Identification of TMEM53 as a novel SADS-CoV restriction factor that targets viral RNA-dependent RNA polymerase

ABSTRACT

Zoonotic transmission of coronaviruses (CoVs) poses a serious public health threat. Swine acute diarrhea syndrome coronavirus (SADS-CoV), originating from a bat HKU2-related CoV, causes devastating swine diseases and poses a high risk of spillover to humans. Currently, licensed therapeutics that can prevent potential human outbreaks are unavailable. Identifying the cellular proteins that restrict viral infection is imperative for developing effective interventions and therapeutics. We utilized a large-scale human cDNA screening and identified transmembrane protein 53 (TMEM53) as a novel cell-intrinsic SADS-CoV restriction factor. The inhibitory effect of TMEM53 on SADS-CoV infection was found to be independent of canonical type I interferon responses. Instead, TMEM53 interacts with non-structural protein 12 (NSP12) and disrupts viral RNA-dependent RNA polymerase (RdRp) complex assembly by interrupting NSP8-NSP12 interaction, thus suppressing viral RdRp activity and RNA synthesis. Deleting the transmembrane domain of TMEM53 resulted in the abrogation of TMEM53-NSP12 interaction and TMEM53 antiviral activity. Importantly, TMEM53 exhibited broad antiviral activity against multiple HKU2-related CoVs. Our findings reveal a novel role of TMEM53 in SADS-CoV restriction and pave the way to host-directed therapeutics against HKU2-related CoV infection.

Introduction

In the past two decades, coronaviruses have caused three major epidemics, including the severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) epidemics and the ongoing SARS-CoV-2 pandemic [1,2]. CoV outbreaks can profoundly and seriously affect the global economy and public health [3]. Zoonotic CoVs are continuously emerging, and therapeutics against CoVs, particularly the newly emerging CoVs, are underdeveloped. Therefore, there is an urgent need to identify potential drug targets for emerging CoVs with zoonotic potential to prepare for future CoV outbreaks.

We previously identified a novel bat-origin CoV, swine acute diarrhea syndrome coronavirus (SADS-CoV), as the causative agent of a large-scale outbreak of severe diarrhea in sucking piglets that emerged in Guangdong province, China, during 2016-2017 [4]. SADS-CoV belongs to the Rhinolophus bat CoV HKU2 species and mainly infects the intestine in vivo, and it has a mortality rate of 90% in piglets of less than five days of age [5]. Although no pig farm workers were SADS-CoV seropositive during the outbreak, we and others found that SADS-CoV has a broad host range and replicates efficiently in cells derived from mammals, especially in primary human lung and intestinal cells, highlighting the potential risk of animal-to-human spillover [6–8]. In addition, after the first outbreak, SADS-CoV re-emerged in 2018, 2019 and 2021, posing a continuous threat to livestock herds and public health [9–11]. No US Food and Drug Administration (FDA)-approved drugs or vaccines against SADS-CoV infection are available. Moreover, the receptor-blockade strategy widely used for SARS-CoV-2 may not apply to SADS-CoV, as its specific receptor has not yet been identified [4,12]. Therefore, understanding the host factors restricting viral infection is imperative for finding potential intervention strategies against SADS-CoV.

In an effort for pandemic preparedness, we previously explored SADS-CoV host dependency factors as potential therapeutic targets and identified zinc finger DHHC-type palmitoyltransferase 17 (ZDHHC17) as a promising drug target against SADS-CoV replication [13]. In the current study, we employed a human cDNA library to screen for host factors that inhibit SADS-CoV infection, and found that transmembrane protein 53 (TMEM53) specifically inhibits viral RNA replication by disrupting non-structural protein (NSP)8-NSP12 interaction. Importantly, TMEM53 showed viral inhibitory activity against a series of closely related bat HKU2-related CoVs, which may cause future zoonoses. Thus, we identified TMEM53 as HKU2-CoV genus-specific inhibitor that may serve as a therapeutic target against SADS-CoV.

Materials and methods

Cells and viruses

HeLa (#CCL-2), baby hamster kidney (BHK-21) (#CCL-10), Vero (#CCL-81) and HEK293T (#ACS-4500) cells were obtained from the American Type Culture Collection (Manassas, VA, USA). Human colon adenocarcinoma cells (HCT-8) were obtained from the National Virus Resource Centre (Wuhan, China). Huh7 cells were provided by Dr. Lin-Fa Wang (Duke-NUS Medical School, Singapore). The cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco) supplemented with 10% (v/v) fetal bovine serum (Gibco) at 37°C in the presence of 5% CO₂ and passaged for subsequent experiments. All cell lines were mycoplasma-negative.

A SADS-CoV isolate (GenBank accession No. MG557844) was propagated and titrated on Vero cells. Bat SARS-related CoV WIV1 and SARS-CoV-2 (IVCAS 6.7512) were acquired in our previous study [14,15], and infection experiments with these viruses were performed in biosafety level 2 and level 3 facilities, respectively, at the Wuhan Institute of Virology. HCoV-OC43 and HCoV-229E were provided by the National Virus Resource Centre (Wuhan, China) and propagated and titrated on HCT-8 and Huh7 cells, respectively. Herpes simplex virus type 1 (HSV-1) was generously provided by Yan-Yi Wang (Wuhan Institute of Virology, Chinese Academy of Sciences). H1N1 influenza A virus strain PR8 and Sendai virus (SeV) were kind gifts from Quanjiao Chen (Wuhan Institute of Virology, Chinese Academy of Sciences).

Antibodies and reagents

The following primary antibodies were used in this study: mouse monoclonal anti-FLAG (M20008; Abmart), mouse monoclonal anti-β-actin (66009-1-lg; Proteintech), rabbit polyclonal anti-TMEM53 (HPA021134; Sigma-Aldrich), mouse monoclonal anti-dsRNA J2 (J2-1702; Scicons), rabbit polyclonal anti-STAT1 (10144-2-AP; Proteintech), mouse monoclonal anti-HA (TA180128; OriGene), rabbit polyclonal anti-HA (H6908; Sigma-Aldrich), and mouse anti-IgG (A7028, Beyotime). Anti-S-tag and anti-NP monoclonal antibodies were prepared in-house. We used the following secondary antibodies: horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (SA00001-1, Proteintech), HRP-conjugated goat anti-rabbit IgG (SA00001-2, Proteintech), goat anti-mouse IgG H&L DyLight 488 (ab96879, Abcam), goat anti-rabbit IgG H&L Cy3 (ab6939, Abcam), and goat anti-mouse IgG H&L Cy3 (SA00009-1, Proteintech). Universal type I interferon (IFN) (11200, PBL) was acquired from PBL Assay Science.

Plasmid construction

FLAG-tagged expression vectors for CD14 molecule (CD14), clusterin (CLU), TMEM53, lymphocyte antigen 6 family member E (LY6E), TMEM53-ΔTM, NSP8, and melanoma differentiation-associated protein 5 (MDA5) were subcloned into pCMV-3Tag-8 using specific primers (Supplemental Table 1). Human influenza hemagglutinin (HA)-tagged expression vectors for NSP7, NSP8 and NSP12 were synthesized (Tsingke Biotechnology) using the codon optimization strategy and were subcloned into pCAGGS-HA. NSP12 truncations (RdRp and ΔRdRp) were amplified using primers (Supplemental Table 1) and cloned into pCAGGS-HA. TMEM53 and LY6E lentiviral vectors were constructed by ligating FLAG-tagged TMEM53 and LY6E amplified by PCR using gene-specific primers (Supplemental Table 1) into pLVX-IRES-Neo. FLAG-tagged lentiviral vectors of TMEM53 orthologs and HA-tagged NSP12 orthologues were synthesized at Tsingke Biotechnology. All constructs were verified by direct sequencing.

cDNA library construction and screen

We followed a previously reported procedure with modification [16]. In brief, Huh7 cells were grown into 15-cm-diameter dishes (1 × 10⁷ cells/dish) for 12 h and then treated with universal type I IFN at 500 U/ml for 6 h. Then, the cells were harvested with TRIzol and mRNA was isolated. According to the manufacturer’s instructions, approximately 2 µg of mRNA was used to generate the cDNA ibrary using the CloneMiner^TM II cDNA Library Construction Kit (A11180, Invitrogen). The lentiviral cDNA library was packaged with psPAX2 and pMD2.G plasmids (at a 4:3:1 ratio) in HEK293T cells. The lentiviral titre was calculated based on visible cell colonies selected with 6 µg/ml blasticidin in HeLa cells. A total of 1.5 × 10⁷ HeLa cells were transduced with cDNA library lentiviruses at a multiplicity of infection (MOI) of 0.3 to avoid a cumulative effect of multiple lentiviral inserts. At 48 h post-transduction, the HeLa cells were treated with 6 µg/ml blasticidin for a 10-day selection before the cDNA library-transduced cells were harvested. For the cDNA library screen, the cDNA library-transduced HeLa cells were reseeded and infected with SADS-CoV (MOI = 0.1) or mock-infected. Twelve days post-infection, surviving cells were collected and expanded for the next round of infection. After three rounds of infection, surviving cells were collected for genomic DNA extraction. Integrated cDNA clones were amplified with forward 5'-CGCAAATGGGCGGTAGGCGTG-3’ and reverse 5'-ACCACTTTGTACAAGAAA-3’ primers and used for next-generation sequencing. After deep sequencing, clean reads were mapped to the human reference genome GRCH38.p13 using HISAT2 v2.2.1 and annotated using StringTie v2.1.5. cDNAs enriched in surviving cells compared to uninfected cells were identified using DESeq2 package v1.34.0 in R v4.1.1.

Lentivirus production and transduction

All lentiviruses were produced by cotransfecting HEK293T cells with lentiviral constructs psPAX2 and pMD2.G (at a ratio of 4:3:1) using Lipofectamine 3000 transfection reagent (L3000008; Invitrogen) according to the manufacturer’s instructions. At 48 h post-transfection, the cell culture supernatant was collected and filtered through a 0.45 μm filter. For lentivirus transduction, HeLa and Huh7 cells were seeded in 6-well plates at a density of 4 × 10⁵ cells/well, cultured overnight, and transfected with 1 ml of packaged lentiviruses. After 48 h of incubation, transduced cells were selected with 6 µg/ml blasticidin, and blasticidin-resistant cells were pooled and expanded until use.

Generation of knockout cell lines

We used the CRISPR-Cas9 system to knock out the TMEM53 and STAT1 genes in the indicated cell lines. Briefly, small guide RNAs (sgRNAs) targeting TMEM53 and STAT1 (Table S1) were annealed and cloned into the lentiCRISPR-v2 vector (Addgene). The lentiviral constructs were cotransfected into HEK293T cells with psPAX2 and pMD2.G using Lipofectamine 3000 for 48 h. Then, HeLa and Huh7 cells were transduced with the above-packaged lentiviruses and selected with 2 mg/ml puromycin for 7 days. Surviving cells were plated in limiting dilutions and individual cells were expanded to extract genomic DNA. sgRNA target sites were amplified using gene-specific primer pairs (Table S1). Gene knockout was validated by sequencing and western blotting.

Immunofluorescence analysis

HeLa cells were seeded on a chamber slide (154526; Thermo Scientific), cultured overnight, and transfected with the indicated plasmids for 48 h. Then, the cells were washed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde at 37°C for 10 min followed by permeabilization with 0.2% Triton X-100 (T9284; Sigma Aldrich) at room temperature for 10 min. The cells were incubated overnight with the indicated primary antibodies (mouse anti-FLAG, 1:200 and rabbit anti-HA, 1:200) at 4°C. After three washes with PBS, and the cells were incubated with secondary antibodies anti-mouse DyLight 488 (1:200) and anti-rabbit Cy3 (1:200) at room temperature for 2 h. Nuclei were stained with 4′,6-diamidino-2-phenylindole. The slides were imaged using a fluorescence microscope (Nikon).

Virion attachment and entry assay

For the virion attachment and entry assay, we used the procedures reported in our previous study [13]. In brief, control cells and Huh7 cells stably expressing TMEM53 or LY6E were seeded in 24-well plates (4 × 10⁵ cells/well) and infected with SADS-CoV at an MOI of 1. Infected cells were placed on ice for 1 h to allow virus attachment but impede its entry. Then, the supernatant was removed followed by three washes with precooled PBS. The cell-bound virus was isolated using TRIzol reagent. The viral RNA was evaluated by quantitative reverse transcription PCR (RT-qPCR). To analyze viral entry, the infected cells were incubated in pre-warmed DMEM at 37°C for another 1 h and then treated with 1 mg/mL pronase dissolved in PBS at 37°C for 10 min to remove non-internalized virus particles from the cell surface. After three washes with PBS, the cells were harvested for RNA extraction and the amount of viral RNA in the cells was measured using RT-qPCR.

Generation of reporter virus and replicons

A yeast-based reverse-genetics platform was employed to construct a SADS-CoV cDNA clone carrying mNeonGreen. In brief, the SADS-CoV (GenBank No. MG557844) genome was divided into eight overlapping cDNA fragments. These fragments were amplified, purified, and cloned into the pGEM-T-EASY vector, which was then digested at its unique BsaI restriction site. The mNeonGreen gene was amplified to replace the NS7a and NS7b genes in the last subclone. After verification by sequencing, all SADS-CoV fragments were digested and repurified and then simultaneously transformed into Saccharomyces cerevisiae harbouring a linearized yeast artificial chromosome (YAC) to assemble the viral cDNA clone. The infectious YAC clone was sequenced and transfected into Huh7 cells using Lipofectamine 3000. After blind passage in Huh7 cells, rescued viruses were harvested.

For replicon constructions, the 5′-terminal sequences of three Rhinacovirus genomes, including SADS-CoV (MG557844), 162140 (MF094688), and HKU2 (NC_009988), were split into multiple overlapping cDNA fragments. The S/3a and E/M open reading frames in the three genomes were deleted and replaced with the neomycin phosphotransferase (Neo) and NanoLuc luciferase (Nluc) genes, respectively. These full-length replicons were generated in one step as described above. Nluc expression and subgenomic N RNA production were measured in BHK-21 cells transfected with the plasmids. The MERS-CoV replicon was generated in our previous study by modified Nluc replacement [17].

RNA extraction and RT-qPCR

Total host RNA and viral RNA were extracted using an RNAsimple Total RNA Kit (DP419; TIANGEN) and Virus DNA/RNA Extraction Kit (RM401; Vazyme), respectively, according to the manufacturer's instructions. To detect the viral RNA minus-strand and plus-strand (+/–vRNA) of SADS-CoV, +/– vRNAs were synthesized using an RdRp-forward and – reverse primer, respectively. RT-qPCRs were run using a HiScript II One Step qRT-PCR SYBR Green Kit (Q221-01; Vazyme) and gene-specific primer pairs (Table S1) in a CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratories). The β-actin gene transcript was used for target gene expression normalization. Relative expression was determined using the 2^-ΔΔCT method.

Luciferase reporter assay

Huh7-TMEM53 and control cells were transfected with 0.1 μg luciferase reporter plasmid (IFN-Luc or IFN-stimulated response element (ISRE-Luc) and 0.01 μg Renilla luciferase control plasmid (pRL-TK) using Lipofectamine 3000 according to the manufacturer’s instructions. At 24 h post-transfection, the cells were infected with SeV (10 HAU/ml) for 24 h before harvest. Luciferase activity was determined using a Dual-Luciferase Reporter Assay System kit (E1960; Promega).

Co-immunoprecipitation (co-IP) and western blotting

HEK293T cells seeded in 10-cm-diameter dishes were cotransfected with the indicated expression vectors (5 μg each) using Lipofectamine 3000. At 48 h post-transfection, the cells were collected and lysed in 500 μl lysis buffer (P0013; Beyotime) on ice for 1 h. The lysates were harvested by centrifugation at 12,000 g, 4°C for 10 min. Protein beads (15920010; Life Technologies) were incubated with the indicated antibodies at room temperature for 1 h, added to the cell lysates, and incubated at 4°C overnight. After four washes with lysis buffer, the precipitated proteins were eluted from the beads and used for immunoblot analysis.

For western blotting, the isolated cellular protein was denatured at 95°C for 10 min. The cell lysates were resolved by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. The membranes were blocked using 5% bovine serum albumin (BSA) dissolved in Tris-buffered saline containing 0.2% Tween-20 (TBST; P1379; Sigma) at room temperature for 2 h. Then, the membranes were incubated with primary antibodies at 4°C overnight. After three washes with TBST, the membranes were incubated with secondary antibodies at room temperature for 1 h. After another three washes with TBST, the proteins on the membrane were visualized using an enhanced chemiluminescence reagent (WBKLS0500; Millipore).

Cell viability assay

Cell viability was assessed using a Cell Counting kit-8 (CCK-8) (HY-K030; MedChemExpress) according to the manufacturer’s instructions. Briefly, cells were grown in 96-well plates at 37°C overnight. After removal of the supernatant at the indicated times, 100 μl of fresh medium containing 10 μl of diagnostic reagent was added to the plates. After a 2-h incubation at 37°C, the absorbance at 450 nm was measured using a microplate reader.

RNA immunoprecipitation assay

The RNA immunoprecipitation assay was performed using an RNA immunoprecipitation kit (Bes5101; BersinBio) according to the manufacturer’s instructions. In brief, 2 × 10⁷ HEK293T cells were transfected with 20 μg FLAG-tagged TMEM53 or MDA5 expression vector using Lipofectamine 3000. At 12 h post-transfection, the cells were infected with SADS-CoV at an MOI of 5 and cultured in DMEM containing 2% Tryptose Phosphate Broth for 48 h. The cells were collected and lysed with polysomal lysis buffer containing protease inhibitor and RNase inhibitor on ice for 10 min. The cell lysates were incubated with DNase salt stock and DNase at 37°C for 10 min. Then, the lysates were incubated with 0.5 M ethylenediaminetetraacetic acid, 0.5 M ethylenediaminetetraacetic acid, and dithiothreitol on ice followed by centrifugation at 16,000 g, 4°C for 10 min. As described above, the harvested lysates were incubated with protein A/G beads pretreated with mouse anti-FLAG or mouse anti-IgG antibodies. After 16 h of incubation, the beads were washed and collected for immunoblot analysis and RNA extraction.

Proximity ligation assay (PLA)

HeLa cells were seeded on a chamber slide (1 × 10⁵ cells/well) and transfected with the indicated plasmids using Lipofectamine 3000 for 48 h. The cells were fixed with 4% paraformaldehyde at 37°C for 10 min, permeabilized with 0.2% Triton X-100 for 10 min, and blocked with 5% BSA at room temperature for 1 h. Following the manufacturer's instructions, the PLA was performed using a Duolink In Situ Red Starter Kit Mouse/Rabbit (Duo92101; Sigma-Aldrich) per the manufacturer’s instructions. Mouse anti-FLAG and rabbit anti-HA antibodies were used as primary antibodies. Images were captured using a fluorescence microscope (Nikon).

Cell-based SADS-CoV RdRp activity assay

We followed the procedure for constructing a SARS-CoV-2 RdRp reporter plasmid to generate the SADS-CoV reporter plasmid [18]. The sense-oriented firefly luciferase gene (FLuc), antisense 3′-untranslated region (UTR) of SADS-CoV, antisense NanoLuc luciferase (NLuc), antisense 5′-UTR of SADS-CoV, and hepatitis delta virus ribozyme sequences were synthesized at Tsingke Biotechnology and cloned into the pcDNA3.1 vector.

For the RdRp activity assay, control cells and Huh7-TMEM53 cells were seeded in 24-well plates and cotransfected with HA-tagged SADS-CoV NSP7, NSP8, and NSP12 expression plasmids using Lipofectamine 3000. The cells were collected at 24 and 48 h post-transfection, and Nluc expression was measured using the Nano-light Luciferase Reporter Assay System (MA0521, Meilunbio). Fluc expression was determined using the Luciferase Reporter Assay System (E1960; Promega) for normalization.

Single-cell RNAseq data analysis

A public single-cell data set was collected from the University of California Santa Cruz single-cell platform. Data of healthy lung from Madissoon et. al. [19] and data of healthy ileum from Wang et. al. [20] were used for analysis. Dimension reduction and cell type annotation were obtained from the original publications. Cells with more than one unique TMEM53 mRNA were considered TMEM53-positive and are highlighted in red in the figures. The single-cell data analysis was performed using the Seurat package v4.0.5 in R v4.1.1.

Statistical analysis

Data are presented as mean ± standard deviation (SD) and were analyzed using the unpaired Student's t-test in the GraphPad software. Differences with a P-value < 0.05 were considered statistically significant.

Results

A gain-of-function screen identifies TMEM53 as a SADS-CoV restriction factor

To identify host factors restricting SADS-CoV infection, we established a screening platform using Huh7 cells treated with universal type I IFN to generate a lentivirus-based cDNA library enriched for antiviral genes. We transduced the pooled lentiviral cDNA library into HeLa cells, which are prone to SADS-CoV infection and exhibit obvious cytopathic effects, followed by antibiotic screening and SADS-CoV infection. After three rounds of SADS-CoV challenge, surviving cells expressing cDNA clones restricting SADS-CoV infection were identified (Figure 1a). Candidate restriction factors of SADS-CoV were enriched based on thresholds of reads count > 10⁴ and Log2 (fold change) > 2. Using this approach, we identified four top hits, including CD14, CLU, TMEM53 and LY6E (Figure 1b). CD14 and LY6E were previously identified as promising targets for SARS-CoV-2 or other CoV infections, supporting the reliability of our screening results [21–23]. For validation, we overexpressed the four candidate genes to assess their ability to modulate SADS-CoV infection (Supplementary Figure S1a). In an overexpression analysis, only TMEM53 and LY6E significantly restricted SADS-CoV infection in Hela cells (Figure 1c), as well as in Huh7 and HEK293T cells (Figure 1d and Supplementary Figure S1b). Given the importance of the transmembrane member family genes [24], we selected TMEM53, a lesser-known gene, for further investigation. Based on publicly available single-cell RNA sequencing data, we found that, compared to in other cell types, TMEM53 is relatively highly expressed in lung type 2 alveolar cells and ileal enterocyte cells, which are typical target cells for CoVs (Supplementary Figure S2).

Figure 1.

Identification and validation of TMEM53 as an antiviral factor against SADS-CoV. (a) Schematic of cDNA library screen. (b) Screen results from HeLa cells infected with SADS-CoV. Genes with a log₂ (fold change) > 2 and reads count > 10⁴ were highlighted in red. (c) Result validation by transient overexpression of FLAG-tagged CD14, CLU, TMEM53 and LY6E in HeLa cells followed by SADS-CoV infection (MOI = 0.1). At 48 h post-infection (hpi), the viral RNA in the supernatant was measured using RT-qPCR targeting at the SADS-CoV NSP12 gene (Table S1). (d) TMEM53 overexpression inhibited SADS-CoV infection (MOI = 0.1) in Huh7 cells. At 48 hpi, the viral RNA in the supernatant (left) and cytoplasm (right) were measured by RT-qPCR. LY6E was used as a positive control in the subsequent experiments. (e-f) Anti-SADS-CoV activity in TMEM53 stably expressing HeLa (HeLa-TMEM53) and Huh7 cells (Huh7-TMEM53). Cells were infected with SADS-CoV at a MOI of 0.1. At 48 hpi, the viral RNA in the supernatant and cytoplasm were determined by RT-qPCR. Cells transduced with lentiviral empty vector were used as control (Ctrl) in the subsequent assays unless otherwise stated. (g) HeLa-TMEM53 and Huh7-TMEM53 cells were infected with SADS-CoV (MOI = 0.1). NP proteins were assessed by immunofluorescence microscopy at 48 h post-infection. (h) Statistical analysis of Figure 1g. The normalized mean fluorescence intensity of NP protein was quantified with Image J software. N = 5 panels per group. (i) The infectious virions secreted from the HeLa– or Huh7– expressing LY6E or TMEM53 and control cells were evaluated by 50% tissue culture infectious dose (TCID₅₀) assays in Huh7 cells. (j-k) TMEM53 inhibited the reporter virus SADS-CoV-mNG infection. Huh7-TMEM53 cells were infected with SADS-CoV-mNG (MOI = 0.1). mNG-positive cells were assessed by immunofluorescence microscopy (j) and flow cytometry (k) at 48 hpi. (l) Crystal violet staining of polyclonal cell populations infected with SADS-CoV (MOI = 0.1) in TMEM53 or LY6E stably expressing Huh7 cells at 48 hpi. (m) TMEM53 knockout promoted SADS-CoV infection in Huh7 cells. Wild type (WT) Huh7 cells and Huh7-TMEM53-KO (KO) cells were infected with SADS-CoV (MOI = 0.1) for 48 h. The viral RNA in the supernatant and cytoplasm were measured by RT-qPCR. All data were representative of three independent experiments with similar results. Data were presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, two-tailed unpaired Student t test.

To minimize variations due to transient transfection, we generated HeLa and Huh7 cells stably expressing TMEM53 and LY6E (Supplementary Figure S3a and S3b). TMEM53 and LY6E remarkably inhibited SADS-CoV infection in both cell types when we tested viral RNA or nucleoprotein (Figure 1e to h). The viral titres in the supernatants of Huh7 and HeLa cells stably expressing TMEM53 or LY6E were significantly lower than those in the supernatants of control cells (Figure 1i). Next, we constructed a SADS-CoV reporter virus tagged with mNeonGreen (SADS-CoV-mNG) (Figure 1j, upper panel). Consistently, the replication of SADS-CoV-mNG was distinctly decreased in TMEM53– or LY6E– expressing Huh7 cells (Figure 1j, k). In parallel, plaque assays revealed that TMEM53– or LY6E– expressing cells remained viable after SADS-CoV infection at MOIs of 0.01 and 0.1 (Figure 1l and Supplementary Figure S3c), indicating that TMEM53 and LY6E expression indeed restricted SADS-CoV infection.

To assess whether endogenous TMEM53 inhibits SADS-CoV infection, we generated TMEM53-knockout Huh7 and HeLa cells using the CRISPR-Cas9 system (Supplementary Figure S3d and S3e). TMEM53 knockout significantly enhanced SADS-CoV infection compared to that in wild-type cells (Figure 1m and Supplementary Figure S3f). To rule out the possible impact of TMEM53 on cell viability, TMEM53– expressing and –knockout Huh7 cells were analyzed using a CCK8 assay. We observed no obvious effect on cell viability (Supplementary Figure S3g and S3h). These results demonstrated that TMEM53 is a bona fide SADS-CoV restriction factor.

TMEM53 inhibits SADS-CoV RNA synthesis independently of interferon responses

Next, we sought to understand how TMEM53 inhibits SADS-CoV infection. First, we measured the effect of TMEM53 on SADS-CoV attachment and entry using a modified method based on our previous study [13]. TMEM53 expression had no obvious effect on SADS-CoV attachment and entry, as demonstrated by the equivalent levels of viral RNA in control and TMEM53-expressing cells (Figure 2a, b). In line with a previous finding that LY6E generally inhibits CoV entry [22,25], LY6E was found to inhibit SADS-CoV entry (Figure 2b). Next, we measured the time-dependent total viral RNA (vRNA) expression in the presence of TMEM53 expression. The increase in vRNA expression was inhibited from 8 h post-infection (Figure 2c), indicating that TMEM53 may inhibit SADS-CoV RNA synthesis. As the increase in total RNA may be from exogenous viral infection at 0-2 h post-infection, we measured viral double-stranded RNA (dsRNA), positive-strand and negative-strand viral RNA (+vRNA and –vRNA) and subgenomic RNA (sgRNA), all of which are intermediate products of viral replication that can only be produced by endogenous viral replication. TMEM53 significantly inhibited viral dsRNA, +vRNA, –vRNA and sgRNA levels in Huh7 cells infected with SADS-CoV (Figure 2d-f). To confirm that TMEM53 targets SADS-CoV replication, we constructed a minimal SADS-CoV replicon expressing an NLuc reporter (SADS-CoV-Nluc) that lacks the viral spike protein and is unable to generate infectious particles (Supplementary Figure S4a). The stability of this reporter was demonstrated by the continuous expression of Nluc and the subgenomic N gene in BHK-21 cells (Supplementary Figure S4b and S4c). As shown in Figure 2g, SADS-CoV-Nluc replication was significantly inhibited in Huh7-TMEM53 cells, whereas LY6E expression had no obvious effects on SADS-CoV-Nluc replication. Together, these data suggested that TMEM53 inhibits SADS-CoV replication by inhibiting viral RNA synthesis.

Figure 2.

TMEM53 inhibited viral RNA synthesis independent of interferon response. (a-b) Huh7-TMEM53, Huh7-LY6E or control cells were infected with SADS-CoV at a MOI of 5. Viral particles attached to the cell surface or internalized were assessed by RT-qPCR. (c) SADS-CoV growth kinetics in Huh7-TMEM53 cells. Control cells and Huh7-TMEM53 were infected with SADS-CoV (MOI = 0.1), and viral RNA in the supernatant was measured by RT-qPCR at the indicated time points. (d) TMEM53 inhibited dsRNA production. Huh7-TMEM53 and control cells were infected with SADS-CoV (MOI = 1), and dsRNA was detected at 48 hpi by immunofluorescence microscopy. Nuclei were stained with blue staining (DAPI). Scale bar, 10 μm. (e-f) Control and Huh7-TMEM53 cells were infected with SADS-CoV (MOI = 0.1), and intracellular + vRNA, – vRNA and sgRNA were measured by RT-qPCR at the indicated time. (g) TMEM53 expression inhibited SADS-CoV replicon activity. Control and Huh7-TMEM53 or Huh7-LY6E Cells were transfected with SADS-CoV replicon and Nluc activity was quantified at the indicated time. (h-k) TMEM53 induction upon SADS-CoV infection and interferon treatment. Huh7 cells were infected with SADS-CoV (MOI = 0.1) (h) or treated with universal type I IFN (500U/ml) (i-k). Cells were harvested at the indicated time and the mRNA expression levels were determined by RT-qPCR. (l) STAT1^{-/ –} Huh7 cells transduced with TMEM53 were infected with SADS-CoV (MOI = 0.05) for 48 h. The viral RNA in the supernatant was analyzed by RT-qPCR. The experiments were independently repeated three times with similar results. Values were presented as mean ± SD. ns: non-significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, two-tailed unpaired Student t test.

TMEM family members, such as TMEM41B, TMEM120A, and TMEM173 (STING), are critically involved in modulating antiviral innate immune responses [26–28]. To understand whether the inhibition of viral RNA synthesis relies on innate immunity, we tested TMEM53 transcriptional regulation during SADS-CoV infection in the presence or absence of host IFN responses. SADS-CoV infection did not induce TMEM53 mRNA expression (Figure 2h). Similarly, direct treatment with type I IFN hardly affected TMEM53 expression (Figure 2i), although it markedly induced the expression of interferon-stimulated genes (ISGs), including OAS1 and Mx1 (Figure 2j, k). To test the possibility that TMEM53 counteracts SADS-CoV infection via upregulating antiviral IFN responses, we examined the effects of TMEM53 expression on IFN-β and IFN stimulated response element (ISRE) promoter activation triggered by SeV, a potent stimulator of type I IFN and ISG induction. As shown, TMEM53 expression did not affect the activation of IFN-β and ISRE promoter reporters (Supplementary Figure S5a and S5b). Consistent herewith, IFNB1 and selected ISGs transcript levels remained unchanged in TMEM53-expressing Huh7 cells infected with SeV compared to those in control cells (Supplementary Figure S5c to S5f). Additionally, TMEM53 still exerted excellent anti-SADS-CoV activity in STAT1 knockout Huh7 cells (Figure 2l). Together, these data indicated that TMEM53 is a cell-intrinsic antiviral effector that is independent of interferon production or signalling, suggesting that it may directly target the viral replication machinery.

TMEM53 suppresses SADS-CoV RdRp activity by interacting with NSP12 and NSP8

We next investigated the potential mechanism underlying the inhibition of viral RNA synthesis by TMEM53. In CoVs, the core polymerase complex formed by NSP12 and the two assistant factors NSP7 and NSP8 is mainly responsible for viral RNA synthesis [29]. As TMEM53 significantly suppressed new viral RNA production, we hypothesized that it might bind the viral RNA template, thus inhibiting RNA-dependent RNA synthesis or suppressing RdRp function.

We first tested whether TMEM53 can bind to viral RNA. We used MDA5, a pattern recognition receptor capable of binding to viral RNA, as a positive control [30,31]. We used the RNA immunoprecipitation method and RT-qPCR to measure the physical interaction between TMEM53 and SADS-CoV vRNA (Figure 3a). The data showed that MDA5 bound to SADS-CoV vRNA upon infection, whereas vRNA binding was not observed for TMEM53, suggesting it does not target vRNA directly (Figure 3b).

Figure 3.

TMEM53 suppressed SADS-CoV RdRp activity. (a) Schematic of RNA immunoprecipitation. (b) Quantifying SADS-CoV RNA in TMEM53, MDA5, and IgG immunoprecipitates by RT-qPCR. FLAG-tagged TMEM53 and MDA5 in IP were shown by western blot. (c) Schematic diagram of the cell-based SADS-CoV RdRp activity assay. (d) The RdRp activity in the cell-based assay system. Huh7 cells were cotransfected with RdRp reporter plasmids and NSP12-HA for 24 or 48 h. The RdRp activity was determined by measuring the Nluc and Fluc values, and a ratio of Nluc/Fluc reflecting RdRp activity was shown. (e) Expression of NSP7 and NSP8 increased RdRp activity. Huh7 cells were co-transfected with RdRp reporter plasmids, the NSP12-HA, NSP7-HA and NSP8-HA. The Nluc and Fluc values were measured at the indicated time points. (f) TMEM53 inhibited SADS-CoV RdRp activity. Control and Huh7-TMEM53 cells were co-transfected with RdRp reporter and RdRp plasmids for 24 h or 48 h before measuring the RdRp activity. All data were representative or presented as the mean ± SD of three independent experiments. ns: non-significant, **P < 0.01, ***P < 0.001, ****P < 0.0001, two-tailed unpaired Student t test.

Next, we assessed whether TMEM53 affects RdRp activity using a modified cell-based reporter system that specifically detects the NSP12-NSP7-NSP8 core polymerase complex activity of SADS-CoV (Figure 3c) [18,32]. In this approach, RdRp activity is reflected by the Nluc expression level, and Fluc was used as an internal control. As expected, overexpression of NSP12 or NSP12-NSP7-NSP8 significantly activated the system as demonstrated by an increase in the Nluc/Fluc ratio (Figure 3d, e). However, the Nluc/Fluc ratio was significantly decreased in TMEM53-expressing cells compared to control cells, indicating dampened RdRp activity in the presence of TMEM53, which was in line with the observation after viral infection (Figure 3f). Together, these results suggested that TMEM53 impairs the RdRp activity of SADS-CoV.

Subsequently, we sought to determine whether TMEM53 suppresses RdRp core polymerase activity by physically interacting with NSP12, NSP8 and NSP7. In co-IP experiments, TMEM53 specifically interacted with NSP12 and NSP8 (Figure 4a, b), but not NSP7 (Figure 4c). In line with this observation, TMEM53 colocalized with NSP12 and NSP8 as indicated by confocal fluorescence microscopy (Figure 4d). Moreover, the physical interaction between TMEM53 and NSP12 or NSP8, but not NSP7, was confirmed by proximity ligation assay (PLA) (Figure 4e and Supplementary Figure S6a). To map the key domains that are required for the interaction of TMEM53 with NSP12, we generated two truncated NSP12 versions having either only the core RdRp domain (NSP12-RdRp) or lacking this domain (NSP12-ΔRdRp) (Figure 4f). Because of the small size of NSP8, no truncated version of this protein was generated. Using co-IP assays, we found that NSP12-ΔRdRp, but not NSP12-RdRp interacted with TMEM53, suggesting that the N-terminal section of NSP12 is important for interaction with TMEM53 (Figure 4g). Based on these data, we hypothesized that TMEM53 disrupts the interaction between NSP12 and NSP8 (Figure 4h and Supplementary Figure S6b). Indeed, a co-IP assay indicated that TMEM53 dramatically inhibited the interaction between NSP8 and NSP12 (Figure 4i). A reciprocal co-IP assay confirmed the disruption of the interaction between NSP8 and NSP12 by TMEM53 (Supplementary Figure S6c). Similarly, using more sensitive PLA assay, we found that TMEM53 expression significantly decreased visible PLA signals compared to those in control cells (Figure 4j-k and Supplementary Figure S6d). As the NSP7-NSP8 heterodimer is also important for boosting RdRp activity [29], we investigated whether TMEM53 affects NSP7-NSP8 interaction. The results showed that NSP7-NSP8 interaction was not affected by TMEM53 expression (Supplementary Figure S6e). Together, these data demonstrated that TMEM53 suppresses SADS-CoV RdRp activity by disrupting NSP12-NSP8 complex formation.

Figure 4.

TMEM53 disrupted NSP8 and NSP12 interactions. (a-c) TMEM53 interacted with NSP12 and NSP8. HEK293T cells were transfected with TMEM53-FLAG and NSP12-HA, NSP8-HA or NSP7-HA for 48 h before harvesting for immunoprecipitation analysis. IgG was used as a control. IgG-L, IgG L chain. (d) HeLa cells were co-transfected with TMEM53-FLAG and NSP12-HA, NSP8-HA or NSP7-HA for 48 h. The colocalization of TMEM53 with NSP12, NSP8 or NSP7 was analyzed by immunofluorescence microscopy. Scale bar, 10 μm. (e) The interactions between TMEM53 and NSP12, NSP8 or NSP7 was revealed by PLA in HeLa cells. The number of PLA red fluorescent dots was counted with Image J software (right panel). N = 5 panels per group. Scale bar, 10 μm. (f) Illustration depicting truncations of NSP12. (g) Domain mapping of TMEM53 for interaction with NSP12 in HEK293T cells. The procedure was the same as in a-c. IgG-H, IgG H chain. (h) The interaction between NSP8 and NSP12 was verified by Co-IP in HEK293T cells. (i) TMEM53 impaired the recruitment of NSP8 to NSP12. HEK293T cells were co-transfected with NSP8-FLAG, NSP12-HA and TMEM53-S-tag or empty vector for 48 h before the immunoprecipitation assays. (j-k) HeLa cells were co-transfected with NSP8-FLAG, NSP12-HA and TMEM53-S-tag or empty vector for 48 h. Cells were fixed and analyzed by PLA assay (j). Scale bar, 10 μm. The number of red fluorescent dots in each panel was quantified with Image J software (k). The PLA controls are shown in Supplementary Figure S6c. The experiments were independently repeated three times with similar results. Values were presented as mean ± SD. ****P < 0.0001, two-tailed unpaired Student t test.

Transmembrane (TM) domain of TMEM53 is critical for antiviral activity

Next, we determined which domain in TMEM53 is important for its anti-SADS-CoV action. Information regarding the functionality of TMEM53 is limited. As TMEM protein family members are characterized by a classic TM domain [24], we constructed a TM-deleted version (TMEM53-ΔTM) to investigate whether the TM domain is required for antiviral function (Figure 5a). The TMEM53-ΔTM mutant had substantially reduced antiviral activity compared to the wild type (Figure 5b, c). This functional loss may have been caused by an alteration in the localization of TMEM53-ΔTM to the Golgi complex (Figure 5d-e and Supplementary Figure S7a-S7d), as ER and Golgi apparatus are critical for coronavirus replication [33]. Further, TMEM53-ΔTM had lost the capability to bind to NSP12 (but not to NSP8) (Figure 5f and g) and the ability to disrupt NSP12-NSP8 interaction (Figure 5h). Finally, we tested whether the TMEM53-ΔTM mutant would impair the RdRp activity of SADS-CoV. As shown in Figure 5i, TMEM53 lacking the TM domain no longer inhibited SADS-CoV RdRp activity. Therefore, the TM domain is likely important for the anti-SADS-CoV activity of TMEM53.

Figure 5.

TM domain is critical for the antiviral activity of TMEM53. (a) Illustration depicting truncations of TMEM53. (b-c) Huh7 cells stably expressing TMEM53-FL or TMEM53-ΔTM were infected with SADS-CoV (MOI = 0.1) for 48 h, and the viral RNA in the supernatant (b) and cytoplasm (c) were determined by RT-qPCR. (d) TMEM53 colocalized with ER and Golgi apparatus. HeLa cells were co-transfected with FLAG-tagged TMEM53 and SecG1β-GFP (ER marker) or B4Gal-Ti-RFP (Golgi apparatus marker) for 48 h, respectively. Cells were harvested and visualized by immunofluorescence microscopy. DAPI was used for nuclei detection. (e) Colocalization of TMEM53-ΔTM with ER and Golgi apparatus. The procedure was the same as in (d). (f-g) The interaction between TMEM53-ΔTM and NSP12 (f) or NSP8 (g) in HEK293T cells. (h) TMEM53-ΔTM did not affect NSP8 and NSP12 interactions. HEK293T cells were co-transfected TMEM53-ΔTM with NSP8 and NSP12 plasmids for 48 h before harvesting for Co-IP assay. (i) TMEM53-ΔTM did not affect SADS-CoV RdRp activity. Control and Huh7– TMEM53-ΔTM cells were co-transfected with RdRp reporter plasmids and RdRp plasmids for 24 h or 48 h before measuring the RdRp activity. The data shown are representative of three independent experiments with similar results. Data were presented as mean ± SD. ns: non-significant, **P < 0.01, two-tailed unpaired Student t test.

TMEM53 is a potent HKU2-related CoVs species-specific antiviral factor

Finally, considering its potency against SADS-CoV, we wanted to determine whether TMEM53 has broad antiviral activity against other viruses. We evaluated its activity against various alpha – and beta-CoVs, including bat SARS-related WIV1, SARS-CoV-2, HCoV-OC43, HCoV-229E, and MERS-CoV, as well as two non-CoVs, HSV-1 and PR8 influenza A virus. The data showed that TMEM53 expression had no obvious inhibitory effects on any of the viruses tested (Figure 6a-d).

Figure 6.

TMEM53 inhibited the HKU2r-CoVs replication. (a) Huh7 cells expressing TMEM53 were infected with SADS-CoV (MOI = 0.1), WIV (MOI = 0.1), SARS-CoV-2 (MOI = 0.05), HCoV-OC43 (MOI = 0.5) and HCoV-229E (MOI = 0.5) for 48 h before harvesting for further analysis. The viral RNA was measured by RT-qPCR. (b) TMEM53 did not affect the replicon activity of MERS-CoV. Control cells and Huh7-TMEM53 cells were transfected with MERS-CoV replicon plasmids for 48 h before harvesting for Nluc value determination. (c-d) TMEM53-Huh7 cells were infected with HSV-1 (MOI = 1) and PR8 (MOI = 1) for 48 h before the harvest. The viral RNA was determined by RT-qPCR. (e-f) TMEM53 inhibited the replicon activity of bat 162140-CoV and HKU2-CoV. Huh7-TMEM53 cells were transfected with replicon plasmids for 162140-CoV or HKU2-CoV. Cells were harvested and the Nluc value was analyzed at the indicated time points. (g) Sequence identity of the NSP12 protein in coronaviruses. (h) TMEM53 interacted with NSP12 protein from HKU2-related CoVs but not SARS-CoV-2. HEK293T cells were co-transfected TMEM53-FLAG and the indicated HA-tagged NSP12 plasmids for 48 h. Cells were harvested for Co-IP assay.

Bat HKU2-related CoVs are a large family of genetically divergent CoVs, from which SADS-CoV originated [4]. Therefore, we assessed whether TMEM53 has broad antiviral activity against this viral species. To this end, we developed bat 162140-CoV and bat HKU2-CoV replicon systems as surrogate models for viral genome replication (Figure 6e, f, upper panels). According to our previous study, 162140-CoV is the closest virus to SADS-CoV, whereas HKU2-CoV is relatively distant [4]. Strikingly, TMEM53 expression significantly inhibited the replicon activity of both 162140-CoV and HKU2-CoV (Figure 6e, f). As TMEM53 exerts antiviral activity by binding to NSP12, the antiviral activity against HKU2-related CoVs can be attributed to their NSP12 sequences being relatively close to that of SADS-CoV, unlike those of SARS-CoV-2, WIV1, and MERS-CoV (Figure 6g). Finally, we assessed the binding activity between TMEM53 and NSP12 from 162140-CoV or SARS-CoV-2. As expected, co-IP data revealed that TMEM53 efficiently bound to NSP12 of HKU2-related CoVs, but not that of SARS-CoV-2 (Figure 6h). Taken together, these results suggested that TMEM53 is a potent HKU2-related CoV species-specific antiviral factor.

Discussion

In this study, we hypothesized that we could find a potential intervention target for SADS-CoV infection by screening host restriction factors against SADS-CoV, a bat-origin swine disease-causing agent with a high risk of animal-to-human spillover [6]. We identified TMEM53 as a novel cell-intrinsic host restriction factor through a cDNA library screen. We found that TMEM53 interacts with NSP12 and NSP8 and inhibits functional RdRp activity via perturbing the interaction between NSP12 and NSP8 (Figure S8). TMEM53 functionality does not rely on viral RNA binding or canonical IFN responses, and TMEM53 exerts an antiviral effect on other bat HKU2-related CoVs in a species-specific manner. Therefore, TMEM53 may be a potential therapeutic target for HKU2-related CoV infections and useful for preparing for future outbreaks of this viral species.

TMEM53 was initially described as a nuclear envelope transmembrane (NET) protein because of its nuclear envelope localization [34]. Two studies revealed its functions in cell cycle regulation and induction of the rare disease, sclerosing bone disorder, respectively [35,36]. As far as we know, this is the first study to characterize the role of TMEM53 as a viral infection inhibitor. As a cell-intrinsic restriction factor, constitutive TMEM53 expression in the lungs and intestine, the two main entry portals for CoVs, may provide a critical first line of host defense against SADS-CoV and related virus infections. Although it showed HKU2-related CoV species-specificity in the current study, TMEM53 may be functional against other viruses not tested in this study. Since SADS-CoV is a swine virus, further study must determine whether TMEM53 from pigs or other species has anti-SADS-CoV activity. Although there are currently no reports of zoonotic infection of humans by SADS-CoV, our findings provide a promising intervention target against potential human outbreaks.

cDNA expression library screening has been widely used to identify host restriction factors for viruses including HIV-1, Ebola, Zika, coxsackievirus, and SARS-CoV-2 [27,37–40]. For SADS-CoV, host factors modulating viral infection remain to be identified and are less well understood, although some host dependency factors such as ZDHHC17 and PLAC8, have been identified through whole-genome CRISPR-Cas9-mediated high-throughput screenings [13,41]. Besides TEME53, our cDNA library screening uncovered LY6E as a top hit. It has previously been shown that LY6E inhibits the entry of multiple CoVs by blocking viral and cellular membrane fusion [22], which supports the reliability of our screening and warrants further study of this protein. Whether other viral inhibitors exist requires further investigation using multi-omics methods.

Upon CoVs infection, the viral RNA is released into the cytoplasm and translated by the host cell machinery into two polyproteins (pp1a and pp1ab). They are cleaved into 16 individual non-structural proteins to establish viral replication and transcription complexes for viral RNA synthesis [42]. Among them, NSP12 is the central component of the replication and transcription machinery of CoVs, together with the NSP7-NSP8 heterodimer and an additional NSP8 subunit, which form the minimal core RdRp complex [43]. We found that TMEM53 interacts with NSP12 and NSP8 and impairs the NSP12-NSP8 interaction, but not NSP7-NSP8 heterodimer formation (Figure 4 and Supplementary Figure S6e). In addition, TMEM53-ΔTM lost the ability to disrupt NSP12-NSP8 interaction and substantially reduced antiviral activity (Figure 5). Therefore, disrupting NSP12-NSP8 complex formation may be crucial for TMEM53 to exert its antiviral effects.

RdRp is an important drug target for RNA viral diseases, and the development of RdRp inhibitors is crucial for clinical therapy [44,45]. Abundant evidence indicates that RdRp inhibitors can be used to treat virus infection [46–48]. We found that TMEM53 distinctly suppresses SADS-CoV RdRp activity by disrupting NSP12-NSP8 complex formation. However, the structural basis of TMEM53-mediated RdRp inhibition remains to be determined. Co-IP results showed that NSP12 interacts with TMEM53 via its N-terminal domain, where the Nidovirus RdRp-associated nucleotidyltransferase (NiRAN) domain is located. The NiRAN domain of NSP12 is vital for viral RNA synthesis because of its GTPase activity [49]. Recent studies have suggested that NSP9 binds to the NiRAN domain and affects RNA capping structure synthesis by affecting the enzymatic activity of the NSP12 NiRAN [50,51]. Further investigations are necessary to determine whether TMEM53 affects the RNA capping process by binding to other NSPs, such as NSP9, and that determining the protein structure of the TMEM53-NSP12 complex will benefit the development of RdRp-targeting inhibitors.

Taken together, our findings unveiled a novel function of TMEM53 in virus restriction and revealed that TMEM53 inhibits SADS-CoV replication by disrupting the formation of a functional core polymerase complex, thereby inhibiting viral RNA synthesis. Additionally, our study established TMEM53 as a potent inhibitor of HKU2-related CoVs and provided insights for antiviral drug development.

Funding Statement

This work was supported by the National Natural Science Foundation of China [grant no. 31830096], China Postdoctoral Science Foundation [grant no. 2021M703453], the National Key R&D Program of China [grant no. 2021YFC2300901 and 2022YFC2305500] and the Self-Supporting Program of Guangzhou Laboratory [grant no. SRPG22-001].

Disclosure statement

No potential conflict of interest was reported by the author(s).