ABSTRACT
Up-regulation of interferon-stimulated genes (ISGs) is key to antiviral states mediated by interferon (IFN) but little is known about activity and underlying mechanisms of most ISGs against Enterovirus 71 (EV71). EV71 causes hand-foot-mouth disease in infants and occasionally severe neurological symptoms. Here we report that the product of L3HYPDH, a newly identified ISG, inhibits the replication of EV71. This anti-EV71 activity was mapped to the C-terminal 60 amino acids region as well as the N-terminal region spanning from amino acid position 61 to 120 of L3HYPDH protein. L3HYPDH was shown to interfere with EV71 propagation at the RNA replication and protein translation levels. Specifically, L3HYPDH impairs translation mediated by the EV71 international ribosome entry site (IRES) but not by the HCV IRES, and this activity is conferred by the C-terminal region of L3HYPDH. Thus, L3HYPDH has antiviral activity against EV71, suggesting a potential mechanism for broad-spectrum antiviral effects of IFN.
IMPORTANCE
Human EV71 can cause hand-foot-mouth disease (HFMD) and even death; however, no effective anti-EV71 treatment is available. Although EV71 suppresses induction of IFN and activation of IFN signaling pathways, type I IFN treatment can enhance the anti-EV71 state. IFN-stimulated genes (ISGs) are critical for innate immune defenses; however, the antiviral activities of many ISGs are not known. EV71 is seldom used for ISGs studies. So understanding the mechanism by which ISGs exert activity against EV71 will help to better understand IFN-triggered antiviral activity and provide new strategies to treat enterovirus infection. L3HYPDH is a newly identified ISG. We report here that L3HYPDH significantly inhibits EV71 replication by repressing RNA replication and protein translation, which suggests a mechanism underlying type I IFN against EV71. This would assist with the development of novel therapeutics to treat HFMD.
INTRODUCTION
Hand-foot-mouth disease (HFMD) is a common viral disease in infants and children across the Asian-Pacific region, characterized by fever, rash, and occasionally severe neurological symptoms (1, 2). Enterovirus 71 (EV71) is a major causative agent of HFMD. Different from many other viruses, EV71 suppresses induction of type I interferons (IFNs) and activation of IFN signaling pathways, and consequently, inhibits host anti-viral defenses (3–5). Nonetheless, EV71-infected cells still respond to type I IFN treatment and display an enhanced antiviral state. For example, in vitro studies showed that some type I IFNs, including IFN-α4, IFN-α6, IFN-α14 and IFN-α16, significantly reduced cytopathic effect (CPE) induced by EV71 infection (6). An IFN-α2b aerosol therapy has been used topically to treat HFMD (7). However, how IFNs suppress EV71 infection is not clear.
IFN-mediated antiviral mechanisms are diverse and complicated. Up-regulation of IFN-stimulated genes (ISGs) has shown to be critical to innate immune defenses against invading pathogens, so studies are underway to assess their functions and underlying mechanisms for developing future antiviral therapies. ISGs are abundant and different cells respond variously to different types of IFNs and the stimulation duration, resulting in diverse ISG expression patterns (8–10). Data of several systematic detections for antiviral activity suggest that different sets of ISGs target different viruses in unique ways, and antiviral roles may be broad or specific, strong or weak (11, 12). However, EV71 is seldom used for characterizing antiviral activity of ISGs, including well-characterized classic ISGs.
EV71 is enterovirus of the Picornaviridae family. Its genomic RNA is about 7400 nt long, single and positive-stranded, and contains only one open reading frame (ORF) flanked with a 5’-untranslated region (5’-UTR) and a 3’-UTR (1). The 5’-UTR contains a cloverleaf structure and an internal ribosome entry site (IRES), responsible for viral RNA replication and translation, respectively (13). The life cycle of EV71 starts with attachment to the host cell surface by recognition of a specific receptor, followed by endocytosis and release of viral RNA into the cytoplasm (14). Then, EV71 IRES initiates viral translation by recruiting host proteins. Synthesized polyproteins are processed into structural and non-structural proteins by its own protease 2A and 3C. When viral proteins accumulate, viral protein 3CD binds to the cloverleaf structure of 5’UTR to stop viral protein synthesis and initiates viral RNA replication. Produced RNAs then direct viral protein synthesis in large quantities. With the assembly of viral RNAs and proteins into virions, the host cell lyses and progeny viruses are released for a new round of infection (15, 16).
Many ISGs are antiviral effectors but these represent a few existing ISGs, and more will be identified. Recently, 91 new ISGs were identified from human immune cell lines after treatment with the consensus interferon (17), but their antiviral activity is unclear. Using a fluorescent activated cell sorting-based strategy for screening, we identified several ISGs with anti-EV71 efficacy (data not shown). One of them is C14orf149, which was identified as a gene encoding a trans-3-hydroxy-L-proline dehydratase and then renamed L3HYPDH (18). Here, we report that this ISG product, L3HYPDH, possesses antiviral activity against EV71, and its mechanism of action was investigated with a series of biochemical and genetic assays.
MATERIALS AND METHODS
Plasmids construction
pCAG-DsRed, a red fluorescent protein-expressing plasmid, has been described previously (19). pWSK-EV71-GFP is an infectious EV71-GFP cDNA clone, with a GFP-coding sequence inserted downstream of EV71 5’UTR and in frame fusion with the downstream VP4, and expression of EV71-GFP is driven by a T7 promoter (20). pcDNA3.1-T7RNP expresses T7 RNA polymerase. These plasmids were kindly provided by Dr. Liguo Zhang at the Institute of Biophysics, Chinese Academy of Sciences (IBP, CAS). A siRNA targeting the coding sequence of L3HYPDH from position 791 to 811 was designed according to the recommendation of Sigma-Aldrich (https://www.sigmaaldrich.com/catalog/genes) and named shRNA149. A pair of complementary oligonucleotides 5’-GATCCCCCAGATGAACAGGTTGACAGAATTCAAGAGATTCTGTCAACC TGTTCATCTGTTTTTA-3’ (sense) and 5’-AGCTTAAAAACAGATGAACA GGTTGACAGAATCTCTTGAATTCTGTCAACCTGTTCATCTGGGG-3’ (antisense) were synthesized with 5’ ends being BglII and HindIII restriction site overhangs. For each oligonucleotide, the target sequence was sense followed by antisense orientations separated by a nine-nucleotide spacer. Oligonucleotides were annealed and then cloned into the BglII and HindIII sites of pSUPER.retro.neo+gfp (Oligoengine, herein abbreviated for pSUPER-GFP) to generate pSUPER-GFP-shRNA149. L3HYPDH wild type (WT) and deletion mutants as indicated in Fig 3 were amplified with PCR using pLPCX-C14orf149 (17), kindly provided by Dr. Guangxia Gao at IBP, CAS, as the template. PCR products of L3HYPDH WT and deleted mutants were digested with BamHI & NotI and KpnI & XbaI, respectively, and inserted into similarly digested pcDNA4-To/myc-His B (Invitrogen), resulting in pcDNA4-L3HYPDH, pcDNA4-L3HYPDHΔN1, pcDNA4-L3HYPDHΔN2, pcDNA4-L3HYPDHΔN3, pcDNA4-L3HYPDHΔC1, pcDNA4-L3HYPDHΔC2, and pcDNA4-L3HYPDHΔC3. psiCHECK2-M was a modified form of psiCHECK-2 (Promega) with deletion of the HSV-TK promoter (Fig 5A). Inverse PCR was performed with high-fidelity DNA polymerase Phusion (ThermoFisher) and a pair of back-to-back primers to amplify the whole plasmid except the HSV-TK promoter sequence. PCR products were self-ligated and resulted in psiCHECK2-M; meanwhile, a SalI and a NotI sites within the back-to-back primers were introduced into the plasmid. EV71-5’UTR and HCV-5’UTR were amplified from pWSK-EV71-GFP and pNL4-3RL-HCV-FL (21) by PCR, respectively. After digestion with SalI and NotI, the PCR products were linked into the similarly digested psiCHECK2-M and resulted in psiCHECK2-M-EV71-5’UTR and psiCHECK2-M-HCV-5’UTR. All primers used are listed in Table S1.
Cell culture and virus preparation
293A, 293A-SCARB2, RD, Vero, Hela, and A549 cells were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco). 293A-SCARB2 (Kindly provided by Dr. Liguo Zhang at IBP, CAS), originated from a 293A cell line and constitutively expresses the main EV71 receptor scavenger receptor class B member 2 (SCARB2). To generate the cell line constitutively expressing tagged L3HYPDH, 293A-SCARB2 cells were transfected with pcDNA4-L3HYPDH as described below and selected with Zeocin (200 μg/ml). Resistant colonies were individually expanded and detected by Western blot. One positive clone was chosen and named 293A-SCARB2-L3HYPDH. This process was applied to the empty vector and resulted in control cell 293A-SCARB2-Ctrl.
EV71-MZ (GenBank accession no. KY582572), isolated from the throat swab of an ICU patient at Meizhou People’s Hospital in 2014, was amplified by successive passages in RD cells until apparent CPE appeared. EV71-GFP was generated by co-transfecting pWSK-EV71-GFP and pcDNA3.1-T7RNAP into 293A-SCARB2 cells as described previously (20). Viral supernatants were titrated using a plague assay, aliquoted, and then used for infection.
Transfection and infection
Depending on the experiments, cells were seeded into a 24-well or 6-well plate or 10 cm dish and were grown to approximately 80% confluence prior to transfection or infection. All plasmid and RNA transfections were carried out by using Lipofectamine TM 2000 (Life Technology) according to the manufacturer’s instructions. After incubation for the indicated time, cells were treated as required.
Viral infection was performed by incubating cells with EV71-GFP or EV71-MZ at a different multiplicity of infection (MOI) for 1 h, with shaking every 15 min, and then the unbound viruses were aspirated. Cells were washed with PBS, added fresh medium and incubated for specific time, followed by FACS assay, RT-qPCR measurement, or supernatant titration.
Plaque assay
The plaque assay was performed as described previously (22). Briefly, RD cells were incubated with viral supernatants undiluted or diluted in 10-fold series for 1 h. Subsequently, the supernatants were aspirated, and cells were covered with DMEM containing 1% methylcellulose (Sigma-Aldrich) and 2% FBS. After incubation for 4 days, cells were fixed with 4% paraformaldehyde (Sigma-Aldrich) and stained with 0.1% crystal violet. Plaques were then quantified via visual scoring.
Fluorescent activated cell sorting (FACS) assay
To measure GFP production from EV71-GFP, 1×106 infected cells were collected and fixed in 4% paraformaldehyde for 15 min. After washing three times with PBS, cells were resuspended in 0.5 ml of PBS for flow cytometry (LSRFortessa, BD). To assess effects of L3HYPDH knockdown by RNAi, the cells were transfected with pSUPER-GFP-shRNA149 or pSUPER-GFP. After incubation for the indicated time, the cells were harvested and washed with PBS. GFP-positive cells were obtained through FACS, and then lysed for Western blot or seeded into a 24-well plate for EV71-GFP infection or reporter plasmid transfection as required.
IFN stimulation
Cells were treated with 1000 IU/ml of recombinant human IFN-α2b (Prospec) for different time, and then total RNAs were isolated and used to measure specific mRNA abundance by RT-qPCR.
In vitro transcription of EV71-GFP and microscope assay of GFP
pWSK-EV71-GFP was linearized XbaI and EV71-GFP RNAs were transcribed using the T7 RiboMax kit (Promega). After transfection into 293A-SCARB2-L3HYPDH and 293A-SCARB2-Ctrl cells, the GFP signal was observed under a fluorescence microscope (System Microscope BX63, Olympus) at the indicated times; total RNAs were isolated for RT-qPCR assay.
RNA isolation and RT-qPCR
Total RNAs were isolated from cells using TRI Reagent (Sigma-Aldrich) according to the manufacturer’s instructions. RT-qPCR was carried out as described previously (23) to measure target mRNA. Briefly, RNAs were treated with DNase using an RQ1 RNase-Free DNase Kit (Promega); cDNAs were synthesized using PrimeScript RT reagent Kit (Takara, Dalian) and then diluted and subjected to quantitative PCR using TransStart Green qPCR SuperMix (TransGen Biotch) in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad). Primers used appear in Table S1.
Immunofluorescence assay (IFA)
Subcellular localization of L3HYPDH proteins and the attachment and endocytosis of EV71 virions were detected using IFA as described previously (24) with some modifications. Briefly, 293A-SCARB2-L3HYPDH and 293A-SCARB2-Crtl were individually seeded onto a coverslip. Polyclonal antibody (PAb) specific to c-myc (Sigma-Aldrich, 1:100) and Alexa Fluor 555-labeled anti-rabbit IgG (ThermoFisher, 1:100) were used as primary and secondary antibody, respectively, to localize the subcellular distribution of the tagged L3HYPDH proteins. Similarly, 293A-SCARB2-L3HYPDH and 293A-SCARB2-Crtl cells were infected with EV71-MZ (MOI, 100) at 4°C for 1 h to allow viral attachment or incubated for an additional 30 min at 37°C to allow viral endocytosis. The Anti-EV71 VP2 monoclonal antibody (MAb) (Millipore, 1:50) and Alex flour 555-labeled anti-mouse IgG (ThermoFisher, 1:100) were used as primary and secondary antibody, respectively, to visualize EV71 virions. Nuclei were stained with DAPI (Roche). Fluorescent images of cells were captured using a Zeiss LSM780 META confocal imaging system.
Western blot
Western blot was performed as described previously (24) with some modifications. Briefly, 48 h after transfection, the cells were lysed with SDS-lysis buffer (30 mM SDS, 50 mM pH 6.8 Tris-HCL, 100 mM DTT and 20 mg/L bromophenol blue) directly and proteins were isolated with 10% SDS-PAGE. The membrane was probed with anti-6×His MAb (Abcam) and anti-GAPDH PAb (Sangon), followed by incubation with HRP-conjugated anti-mouse IgG and anti-rabbit IgG (Santa Cruz Biotechology), respectively. Proteins were visualized with ECL.
Luciferase activity assay
Cell lysate was prepared by using passive lysis buffer (Promega). Firefly and renilla luciferase activities were measured using a Dual Luciferase Assay kit (Promega) according to the manufacturer’s instructions.
Statistical analysis
All the experiments involving counting or calculation were performed independently at least three times and data are means ± standard deviation (SD). A Student’s two-tailed t test was used for statistical analysis by using GraphPad Prism 6.1 (GraphPad 6 Software, San Diego, CA). P< 0.05 was considered statistically significant.
RESULTS
Over-expression of L3HYPDH inhibits EV71-GFP replication
The activity of L3HYPDH against EV71 was detected in 293A-SCARB2 cells using a FACS-based assay (Fig 1A). 293A-SCARB2 cells were used to facilitate EV71 infection. After co-transfection with pCAG-DsRed and pcDNA4-L3HYPDH at a ratio of 1:3, the cells were infected with EV71-GFP. All DsRed-expressing cells were assumed to express L3HYPDH, and viral replication in the DsRed-positive populations was quantified by FACS assay of GFP at 18 h post infection (Fig 1B). Cells overexpressing L3HYPDH expressed approximately 90% less GFP than control cells (Fig 1C). This result indicated that over-expression of exogenous L3HYPDH inhibits the expression of GFP, suggesting potential antiviral action against EV71 replication.
Expression of endogenous L3HYPDH suppresses EV71 replication
The expression of endogenous L3HYPDH and its response to IFN-α2b treatment were examined in different cell lines using RT-qPCR. Total RNAs were isolated from 293A, 293A-SCARB2, Vero, A549, RD, and Hela cells in the absence or presence of IFN-α2b for different times as indicated (Fig 2A, 2B). RT-qPCR assay showed that the basal mRNA level of L3HYPDH was slightly higher in 293A, 293A-SCARB2 and Vero cells than in RD and A549, and the lowest in Hela cells (Fig 2A). Upon exposure to IFN-α2b, the level of L3HYPDH mRNA was up-regulated and peaked at about 18 h in 293A, 293A-SCARB2 and A549 and at about 12 h in Vero cells. Peak levels were 3 to 5 times higher than the basal level (Fig 2B). In contrast, the mRNA level changed little in Hela cells and even decreased a little in RD cells (Fig 2B). These results indicate that IFN-α2b stimulates the expression of L3HYPDH. However, the expression of L3HYPDH and its response to IFN-α2b differ in different cells, which is common for ISGs.
A shRNA specific to L3HYPDH, designated as shRNA149, was designed and transcribed from pSuper-GFP-shRNA149. Its knockdown efficiency was detected in 293A-SCARB2 cells by co-transfecting pcDNA4-L3HYPDH together with pSUPER-GFP-shRNA149 or with pSUPER-GFP as a control. Western blot analysis showed that the tagged L3HYPDH protein level decreased dramatically in the presence of shRNA149 (Fig 2C). To determine if the endogenous L3HYPDH could suppress EV71 replication, 293A-SCARB2 cells were transfected with pSUPER-GFP-shRNA149, and the GFP-positive cells were sorted by FACS, followed by EV71-MZ infection. RT-qPCR assay revealed that shRNA149 reduced L3HYPDH mRNA level by more than 80% (Fig 2D) and increased EV71 mRNA level from 1 to 1.7 (Fig 2E), indicating that the expression of endogenous L3HYPDH impaired EV71 replication. In this way, L3HYPDH possesses antiviral activity and is involved in the inherent cellular suppression on EV71 replication.
Determination of the amino acid sequences essential for anti-EV71 activity of L3HYPDH
The region critical to L3HYPDH action against EV71 was mapped by serial deletions combination with the FACS-based assay. Three N-terminal and three C-terminal progressive deletion mutants of L3HYPDH are schematically shown in Fig 3 (middle panel), designated ΔN1, ΔN2, ΔN3, ΔC1, ΔC2, and ΔC3. Their coding sequences were cloned in fusion with a myc-6×His tag at the C-terminus as with L3HYPDH WT. The resulting plasmids were individually transfected into 293A-SCARB2 cells together with pCAG-DsRed followed by EV71-GFP infection as described in Fig 1A. Western blot showed that the protein levels of these truncated mutants were somewhat lower than that of WT, but still comparable (Fig 3, lower panel). FACS assay showed that L3HYPDHΔN2 lacking the amino acids from position 1 to 120 significantly impaired the antiviral activity in comparison with WT, while L3HYPDHΔN1 lacking the amino acids from position 1 to 60 only slightly weakened the antiviral activity. L3HYPDHΔC1 lacking the C-terminal 60 amino acids from 295 to 354 also weaken the antiviral activity somewhat, while further deletion did not heighten this impairment. These results indicate that the amino acid sequences from position 61 to 120 and from 295 to 354 are both required for the development of anti-EV71 activity of L3HYPDH.
EV71 replication is suppressed in the cell line 293A-SCARB2-L3HYPDH expressing L3HYPDH constitutively
To facilitate investigation of the antiviral mechanism, the cell line 293A-SCARB2-L3HYPDH expressing L3HYPDH constitutively and the corresponding control cell line 293A-SCARB2-Ctrl were generated. These two cell lines were infected with EV71-GFP at an MOI of 0.1. FACS assay showed that the GFP production in 293A-SCARB2-L3HYPDH decreased to about 84% of control levels (Fig 4A). Upon infection with EV71-MZ, a clinical isolate of EV71, there was significantly less viral multiplication in 293A-SCARB2-L3HYPDH than in the control cell (Fig 4B). IFA showed that L3HYPDH proteins were mainly located in the cytoplasm (Fig 4C), consistent with its anti-EV71 action. Therefore, the cell 293A-SCARB2-L3HYPDH displays remarkable anti-EV71 activity due to the over-expression of L3HYPDH, and thus can be exploited to uncover the underlying antiviral mechanism.
L3HYPDH interferes with the synthesis of viral RNA and proteins
The effects of L3HYPDH on different life stages of EV71 replication were examined in 293A-SCARB2-L3HYPDH cells. Based on the knowledge that EV71 is only adsorbed on the cell surface and could not finish endocytosis at 4℃, 293A-SCARB2-L3HYPDH and the control cells were incubated with EV71-MZ for 1 h at 4℃ followed by IFA with the antibody specific to EV71 VP2. As shown in Fig 5A, massive number of virions distributed on the outer surfaces of both cell lines, showing no difference in numbers, indicating that L3HYPDH could not interfere with EV71 attachment. After attachment at 4℃, the viruses were further incubated with the cells for an additional 30 min at 37℃ to complete endocytosis. IFA showed that many viruses entered both cell lines, showing little difference (Fig 5B), indicating that L3HYPDH had no effect on EV71 endocytosis. In this way, these results demonstrate that L3HYPDH does not impede the viral attachment and endocytosis.
The effects of L3HYPDH on the synthesis of viral RNA and proteins were investigated by monitoring changes in their levels over time. 293A-SCARB2-L3HYPDH and 293A-SCARB2-Ctrl cells were infected with EV71-MZ. Total RNAs were isolated at different times post-infection and the viral RNA abundance was measured by RT-qPCR. The increase in EV71 RNA levels over time in 293A-SCARB2-L3HYPDH cells was much lower than in the control cells (Fig 5C). Due to the tight cross-talk between viral translation and viral RNA synthesis, these results suggested that L3HYPDH might inhibit the synthesis of viral RNA, proteins, or both. To further confirm this assumption, EV71-GFP RNAs were transfected into 293A-SCARB2-L3HYPDH and the control cell, and then the viral RNA and GFP proteins were measured and compared at different times after transfection. Microscope and RT-qPCR analyses showed that both the number of GFP-positive cells and the viral RNA level were much lower in 293A-SCARB2-L3HYPDH cells than in the control cells (Fig 5D, 5E). These results provide more evidence that L3HYPDH might suppress viral RNA replication, viral protein synthesis, or both.
L3HYPDH impairs the translation mediated by EV71-5’UTR
The repression of L3HYPDH on viral protein synthesis was investigated using a bicistronic reporter system. As shown in Fig 6A, psiCHECK-2-based reporter plasmids were constructed with the HSV-TK promoter deleted to generate the control (psiCHECK2-M) or replaced with EV71-5’UTR or HCV-5’UTR, which contains EV71 IRES or HCV IRES, respectively. pcDNA4-L3HYPDH or the empty vector was transfected into 293A cells together with one of the three reporter plasmids at a ratio of 3:1, and then the luciferase activity and mRNA level were measured after incubation for 48 h. For these reporters, the mRNA level ratio of Fluc/Rluc in L3HYPDH-overexpressed cells was equal to that in the empty vector-transfected cells as revealed by RT-qPCR (Fig 6B). However, the luciferase activity ratio (Fluc/Rluc) showed variability (Fig 6C). Whether L3HYPDH was over-expressed or not, the Fluc/Rluc ratio of the control reporter was extremely low due to the absence of IRES; the ratio of the EV71-5’UTR-containing reporter reduced by 29% upon overexpression of L3HYPDH; while the ratio of the HCV-5’UTR-containing reporter changed little. We here proposed that L3HYPDH could specifically inhibit the reporter translation mediated by EV71 IRES. RNAi assay further provided evidence for this speculation. 293A-SCARB2-L3HYPDH cells were transfected with the shRNA149-expressing plasmid or the empty vector. Then the GFP-positive cells were isolated and transfected with the reporter plasmids. Compared to the control cells, the Fluc/Rluc ratio of the EV71-5’UTR-containing reporter in 293A-SCARB2-L3HYPDH cells increased moderately upon L3HYPDH knockdown (Fig 6D).
Given that the amino acid sequence from position 61 to 120 and the C-terminal 60 amino acids together contribute to anti-EV71 activity, we examined whether these sequences were involved in the action of inhibiting the EV71 IRES-mediated translation. 293A cells were transfected with the plasmids expressing L3HYPDH WT or deletion mutants together with the EV71-5’UTR-containing bicistronic reporter plasmids. Reporter assay revealed that, compared to the empty vector, the expression of WT, ΔN1, ΔN2 and ΔN3 reduced the Fluc/Rluc ratio by approximately 25% in 293A cells, consisting with the result shown above; while the expression of ΔC1, ΔC2, and ΔC3 had little effect on the ratio by comparison with control cells (Fig 6E). In combination with the results shown in Fig 3, these data indicate that the C-terminal sequence containing 60 amino acids from 295 to 354 is required for L3HYPDH to inhibit EV71-IRES-mediated translation.
Altogether, these results indicate that L3HYPDH can specifically impair the translation initiated by EV71-5’UTR, and the C-terminal region is responsible for this inhibiting activity.
DISCUSSION
In this work, we report that the recently identified ISG product L3HYPDH has antiviral activity against EV71 according to RNAi knockdown and over-expression experiments. Over-expression of L3HYPDH repressed GFP production of EV71-GFP (Fig 1B, 4B) and caused significant inhibition of propagation of the clinical isolate EV71-MZ (Fig 4C). L3HYPDH knockdown increased EV71 mRNA in 293A-CARB2 cells (Fig 2E), highlighting that this gene is an important ISG with antiviral activity. Additionally, our data showed that IFN-α2b treatment was less effective against EV71 in cell culture when expression of L3HYPDH was depressed by RNAi (Fig S1). Therefore, L3HYPDH is key to antiviral activity of IFN-α2b against EV71. The potential activity of L3HYPDH against other viruses is not known. Given that different viruses are usually targeted by unique sets of ISGs (11), an extensive investigation on L3HYPDH will help to further elucidate the mechanism of IFN-mediated innate immunity against invading viruses.
Our data show that L3HYPDH may interfere with EV71 replication at post-entry stage (Fig 4). Bicistronic reporter assays confirmed that expression of L3HYPDH inhibited translation initiated by EV71 IRES (Fig 6B), however, the reporter protein was less reduced than EV71 RNA and virus-carrying GFP production during the first round of infection (Fig 1B, 4A, 5C-E). These inconsistences suggest that L3HYPDH hampers EV71 replication at steps other than translation. Although inhibition of viral RNA replication is likely, other potential effects on viral RNA stability, viral assembly and viral release cannot be excluded. Therefore, L3HYPDH inhibits EV71 replication at least at two levels, and these data are in agreement with previous studies indicating that many ISGs block viral replication at multiple stages of the viral life cycle (25–27). Considering that a range of proteins are involved in the viral RNA replication and translation process, we performed co-immunoprecipitation and tandem affinity purification combination mass spectrometry to screen for proteins interacting with L3HYPDH. Neither viral nor host proteins were identified (data not shown). These results suggest that the association of L3HYPDH proteins with other proteins should be transient or weak. L3HYPDH might also function by binding to the viral RNA directly; however, no known RNA-binding domains were predicted with online software (data not shown).
Viral translation is completely host cell-dependent. To maximize efficiency, different viruses evolved many strategies to facilitate selective translation of viral mRNAs over host transcripts (15, 28–30). Among these, the IRES-mediated translation initiation is necessary for picornavirus and hepacivirus to replicate (13, 31). Reporter assays showed that expression of L3HYPDH impaired initiation of translation mediated by EV71 IRES but not HCV IRES (Fig 6B, 6C). These two IRES differ in nucleotide length and structure as well as in host factors required for translation initiation and regulation (32). A potential target of L3HYPDH should be involved in EV71-5’UTR-mediated translation. Given that an ISG may interfere with different stages of different viral life cycles, whether L3HYPDH has the activity against HCV is unclear. Meanwhile, despite being present in all picornaviruses, IRES is diverse in length and structure and requires different host factors to function (33–35). Whether L3HYPDH can inhibit other genuses of picornavirus by interfering with IRES-mediated translation is not clear is not clear, but this is worthy of study.
L3HYPDH is a trans-3-hydroxy-L-proline dehydratase, and specifically catalyzes the dehydration of dietary trans-3-hydroxy-L-proline and from degradation of proteins such as collagen IV that contain it. This dehydratase contains two active sites, a Cys residue at the 104 position and a Thr residue at the 273 position (18). Interestingly, the region required for anti-EV71 activity was mainly mapped to the amino acid sequence from position 61 to 120 of L3HYPDH protein (Fig 3), which contains the Cys104 active site. Whether this proline dehydratase activity is involved in the anti-EV71 activity is not known. L3HYPDH functions as an anti-EV71 effector. Understanding ISG products and antiviral spectra, as well as their mechanisms of action and biological function will help create novel therapeutics for HFMD.
Knockdown of L3HYPDH impairs anti-EV71 efficacy of IFN-α2b
IFN-α2b treatment combination with RNAi assay were performed to further determine if the endogenous L3HYPDH could suppress EV71 replication, 293A-SCARB2 cells were transfected with pSUPER-GFP-shRNA149, and the GFP-positive cells were sorted by FACS and divided into two parts for IFN-α2b treatment and mock-treatment, followed by EV71-MZ infection. RT-qPCR assay revealed that the IFN treatment caused a triple increase of the endogenous L3HYPDHmRNA level (Figure S1A) and about 40% reduction in the EV71 RNA abundance in 293A-SCARB2 cells (Figure S1B), further validating the perspective that type I IFN is capable of inhibiting EV71 infection. When the expression of L3HYPDH was reduced by approximate 80% by RNAi, IFN-α2b treatment was less effective against EV71 replication and the viral RNA level increased from 0.6 to 0.78 (Figure S1A, S1B). Although the increase in viral yield was not significant, the results indicated that the L3HYPDH products play an irreplaceable role in the anti-EV71 action intrigued by IFN-α2b, suggesting L3HYPDH as an important ISG possessing antiviral activity.
ACKNOWLEDGEMENTS
This work was supported by the grants from Guangdong Innovative Research Team Program (2009010058), the National Key Program for Infectious Disease of China (2012ZX10001003), and Science and Technology Planning Project of Guangdong Province (A2016467 to X. Meng). We thank Dr. Liguo Zhang for providing the plasmids pCAG-DsRed, pWSK-EV71-GFP, pcDNA3.1-T7RNP and the cell line 293A-SCARB2. We also thank Dr. Guangxia Gao for providing the plasmids pLPCX-C14orf149 and pNL4-3RL-HCV-FL.