LncRNA NRAV promotes respiratory syncytial virus (RSV) replication by regulating vesicle transportation via targeting miR-509-3p/Rab5c axis.

Respiratory syncytial virus (RSV) is an enveloped RNA virus which is responsible for approximately 80% of lower respiratory tract infections in children. Current lines of evidence have supported the functional involvement of long non-coding RNA (lncRNA) in many viral infectious diseases. However, the overall biological effect and clinical role of lncRNAs in RSV infection remain unclear. In this study, lncRNAs related to respiratory virus infection were obtained from lncRNA database. And we collected 144 clinical sputum specimens to identify lncRNAs related to RSV infection. qPCR detection indicated that the expression of lncRNA NRAV in RSV-positive patients was significantly lower than that in uninfected ones, but lncRNA PRINS, NEAT1, and NEST showed no difference in vivo and in vitro Meanwhile, over expression of Negative Regulator of Antiviral Response (NRAV) promoted RSV proliferation in A549 and BEAS-2B cells, and vice versa, indicating that the down-regulation of NRAV was part of the host antiviral defense. RNA FISH confirmed that NRAV was mainly located in the cytoplasm. Through RNA sequencing we found that Rab5c, which is a vesicle transporting protein, showed the same change trend as NRAV. Subsequent investigation revealed that NRAV was able to favor RSV production indirectly by sponging miR-509-3p so as to release Rab5c and facilitate vesicle transportation. The study provides a new insight into virus-host interaction through non-coding RNA, which may contribute to exploring potential antivirus target for respiratory virus.IMPORTANCE Mechanism of interaction between RSV and host non-coding RNAs has not been fully understood. In this study, we found that the expression of lncRNA NRAV was reduced in RSV-infected patients, and over-expression of NRAV facilitated RSV production in vitro, suggesting that the reduction of NRAV in RSV infection was part of the host antiviral response. We also found that NRAV competed with vesicle protein Rab5c for miR509-3p in cytoplasm, to promote RSV vesicle transport and accelerate RSV proliferation, thereby improving our understanding of the pathogenic mechanism of RSV infection.

R espiratory syncytial virus (RSV), an enveloped single-negative-strand RNA virus (1), is one of the main pathogens causing acute respiratory infection (2) and asthma (3). Globally, RSV infection accounts for approximately 3.2 million hospitalizations annually among children younger than 5 years (4). Despite years of research, the mechanism of interaction between RSV and host cells has not been fully elucidated, so much so that RSV infection is still prevalent, and a side effect-free vaccine has not been successfully developed (5).
Transcriptome studies and genomic microarrays have shown that more than 90% of genomes are transcribed to noncoding RNAs (6). Depending on its distinctive long transcript, secondary structure, and subcellular location (7), long noncoding RNA (lncRNA) can regulate gene expression at different levels (8). Numerous reports have shown that lncRNAs play important roles in the process of viral infection (9,10). For example, following human immunodeficiency virus (HIV) infection, late protein Vpu induces host cells to express lncRNA noncoding repressor of nuclear factor of activated T cells (NFAT) (NRON) at a high level; afterward, NRON binds to the NFAT scaffold complex, resulting in transcriptional inhibition of HIV (11). Hepatitis C virus (HCV) can upregulate the expression of lncRNA Lethe, and Lethe inhibits the transcriptional activity of NF-B, leading to the promotion of HCV replication (12).
As one of the leading pathogens in the respiratory tract, RSV will inevitably cause changes in the expression of some noncoding RNAs in host cells during infection. Here, we explored the probable host lncRNAs involved in RSV infection. Based on the results from clinical samples and a subsequent RNA-seq assay, we focused on lncRNA negative regulator of antiviral response (NRAV) and revealed the cross talk among NRAV, microRNA miR-509-3p, and Rab5c in cytoplasm, shedding new light on the regulation of noncoding RNA in RSV infection.

RESULTS
NRAV was downregulated in RSV-infected patients and mainly located in the cytoplasm. To identify host lncRNAs that were influenced by RSV infection, we retrieved four lncRNAs associated with RNA virus or respiratory virus infection in ln-cRNAdb database, i.e., NEAT1 (GenBank accession number NC_000011.10), PRINS (GenBank accession number HG975433.1), NeST (GenBank accession number NR_104124.1), and NRAV (GenBank accession number NR_038854). The relative expression levels of these lncRNAs were determined by reverse transcription-quantitative PCR (qRT-PCR) in 144 clinical sputum specimens (RSV -, n ϭ 75, RSV ϩ , n ϭ 69) ( Table 1). The results showed that NRAV expression in the RSV-infected group was lower than that in the uninfected groups (Fig. 1A), while NEAT1, PRINS, and NeST were not detected in these sputum samples.
We then observed lncRNA expression levels in vitro. Regarding the different cell types in sputum, Guiot reported that there were macrophages (49%), squamous or epithelial cells (23%), lymphocytes (3%), eosinophils (1%) and so on, which could be divided into immune cells and epithelial cells (13). Considering that the proportion of lymphocytes and eosinophils in sputum is much lower than other cell types, we selected macrophages and epithelial cells to evaluate NRAV levels. For epithelial cells, A549 or BEAS-2B cells were infected with RSV for 72 h. Total RNA was collected at indicated time points since infection, and qRT-PCR was performed. NRAV was reduced in A549 (Fig. 1B) or BEAS-2B (Fig. 1C) cells during RSV infection, but PRINS, NEAT1, and NeST showed no difference ( Fig. 1E to G). Similarly, for macrophages, the M⌽-THP-1cells (14) were infected and detected in the same manner. However, the expression of NRAV was decreased slightly in M⌽-THP-1cells, and the difference was not statistically significant (Fig. 1D). As a result, NRAV was significantly downregulated in epithelial cells after RSV infection, and we mainly focused on A549 and BEAS-2B in vitro for the following study.
Subsequently, we explored the conservation of NRAV in the LNCipedia database (https://lncipedia.org/db/transcript/NRAV:3), and NRAV shared no locus conservation with mouse. In addition, RNA FISH was performed in A549 to determine the subcellular location of NRAV. Data showed that NRAV was located in both the cytoplasm and nucleus but mainly in the cytoplasm (Fig. 1H).
NRAV promoted RSV replication in vitro. The full length of the NRAV sequence has already been reported (15). To evaluate the influence of NRAV on RSV proliferation, we constructed pcDNA3.1-NRAV containing the whole length of NRAV to express NRAV ectopically in vitro, and qRT-PCR was performed to determine NRAV expression ( Fig. 2A). Both NRAV-overexpressed cells and the control group were infected by RSV (multiplicity of infection [MOI] ϭ 3), and cell lysates were harvested at indicated time points to evaluate RSV replication by virus plaque assay (Fig. 2B), immunoblotting (Fig. 2C), and qRT-PCR ( Fig. 2D and E). Results implied that overexpression of NRAV induced higher RSV titer, more F protein expression, and higher RSV gene levels.
Knockdown of NRAV weakened RSV proliferation in vitro. To further validate the functionality of NRAV on RSV proliferation, we generated NRAV-specific small interfering RNA (siRNA) to silence NRAV ectopically in vitro, and qRT-PCR was performed to determine NRAV expression (Fig. 3A). Both NRAV knocked-down cells and the control group were infected by RSV (MOI ϭ 3). Cell lysates were harvested at indicated time points to analyze RSV proliferation by virus plaque assay (Fig. 3B), immunoblotting (Fig.  3C), and qRT-PCR ( Fig. 3D and E). The results showed that downregulation of NRAV induced lower RSV titer, less F protein expression, and lower RSV gene levels.
NRAV was positively correlated with Rab5c and was a direct target of miR-509-3p. To investigate which mRNAs were affected by NRAV expression, genome-wide mRNA sequencing was performed in cells as described below. In the overexpression group, pcDNA3.1-NRAV was transfected into A549 cells (n ϭ 3), while pcDNA3.1 empty vector was transfected as the control (n ϭ 3). In the knockdown group, NRAV siRNA (si-NRAV) (n ϭ 1) and negative-control siRNA (si-control) (n ϭ 1) were transfected into A549 cells. Data implied that there were 15 upregulated mRNAs in NRAV-overexpressed cells compared with the controls (see Table S1 in the supplemental material), and a total of 130 downregulated mRNAs were found in NRAV knocked-down cells (Table S2) (P Ͻ 0.05; fold change, Ն2). Interestingly, in the overlap of the two comparisons, we found only one molecule, Rab5c (GenBank accession number NM_201434.3), which was in the same trend as NRAV (Fig. 4A). To further confirm the mRNA sequence result, Rab5c levels were determined by qRT-PCR in clinical sputum specimens (RSV -75, RSV ϩ 69), which were the same samples used in Fig. 1A, and Rab5c expression in RSVinfected samples was lower than that in uninfected ones (Fig. 4B). Moreover, we examined Rab5c expression by qRT-PCR (Fig. 4C) or immunoblotting ( Fig. 4D and E) in vitro, and RSV infection induced lower Rab5c expression than in the controls. The above results indicated that Rab5c was positively correlated with NRAV during RSV infection. As an early endosomal marker protein, Rab5 is mainly involved in cellular vesicle transport and endocytosis (16,17). The Rab5 family contains Rab5a, b, and c, which share similar function and location (18,19). It has been reported that RSV entered respiratory epithelial cells in the form of macropinocytosis, which was then transformed by a Rab5-positive macropinosome. The RSV expression level was decreased significantly when the cells were transfected with Rab5 dominant negative mutant (20), suggesting that the Rab5 molecule was directly involved in the intracellular transmission of RSV virus. Furthermore, we performed an immunofluorescence assay, and confocal microscopy showed that RSV could be colocalized with Rab5positive vacuoles in the cytoplasm (Fig. 4F), which was consistent with Krzyzaniak's report (20).
The function of lncRNA is closely related to its subcellular location (7). Based to the results showing that NRAV was mainly located in the cytoplasm and influenced Rab5c expression in the same trend, we speculated that Rab5c might be the target molecule of NRAV in respiratory epithelial cells.
Numerous cytoplasmic lncRNAs have been reported to perform as competing endogenous RNAs (ceRNAs) by binding common microRNAs (miRNAs) (21). We analyzed the target miRNA of NRAV and Rab5c via bioinformatics (lncrnadb and RegRNA 2.0). The results suggested that both NRAV and Rab5c had the potential to bind miR-509-3p; that is, they might be connected by the same miRNA response element (miR-509-3p) in cytoplasm. What's more, miR-509-3p had been reported to target the 3= UTR of Rab5c (22).
Additionally, we evaluated the expression of miR-509-3p in clinical sputum specimens ( Fig. 4G) which were the same samples used in Fig. 1A, and the expression of miR-509-3p was further examined in A549 and BEAS-2B cells (Fig. 4H). Data showed that infection of RSV induced higher expression of miR-509-3p in vivo and in vitro. To investigate the direct association between NRAV and miR-509-3p, we generated a biotin-labeled NRAV probe and a negative-control (NC) probe. A549 cell lysates were incubated with the probes overnight. The specific binding was confirmed by affinity pulldown of endogenic miR-509-3p and miR-653-5p (used as a control). MiR-509-3p and miR-653-5p were then extracted and assessed by qRT-PCR. Meanwhile, the NRAV level was examined by qRT-PCR to evaluate the efficacy of NRAV pulldown (Fig. 4I). Data showed that endogenic miR-509-3p, other than miR-653-5p, could be pulled down by biotin-labeled NRAV.
To further validate the binding, miR-509-3p mimics and inhibitor were generated and verified by qRT-PCR in vitro ( Fig. 4J and K). Luciferase reporters containing NRAV (1,471 to 1,530 bp) wild-type or mutated miR-509-3p binding sites were constructed (Fig. 4L) and cotransfected with miR-509-3p mimics or the mimic control. It was indicated that overexpression of miR-509-3p weakened the luciferase activities of wild-type pmirGLO-NRAV but not those of mutant type or empty vector ( Fig. 4M and N). We constructed pcDNA3.1-NRAVmut containing full-length wild-type NRAV with point mutations in miR-509-3p binding sites to express NRAVmut ectopically. Expression of miR-509-3p was decreased in the pcDNA3.1-NRAV (wild-type) group but not in empty vector or mutant vectors (Fig. 4O).
The results implied that NRAV was positively correlated with Rab5c in vivo and in vitro, and NRAV was physically associated with miR-509-3p.
On the other hand, in the silencing experiment, cells were transfected with si-control or si-NRAV. In the rescue group, si-NRAV and miR-509 inhibitor were cotransfected into cells. From the next day on, the cells were infected by RSV (MOI ϭ 3) for 48 h. Data suggested that silencing NRAV reduced RSV production, and miR-509-3p inhibitor abolished this decrease in RSV gene expression ( Fig. 5F and G), F protein level ( Fig. 5H and I), and viral titer (Fig. 5J). Accordingly, the results revealed that miR-509-3p was able to reverse the effect of NRAV on RSV replication through miR-509-3p/NRAV interaction.
Subsequently, we examined such a relationship through immunoblotting. Overexpression of NRAV, but not NRAVmut, increased Rab5c expression, and miR-509-3p mimics reversed the high Rab5c expression in NRAV overexpressed cells. In NRAVmut overexpressed cells, miR-509-3p mimics made Rab5c expression lower than the empty control group (Fig. 6C and D). Knocking down of NRAV weakened Rab5c expression, which was abolished by miR-509-3p inhibitor ( Fig. 6E and F). Meanwhile, miR-509-3p mimics reduced Rab5c protein levels in vitro, and overexpression of NRAV, but not NRAVmut, reversed this decrease (Fig. 6G to H). The results suggested that NRAV was positively associated with Rab5c via competitive binding of miR-509-3p in respiratory epithelial cells, which was called the NRAV/miR-509-3p/ Rab5c axis.
The effect of NRAV on RSV was realized by Rab5c. To verify the role of Rab5c in the NRAV/miR-509-3p/Rab5c axis during RSV infection, we constructed a constitutive active (C/A) mutant and a dominant negative (D/N) mutant of Rab5c, i.e., Rab5cQ80L and Rab5cS35N (23). A549 or BEAS-2B cells were transfected by appointed plasmids, and from the next day on, the cells listed above were infected by RSV (MOI ϭ 3) for 48 h. Data showed that overexpression of both NRAV and Rab5cQ80L increased RSV gene levels ( Fig. 7A to D), F protein expression ( Fig. 7E and F), and virus titer (Fig. 7G). Moreover, Rab5cS35N reversed such rises in NRAV overexpressed cells. In NRAVmut overexpressed cells, Rab5cS35N made RSV expression lower than the empty control group (Fig. 7A to G). In addition, both si-NRAV and Rab5cS35N reduced RSV gene expression (Fig. 7H to K), F protein level ( Fig. 7L and M), and viral titer (Fig. 7N), and Rab5cQ80L rescued these reductions in NRAV knocked-down cells (Fig. 7H to N).
All the results revealed that NRAV influenced the expression of RSV via the NRAV/ miR-509-3p/Rab5c axis so that it regulated the RSV infection process through Rab5cpositive vesicle transport (Fig. 8).

DISCUSSION
In this study, NRAV was downregulated in RSV-infected samples. In addition, NRAV promoted RSV proliferation by competitively binding miR-509-3p, enhancing expression of Rab5c, and accelerating intracellular vesicle transport of RSV.
Over the past decade, researchers have identified differentially expressed lncRNAs in diverse viral infections by whole-transcriptome sequencing and established several lncRNA databases (24), such as NONCODE (http://www.noncode.org) and lncRNAdb. Given that RSV is an RNA virus (25) that infects the respiratory tract (26), we have found four candidate lncRNAs which can be influenced by RNA virus or respiratory virus via bioinformatics retrieval, including NEAT1, PRINS, NeST, and NRAV. We designed and synthesized primers for subsequent validation.
In order to confirm RSV infection, sputum specimens of pediatric inpatients were collected and detected by a multiple detection kit for 13 respiratory pathogens. The RNAs prepared from sputum specimens were randomly collected from 75 full-negative specimens and 69 RSV-single-positive specimens to verify bioinformatics results. Detected by qRT-PCR, however, only NRAV was differentially expressed in RSV-infected clinical samples and cell lines. Thus, we focused on NRAV for functional study. On the other hand, we will explore NRAV levels in patients infected with other respiratory viruses in a future study.
In this study, functional experiments were carried out in vitro, because NRAV had no conserved locus in mouse. Results indicated that RSV production could be improved when NRAV was overexpressed, and replication of RSV was inhibited when silencing NRAV. The experiment was carried out in A549 and BEAS-2B cells in functional experiments. The A549 cell line is derived from human adenocarcinoma type II alveolar cells, and BEAS-2B was developed from primary human bronchial epithelial cells transformed with adenovirus type 12 (Ad12)-SV40. Different characteristics were observed in these two cell lines during RSV infection. Hillyer (27) has reported that A549 cells are permissive to RSV infection and produce proinflammatory genes after RSV infection. Instead, BEAS-2B cells could restrict the infection and produce molecules characteristic of an antiviral reaction. Consequently, A549 cells were highly susceptible to RSV infection, whereas BEAS-2B cells limited infection to small individual foci or cells during RSV infection. Similar to those results, in our study, compared with BEAS-2B, when A549 cells were infected with RSV, the increase of RSV gene expression was much more obvious. Furthermore, overexpression of NRAV led to robust enhancement of RSV genes in A549, and the degree of increase was significantly higher than that in BEAS-2B cells (Fig. 5A and B).
During the pcDNA3.1-based overexpression or siRNA-based knockdown experiment, it was interesting that the expression features of RSV genes were quite different between the empty vector (EV) group (Fig. 2) and the si-control group (Fig. 3). It is widely acknowledged that transfection of an NC siRNA or empty vector DNA might have a negligible impact on cells. In effect, however, Stepanenko and Heng have reported that a range of molecular manipulations from the NC siRNA or empty vector will probably influence the biological system, resulting in types of nonspecific effects on cells in the NC group. Furthermore, during the transient transfection, due to differences in nucleic acid type, total base amount, and molecular structure, the nonspecific impact from empty vector backbone RNA (pcDNA3.1, 5,428 bp) and siRNA (double-strand short interfering RNA, 21 bp) could differ from each other (28), which might induce the expression features of RSV genes that were quite different in the EV and siNC groups.
In order to pinpoint the downstream target of NRAV, four groups of A549 cells were used for RNA-seq, including over-NRAV versus the empty vector group (n ϭ 3) and  The subcellular localization of lncRNA is one of the leading determinants of its molecular function (29). Based on the results that NRAV was mainly located in the cytoplasm subcelluarly and showed the same trend with Rab5c, we tentatively put forward the idea that NRAV and Rab5c might share the same miRNA in cytoplasm, which is a major miRNA pool. In the light of bioinformatics analysis and a previous report, miR-509-3p has been reported to be able to bind Rab5c (22). Meanwhile, NRAV was predicted to bind to miR-509-3p's seed region (Fig. 4L). Subsequently, we combined RNA pulldown (Fig. 4I) and dual luciferase reporter assays ( Fig. 4M and N) to identify the NRAV/miR-509-3p association. In addition, we found that the effect of NRAV on RSV could be reversed by miR-509-3p.
The total RNA and protein in Fig. 5F to I were all extracted at 48 h post-RSV infection. It is interesting that RSV F mRNA expression in Fig. 5F and G were not consistent with their protein levels in Fig. 5H and I. In order to clarify this phenomenon, we separately extracted total RNA at 24, 36, and 48 h to detect the expressions of the F gene. The results showed that whether the RNA was extracted 24, 36, or 48 h postinfection, silencing NRAV reduced RSV F expression, and miR-509-3p inhibitor could rescue this decrease. With the extension of time, however, the expression of the F gene changed dynamically in the rescue group. When the cells were infected for 24 or 36 h, miR-509-3p inhibitor could significantly rescue the expression of F in mRNA level (data not shown). Nevertheless, at 48 h postinfection, the recovery of F expression was not as high as that at 24 or 36 h. Interestingly, in the rescue group, the mRNA level of the F gene at 24 h postinfection was somewhat consistent with F protein at 48 h after infection. As a result, when both RNA and protein were extracted at 48 h postinfection,  On the other hand, dual luciferase reporter assay ( Fig. 6A and B) and immunoblotting (Fig. 6C to H) were performed to verify that Rab5c could be positively influenced by NRAV, and similarly, the influence could be abolished by miR-509-3p. However, both RSV and Rab5c levels have not completely returned to NC levels. It is probable that NRAV is able to regulate virus proliferation in multiple ways, for example, an NRAV/ histone modification/interferon-stimulated gene (ISG) mechanism (15).
It has been reported that Rab proteins, such as Rab5 and Rab11, play a critical role in the entry, assembly, and excretion of RSV. Krzyzaniak inferred that host cell entry and intracellular transmission of RSV depended on Rab5-positive macropinosomes (20). Brock and Utley reported that Rab11 and Rab11FIPs were essential in RSV assembly and budding (30,31). In our study, A549 cells were infected with RSV (MOI ϭ 3) for 1 h, which was the average time for internalization (20). Confocal microscopy showed that some of the RSV was colocalized with Rab5c-positive vacuoles in the cytoplasm (Fig. 4F). These results indicated that through targeting the miR-509-3p/Rab5c axis, NRAV might promote RSV infection at the entry and intracellular transmission stage.
In our previous study, we found that RSV induced autophagy in host cells to facilitate RSV replication (32). Meanwhile, it had been reported that Rab5, which was indispensable for the recruitment of Rab7 to PI3P-positive autophagosomes, was one of the important factors to maintain unobstructed autophagy flow (33). Subsequently, we will further explore whether the effect of autophagy on RSV proliferation is mediated by the NRAV/miR-509-3p/Rab5c axis.
In order to explore whether NRAV/Rab5c was correlated with the progress and prognosis of RSV infection, we defined the length of hospital stay as an index to reflect the progress and prognosis of RSV infection. Based on the median expression of NRAV, we defined the RSV-infected patients as high NRAV group (n ϭ 35) and low NRAV group (n ϭ 34). Then, we compared NRAV level and the length of hospital stay, and no correlation was found. In addition, we observed no correlation between Rab5c expression and the length of hospital stay (data not shown). As a result, it is unclear whether this insignificance is due to the small number of RSV-positive cases in this study or if there is truly no prognostic difference. More patients should be included in our future study.
Overall, we found that RSV could decrease NRAV expression in patients' sputum specimens. Overexpression of NRAV promoted RSV proliferation in vitro and vice versa. NRAV could act as a ceRNA in the NRAV/miR-509-3p/Rab5c axis during RSV infection, thus promoting RSV vesicle transport and accelerating RSV entry (Fig. 8), suggesting that the downregulation of NRAV in RSV infection was part of the host antiviral defense. The results may facilitate improvement in exploring a potential noncoding RNA target for diagnosis and treatment of respiratory virus infection.

MATERIALS AND METHODS
Patients. This study was conducted in accordance with the principles of the Declaration of Helsinki. Between 19 September 2018, and 25 February 2019, sputum specimens of pediatric inpatients were randomly collected with a sputum aspirator based on National Clinical Laboratory Procedures. Specimens were detected using a multiple detection kit for 13 respiratory pathogens (Health Gene Tech, Ningbo, China) in the Second Hospital of Hebei Medical University. Total RNA was collected and qualified according to the kit; 13 common respiratory pathogens were detected in sputum samples by RT-PCR and capillary electrophoresis, including influenza A virus H1N1 and H3N2, parainfluenza viruses, human metapneumovirus, influenza B viruses, respiratory syncytial virus, coronavirus, rhinoviruses, bocaviruse, Chlamydia, Mycoplasma pneumoniae, and adenoviruses. Seventy-five samples were randomly selected as negative controls, and the detected results of thirteen pathogens in these samples were all negative. For the experimental group, 69 samples were randomly selected, in which, of the 13 pathogens, only RSV was positive. The patients' information is shown in Table 1. Written informed consent was obtained from all parents or guardians.
Monocytes (THP-1) were treated with phorbol 12-myristate 13-acetate (PMA, 15 ng/ml; Multiscience) and were differentiated into macrophages M⌽-THP-1 after 12 h. Viral infection and virus titer assay. A549 or BEAS-2B cells were washed and incubated with RSV at a certain multiplicity of infection (MOI) or mock-infected (with medium alone) for 2 h in serum-free RPMI 1640. The supernatant was aspirated after adsorption, and cells were cultured with RPMI 1640 supplemented with 2% (vol/vol) FBS for the indicated times.
Cell lysates were harvested at the indicated time points. RSV titers were detected by plaque-forming assay (35) and expressed in PFU/milliliter cell lysates.
RNA extraction and reverse transcription quantitative PCR (qRT-PCR). Total RNA was isolated with RNAiso Plus reagent (9101; TaKaRa). First-strand cDNA was generated with a PrimeScript RT reagent kit (047A; TaKaRa). Real-time PCR was performed in an ABI Prism 7500 sequence detection system (Applied Biosystems) using PowerUp SYBR green master mix (A25742; Applied Biosystems). ACTB was employed as an endogenous control for lncRNA and mRNA. The primers are shown in Table 2 and were synthesized by Invitrogen.
RNA fluorescent in situ hybridization. NRAV (GenBank accession number NR_038854) probe mix and NC probe are shown in Table 2. All the probes were labeled with CY3 fluorescent dye. RNA fluorescent in situ hybridization (RNA-FISH) was performed using a fluorescent in situ hybridization kit (Gene Pharma, China) following the manufacturer's instructions. Fluorescence detection was performed with a confocal laser-scanning microscope (Leica TCS SP5).
Plasmids and small RNAs. The full lengths of wild-type NRAV and NRAVmut with point mutations in miR-509-3p binding sites (Gene Pharma, Suzhou, China) were synthesized and subcloned into the The sequences of si-NRAV, the miR-509-3p mimics and inhibitor, and the miRNA control are shown in Table 2.
Cell transfection. Cells were seeded in six-well plates. Plasmids or miRNAs were transfected into cells with Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions.
RNA-seq analysis. Total RNA of A549 cells was isolated and employed for RNA-seq analysis. Construction of the cDNA library and sequencing were performed by Shanghai Biotechnology Corporation using an Illumina HiSeq 2000 system. High-quality reads were mapped with Homo sapiens genome hg38, using Hisat2 version 2.0.4. The expression level of each gene was standardized to fragments per kilobase of exon model per million mapped reads (FPKM) using StringTie version 1.3.0 and trimmed mean of M values (TMM) (P Ͻ 0.05, fold change, Ն 2).
RNA pulldown assay. The NRAV probe mix and NC probe were the same as those used in RNA-FISH, and the probes were modified with 3=, 5= biotins. RNA pulldown was performed using an RNA pulldown kit (Gene Pharma, Suzhou, China) following the manufacturer's instructions. NRAV and miRNAs bound to the pulldown beads were examined using qRT-PCR. The primers of hsa-miR-509-3p (MIMAT0002881) and hsa-miR-653-5p (MIMAT0003328) are presented in Table 2.
Statistical analysis. Data from three independent experiments were expressed as the mean Ϯ standard deviation (SD). SPSS version 16.0 software was used for statistical analyses. The comparisons were performed by using Student's t test or one-way analysis of variance (ANOVA). P Ͻ 0.05 was considered to indicate a statistically significant difference.
Data availability. The RNA-seq data set is available in the Gene Expression Omnibus (GEO) and can be accessed via the GSE series number GSE145371.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.