HERP Binds TBK1 To Activate Innate Immunity and Repress Virus Replication in Response to Endoplasmic Reticulum Stress

A wide range of pathogens is sensed by distinct pattern recognition receptors (PRRs) of the innate immune system, leading to activation of the specific signaling pathways, which results in the stimulation of antimicrobial proteins. Three major classes of PRRs that sense virus infection have been reported: 1) TLRs, including TLR3/7/8/9 located at the endosome; 2) RIG-I–like receptors (RLRs) located at the mitochondrial membrane; and 3) NOD-like receptors found in lymphocytes, macrophages, and dendritic cells (1, 2). Besides the PRRs response to non–self-cytosolic RNA, physiological endoplasmic reticulum (ER) stress also invokes innate immune signaling in response to invading viral RNA. The ER plays crucial roles in the cellular homeostasis via processing and folding proteins. Simultaneously, pathogens can interact with the ER by regulating phagosomes, stimulating the downstream components such as TANK-binding kinase 1 (TBK1)/inducible IκB kinase (IKKi), IFN regulatory factors (IRFs), and NF-κB, which further induces the IFNs and expression of IFN-stimulated genes (ISGs) to inhibit the viral replication and eradicate virus-infected cells (1, 3–5). The IFN-mediated pathways play a crucial role for the proper cellular response against various viral infections. The antiviral responses mediated by type-III IFNs are very similar to those of type-I IFNs, via a distinct receptor complex (6, 7). Consistent with type-I IFNs, the transcription of IFN-λ1 requires the activation of IRF3/7 and NF-κB pathways (8).

Homocysteine-induced ER protein (HERP) is an ER-resident membrane protein, which has a ubiquitin-like (UBL) domain at its N terminus and transmembrane domain(s) at the C terminus (9). HERP is upregulated in response to ER stress and is induced by the unfolded protein response. HERP contains an ER stress response element in its promoter region, and an N-terminal UBL domain which interacts with the ER-associated degradation system to gain a ubiquitylation function (10–12). HERP is involved in the improvement of folding capacity as well as in the protein-loading capacity of the ER, which plays a critical role in several diseases mechanisms in patients with diabetes and Parkinson’s disease (13, 14). However, the role of HERP in the regulation of the innate immune response upon virus infection is unknown.

Enterovirus 71 (EV71) is a small, nonenveloped, positive-strand RNA virus, belonging to Picornaviridae family. It was recognized in the United States in 1974 and has since spread worldwide (15, 16). EV71 infection leads to hand, foot, and mouth disease, and may also cause neurological diseases (17, 18). Our previous study suggested that the ER is an essential organelle for EV71 replication (19), and it was also reported that EV71 induces ER stress which leads to modulation of viral replication (20, 21). An IFN response may be a consequence of this ER stress (22). Several ER stress–induced or ER-associated proteins are involved in the regulation of the IFN pathway, influencing viral replication in many ways (23, 24). However, the mechanism of how ER stress–associated HERP mediates IFN regulation during EV71 replication remains unclear.

In this study, we demonstrated that EV71 infection upregulates the expression of HERP both in vitro and in vivo, and further revealed that the ER stress response mediated by mitochondrial antiviral signaling protein (MAVS) enhances HERP expression. HERP subsequently induces the production of type-I IFNs (IFN-α and IFN-β) and type-III IFN (IFN-λ1) by regulating the phosphorylation and nuclear translocation of IRF3 and NF-κB. HERP binds TBK1 to amplify the MAVS signaling and facilitate the expression of IFNs, which leads to the repression of the replication of RNA viruses, including EV71, Sendai virus (SeV), influenza A virus (IAV), and vesicular stomatitis virus (VSV). These findings provide new insights into the mechanism underlying the replication of RNA viruses and the production of IFNs upon viral infection, and also demonstrate a new role for HERP in the regulation of host innate immunity in response to ER stress during viral infection.

The BALB/c mice in this study were purchased from Shanghai Laboratory Animal Center. All mice were housed under specific pathogen-free conditions in individually ventilated cages. One-day-old suckling mice were i.p. injected with 1 × 10⁷ PFUs of MA-EV71 (mouse-adapted EV71 strain) in 50 μl PBS. In the EV71-infection mice model, mice were sacrificed at 10 d postinfection. Simultaneously, PBMCs from mice were harvested and subjected to RNA extraction. Animal care and sacrifice were conducted according to methods approved by the Animal Care and Use Committee, the Center for Animal Experiments of Wuhan University.

Peripheral blood specimens were randomly obtained from 24 EV71-infected patients from May 2 to May 10, 2017, with matched sex and age (Table I). All participants were diagnosed with EV71 infection by the presence of EV71 RNA with the specific EV71 VP1 primer. Prior to this study, patients did not suffer any concomitant disease at the time of sampling, did not show any serological markers suggestive of autoimmune disease, and also had not received any antiviral or immunomodulatory therapy.

The Institutional Review Board of the College of Life Sciences, Wuhan University, approved the collection of blood samples for this research, in accordance with the guidelines for the protection of human subjects. Written informed consent was obtained from each participant.

Human rhabdomyosarcoma (RD) cells, human embryonic kidney 293T (HEK293T) cells, and human acute monocytic leukemia (THP-1) cells were purchased from American Type Culture Collection and cultured in Life Technologies MEM, DMEM, and RPMI 1640 medium (Invitrogen, Carlsbad, CA), respectively, and supplemented with 10% FBS (Life Technologies), 100 U/ml penicillin, and 100 mg/ml streptomycin sulfate at 37°C in a 5% CO₂ incubator. MAVS^−/− HEK293T cells were kindly provided by Dr. D. Chen of Peking University, Beijing, China.

To isolate primary mouse bone marrow–derived macrophages (BMDMs), 6-wk-old male wild-type (WT) or IFNAR1^−/− A129 mice (a gift from Dr. C. Qin, Academy of Military Medical Science, Beijing, China) were sacrificed and pelvic and femoral bones were collected to separate bone marrow. After RBC lysis buffer treatment, bone marrow cells were cultured in RPMI 1640 medium with 10% FBS followed by 100 ng/ml mouse GM-CSF (PeproTech) induction for 3 d until cells became attached.

The growth and virus titration of EV71 (Xiangyang-Hubei-09, GenBank accession number JN230523.1), SeV (a gift from Dr. H. Shu of Wuhan University, Wuhan, China), IAV H3N2 (A/HongKong/498/97, provided by the China Center for Type Culture Collection), and VSV-GFP (a gift from Dr. M. Chen of Wuhan University, Wuhan, China) were performed as described previously (25, 26). Cells were infected with viruses at the indicated multiplicities of infection (MOIs) and unbound virus was washed away 2 h later, and cells were then incubated at 37°C.

The full-length human HERP gene (GenBank accession number NM_001010989; https://www.ncbi.nlm.nih.gov/nuccore/NM_001010989.2) was amplified from cDNA isolated from HEK293T cells and inserted into the pCMV vector (pCMV-FLAG2B) (Invitrogen) to generate pCMV-HERP. The full-length mouse HERP gene was amplified from cDNA isolated from Raw264.7 and inserted into pLenti-CMV-3×FLAG, which was derived from the backbone of pLenti-CMV-MCS-GFP-SV-puro (plasmid number 73582; Addgene, Cambridge, MA).

Truncated HERP genes were cloned into the pCMV vector to encode FLAG-ΔUBL (aa 88–391), FLAG-ΔC (aa 1–240), FLAG-ΔCΔUBL (aa 88–240) and FLAG-UBL (aa 1–87). The primers used in this study are listed in Supplemental Table I.

The DNA fragments of TLR7, MyD88, IRAK1, TNFR-associated factor 3 (TRAF3), TRAF6, TAB1, TAK1, and TBK1 genes were ligated into the eGFP-C1 (Takara) vector to generate plasmids expressing GFP-tagged proteins.

The pGL3–5×NF-κB luciferase reporter was provided by Prof. H. Shu (Wuhan University), and pGL3–IFN-stimulated response element (ISRE), pGL3–IFN-β, and pGL3–IFN-λ1 promoter luciferase reporters were provided by Prof. Y. Zhu (Wuhan University).

Recombinant lentiviruses carrying the mouse HERP-expressing gene (Lenti-mHERP) or the empty vector (3×FLAG tag; Lenti-3F) were generated as follows: three plasmids, pLenti-CMV-mHERP or -3×FLAG, psPAX2, and pMD2.G (Addgene) were cotransfected into HEK293T cells in the ratio 4:3:1. At 72 h posttransfection, lentiviral particles in harvested media from HEK293T cells were transferred to BMDMs for additional 4-d culture until further assay.

IFN-α purchased from (3SBio, Shenyang, China) was added at a final concentration of 300 U/ml. IFN-λ1 purchased from Proteintech (Chicago, IL) was used at a final concentration of 50 ng/ml. The synthetic analogue of dsRNA, polyinosinic-polycytidylic acid [poly(I:C)] (InvivoGen, San Diego, CA) was used at a final concentration of 200 ng/ml.

Rabbit anti–EV71-VP1 Ab was obtained from Abnova (Taiwan, China). Rat anti-FLAG and rat anti-CD68 Abs were purchased from BioLegend (San Diego, CA). Rabbit anti–EV71-3C Ab was raised against residues 76–88 of 3C protein (Abgent, Suzhou, China). Goat anti-HERP, rabbit anti-IRF3, anti-IRF7, anti–NF-κB p65, anti–NF-κB p50, and anti–Lamin A Abs were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti-MAVS, anti-TBK1, anti–phosphorylated TBK1 (p-TBK1, Ser 172), anti–phosphorylated NF-κB p65 (p-p65, Ser 536), and anti–phosphorylated IRF3 (p-IRF3, Ser 386) Abs were purchased from Cell Signaling Technology (Beverly, MA). Rabbit anti-IFNAR1 Ab was purchased from Abcam (Cambridge, MA). Mouse anti–dsRNA mAb J2 was obtained from Scicons. Mouse anti-GFP, anti-Myc, anti–β-actin, and anti-GAPDH Abs were purchased from Proteintech Group.

The proteins in the supernatant of transfected cells were measured with human IFN-α and IFN-λ1 immunoassays (eBioscience, Santa Diego, CA) following the manufacturer’s instructions.

Cells were harvested and whole cell lysates were prepared on ice by lysing cells in PBS (pH 7.4) containing 0.01% Triton X-100, 0.01% EDTA, and 10% protease inhibitor mixture (Roche, Indianapolis, IN). Lysate was successively incubated with mouse anti-Myc or normal mouse IgG Abs for 4 h and then incubated with protein G agarose (GE Healthcare) for another 1 h. To separate and collect the cytosolic and nuclear protein fractions, cells were washed and resuspended in hypotonic buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 0.5 mM DTT, 10% protease mixture inhibitor) for 15 min on ice. Nuclei were pelleted by centrifugation at 12,000 rpm for 1 min, and the pellets and cytosolic protein–containing supernatants were collected.

The protein concentration of each sample was determined with the Bio-Rad (Hercules, CA) Protein Assay. Protein samples were subjected to 12% SDS-PAGE and then transferred to a nitrocellulose membrane (Amersham, Piscataway, NY). Blots were developed with the SuperSignal Chemiluminescent reagent (Pierce, Rockford, IL) and analyzed using a Luminescent Image Analyzer (Fujifilm LAS-4000; Fujifilm, Tokyo, Japan). Between the blotting of Abs and phosphorylation-specific Abs, the membrane was stripped in 0.2% SDS, 62.5 mM Tris-HCl, pH 6.8, and 0.1 mM 2-ME for 30 min at 50°C.

HEK293T cells and RD cells were plated at the appropriate density (1 × 10⁶ cells per well of a six-well plate), depending on the experiment, and were grown to 80% confluence prior to transfection. Cells were transfected with plasmids using Lipofectamine 2000 (Invitrogen) for 24 h and then stimulated or harvested. The dual-luciferase reporter assay system (Promega, Madison, WI) was used to measure the luciferase activity of each sample 48 h after transfection, and Renilla luciferase reporter vector pRL-TK activities were determined as internal controls for transfection efficiency. Assays were performed in triplicate, and the results were expressed relative to the vector or control samples, which were set at 100%. For knockdown of HERP, the small interfering (siRNA) specific to HERP (siR-HERP) or control siRNA (siR-Ctrl) was transfected into THP-1 cells using INTERFERin (Polyplus Transfection Company, Illkirch, France) or HEK293T cells using Lipofectamine 2000, respectively.

The siRNA specific to the negative control and HERP were synthesized by RiboBio (Guangzhou, China). siR-Ctrl targeted the sequence 5′-TTCTCCGAACGTGTCACGT-3′, siHERP#1 targeted the sequence 5′-GGGAAGTTCTTCGGAACCT-3′, and siHERP#2 targeted the sequence: 5′-TGTTCAAATTACACTAAGT-3′.

Quantitative real-time PCR (qRT-PCR) amplifications were performed to determine relative mRNA levels in the presence of specific primers. Total cellular RNA was isolated with TRIzol reagent (Invitrogen). Cellular RNA samples were reverse transcribed with random primers. Detection was performed in a LightCycler480 (Roche). The data represent absolute mRNA copy numbers normalized to GAPDH, which was used as a reference gene. The relative fold expression values were determined applying the ΔΔ cycle threshold method. The sequences of real-time PCR primers are listed in Table II.

Tissue samples were mounted on slides from paraffin blocks (5-μm sections), deparaffinated three times in xylene for 5 min, and hydrated in a methanol gradient (100, 95, 70, and 50%). H₂O₂ (3%) and 10 mM citrate buffer (pH 6) were used for Ag retrieval. Unspecific peroxidase activity was blocked for 30 min with 5% BSA. The slides were incubated with the primary Ab overnight at 4°C. Following this, the slides were washed with PBS for 10 min. The incubation of biotinylated secondary Ab was performed initially for 30 min and subsequently with the avidin biotin complex kit (Dako) for an additional 30 min. For detection, 3,3′-diaminobenzidine tetrahydrochloride hydrate with 5% H₂O₂ was used. The slides were counterstained with hematoxylin to stain nuclei. Immunohistochemistry (IHC) was conducted at each site using specific Abs and samples were visualized by microscopy (Olympus, Tokyo, Japan).

HEK293T cells were prepared on 20-mm covered glass bottom dishes. After being fixed and permeabilized, the cells were blocked in PBS containing 5% BSA for 1 h, and incubated with Abs for 3 h at 37°C. The samples were incubated with TRITC-conjugated donkey anti-goat, FITC-conjugated donkey anti-rabbit, or AMCA-conjugated donkey anti-mouse IgG (Proteintech Group) for 45 min at room temperature. To stain nuclei, 1 μg/ml DAPI (Roche) methanol solution was added and samples were incubated for 15 min at room temperature. After washing with PBS, samples were visualized by confocal laser scanning microscopy (Fluoview FV1000; Olympus).

All experiments were reproducible, and each set of experiments was repeated at least three times with similar results. Statistical significance for comparison of two means was assessed by an unpaired Student t test. For dose-dependent experiments or multiple comparisons, a one-way ANOVA test was used followed by a post hoc test (Dunnett or Tukey test). Analyses were performed using the Prism 5 software (GraphPad). Means were illustrated using a histogram with error bars representing ± SD, and statistical relevance was evaluated using the following p values: *p < 0.05, **p < 0.01.

The expression of HERP protein is elevated in response to ER stress in physiological or pathological conditions (9). EV71 infection induces ER stress and ER-resident chaperone proteins (20). To explore the effect of EV71 infection on the expression of HERP, human RD cells were mock infected or infected with EV71 or UV-inactivated EV71. The results showed that HERP mRNA (Fig. 1A) and protein (Fig. 1B) were significantly induced by EV71 at 8 h postinfection (h p.i.). In addition, in RD cells infected with EV71 at different MOI, HERP mRNA (Fig. 1C) and HERP protein (Fig. 1D) were elevated by EV71 in a dose-dependent manner. Therefore, HERP expression is induced by EV71 infection in vitro.

The effect of EV71 infection on HERP was further evaluated in BALB/c mice. We conducted IHC and qRT-PCR approaches to examine targeted proteins and RNAs in EV71-infected mice. The viral VP1 protein was detected in the spleen of infected mice (Fig. 1E, left panel). HERP protein was significantly upregulated in the spleen of EV71-infected mice (Fig. 1E, middle panel). Importantly, the number of CD68⁺ cells were sharply increased in the spleen of EV71-infected mice (Fig. 1E, right panel), suggesting a strong immune response during EV71 infection (27). We next tested mRNA levels of HERP and type-I IFN in mouse PBMCs. qRT-PCR results revealed that HERP and IFN-β mRNA were significantly induced in PBMCs of EV71-infected mice (Fig. 1F). Moreover, the Pearson correlation analysis showed that the HERP mRNA level was positively correlated with the IFN-β mRNA level in PBMCs of EV71-infected mice (Fig. 1G). To verify the correlation between HERP and IFNs, we determined the mRNA levels of HERP, IFN-β, and IFN-λ1 in PBMCs of 24 EV71-infected patients (Table I). Consistently, the Pearson correlation analysis showed that the HERP mRNA level was positively correlated with the IFN-β or IFN-λ1 mRNA level (Fig. 1H), indicating a potential role of HERP in IFN action. Moreover, we also found the EV71 VP1 RNA level was positively correlated with the HERP mRNA level in both EV71-infected PBMC from mice (Supplemental Fig. 1A) and patients (Supplemental Fig. 1B). Taken together, EV71 infection induces HERP expression both in vitro and in vivo.

ER stress–induced or ER-associated proteins were involved in viral replication through the regulation of the IFN pathway in various ways (23, 24). Because HERP is induced by EV71, the effect of HERP on EV71 replication was examined. RD cells were transfected with pCMV-HERP and infected with EV71. The results showed that EV71 VP1 RNA (Fig. 2A) and VP1 and 3C proteins (Fig. 2B) were downregulated in the presence of HERP in a dose-dependent fashion. Moreover, RD cells were transfected with FLAG-HERP and infected with EV71 over different times. The results revealed that EV71 VP1 RNA (Fig. 2C) and the VP1 and 3C proteins (Fig. 2D) were upregulated as EV71 infection time increased, but downregulated in the presence of HERP in a time-dependent manner. Thus, we revealed that HERP dramatically represses EV71 replication.

To further confirm the inhibitory effect of HERP on EV71 replication, two independent siRNAs specific to HERP (siR-HERP#1 and siR-HERP#2) were constructed and the efficiency of siRNA knockdown was evaluated (Supplemental Fig. 2A). RD cells were transfected with siRNAs specific to HERP and infected with EV71. The results showed that EV71 VP1 RNA and protein were elevated when HERP was knocked down (Fig. 2E, 2F). The mRNA levels of IFN-α, IFN-β, and IFN-λ1 were significantly downregulated in the presence of siR-HERP#1 and siR-HERP#2 (Supplemental Fig. 2B), suggesting that HERP regulates the expression of IFNs. Thus, we demonstrated that HERP attenuates EV71 replication, which may rely on the regulation of IFN action.

Considering the inhibition of HERP on EV71 replication, we subsequently determined the effects of IFN-α and IFN-λ1 on the replication of EV71. RD cells were treated with IFN-α at different concentrations and infected with EV71. The results revealed that EV71 VP1 RNA was reduced after the treatment of IFN-α in a dose-dependent manner (Fig. 3A). Moreover, EV71 RNA (Fig. 3B), protein (Fig. 3C), and virus copy number (Fig. 3D) were repressed by IFN-α and downregulated by IFN-λ1, which ranged from 5 to 50 ng/ml, in dose-dependent fashions. Notably, IFN-λ1 repressed EV71 copy number in a time-dependent manner (Fig. 3E), which is consistent with the previous report (28), confirming that IFN-α and IFN-λ1 effectively repress EV71 replication. To further verify the correlation between HERP and IFNs in the regulation of antiviral activity, RD cells were treated with IFN-α or IFN-λ1, transfected with siR-Ctrl or siR-HERP (siR-HERP#1), and infected with EV71. The results showed that EV71 VP1 RNA was significantly reduced by IFN-α and IFN-λ1, but this reduction was recovered in the presence of siR-HERP (Fig. 3F), implicating that type-I and -III IFNs were involved in the suppression of EV71 replication through HERP regulation.

To examine the role of HERP in the regulation of IFNs, we used poly(I:C) as a positive stimulus. HEK293T cells were transfected with pCMV-HERP for 8, 12, or 24 h. The mRNA levels of IFN-α, IFN-β, and IFN-λ1 were stimulated by poly(I:C) as expected, and strongly induced by HERP overexpression in a time-dependent manner (Fig. 4A). The ELISA results showed that the secreted protein levels of IFN-α and IFN-λ1 were increased by HERP overexpression in RD cells (Fig. 4B) and HEK293T cells (Fig. 4C). We also noticed that, compared with type-III IFNs (IFN-λ1), the induction of type-I IFNs (IFN-α/β) was more sensitive to the accumulation of HERP. Taken together, we demonstrated that HERP promotes the production of IFN-α, IFN-β, and IFN-λ1.

The mechanism involved in the regulation of IFNs mediated by HERP was then investigated. HEK293T cells were cotransfected with pCMV-HERP and individual reporter plasmids pGL3-ISRE, pGL3–IFN-β, pGL3–IFN-λ1, and pGL3–NF-κB. Luciferase activity assays showed that the activities of ISRE, the IFN-β promoter, the IFN-λ1 promoter, and NF-κB were significantly activated in the presence of HERP (Fig. 4D). Considering HERP is associated with the ER stress response to viral infection, HEK293T, THP-1, and RD cells were infected with SeV. The mRNA levels of IFN-α, IFN-β, and IFN-λ1 were significantly induced by SeV in RD cells (Supplemental Fig. 3A), HEK293T cells (Supplemental Fig. 3B), and THP-1 cells (Supplemental Fig. 3C); indicating that these cell lines are sensitive to SeV infection. To test whether the induction of IFNs mediated by EV71/SeV infection or poly(I:C) stimulation results from ER stress, we detected XBP1 mRNA splicing status as the marker in the process (29). XBP1 mRNA was spliced during EV71 and SeV infection, but not by poly(I:C) stimulation (Fig. 4E), demonstrating that ER stress is caused by EV71 or SeV infection rather than poly(I:C) stimulation. In HEK293T cells, NF-κB activity was activated by SeV infection and further enhanced in the presence of HERP (Fig. 4F).

The function of HERP in the production of IFNs was further evaluated by using siRNAs specific to HERP. The results showed that siR-HERP attenuated the activities of IFN-β, IFN-λ1, and NF-κB induced by EV71 or SeV infection; but it had no effect on the activities of IFN-β and IFN-λ1 stimulated by poly(I:C) as well as NF-κB stimulated by the TNF-α protein (Fig. 4G). THP-1 cells were then transfected with siR-HERP, infected with EV71 or SeV, or treated with poly(I:C). The efficiency of siRNA knockdown was evaluated (Supplemental Fig. 2C). IFN-α mRNA was induced by SeV but not by EV71 infection (Fig. 4H, left panel), whereas IFN-β mRNA (Fig. 4H, middle panel) and IFN-λ1 mRNA (Fig. 4H, right panel) were induced by both EV71 and SeV infection, and these inductions were attenuated in the presence of siR-HERP. The mRNAs of IFN-α, IFN-β, and IFN-λ1 were stimulated by poly(I:C), but these activations were not influenced by siR-HERP (Fig. 4H). These results suggested that HERP accelerates the production of IFNs depending on ER stress.

The activation of multiple intracellular signaling cascades, including IRF3/7 and NF-κB, are required for the induction of IFNs (8). In this study, the effect of HERP on the functions of IRF3 and NF-κB was investigated. RD cells were transfected with pCMV-HERP over different times. IRF3, NF-κB p50, and NF-κB p65 proteins were not affected in the whole cell lysis (Fig. 5A, left panel), but were reduced in cytoplasmic extracts (Fig. 5A, middle panel); whereas they were increased in nuclear extracts (Fig. 5A, right panel) after HERP overexpression, suggesting that HERP facilitates the nuclear translocations of IRF3 and NF-κB.

Since phosphorylation of IRF3 and NF-κB are essential for the regulation of IFNs, which is modulated by TBK1 activation (30), we determined the effect of HERP on phosphorylation of TBK1, IRF3, and NF-κB during virus infection. Cells were transfected with siR-HERP and infected with SeV or EV71 over different times. Phosphorylation of IRF3, NF-κB p65, and IκBα were enhanced by SeV (Fig. 5B) or EV71 infection (Fig. 5C); but these activations were attenuated in the presence of siR-HERP (Fig. 5B, 5C), suggesting that HERP is required for virus-induced phosphorylation of IRF3 and NF-κB. Taken together, we demonstrated that HERP facilitates virus-induced phosphorylation of IRF3 and NF-κB, resulting in the accumulation of antiviral proteins.

The regulation of IFNs relies on the assembly of a series of components in the cytosol, including the TRAF3–TBK1–IκB kinase ε (IKKε) complex (TRAF3–TBK1–IKKε), to activate IRFs and NF-κB (31). The role of HERP in the regulation of the IFN-associated TRAF3–TBK1–IKKε complex was assessed. HEK293T cells were cotransfected with the plasmid expressing Myc-tagged HERP and GFP-tagged TRAF3, IKKε, TBK1, IRAK1, TRAF6, TAK1, and TAB1, respectively. Coimmunoprecipitation (Co-IP) results showed that HERP interacted with TBK1, but not TRAF3, IKKε, IRAK1, TRAF6, TAK1, or TAB1 (Fig. 6A). To further verify the interaction between HERP and TBK1, HEK293T cells were cotransfected with plasmids expressing Myc-tagged HERP and GFP-tagged TBK1. Co-IP results confirmed that HERP interacted with TBK1 (Fig. 6B). In primary mouse BMDMs, endogenous HERP interacted with TBK1, and the interaction between them was remarkably enhanced after SeV infection (Fig. 6C).

In addition, the cellular localizations of HERP and TBK1 were examined by confocal microscope. In HEK293T cells, TBK1 specifically colocalized with HERP in the cytoplasm (Fig. 6D), suggesting that HERP interacts with TBK1. ER is a pivotal organelle where RNA virus replication commonly occurs, including that of EV71 (32). We further examined the colocalization among HERP, ER, and viral RNA during EV71 infection. Most HERP was located in the ER as expected, and HERP and virus dsRNA were strongly colocalized at foci in the specialized area (viral replication complex) of the cytosol after EV71 infection (Fig. 6E), which suggests that HERP is augmented and aggregated in ER during virus infection. The colocalization of HERP and TBK1 was also captured in vivo. Compared to mock-infected mice, HERP strongly colocalized with TBK1 in the spleen of EV71-infected mice (Fig. 6F). Taken together, we demonstrated that HERP interacts with TBK1.

MAVS mediates the immune response by interacting with TBK1 and is also linked to ER stress, resulting in facilitating the nuclear translocation of IRF3 (33). To investigate whether HERP-induced IFN production was mediated by MAVS, MAVS^−/− HEK293T cells were used (Fig. 7A). We initially assessed the role of MAVS in virus-induced ER stress. The spliced XBP1 mRNA was detected by SeV infection soon after 4 h in WT cells, whereas it was hardly detected in MAVS^−/− cells (Fig. 7B), indicating that MAVS is required for virus-induced ER stress. Since virus-induced HERP depends on ER stress, WT and MAVS^−/− cells were infected with SeV. qRT-PCR results revealed that HERP mRNA was sharply induced in WT cells by SeV infection, whereas such induction was strongly impaired in MAVS^−/− cells (Fig. 7C, left panel). Correspondingly, SeV-induced IFN-α, IFN-β, and IFN-λ1 mRNA levels were significantly restrained in MAVS^−/− cells (Fig. 7C), suggesting that MAVS modulated the HERP-induced IFN expression. Then, we conducted HERP overexpression in WT and MAVS^−/− cells. HERP promoted IFN-α, IFN-β, and IFN-λ1 mRNA expression in WT cells, but such induction was obviously reduced in MAVS^−/− cells (Fig. 7D), which is again consistent with the evidence that MAVS modulated HERP-induced IFN expression. Because HERP binds TBK1 in IFN regulation, the role of MAVS in the HERP–TBK1 interaction was also clarified. In MAVS^−/− cells, Co-IP assays revealed that HERP still interacted with TBK1 (Fig. 7E), indicating that HERP binds TBK1 independent of MAVS. Moreover, in EV71-infected WT cells, upregulated HERP enhanced the interaction with TBK1, but the interaction was reduced in EV71-infected MAVS^−/− cells (Fig. 7F). However, we failed to observe the association between HERP and IRF3, even after EV71 infection (Fig. 7F), implying that HERP is not a direct carrier assembling with TBK1 and IRF3. Altogether, these data demonstrated that the role of HERP in IFN regulation relies on MAVS signaling and HERP amplifies signaling from the RLR–MAVS pathway.

HERP contains a UBL domain at its N terminus that has important functions in interacting with other proteins in ubiquitylation and degradation (34). To determine which domain was involved in the antiviral function of HERP, four plasmids encoding truncated FLAG-tagged HERP proteins, HERPΔUBL, HERPΔC, HERPΔCΔUBL, and HERP-UBL, were constructed based on HERP functional domains (Fig. 8A). HEK293T cells were cotransfected with pGFP-TBK1 and pFLAG-tagged HERPΔUBL, pFLAG-tagged HERPΔC, pFLAG-tagged HERPΔCΔUBL, or pFLAG-tagged HERP-UBL, respectively. HERPΔUBL interacted with TBK1, but HERPΔC, HERPΔCΔUBL, and HERP-UBL failed to interact with TBK1 (Fig. 8B), indicating that the C-terminal domain of HERP is involved in the interaction of HERP with TBK1. We then determined whether the HERP C terminus domain is required for the regulation of expression of IFNs. The mRNA levels of IFN-α, IFN-β, and IFN-λ1 were significantly enhanced by HERP or HERPΔUBL, but not by HERPΔC, HERPΔCΔUBL, or HERP-UBL (Fig. 8C), confirming that the HERP C terminus, including the transmembrane domain, is required for the activation of production of IFNs.

Since HERP inhibits EV71 replication by activating IFNs and antiviral proteins, the functional domain of HERP that is involved in such regulation was examined. HEK293T cells were transfected with each truncated HERP-expressing plasmid and infected with EV71. EV71 VP1 RNA was significantly attenuated by HERP and HERPΔUBL, reduced by HERPΔC and HERPΔCΔUBL, but was not affected by HERP-UBL (Fig. 8D); suggesting that the HERP-UBL domain is not involved in the inhibition of EV71 replication. The effects of HERP on the replications of other RNA viruses were also investigated. HEK293T cells were transfected with pCMV-HERP and infected with SeV or IAV. SeV N RNA (Fig. 8E, left panel) and IAV N RNA (Fig. 8E, right panel) were significantly attenuated by HERP overexpression, indicating that HERP represses the replication of SeV and IAV, which suggests that HERP has a broad antiviral effect. We further assessed whether HERP antiviral activity relied on the action of IFNs (type I and IFN-λ). Given that mouse cells lack the IFN-λ1 gene, most of the mouse cells show no direct antiviral effect to IFN-λ protein, including dendritic cells and macrophages (35). The primary BMDMs from WT and IFNAR1^−/− mice were overexpressed with HERP. HERP obviously inhibited both EV71 (Fig. 8F) and VSV (Fig. 8G) replication in WT cells, but the inhibitory effect was abolished in IFNAR1^−/− cells, indicating that type-I IFN played an absolutely dominant role in HERP antiviral activity in mouse cells, and that HERP has antiviral activity that is dependent on IFNs. Taken together, we revealed that HERP interacts with TBK1 through its C-terminal domain, but not the UBL domain, resulting in the induction of IFNs and the repression of RNA virus replication (Fig. 9).

EV71 infection induces ER stress and activates ER-resident chaperone proteins, binding protein (BiP or Grp78) and calreticulin (CRT) (20, 21). We previously reported that EV71 replication occurred in the ER-derived, membrane-associated complex (19). In this study, we reveal that EV71 infection induces HERP expression both in vitro and in vivo. Although HERP is involved in various cellular processes and has multiple functions, including protein ubiquitylation (34, 36), dyslipidemia or diabetes (14, 37), and Parkinson’s disease (13), little was known about the role of HERP in viral infection and pathogenesis. We demonstrate that HERP plays an inhibitory role in viral replication.

IFN-mediated pathways are crucial in cellular responses against viral infection. IFN was first discovered as an agent to treat IAV infection (38, 39). Type-I IFNs were previously reported to protect mice from EV71 infection (40). We speculate that HERP may regulate viral replication through regulating the actions of IFN pathways, and further reveal that HERP is able to facilitate the expression of IFNs (IFN-α, IFN-β, and IFN-λ1), resulting in the repression of viral replication.

Transcription of IFN-λ1, similar to type-I IFNs, requires the activation of the IRF3/7 and NF-κB pathways (30). We demonstrate that HERP binds TBK1 to promote the phosphorylation and translocation of IRF3 and NF-κB to the nucleus and facilitate the production of IFNs, suggesting that HERP is required for the production of type-I and -III IFNs. During this activation, HERP enhances the assembly of TBK1-related components in the cytosol and amplifies the signaling of the RLR–MAVS pathway, which serves as an ER IFN stimulator. Since classical ER stress is indeed MAVS dependent (33), we suggest that MAVS-mediated ER stress reinforces IFN production under either physiological (e.g., cell stress response) or pathological (e.g., virus infection) conditions. HERP is induced in response to ER stress, and thus it may act as an important link between MAVS-mediated ER stress and IFN production. As an ER stress–inducible factor, HERP has been identified to be involved in the anti-dsDNA response in human lupus (41). In this study, we demonstrate that HERP is a potential trigger for viral dsRNA in the ER-derived replication complex coupled with MAVS-mediated ER stress. HERP attenuates the replication of several RNA viruses, indicating that HERP may have a broad antiviral effect.

HERP has a UBL domain at its N terminus with important functions (9). The UBL domain is not involved in the antiviral function of HERP, indicating that HERP attenuates viral replication through activation of IFNs without the ubiquitination function mediated by UBL domain. The C-terminal domain, including the transmembrane domain of HERP, is required for the binding of HERP with TBK1 and its antiviral effect, indicating that such biological function may rely on the membrane localization of HERP.

In conclusion, we identify a previously unrecognized function of HERP in viral infections and reveal a new mechanism underlying the regulation of RNA virus replication (Fig. 9). Virus infection alters the ER membrane and induces ER stress mediated by MAVS through the RLR pathway, resulting in the activation of HERP, an ER-resident membrane protein. HERP subsequently binds TBK1, which plays a central role in innate immunity by serving as an integrator of MAVS signaling induced by receptor-mediated pathogen detection and as a modulator of IFNs. The interaction of HERP with TBK1 facilitates the phosphorylation and nuclear translocation of IRF3 and NF-κB and increases the production of IFNs and ISGs, which leads to inhibition of replication of RNA viruses. Our findings provide new insights into the regulation of viral infection and replication and may contribute to the development of new antiviral agents.

We thank Dr. Danying Chen of Peking University, Beijing, China, for kindly providing MAVS^−/− HEK293T cells; Dr. Chengfeng Qin of the Academy of Military Medical Science, Beijing, China, for IFNAR1^−/− A129 mice; Dr. Hongbing Shu of Wuhan University, China, for SeV; and Dr. Mingzhou Chen of Wuhan University, China, for VSV-GFP.

The authors have no financial conflicts of interest.