfected with 30, 60, or 90 nM miR-30a/b/c mimics, together with SOCS1 or SOCS3 3 0 UTR, and 24 hr after transfection, cells were lysed for luciferase activity analysis. (f and g) The 293T cells were transfected with 100 nM miR-30a/b/c inhibitor or NC inhibitor (NC-inhibitor), together with SOCS1 or SOCS3 3 0 UTR. Twenty-four hours later, the cells were lysed for luciferase activity analysis. A549 cells were transfected with 80 nm miR-30a/b/c mimics or 100 nM miR-30a/b/c inhibitors, and 36 hr after transfection, the mRNA and protein levels of SOCS1 and SOCS3 were determined by qRT-PCR (h and i) and western blot analyses (j and k), respectively. The values are shown as the mean and SD and are representative of at least three independent experiments. Data were analysed using Student's t test. ***p < .001; **p < .01; *p < .05 F I G U R E 5 MiR-30 targets NEDD4. (a) Analysis of NEDD4 3 0 UTR potential binding sites for miR-30c; three putative binding sites were predicted (942-949, 960-967, and 1116-1122 in the 3 0 UTR of NEDD4 mRNA), and nucleotides in red show mutations of the 3 0 UTR of NEDD4, which is complementary to the seed region of miR-30c. (b) Effects of miR-30a/b/c on firefly luciferase activity from reporter constructs containing NEDD3 3 0 UTR WT or sequences with mutations in three putative positions (NEDD4-Mut1, Mut2, and Mut3). First, 293T cells were co-transfected with miR-30a/b/c mimics and NEDD4 3 0 UTR or 3 0 UTR mutants, and 24 hr later, the cells were lysed for luciferase activity analysis. (c) Dose-dependent inhibitory effects of miR-30a/b/c on the NEDD4 3 0 UTR. First, 293T cells were transfected with 30, 60, or 90 nM miR-30a/b/c, together with the NEDD4 3 0 UTR. Twenty-four hours later, luciferase activity was analysed. (d) MiR-30a/b/c inhibitors increased the NEDD4 3 0 UTR activity. First, 293T cells were transfected with 100 nM miR-30a/b/c inhibitors and NEDD4 3 0 UTR. Twenty-four hours later, luciferase activity was analysed. A549 cells were transfected with 80 nM miR-30a/b/c mimics or 100 nM miR-30a/b/c inhibitors, and 36 hr later, the mRNA and protein levels of NEDD4 were determined by qRT-PCR (e and f) and western blot analyses (g). The values are shown as the mean and SD and are representative of at least three independent experiments. Data were analysed using Student's t test. ***p < .001; **p < .01; *p < .05 with reporter constructs, together with miR-30a/b/c/d/e or NC mimics, followed by luciferase activity assays at 24 hr post-transfection. The result showed that miR-30 mimics transfection did not affect the luciferase expression of reporter constructs that contained the putative target site from the M gene ( Figure 3b ). As a positive control, miR-30 mimics transfection significantly reduced the luciferase expression of reporter constructs containing the 3 0 UTR of p53 compared with NC mimics transfection, which has been reported to be targeted by miR-30 family (J. . Therefore, the inhibitory effect of miR-30 on IAV proliferation is not due to the directly targeting viral genome RNA, and likely involves the miR-30-mediated regulation of host genes.

2.4 | SOCS1 and SOCS3 are potential targets of miR-30

To determine targeted host genes by miR-30, we conducted a genome-wide computational prediction. To minimise the number of false-positive predictions, we performed the analysis using four different prediction algorithms, and only interactions predicted by all the algorithms were chose. The algorithms applied are based on evolutionary conservation of sites (TargetScan and MirTarget2) and seed complementarity (microRNA.org and DIANA-MicroT). This led us to focus on SOCS1 and SOCS3, which were reported to be involved in the antiviral response (Alexander, 2002; Carow & Rottenberg, 2014; Pothlichet, Chignard, & Si-Tahar, 2008) .

Our analysis identified one putative binding site for miR-30 in the 3 0 UTR of human SOCS1 and SOCS3. To examine if miR-30 can directly interact with these sites and inhibit the expression of the targeted genes, we performed luciferase reporter assays. We first constructed reporter plasmids by cloning the 3 0 UTRs of the human SOCS1 and SOCS3 genes into the pmirGLO luciferase reporter vector, and mutant SOCS1 and SOCS3 3 0 UTR luciferase reporter vectors containing the target sequence mutations were generated as controls ( Figure 4a ). When the reporter vectors were co-transfected into 293T cells with miR-30a/b/c mimics, we observed that miR-30a/b/c mimics substantially decreased the luciferase activities of the SOCS1 3 0 UTR F I G U R E 6 MiR-30c promotes JAK-STAT signalling pathway activation. A549 cells were transfected with 80 nM NC or miR-30c mimics, and 36 hr later, cells were treated with 500 U/ml human rIFN-β for 30 min (a), transfected with 200 ng poly(I:C) for 4 hr (b), or infected with 5 MOI of H5N1 influenza virus for 4 hr (c). Cells were lysed, and cell extracts were analysed by western blot analysis. A549 cells were transfected with miR-30c mimics; 36 hr later, cells were treated with 100 U/ml rIFN-β for 3 hr, transfected with 100 ng/ml poly(I:C) for 5 hr, or infected with 3 MOI of H5N1 influenza virus for 8 hr. RNA was extracted for mRNA analyses of IFN-β, IL-6 (d), MX1, IFIT3, and CXCL10 (e) by qRT-PCR. Data are shown as the mean and SD and as one representative of three independent experiments. Data were analysed using Student's t test. #: nonsignificant; ***p < .001; **p < .01; *p < .05 To further determine whether miR-30a/b/c can target SOCS1 and SOCS3, we examined the expression of endogenous SOCS1 and SOCS3 in A549 cells transfected with mimics or inhibitors of miR-30a/b/c. As shown in Figure 4 , miR-30a/b/c mimic transfection significantly inhibited SOCS1 (Figure 4h ) and SOCS3 mRNA levels (Figure 4i) , whereas SOCS1 and SOCS3 expression could be restored by inhibitors. As expected, the SOCS1 and SOCS3 protein levels were also significantly decreased following miR-30a/b/c overexpression ( Figure 4j ). MiR-30a/b/c overexpression also significantly reduced SOCS1 and SOCS3 expression, including mRNA and protein level in HAECs ( Figure S5A ,B). In contrast, miR-30a/b/c inhibitors increased the SOCS1 and SOCS3 protein levels ( Figure 4k ). These data indicated that miR-30a/b/c functioned similarly in regulating SOCS1 and SOCS3 expression, which may be due to their identical seed region and high sequence similarity ( Figure S1 ). Above all, these results demonstrated that miR-30 negatively regulates SOCS1 and SOCS3 at the transcriptional level by targeting SOCS1 and SOCS3 3 0 UTRs.

Our in silico analysis also identified three putative seed-matched sequences (positions 942-949, 960-967, and 1116-1122) for miR-30 (with miR-30c as a representative) in the 3 0 UTR of NEDD4 mRNA ( Figure 5a ). NEDD4 is an E3 ubiquitin ligase that was reported to promote IAV infection by downregulating the antiviral protein IFITM3 expression (Chesarino et al., 2015) . Thus, we investigated whether miR-30 can target NEDD4. The luciferase reporter assay demonstrated that overexpression of miR-30a/b/c significantly decreased the luciferase activity of the NEDD4-WT 3 0 UTR (Figure 5b) , and dosedependent inhibitory effects were observed (Figure 5c ). In addition, the luciferase activity was slightly inhibited by miR-30a/b/c inhibitor transfection (Figure 5d ). When miR-30a/b/c were transfected together with the NEDD4 3 0 UTR containing a mutation in position 960-967 (NEDD4-Mut2), the inhibitory effect of miR-30a/b/c was not observed; miR-30a/b/c mimic transfection also substantially inhibited the luciferase activities of NEDD4-Mut1 and NEDD4-Mut3, which contained mutations at 942-949 and 1116-1122 in the predicted binding sites, respectively (Figure 5b) . Therefore, these results showed that miR-30 directly inhibited the NEDD4 3 0 UTR at position 960-967 of the predicted target site.

To further demonstrate that miR-30 targets NEDD4 mRNA, we assayed the NEDD4 mRNA levels in A549 cells following transfection of miR-30a/b/c mimics or inhibitors. We found that miR-30a/b/c mimic transfection significantly reduced NEDD4 mRNA (Figure 5e ), whereas the inhibitors significantly increased NEDD4 mRNA (Figure 5f ). Similarly, the protein level of NEDD4 was reduced by miR-30a/b/c mimics and was increased by miR-30a/b/c inhibitors ( Figure 5g) . Besides, miR-30a/b/c overexpression also significantly inhibited NEDD4 expression in HAECs ( Figure S5C ). These results demonstrated that miR-30a/b/c directly suppressed NEDD4 gene expression at the transcriptional level via targeting the NEDD4 3 0 UTRs.

2.6 | MiR-30 promotes activation of the JAK/STAT signalling pathway SOCS1 and SOCS3 were reported to block the JAK/STAT signal pathway (Alexander & Hilton, 2004; Croker, Kiu, & Nicholson, 2008) ; thus, we first examined the activation of JAK/STAT in response to type I IFN following miR-30 overexpression. Among miR-30a/b/c, miR-30c F I G U R E 7 MiR-30c is positively correlated with IFITM3 protein levels. (a) A549 cells were transfected with control vector (pCAGGS/ HA) or 0.5, 1, and 1.5 μg PCAGGS/HA-NEDD4, and 24 hr after transfection, cells were lysed for IFITM3 detection by western blot analysis. A549 cells were transfected with 0, 40, 60, and 100 nM miR-30c mimics or negative control mimics (NC) (b) or 0, 60, 80, and 100 nM miR-30c inhibitors (miR-30c IN) or control inhibitor (NC-IN) (c), and 24 hr later, the cells were lysed for IFITM3 and NEDD4 detection by western blot analysis. Data are shown as the mean and SD and as one representative of three independent experiments was selected as a representative to further investigate its function.

Exogenous addition of rIFN-β activated STAT1 in A549 cells, as shown by the enhanced p-STAT1 levels ( Figure 6a ). However, p-STAT1 was elevated by miR-30c mimic transfection compared with NC transfection (Figure 6a) . Similarly, we observed increased p-STAT1 levels after stimulation by poly(I:C) in miR-30c-overexpressing A549 cells compared with NC cells (Figure 6b) . We inspected the effect of the p-STAT1 level upon IAV infection in miR-30c-overexpressing A549 cells, and the data showed that miR-30c also promoted STAT1 activation (Figure 6c ).

The transcription of ISGs is initiated upon JAK/STAT pathway activation. Therefore, we investigated whether miR-30c affected the expression of ISGs. We first examined IFN-β and IL-6 mRNA production in A549 cells activated by poly(I:C) and influenza virus, and then we observed no remarkable difference in IFN-β and IL-6 expression when miR-30c was overexpressed (Figure 6d) . However, the induction of ISGs, such as MX1, IFIT3, and CXCL10, was increased in miR-30coverexpressing A549 cells treated by poly(I:C) transfection, influenza virus infection, and rIFN-β (Figure 6e) . These results indicate that miR-30c increases cellular sensitivity to type I IFN by triggering the type I IFN-mediated signalling pathway but does not significantly affect IFNβ production.

As NEDD4 was reported to decrease IFITM3 levels by ubiquitinating IFITM3, we hypothesised that miR-30 could modulate baseline IFITM3 levels. We first determined the effect of NEDD4 on turnover of IFITM3. As shown by the data, NEDD4 overexpression significantly decreased the IFITM3 levels ( Figure 7a ).

When miR-30c was overexpressed, increased IFITM3 and decreased NEDD4 protein levels were observed ( Figure 7b ); when A549 cells were transfected with miR-30 inhibitors, IFITM3 levels were decreased, accompanied by an increase in NEDD4 (Figure 7c ). These results suggested that miR-30c was positively correlated with the antiviral protein IFITM3.

F I G U R E 8 SOCS1, SOCS3, and NEDD4 facilitate influenza virus replication. A549 cells were transfected with SOCS1 (a), SOCS3 (b), or NEDD4 (c) overexpression vector, and 24 hr after transfection, the cells were infected with 0.2 MOI of H5N1 influenza virus (HM/H5N1). After 36 hr, the supernatant was collected, and the cells were lysed for viral titre and NP protein determination by plaque and western blot analyses, respectively. A549 cells were transfected with 60 nM SOCS1 or SOCS3 siRNAs (d) or NEDD4 siRNAs (e); 24 hr later, A549 cells were infected with 0.2 MOI of HM/H5N1, and after 36 hr, the supernatant was collected, and the cells were lysed for viral titre and NP protein determination. The western blot results are shown as one representative of three independent experiments, and viral titres data are expressed as the mean and SD of three independent experiments. Data were analysed using Student's t test. **p < .01; *p < .05 2.8 | Antiviral function of miR-30 is mainly through targeting SOCS1, SOCS3, and NEDD4

To elucidate the roles of SOCS1, SOCS3, and NEDD4 in the antiviral function of miR-30, we carried out experiments of SOCS1, SOCS3, and NEDD4 RNA interference knockdown and overexpression. We confirmed that overexpression of SOCS1, SOCS3, and NEDD4 effectively promoted IAV replication in A549 cells, as shown by the NP protein level and viral titre (Figure 8a-c) , whereas, depletion of SOCS1, SOCS3, and NEDD4 by siRNAs significantly inhibited IAV replication in A549 cells (Figure 8d,e) , which resembled the effect of miR-30c overexpression.

When SOCS1 and SOCS3 were knocked down at the same time, the IAV replication was also obviously inhibited compared with control; when the miR-30c was overexpressed in SOCS1-and SOCS3-depleted cells, IAV titre was further significantly reduced compared with control ( Figure 9a ). This could be due to that miR-30c overexpression could reduce NEDD4 expression. However, when SOCS1, SOCS3, and NEDD4 were all knocked down at the same time, miR-30c overexpression could not further suppress IAV replication ( Figure 9b) . The results showed that SOCS1, SOCS3, and NEDD4

were beneficial for influenza virus replication, and antiviral function of miR-30 is mainly through targeting SOCS1, SOCS3, and NEDD4.

In addition to cellular gene expression regulation, miRNAs have been demonstrated to be critical in virus infection, either by targeting viral gene transcripts or by regulating host antiviral response. Thus, the finding that viruses, including influenza virus, modulate the expression of specific cellular miRNAs is not surprising. The miR-30 family was shown to participate in development, apoptosis, and autophagy (Agrawal, Tran, & Wessely, 2009; Pan et al., 2013) .

Previously, several groups reported that the expression of miR-30 family members was reduced by influenza virus invasion (Y. Y. Li et al., 2011; Tambyah et al., 2013) . However, the exact function of miR-30 in influenza virus infection has not been determined. Herein, we showed that miR-30 family members were downregulated by IAV infection. The regulation of miR-30 family members by IAV infection appears to be independent of subtype, as the H1N1 and H9N2 subtypes also altered the miR-30 levels ( Figure S2 ). Interestingly, miR-30 was not affected by VSV infection and poly(I:C) stimulation in A549 cells. Previous studies reported that miR-30a and miR-30d were reduced upon West Nile virus infection in HEK293 cells, while miR-30c was increased by porcine reproductive and respiratory syndrome virus (PRRSV) in porcine alveolar macrophages . It is likely that miR-30 family members can be differentially modulated by various stimulants. Notably, because heatinactivated virus failed to alter the levels of miR-30 family members, and individual overexpression of a virus-coding protein did not influence miR-30 expression (data not shown), we speculated that the downregulation of miR-30 is dependent on influenza virus replication.

However, our attempt to discover the mechanism by which influenza virus infection decreases miR-30 levels failed. We hypothesised that influenza virus infection may affect the transcription of miR-30 by acting on promoters or through oxidative stress induced by virus infection because influenza virus infection was shown to lead to oxidative stress, which was suggested to reduce the expression of miR-30 family members in a previous study (J. . This mechanism requires further investigation.

We further demonstrated that miR-30 family members were inversely correlated with influenza virus proliferation. MiR-30 was reported to participate in various biological processes by targeting different host genes, such as p53, ATG5, BECIN1, and Xlim1/Lhx1 (Agrawal et al., 2009; Yu et al., 2012) . In this report, we demonstrated that miR-30 directly targets SOCS1, SOCS3, and NEDD4, which are negative regulators of the antiviral response F I G U R E 9 Antiviral function of miR-30c is mainly dependent on SOCS1, SOCS3, and NEDD4 expression. A549 cells were transfected with 50 nm siRNA of SOCS1 and 50 nm siRNA of SOCS3 at the same time (a) or transfected with 50 nm siRNA of SOCS1, 50 nm siRNA of SOCS3, and 50 nm siRNA of NEDD4 at the same time (b). After 12 hr, 80 nm miR-30c mimics or miRNA control (miR-NC) was transfected; 24 hr later, A549 cells were infected with 0.2 MOI of HM/H5N1, and after 30 hr, the supernatant were collected for viral titres determination, and cells were lysed for NP test by western blot. Data are shown as the mean and SD and as one representative of three independent experiments. Data were analysed using Student's t test. #: non-significant; **p < .01 pathway. Interestingly, a previous study reported that miR-30 can target JAK1 to positively modulate the IFN response in porcine Marc-145 and PAM cells . However, in our experiments, miR-30 did not target JAK1 in A549 cells, as shown by dual-luciferase reporter assays and real-time polymerase chain reaction (RT-PCR; data not shown). This discrepancy could be due to the different physiological conditions of the different cell types used in our experiment.

It seems that the main targets for one specific miRNA often variable under different physical or pathological conditions or in different cell types to play diverse functions.

Influenza virus has devised multiple strategies to attenuate the type I IFN-mediated antiviral response. Viral coding proteins (NS1, PB2, and PB1-F2) were reported to disrupt the signalling pathways of type I IFN synthesis. In addition to the blockade of IFN production, influenza virus can also inhibit JAK/STAT pathway activation to prevent ISG expression. The influenza virus NS1 protein can inhibit JAK/STAT signalling pathway activation by increasing SOCS1 and SOCS3 expression; in addition, influenza viral proteins antagonise the functions of ISG, such as PB2, which was reported to inhibit mouse Mx1 function (Stranden, Staeheli, & Pavlovic, 1993) 

Lipofectamine 2000 was obtained from Invitrogen (Carlsbad, CA). 

The 3 0 UTRs of human SOCS1, SOCS3, and NEDD4 were cloned from the genomic DNA of A549 cells and inserted into the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Promega). The mutant 3 0 UTRs of human SOCS1, SOCS3, and NEDD4 pmirGLO (SOCS1 3 0 UTR Mut, SOCS3 3 0 UTR Mut, and NEDD4 3 0 UTR Mut, respectively) were constructed by PCR using specific primers. Human SOCS1 and SOCS3 were cloned from the cDNA of A549 cells into pCMV3-N-Flag, and human NEDD4 was inserted into pCAGGS-HA. All primers for cloning are shown in Table S2 .

Lipofectamine 2000 and Lipo8000 were applied to transfect small RNAs and DNA constructs, respectively. For qRT-PCR determination of mRNA or miRNA, the total RNA of cells subjected to different treatments were extracted by Trizol, and a total of 1 μg RNA was reversely transcribed, as described previously (Lin et al., 2015) . The miRNAs and mRNA levels were normalised to that of U6 and GAPDH, respectively. The relative expressions of miRNAs and mRNAs were determined via SYBR Green-based qRT-PCR under an ABI ViiA 7 PCR system. The qRT-PCR primers used here are shown in Table 1 .

The miR-30 family member targets were predicted and selected using MicroRNA.org (Betel, Wilson, Gabow, Marks, & Sander, 2008) , DIANA-MicroT (Alexiou et al., 2010) , TargetScan (Lewis, Burge, & Bartel, 2005) , MirTarget2 (Wong & Wang, 2015) , and RNA22 V2 (Miranda et al., 2006) . MicroRNA and DIANA-MicroT analyses were based on seed complementarity, with target predict score threshold 0.7; TargetScan predicts targets according to aggregate P ct score of the longest 3 0 UTR isoform, which ranks based upon the confidence that targeting is evolutionary conserved; MirTarget2 analysis predicts gene targets with more than 60 target prediction score; RNA22 analysis is performed under sensitivity more than 63% and specificity more than 61%.

To validate the targeted genes by miR-30, luciferase reporter vectors containing wild-type SOCS1, SOCS3, and NEDD4 3 0 UTR or their mutants in seed regions were co-transfected with control mimics, miR-30a/b/c mimics, NC inhibitor, or miR-30a/b/c inhibitors into 293T cells. Twenty-four hours later, cells were lysed, and supernatants were collected for luciferase activity test through the Dual-Luciferase test reagent provided by Promega. All obtained luciferase values were normalised against those of the Renilla luciferase control. 

A549 cells that were transfected with oligonucleotides or plasmids for 24 hr were challenged with influenza viruses at the indicated multiplicity of infection (MOI). After 1 hr of viral adsorption at 37 C, the cells were washed twice with phosphate buffered saline (PBS) and the supernatants were replaced with F12 medium containing 1% FBS. After that, the cells were cultured in a 37 C humidified incubator with 5% CO 2 . At the indicated time post-infection, the supernatants were collected, and the viral titres were analysed by plaque assays in MDCK cells. Buffered Saline Tween (TBST) buffer for about 1 hr at room temperature, washed once for 5 min, and then incubated with indicated primary antibodies for 2 hr at room temperature. After incubating with HRP-conjugated anti-rabbit or anti-mouse secondary antibodies for 1 hr at room temperature, signals were visualised by ECL reagent (Advansta, San Jose, CA). The densitometry of protein expression was quantified relative to GAPDH by ImageJ software; in the densitometry analysis, control was set as 1.00.

The results are expressed as the mean ± SD. Data analysis was performed using Student's t test or two-way analysis of variance (ANOVA). Differences between means were considered significant at p values of <.05.

The miR-30 miRNA family regulates Xenopus pronephros development and targets the transcription factor Xlim1/Lhx1

Suppressors of cytokine signalling (SOCS) in the immune system

The role of suppressors of cytokine signaling (SOCS) proteins in regulation of the immune response

The DIANA-mirExTra web server: From gene expression data to microRNA function

The functions of animal microRNAs

MicroRNAs: Genomics, biogenesis, mechanism, and function

The microRNA.org resource: Targets and expression

Interferons at age 50: Past, current and future impact on biomedicine

SOCS3, a major regulator of infection and inflammation

E3 ubiquitin ligase NEDD4 promotes influenza virus infection by decreasing levels of the antiviral protein IFITM3

SOCS regulation of the JAK/STAT signalling pathway

How do viruses avoid inhibition by endogenous cellular microRNAs?

MicroRNA 34a contributes to virus-mediated apoptosis through binding to its target gene Bax in influenza A virus infection

A three-dimensional model of suppressor of cytokine signalling 1 (SOCS-1)

Micro-RNA-146a feedback inhibits RIG-I-dependent Type I IFN production in macrophages by targeting TRAF6, IRAK1, and IRAK2

MicroRNA-33a disturbs influenza A virus replication by targeting ARCN1 and inhibiting viral ribonucleoprotein activity

The microRNA miR-485 targets host and influenza virus transcripts to regulate antiviral immunity and restrict viral replication

Influenza virus non-structural protein 1 (NS1) disrupts interferon signaling

Functions of the influenza A virus NS1 protein in antiviral defense

Suppressors of cytokine signalling: SOCS

Stats: Transcriptional control and biological impact

Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets

miR-30 regulates mitochondrial fission through targeting p53 and the dynaminrelated protein-1 pathway

MicroRNA expression and virulence in pandemic influenza virus-infected mice

Differential microRNA expression and virulence of avian, 1918 reassortant, and reconstructed 1918 influenza A viruses

Investigation of pathogenesis of H1N1 influenza virus and swine Streptococcus suis serotype 2 co-infection in pigs by microarray analysis

Cellular microRNA let-7c inhibits M1 protein expression of the H1N1 influenza A virus in infected human lung epithelial cells

A pattern-based method for the identification of microRNA binding sites and their corresponding heteroduplexes

The JAK-STAT signaling pathway: Input and output integration

Influenza C virus NS1 protein counteracts RIG-I-mediated IFN signalling

MiR-30-regulated autophagy mediates angiotensin II-induced myocardial hypertrophy

Influenza A virus inhibits type I IFN signaling via NF-kappaB-dependent induction of SOCS-3 expression

Cutting edge: Innate immune response triggered by influenza A virus is negatively regulated by SOCS1 and SOCS3 through a RIG-I/IFNAR1-dependent pathway

Species-specific inhibition of RIG-I ubiquitination and IFN induction by the influenza A virus NS1 protein

miR-144 attenuates the host response to influenza virus by targeting the TRAF6-IRF7 signaling axis

Function of the mouse Mx1 protein is inhibited by overexpression of the PB2 protein of influenza virus

microRNAs in circulation are altered in response to influenza A virus infection in humans

Acetylation-dependent signal transduction for type I interferon receptor

The receptor of the type I interferon family

Influenza virus protein PB1-F2 inhibits the induction of type I interferon by binding to MAVS and decreasing mitochondrial membrane potential

The influenza virus protein PB1-F2 inhibits the induction of type I interferon at the level of the MAVS adaptor protein

MicroRNA function in animal development

miRDB: An online resource for microRNA target prediction and functional annotations

Hemagglutinin of influenza A virus antagonizes type I interferon (IFN) responses by inducing degradation of type I IFN receptor 1

Inducible microRNA-3570 feedback inhibits the RIG-I-dependent innate immune response to rhabdovirus in teleost fish by targeting MAVS/IPS-1

microRNA 30A promotes autophagy in response to cancer therapy

MicroRNA-30c modulates type I IFN responses to facilitate porcine reproductive and respiratory syndrome virus infection by targeting JAK1

Identification of cellular microRNA-136 as a dual regulator of RIG-I-mediated innate immunity that antagonizes H5N1 IAV replication in A549 cells

Human microRNA-30 inhibits influenza virus infection by suppressing the expression of SOCS1, SOCS3, and NEDD4

The authors declare no conflict of interest.

M.J. and X.L. designed the study; X.L., S.Y., P.R., and X.S. conducted the experiments; X.L. wrote the manuscript; M.J. and X.L. analysed the data.

https://orcid.org/0000-0003-2811-562X