N-Acetyltransferase 8 Promotes Viral Replication by Increasing the Stability of Enterovirus 71 Nonstructural Proteins

NAT8 is a vital host factor for EV71 infection in SK-N-SH cells.

To identify the mechanisms by which EV71 hijacks the cellular machinery to promote replication and spread, we performed genome-wide CRISPR/Cas9 screening in SK-N-SH cells (Fig. 1A). Using the MAGeCK algorithm, we found strong enrichment of guide RNAs (gRNAs) targeting known host factors, including entry receptor SCARB2 and replication host factor ATG4B, suggesting that the screening was robust. Among the screened genes (see Table S1 in the supplemental material), the NAT8 gene attracted our interest for its special function and localization. NAT8 is an acetyltransferase located in the ER, where the enterovirus replication organelles form (12, 15). To verify the role of NAT8 in EV71 infection, we designed three single guide RNAs (sgRNAs) targeting NAT8. Immunoblotting results showed that all three sgRNAs could target NAT8 efficiently (Fig. 1B). When control and NAT8 knockout (KO) SK-N-SH cells were infected with EV71, cytopathic effect (CPE) was observed in control cells at 24 h postinfection (hpi), whereas it was undetectable in NAT8 KO cells (Fig. 1C). In consonance with this, cell viability of NAT8 KO cells was significantly higher than that of control cells after infection (Fig. 1D), indicating that NAT8 is required for EV71-induced cell death. Viral RNA and VP1 expression were also dramatically reduced in NAT8 KO cells (Fig. 1E and F), suggesting that NAT8 is essential for EV71 propagation.

In order to confirm the specificity of the KO cell lines, sgRNA-resistant NAT8 was introduced back into KO cell lines to restore NAT8 expression (Fig. 1G). Exogenous expression of NAT8 also restored VP1 expression (Fig. 1G) and reduced cell viability significantly when cells were infected with EV71, as expected (Fig. 1H). All of these results suggested that NAT8 is an essential host factor required for EV71 infection.

NAT8 is required for EV71 replication.

To gain more mechanistic insight into NAT8, we sought to determine at which stage of the viral life cycle NAT8 functions. For this purpose, we infected wild-type and NAT8 KO cells with a full-length virus with a NanoLuc (NL) luciferase reporter (NL-EV71) (16), and luciferase activity was measured at various time points postinfection. Comparable luciferase activity was observed at 4 hpi, while the luciferase activity was significantly lower in NAT8 KO cells at 8 hpi (Fig. 2A), suggesting that NAT8 functions at a later stage of viral infection (viral replication, virus assembly, or virus release). To confirm this, we tested whether NAT8 could influence virus attachment and internalization. Figure 2B shows that comparable amounts of virus were attached to control and NAT8 KO cells, and similar results were obtained in virus internalization assays (Fig. 2C), suggesting that NAT8 is not required for virus entry. To test the role of NAT8 in virus replication, we employed a subgenomic replicon system (SGR) containing a Renilla luciferase reporter that bypassed the virus entry process (16). When cells were transfected with this SGR RNA, NAT8 KO cells generated luciferase amounts comparable to those of control cells at 4 h posttransfection. However, when luciferase levels were determined at 8 h posttransfection or beyond, dramatic decreases were observed for NAT8 KO cells, suggesting that NAT8 is important for viral genome replication (Fig. 2D).

Because NAT8 mainly localizes to the ER, where EV71 replication occurs, we then evaluated the subcellular localization of NAT8 during EV71 infection. SK-N-SH cells overexpressing NAT8-Flag were immunolabeled with anti-double-stranded RNA (dsRNA) antibody for viral RNA and anti-Flag antibody for NAT8. Figure 2E shows that NAT8 protein colocalized with EV71 dsRNA in the cytoplasm, suggesting that NAT8 functions at the places of viral replication. All of these results revealed that NAT8 mainly participated in viral replication during EV71 infection.

Acetyltransferase activity is required for NAT8 to support EV71 infection.

NAT8 is a specific ER-based acetyltransferase, which transfers the acetyl group from the donor to the acceptor. To test whether the acetyltransferase activity of NAT8 is required for viral infection, we made a catalytically dead mutant of NAT8. A previous study reported that a mutation affecting a conserved residue (R149K) abolished the enzymatic activity of NAT8 while barely interfering with the expression of the recombinant protein (17). We then tested whether the impaired infection caused by NAT8 deficiency could be rescued by complementation with NAT8/R149K. Our results showed that, while wild-type NAT8 could rescue EV71 infection and cause cell death, NAT8/R149K failed to support EV71 infection (Fig. 3A and B). These results suggest that the catalytic activity of NAT8 is required to support EV71 infection.

To further prove that the enzymatic activity of NAT8 is required for EV71 infection, we treated NL-EV71-infected SK-N-SH cells with a specific NAT8 inhibitor, i.e., compound 9, which has been reported to inhibit NAT8 acetyltransferase activity in vitro and in vivo (18). The results showed that the luciferase activity decreased in a dose-dependent manner, while cell viability was largely not affected (Fig. 3C). Consistent with this, viral RNA was decreased in a dose-dependent manner while NAT8 mRNA was unaffected when virus-infected cells were treated with compound 9 (Fig. 3D). However, when infected cells were tested for VP1 expression, we found both VP1 protein and NAT8 protein were inhibited after compound 9 treatment (Fig. 3E). The discrepancy between NAT8 mRNA and protein levels after compound 9 treatment was suggested by a previous report (18). Overall, these results suggest that compound 9 could inhibit EV71 infection by both inhibiting NAT8 acetyltransferase activity and downregulating its expression.

NAT8 is required for EV71 infection in different cell lines.

EV71 has been reported to infect different cell lines from various tissue origins (19), and we next tested whether NAT8 is required for EV71 infection in different cell lines. When SH-SY5Y, HUVEC, Vero, RD, 293T, Caco2, or HepG2 cells were infected with NL-EV71 and treated with compound 9, dose-dependent inhibition was observed for all cell lines tested (Fig. 4A to G, upper). Similarly, when these cells were infected with EV71 and VP1 expression was examined, compound 9 could inhibit VP1 expression in a dose-dependent manner (Fig. 4A to G, lower), suggesting NAT8 might play important roles in different cell lines with EV71 infection.

NAT8B and AT-1 are also required for EV71 infection.

NAT8 forms homodimers or heterodimers with another acetyltransferase, NAT8B, and also requires the acetyl-CoA transporter AT-1 to exert this N-lysine acetylation activity (12, 20). We next wanted to test whether these two proteins are also required for EV71 infection. Since NAT8 and NAT8B are paralogs and have a high level of sequence similarity, we first tested whether the sgRNAs we used to generate KO cells had off-target effects. Figure 5A shows that NAT8 sgRNAs inhibit only NAT8 expression, without affecting NAT8B expression. Similarly, NAT8B sgRNAs KO only NAT8B expression, suggesting that the specificity of the sgRNAs we use is robust. Next, when NAT8B KO or AT-1 KO SK-N-SH cells were infected with EV71, viral VP1 protein was dramatically reduced (Fig. 5B). In addition, compared to control cells, CPE was undetectable in NAT8B or AT-1 KO cells (Fig. 5C), and cell viability was significantly increased when these cells were infected with EV71 (Fig. 5D). These results suggested that NAT8B and AT-1 are also important for EV71 infection.

NAT8 interacts with viral NS proteins.

Because NAT8 resides mainly on ER and is colocalized with viral dsRNA, we speculated that NAT8 might interact with NS proteins of EV71 to promote viral replication. To test this hypothesis, we first cotransfected 293T cells with the plasmid expressing NAT8-Flag and plasmids expressing green fluorescent protein (GFP)-tagged EV71 NS proteins 2B, 2C, 3AB, 3C, or 3D. OST48, which has been reported to interact with NAT8 (20), was used as a positive control. Figure 6A shows that NS proteins 2B, 3AB, and 3C and OST48 coimmunoprecipitated with NAT8-Flag, while 2C and 3D did not, indicating that NAT8 could interact with EV71 2B, 3AB, and 3C proteins. To test whether NAT8 interacts with NS proteins during EV71 infection, SK-N-SH cells expressing NAT8-Flag were infected with EV71. Figure 6B shows that NS protein 3AB or 3C coimmunoprecipitated with NAT8. NS protein 2B was not tested due to the lack of suitable antibody. These results suggest that NAT8 interacts with viral NS proteins during EV71 infection.

We then immunostained NAT8 with 2B, 3AB, or 3C proteins in cotransfected 293T cells and found that they were perfectly colocalized (Pearson correlation coefficients of 0.89, 0.94, and 0.92, respectively) (Fig. 6C). Furthermore, colocalization of NAT8 with 3AB or 3C was also observed in NAT8-Flag-expressing SK-N-SH cells infected with EV71 (Pearson correlation coefficients of 0.83 and 0.72, respectively) (Fig. 6D), further suggesting that NAT8 could interact with EV71 NS proteins.

NAT8 enhances the stability of NS proteins.

To test the effect of NAT8 on NS protein expression, 293T cells were transfected with expression vectors for GFP-NS proteins with or without NAT8 coexpression, and then the GFP-NS protein expression level was examined. Figure 7A shows that the expression of 2B, 3AB, and 3C proteins was upregulated in the presence of NAT8 proteins. On the other hand, when 293T cells transfected with expression vectors for GFP-NS proteins were treated with compound 9, which inhibited NAT8 function, the GFP-NS protein expression level was decreased in a dose-dependent manner (Fig. 7B), suggesting that NAT8 could upregulate NS protein expression.

Previous studies reported that CD133 protein expression is positively regulated by NAT8 via increased stability (21), and we then wanted to test whether NAT8 increased the expression of 2B, 3AB, and 3C by enhancing the protein stability. To test this possibility, the amounts of 2B, 3AB, and 3C were determined in the presence of the protein synthesis inhibitor cycloheximide with or without NAT8 coexpression in 293T cells. In control cells, the amounts of 2B, 3AB, and 3C decreased significantly over 8 h of cycloheximide treatment (Fig. 8A); in contrast, proteins maintained similar amounts in the presence of NAT8 expression (Fig. 8A), suggesting that NAT8 could enhance 2B, 3AB, or 3C stability. Similar results were obtained when EV71-infected SK-N-SH cells were treated with cycloheximide; overexpression of NAT8 significantly increased viral 3AB and 3C stability, compared to control cells (Fig. 8B). On the other hand, when EV71-infected SK-N-SH cells were treated with compound 9 (10 μM) in the presence of cycloheximide, accelerated degradation of 3AB or 3C was observed, compared to control cells (Fig. 8C), Taken together, these results indicated that NAT8 increased the protein stability of viral NS proteins.

NAT8 stabilizes NS proteins by inhibiting autophagy-lysosome degradation.

The ubiquitin-proteasome system and autophagy-lysosome degradation are the two major intracellular protein degradation pathways in eukaryotes (22, 23). Proteins degraded by proteasomes are modified by ubiquitin chains, and this pathway could be blocked by MG132 treatment, whereas autophagy-mediated protein degradation relies on fusion with lysosomes and thus can be blocked by bafilomycin A1 (BafA1) treatment (Fig. 9A). To test which pathway was involved in NS protein degradation during NAT8 inhibition, 293T cells were transfected with 2B-, 3AB-, or 3C-expressing construct and then treated with compound 9 in the presence of MG132 or BafA1. We found that BafA1antagonized the protein degradation by compound 9 treatment, while MG132 did not (Fig. 9B to D), suggesting that NAT8 increased the stability of 2B, 3AB, and 3C proteins by inhibiting autophagy-lysosome pathway-dependent degradation.

Cells and reagents.

SK-N-SH and SH-SY5Y cells were maintained at 37°C in modified Eagle’s medium (MEM) (Gibco) containing 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 1 μg/mL streptomycin, in the presence of 5% CO₂. 293T, RD, Caco2, HepG2, and Vero cells were maintained at 37°C in Dulbecco’s MEM (DMEM) (Gibco) supplemented with 10% FBS, 100 U/mL penicillin, and 1 μg/mL streptomycin, in the presence of 5% CO₂. HUVEC cells were maintained at 37°C in endothelial cell medium (ECM) (ScienCell) containing 10% FBS, 100 U/mL penicillin, and 1 μg/mL streptomycin, in the presence of 5% CO₂. The following antibodies were used in this study: NAT8 (ABclonal), NAT8B (Bioss), SLC33A1 (Bioss), β-actin (ABclonal, Woburn, MA), GFP (Thermo Fisher scientific), and Flag (Origene) antibodies. EV71 VP1, 3AB, and 3C antibodies were purchased from GeneTex. dsRNA antibody (J2) was from English and Scientific Consulting (Hungary). 4′,6-Diamidino-2-phenylindole (DAPI)- or Alexa Fluor-conjugated secondary antibodies were purchased from Thermo Fisher Scientific. Compound 9 was kindly provided by Luigi Puglielli (University of Wisconsin-Madison).

Viruses, reporter viruses, and infectious clones.

The EV71 strain and the full-length EV71 with a luciferase reporter (NL-EV71) and subgenomic replicon SGR-EV71 were described previously (16). They were propagated and titrated on RD cells.

Constructs.

The cDNA encoding human NAT8 was obtained by reverse transcription (RT) of total RNA from SK-N-SH cells, followed by PCR using the specific primers listed in Table S2 in the supplemental material. The cDNAs were subcloned into a lentiviral expression vector (pHW200) with a C-terminal Flag tag with MluI and BsiwI sites. NAT8sg3Mut and NAT8R149K mutant were obtained with overlap extension PCR amplification using the primers listed in Table S2 and then cloned were into the same vector between the MluI and BsiwI sites. To construct plasmids expressing 2B, 2C, 3AB, 3C, and 3D, the fragments were amplified from EV71 cDNA and then cloned into the lentiviral expression vector with a C-terminal GFP tag with MluI and BsiwI sites; the primers are listed in Table S2. All constructs were confirmed by sequencing.

Lentiviral packaging and transduction.

Lentivirus stock was prepared as described previously (32). Briefly, 293T cells were cotransfected with psPAX2, pMD2.G, and pLentiCrispr (for KO) or lentiviral expression vector (for overexpression). The sgRNA sequences used are listed in Table S3. Forty-eight hours posttransfection, lentivirus was harvested from the supernatant, filtered through a 0.45-μm filter, and used for transduction. Lentiviral transduction was performed on designated cell lines, and stable cell lines were generated with puromycin or blastcidin S (InVivoGen) selection.

Genome-scale CRISPR/Cas9 KO screens.

The pooled CRISPR screening was conducted as described previously (33). Briefly, the human GeCKOv2 plasmid library from Addgene was first prepared as lentiviral pools, as described above. SK-N-SH cells transduced with a lentiviral pool at a multiplicity of infection (MOI) of 0.3 were then selected with 1 μg/mL puromycin for stable cell lines. Cells were then infected with EV71 at an MOI of 0.1 for 2 weeks. Genomic DNA from surviving cells was prepared with the Quick-gDNA midikit (Zymo Research, Irvine, CA), and the integrated sgRNAs were amplified using PCR. Next-generation sequencing was performed at Oebiotech (Shanghai, China) on a MiSeq instrument (Illumina, CA), and data were analyzed with the MAGeCK algorithm (34).

Western blotting.

Cells with proper treatment were harvested with LDS sample buffer (Thermo Fisher Scientific), followed by separation on a 10% SDS-PAGE gel. Proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane before they were incubated with primary antibodies. Following a Tris-buffered saline with Tween20 (20mM Tris-Cl [pH 7.5], 150mM NaCl, 0.05% Tween20) (TBST) wash, corresponding secondary antibodies were added and bands were visualized with SuperSignal West Femto maximum sensitivity substrate (Thermo Fisher Scientific).

RNA extraction and RT-qPCR.

Different cells with or without infection were harvested, and total cellular RNA was isolated with the GeneJET RNA purification kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. cDNA was synthesized using a high-capacity cDNA RT kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. cDNA was diluted 1:20 and used directly for quantitative PCR (qPCR) analysis using specific primers and the iTaq universal SYBR green master mix (Bio-Rad). Primer sequences are listed in Table S4.

Luciferase assay.

Cells infected with NL-EV71 were harvested at 24 hpi, and luciferase signals were determined with the Nano-Glo luciferase assay system (Promega Corp.) according to the manufacturer’s instructions. For cells transfected with SGR-EV71, viral RNA was in vitro transcribed as described previously (16) and was transfected into SK-N-SH cells or NAT8 KO SK-N-SH cells using the TransIT-mRNA transfection reagent (Mirus Bio) according to the manufacturer’s instructions. Renilla luciferase was measured using a Renilla luciferase assay system (Promega Corp.) at the indicated time points. All luciferase was measured with a BioTek Neo2 microplate reader.

Viral binding and internalization assays.

Control or NAT8 KO SK-N-SH cells were incubated with EV71 (MOI of 5) at 4°C for 1 h (binding assay) or incubated at 4°C for 1 h followed by 1 h of incubation at 37°C (internalization assay). Cell media were then removed, and cells were washed five times with cold phosphate-buffered saline (PBS). Total RNA was extracted, and viral RNA was determined by qPCR as described above.

Compound 9 treatment assay.

Cells were seeded on a 24-well plate 1 day before they were infected with EV71 or NL-EV71. Four hours later, the cells were treated with compound 9 at the indicated concentrations. Cells were harvested for immunoblotting, RT-qPCR, or the luciferase assay as described above.

Protein degradation assay.

293T cells were transfected with EV71 NS-GFP proteins, and compound 9, MG132 (10 μM), or BafA1 (10 nM) was added at 12 h posttransfection and maintained for 24 h. Cells were harvested for immunoblotting as described above. For EV71-infected cells, SK-N-SH cells were either treated with 10 μM compound 9 or overexpressed NAT8-Flag. Twenty-four hours after EV71 infection, cells were treated with 125 μM cycloheximide and harvested at 0, 4, or 8 h for immunoblotting.

Coimmunoprecipitation.

For 293T cells, cells were first cotransfected with NAT8-Flag- and EV71 NS-GFP-encoding plasmids. For SK-N-SH/NAT8-Flag cells, cells were infected with EV71 at an MOI of 5. Twenty-four hours later, the cells were harvested and washed twice with cold PBS, followed by disruption with immunoprecipitation buffer (50 mM Tris [pH 7.4], 150 mM NaCl, 1% Nonidet P-40, 0.1% SDS) at 4°C for 15 min. Cell lysates were clarified by centrifugation at 1,000 × g for 5 min at 4°C. Anti-Flag antibody was mixed with the supernatant and incubated at 4°C for 1 h; the precleared beads with protein A/G were then added to the mixture and incubated at 4°C for 1 h on a roller. The reaction mixtures were then washed four times with cold wash buffer (0.05% Triton X-100) and subsequently analyzed by immunoblotting with the indicated antibodies.

Confocal microscopy.

Immunofluorescence staining was carried out as described previously (16). Briefly, cells were seeded on coverslips coated with poly-d-lysine 1 day before they were infected with EV71 (MOI of 5) and incubated for another 24 h. Cells were then fixed with ice-cold methanol for 10 min, followed by blocking with 2% FBS for 1 h. Cells were then stained with the corresponding antibodies. The images were acquired with a Nikon C2 confocal microscope.

Statistics.

Unless otherwise indicated, all values are means ± standard deviations and represent the results of a minimum of three independent experiments. The two-tailed Student's t test was used to compare the means of control and experimental groups.