African swine fever virus pI73R hijacks host mRNA to suppress interferon beta and reveals a druggable target for antiviral intervention

ABSTRACT

African swine fever (ASF) is an acute transmissible disease caused by the African swine fever virus (ASFV), which has evolved multiple mechanisms to circumvent the host immune response. This research reveals that ASFV pI73R (I73R protein) downregulates cGAS-STING-mediated interferon beta (IFN-β) activation. ASFV pI73R binds to TBK1 and IRF3 mRNAs, promoting their nuclear retention and attenuating IFN-β production. Furthermore, silencing I73R enhances IFN-β mRNA levels and increases TBK1 and IRF3 protein levels in primary porcine alveolar macrophages (PAMs) exposed to ASFV. Unlike other reported mechanisms of inhibition of IFN-β by other ASFV proteins, pI73R inhibits IFN-β through its RNA binding function. Using structure-based virtual screening, we identified a small-molecule compound, STL527159, which disrupts the pI73R-RNA interaction, thereby restoring nuclear export and translation of the target TBK1 and IRF3 mRNAs. This compound markedly reduced ASFV infection in vitro. These findings provide new perspectives on the immune evasion strategies used by ASFV and offer a novel theoretical foundation for the development of antiviral drugs and novel vaccine candidates against ASFV.

Introduction

African swine fever (ASF), a highly transmissible disease triggered by the African swine fever virus (ASFV), results in nearly 100% fatality rates in domesticated pigs [1]. ASFV was initially documented in Kenya in 1921 [2], and outbreaks have since spread throughout Africa, Europe, and East Asia [3]. Currently, ASF poses a considerable threat to several Asian nations including China, Vietnam, and Korea, resulting in substantial economic damages [4–6].

Confronted with this persistent threat, identifying viable drug targets and generating potent antiviral agents are vital for the prevention and control of ASFV. Numerous inhibitors of ASFV proteins have been identified. For instance, genistein [7,8] and arctiin [8] inhibit topoisomerase II activity, thereby suppressing ASFV replication in vitro. Additionally, phosphonoacetic acid and cytosine arabinoside inhibit ASFV by suppressing the synthesis of virus-specific DNA polymerases [9]. Although numerous anti-ASFV drugs have been reported [10,11], no effective drugs are currently available. The primary reason for this is the lack of drug targets and unclear mechanisms limiting drug development [12,13]. Therefore, additional drug targets need to be identified.

An extensive understanding of the virus is required to identify such targets. ASFV is a large DNA virus belonging to the Asfarviridae family that contains 150–167 open reading frames (ORFs) and encodes 150–200 viral proteins [14,15]. Certain proteins are essential for the process of viral replication, while others are dedicated to counteracting key host defence pathways, such as the cytosolic DNA-sensing cGAS-STING pathway which is pivotal for initiating interferon-beta (IFN-β) production [16–18].

ASFV encodes multiple antagonists that disrupt this pathway. Many ASFV proteins contribute to cGAS-STING pathway downregulation to evade immune clearance [19]. For example, ASFV pI215L suppresses cGAS-STING signalling by inhibiting K63-linked ubiquitination of TBK1 [20]. Moreover, ASFV L83L promotes autophagy-lysosomal degradation of STING to inhibit IFN-β [21], and pE120R inhibits IFN-β production by binding to IRF3 and preventing its activation [22]. Although numerous ASFV proteins have been recognized as antagonists of the cGAS-STING pathway, whether other viral proteins participate in host immune evasion remains unknown.

ASFV I73R is an early gene that encodes 72 amino acids and has been identified as a key virulence factor [23]. I73R contains a Zα domain and functions as a nucleic acid-binding protein capable of interacting with a broad range of substrates, including ssRNA, dsRNA, ssDNA and dsDNA [23–25]. pI73R retains cellular mRNAs within the nucleus to suppress the TNF-α pathway [23]. Furthermore, they interact with host proteins to regulate host gene expression and protein synthesis [26]. pI73R also inhibits cGAS-STING-mediated activation of the IFN-β by interfering with IRF3 and NF-κB transcription factors, and this inhibitory effect may be independent of its Z-DNA binding activity [27]. However, its specific mechanism of inhibiting IFN-β has not been elucidated.

Here, we determined that ASFV pI73R suppressed cGAS-STING-mediated IFN-β activation. Mechanistically, pI73R bound to TBK1 and IRF3 mRNAs, thereby promoting their nuclear retention and attenuating IFN-β production. I73R knockdown enhanced IFN-β mRNA levels and increased TBK1 and IRF3 protein levels. Given the critical role of pI73R in RNA binding, there remains a need to develop drugs specifically targeting pI73R, thereby providing evidence for the feasibility of this target. Through structure-based virtual screening of the ZINC natural product database against the essential RNA-binding interface of pI73R, followed by surface plasmon resonance (SPR) measurement validation, we identified a lead compound, STL527159. This small molecule disrupts the pI73R-RNA interaction, thereby restoring nuclear export and translation of TBK1 and IRF3 mRNAs. STL527159 significantly inhibited ASFV replication. Our findings provide new perspectives on the immune evasion strategies used by ASFV, and offer a theoretical foundation for the development of antiviral drugs and novel vaccine candidates against ASFV.

Materials and methods

Cells, virus strain, and drugs

HEK293 T, HeLa, and porcine alveolar macrophage cell line 3D4/21 were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, USA) supplemented with 10% foetal bovine serum (FBS) and 1% penicillin–streptomycin (Gibco, USA). Primary porcine alveolar macrophages (PAMs) and the genotype II ASFV China/LN/2018/1 (GenBank accession number OP856591.1) were provided by China Animal Health and Epidemiology Center (Qingdao, China). All ASFV China/LN/2018/1 relevant experiments were conducted at the China Animal Health and Epidemiology Center’s biological safety level three (BSL-3) laboratory. The test compounds (STK924213, STL527159, STK923882, STL534844, STK923899, STK923782, STK246933, and STK924221) were selected based on virtual screening of the ZINC natural product database and subsequent structural searches within the Vitas-M compound library (TargetMol, USA). Each compound was initially dissolved in dimethyl sulfoxide (DMSO) to prepare a 50 mM stock solution, which was subsequently diluted in cell culture medium as required for the experiments. Detailed information regarding the other reagents is provided in Supplementary Table S1.

Plasmids

In total, 144 ASFV genes were cloned into the pCMV-Myc vector. For example, the ASFV I73R gene was amplified from ASFV China/LN/2018/1 and cloned into the pCMV-Myc vector, named pCMV-Myc-I73R. The primer sequences used for amplification are listed in Supplementary Table S2. All the constructed eukaryotic expression plasmids were validated for successful protein expression using an indirect immunofluorescence assay. The plasmids were constructed and maintained in our laboratory. The pCMV-HA plasmids containing cGAS, STING, TBK1, and IRF3 were synthesized by Sangon Biotech (Shanghai, China). IFN-β, ISRE, and TK-Renilla reporter plasmids were preserved in our lab. The truncated forms of I73R, including I73R-1, I73R-2, I73R-3, and I73R-4, were inserted into pcDNA3.1-C-EGFP, which was synthesized by Sangon Biotech.

Dual-luciferase reporter assay

HEK293 T cells were co-transfected with 100 ng IFN-β or ISRE reporter plasmid, 10 ng Renilla-TK reporter plasmid, 10 ng pCMV-HA-cGAS, 10 ng pCMV-HA-STING, together with 100 ng ASFV ORFs eukaryotic expression plasmids or empty vector for 24 h. Cells were harvested and lysed to measure luciferase activities via Dual-luciferase Reporter Assay kit. To determine transfection efficiency, firefly luciferase activity was normalized to Renilla luciferase activity. The cell lysates were used for protein expression analysis.

Enzyme-linked immunosorbent assay (ELISA)

HEK293 T cells were transfected with pCMV-HA-cGAS, pCMV-HA-STING, and pCMV-Myc-I73R or empty vector for 24 h. Alternatively, 3D4/21 cells were transfected with pCMV-Myc-I73R or empty vector for 24 h and then stimulated with poly(dA:dT) for 12 h. The concentration of IFN-β in cell culture supernatants was quantified using the commercial ELISA kits (Table S1), according to the manufacturer’s instructions. Briefly, samples and standards were added to antibody-precoated wells and incubated at room temperature for 60 min. After washing, the biotinylated detection antibody was added and incubated for 60 min, followed by incubation with a streptavidin-horseradish peroxidase (HRP) conjugate for 20 min. Following the TMB substrate reaction, absorbance was measured at 450 nm. The IFN-β concentration in each sample was determined by interpolation from the standard curve.

RNA extraction and quantitative reverse transcriptase PCR (qRT-PCR)

HEK293 T cells were transfected with pCMV-HA-cGAS, pCMV-HA-STING, and pCMV-Myc-I73R or empty vector for 24 h. Cells were then lysed with TRIzol, followed by phase separation with chloroform. The aqueous phase containing RNA was collected, and the RNA was precipitated with isopropanol, washed with ethanol, and dissolved in RNase-free water. Subsequently, RNAs was reverse-transcribed into cDNA using an Evo M-MLV RT kit. The Applied Biosystems QuantStudio System (CA, USA) was used to perform qRT-PCR with cDNAs and SYBR Green Premix Pro. β-actin was used as an internal reference. The primer sequences are listed in Supplementary Table S2.

Western blotting

HEK293 T cells were transfected with pCMV-HA-cGAS, pCMV-HA-STING, and pCMV-Myc-I73R or empty vector for 24 h. Cells were collected and lysed using cell lysate buffer. The clarified lysates were denatured in 5× loading buffer at 100°C for 10 min. Samples were separated by sodium sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. After blocking, the membranes were incubated overnight at 4 °C with specific primary antibodies, including those against phospho-TBK1 (1:1000), phospho-IRF3 (1:1000), TBK1 (1:1000), and IRF3 (1:1000). This was followed by incubation with the appropriate horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse antibodies. The protein bands were visualized using an Amersham Imager 600 (GE Healthcare, USA) with an enhanced chemiluminescence substrate. Detailed information on the antibodies used is provided in Supplementary Table S3.

Immunofluorescence microscopy

HeLa cells were transfected with pCMV-Myc-I73R or empty vector for 24 h and then stimulated with poly(dA:dT) for 12 h. The cells were fixed with 4% paraformaldehyde for 10 min and permeabilised with 0.1% Triton X-100 for 10 min. After blocking, the cells were incubated with the anti-Myc antibody (1:1000) overnight at 4℃. After washing, the cells were incubated with Alexa Flour 633 goat anti-mouse IgG (H + L) (1:500) and CoraLite® plus 488-conjugated IRF3 polyclonal (1:200) antibodies for 2 h. Subsequently, DAPI (1:10000) was used to stain the cells nuclei at 37℃ for 10 min. A Leica Stellaris confocal microscope (Solms, Germany) was used to visualize subcellular co-localization.

PAMs were cultured in 24-well plates (5 × 10⁵ cells/well) and infected with ASFV at a multiplicity of infection (MOI) of 0.1. The Cells were then exposed to the test compounds at different doses. The cells were fixed, permeabilised, and blocked as described above at 72 hpi. After incubation with ASFV p30 antibody (1:1000), Alexa Flour 488 goat anti-mouse IgG antibody (1:500) and DAPI, echo revolve generation 2 fluorescence microscope was used for imaging.

RNA immunoprecipitation (RIP)

RNA was immunoprecipitated using a commercial RIP kit (Table S1). Briefly, after transfection with pCMV-Myc-I73R for 24 h, HEK293 T cells were lysed in the kit-provided RNA lysis buffer supplemented with RNase and Protease inhibitor cocktail. The cell lysates were incubated with anti-Myc magnetic beads or control IgG beads overnight at 4°C. After washing with RIP wash buffer, the RNA bound to the beads was eluted and isolated using TRIzol. The levels of the target mRNAs in the immunoprecipitated RNA were determined by qRT-PCR, as described in the previous section.

Fluorescence in situ hybridization (FISH)

A FISH kit and Cy3-labeled mRNA FISH Probe Mix for TBK1, IRF3, GAPDH or IKKβ were bought from RiboBio (Guangzhou, China) to conduct FISH assay. Briefly, HEK293 T cells were transfected with pCMV-Myc-I73R or empty vector for 24 h. Cells were then fixed with 4% paraformaldehyde for 10 min and permeabilised with 0.5% Triton X-100 for 5 min. After blocking in pre-hybridization buffer for 30 min, cells were incubated with TBK1, IRF3, GAPDH or IKKβ probe overnight at 37℃. After washing by hybridization wash, cells were incubated with anti-Myc antibody (1:1000) overnight at 4℃, followed by the Alexa Fluor 488-conjugated goat anti-mouse IgG secondary antibody (1:500). After DAPI counterstaining, the cells were imaged using a confocal microscope.

RNA interference

PAMs were cultured in 24-well plates (5 × 10⁵ cells/well) and transfected with 200 nM of small interfering RNAs (siRNAs) targeting I73R (siI73R) or a non-targeting control siRNA (siNC) for 12 h. Subsequently, cells were infected with ASFV at an MOI of 0.1 for 48 h. Cells were harvested for qRT-PCR and western blotting. The siRNAs listed in Supplementary Table S4 were synthesized by Angel Biotech (Suzhou, China).

Structure-based virtual screening

The crystal structure of pI73R (PDB ID: 7VIV), resolved from the genotype II ASFV Georgia 2007/1 isolate, was used as the template for docking [25]. The protein sequence was identical to that of the ASFV China/LN/2018/1 strain used in this study. The dimeric form of the protein is retained. The binding pocket was defined as centred on three key residues (Asn44, Tyr48, and Trp68) [25], with a grid box of dimensions 26 × 26 × 26 ų centred at coordinates (−13.077, 9.282, – 9.939). Small-molecule ligands undergo structural processing before docking [28]. Subsequently, 127,694 compounds sourced from the ZINC natural product database were docked to pI73R using the AutoDock Vina. Protein–ligand interaction analysis was performed using PyMOL and LigPlus.

Surface plasmon resonance measurements

The I73R was inserted into the pCold vector for expression analysis. The recombinant plasmid was transfected into E. coli BL21(DE3) cells. Protein expression was induced by IPTG, and the bacterial cells were harvested and lysed by sonication on ice in a lysis buffer. His-tagged pI73R was purified by Ni-NTA affinity chromatography and dialyzed in PBS prior to use in subsequent experiments.

The binding kinetics between pI73R and STL527159 was determined using a Biacore T200 instrument. His-tagged pI73R was loaded onto a Ni-NTA biosensor chip (Cytiva, USA) and allowed to saturate for 90 s. For KD analysis, compound solutions with a two-fold dilution from 25 to 1.5625 μM were flowed over the chip. Each analysis cycle consisted of a 60 s baseline measurement, 120 s association stage, and 120 s dissociation stage in a running buffer. The assay was performed in a running buffer containing PBS supplemented with 5% DMSO. Response units were measured for each concentration during the equilibrium phase. For data analysis, the response from the reference surface (blank Ni-NTA channel without an immobilized ligand) was subtracted from that of the active channel to obtain specific binding signals. The binding kinetics were then analysed using Biacore T200 Evaluation Software.

Cytotoxicity assay

The cytotoxic effects of STL527159 on HEK293 T cells and PAMs were assessed using a CCK-8 assay. In 96-well plates, cells (2 × 10⁵ cells/ well) were seeded, and STL527159 (0, 2.5, 5, 7.5, 10, 12.5, 25, and 50 μM) were administered. After 72 h, 10% CCK-8 solution was added to the cells. The absorbance was quantified at 450 nm. The STL527159 concentration that induced 50% cytotoxicity (CC₅₀) was calculated from the absorbance data.

Subcellular fractionation

PAMs were seeded in 6-well plates (4 × 10⁶ cells/well) and divided into three groups for subsequent RNA fractionation: mock, ASFV, and ASFV + STL527159. The latter two groups were infected with ASFV China/LN/2018/1 (MOI = 1) for 2 h. Following infection, the culture medium was replaced: the mock and ASFV groups received fresh medium containing DMSO, while ASFV + STL527159 group received medium containing 10 µM STL527159. The cells were then incubated for 24 hpi. Cytoplasmic and nuclear RNA were separately purified from each group using commercial cytoplasmic and nuclear RNA purification kits. The subcellular mRNA levels of TBK1, IRF3, and IKKβ were quantified by qRT-PCR. GAPDH and U6 were used as reference genes for cytoplasmic and nuclear fractions, respectively.

Antiviral activity assay

PAMs were cultured in 24-well plates (5 × 10⁵ cells/well) and infected with ASFV China/LN/2018/1 at an MOI of 0.1 for 1 h. The viral inoculum was removed and replaced with a fresh medium containing different concentrations of the test compounds. Following a further 72 h incubation period, the cells and supernatants were harvested for titration. Viral titres were determined by hemadsorption (HAD) assay, while viral DNA copies and protein expression levels were analysed by quantitative PCR (qPCR), indirect immunofluorescence assay (IFA), and western blotting.

Viral DNA extraction and quantitative PCR (qPCR)

Following the compound treatment described above, ASFV genomic DNA was extracted from infected PAM cell cultures using the TIANamp Genomic DNA Kit. The ASFV B646 gene (p72) was assessed using the qPCR Master Mix on a QuantStudio system. The viral genome copies in these samples were quantified using p72 gene numbers and calculated according to a standard curve. The primer sequences are listed in Supplementary Table S2.

HAD assays

Following the compound treatment and infection protocol detailed above, both infected cells and their culture supernatants were collected and subjected to ten-fold serial dilutions. Virus titration was performed on PAMs using a previously described method [29,30]. ASFV titres were determined based on HAD of erythrocytes surrounding the infected cells. HAD was monitored for 7 days, and 50% HAD doses (HAD₅₀) were determined using the Reed and Muench method [29,30].

Direct inactivation and time of addition assays

Direct interaction and time-of-addition assays were conducted as previously described [31]. For direct interaction experiment, ASFV (MOI = 0.1) was incubated with 10 μM STL527159 for 1 h, and the mixture was diluted 20-fold and added to PAMs.

For time-of-addition assay, pre-treatment involved a 1 h incubation with 10 μM STL527159 before ASFV infection (MOI = 0.1). During-treatment was performed by co-incubating cells with ASFV (MOI = 0.1) and STL527159 (10 μM) for 1 h. In post-treatment, cells were infected with ASFV (MOI = 0.1) for 1 h at 4°C (post-1) or 37°C (post-2), followed by treatment with STL527159 (10 μM). At 72 hpi, viral B646L gene copy number and protein expression were analysed by qPCR and western blotting.

Statistical analysis

Unpaired Student’s t-test were used to analyse the differences between two groups using GraphPad Prism 8 software (Inc., San Diego, CA, USA). Statistical significance is presented as *P < 0.05, **P < 0.01, and ***P < 0.001.

Results

ASFV pI73R inhibits cGAS-STING-mediated IFN-β activation

To identify viral proteins inhibiting cGAS-STING-mediated IFN-β activation, we screened 144 ASFV genes in HEK293 T cells and identified 45 candidates. Among the suppressor proteins, pI73R was the most effective (Figure S1).

To investigate the effect of pI73R, HEK293 T cells were co-transfected with pCMV-HA-cGAS, pCMV-HA-STING, pCMV-Myc-I73R or empty vector, together with IFN-β or ISRE reporter plasmid and Renilla-TK reporter. The promoter activity was also assessed. Results showed that cGAS and STING led to IFN-β and ISRE activation, whereas pI73R overexpression suppressed their activation (Figure 1A and B). Consistent with the reporter assay results, qRT-PCR analysis revealed that pI73R also decreased IFN-β, ISG54, and ISG56 mRNA levels (Figure 1D). At the protein level, ELISA confirmed that pI73R reduced IFN-β protein secretion by approximately two fold (Figure 1C).

Figure 1.

ASFV pI73R inhibits cGAS-STING-mediated IFN-β activation. (A-B) HEK293 T cells were co-transfected with 10 ng pCMV-HA-cGAS, 10 ng pCMV-HA-STING, 100 ng pCMV-Myc-I73R or empty vector, together with 100 ng IFN-β (A) or ISRE (B) reporter plasmid and 10 ng Renilla-TK reporter. Luciferase activities were detected at 24 h after transfection (hpt). Firefly luciferase activity was normalized to Renilla luciferase activity to account for transfection efficiency. (C-D) HEK293 T cells were transfected with 10 ng pCMV-HA-cGAS, 10 ng pCMV-HA-STING, and 100 ng pCMV-Myc-I73R or empty vector for 24 h. (C) IFN-β protein levels in cell culture supernatants were determined by ELISA. (D) IFN-β, ISG54, and ISG56 transcription were measured by qRT-PCR. (E-G) 3D4/21 cells were transfected with 1 μg pCMV-Myc-I73R or empty vector, together with 100 ng IFN-β reporter and 10 ng Renilla-TK reporter for 24 h, followed by transfection with poly(dA:dT) (1 μg/mL) for 12 h. (E) Luciferase activities were determined using a double-luciferase reporter assay. (F) IFN-β protein were analyzed by ELISA. (G) IFN-β mRNA levels were detected by qRT-PCR. Firefly luciferase activity was normalized to Renilla luciferase activity to account for transfection efficiency. Western blotting was used to identify protein expression. Three separate experiments are represented by the data (mean ± SEM). **P < 0.01 or ***P < 0.001.

Further verification in 3D4/21 cells showed that pI73R significantly inhibited IFN-β promoter activation induced by poly(dA:dT) (Figure 1E) or 2'3'-cGAMP (Figure S2A). Furthermore, pI73R reduced IFN-β protein levels (Figure 1F, Figure S2B) and transcription of IFN-β (Figure 1G, Figure S2C) induced by both stimulants. This indicated that pI73R inhibited cGAS-STING- mediated IFN-β production.

ASFV pI73R inhibits TBK1 and IRF3 phosphorylation and blocks IRF3 nuclear translocation

The activation of IFN-β requires IRF3. Phosphorylation of TBK1 is essential for its activation, which triggers IRF3 phosphorylation, homodimerization, and nuclear translocation [32]. To determine whether pI73R inhibits TBK1 or IRF3 phosphorylation, HEK293 T cells were transfected with pCMV-HA-cGAS, pCMV-HA-STING, along with pCMV-Myc-I73R or empty vector. Both TBK1 and IRF3 were phosphorylated by cGAS and STING. ASFV pI73R decreased IRF3 and TBK1 phosphorylation (Figure 2A and B). In 3D4/21 cells, we further found that the presence of pI73R inhibited TBK1 and IRF3 phosphorylation induced by poly(dA:dT) (Figure 2C and D) or 2’3’-cGAMP (Figure S2D and E). In the unstimulated state, pI73R decreases TBK1 and IRF3 phosphorylation.

Figure 2.

ASFV pI73R inhibits phosphorylation of TBK1 and IRF3 and blocks the nuclear translocation of IRF3. (A-B) HEK293 T cells were transfected with pCMV-HA-cGAS, pCMV-HA-STING, and pCMV-Myc-I73R or empty vector for 24 h. (A) Cell lysates were analyzed by western blotting to detect phosphorylated TBK1 and IRF3. (B) The ratios of p-TBK1 and p-IRF3 was quantified using Image J. (C-D) 3D4/21 cells were transfected with pCMV-Myc-I73R or empty vector for 24 h, followed by transfection with poly(dA:dT) (1 μg/mL) for 12 h. (C) Cell lysates were analyzed by western blotting to detect phosphorylated TBK1 and IRF3. (D) The ratios of p-TBK1 and p-IRF3 was quantified using ImageJ. (E-F) HeLa cells were transfected with pCMV-Myc-I73R or empty vector for 24 h and then stimulated with poly(dA:dT) for 12 h. (E) Laser confocal microscopy was used to visualize IRF3 (green), pI73R (red), and nuclei (blue). Bars, 50 μm. (F) Quantification of IRF3 nuclear translocation involved counting 100 cells across multiple zones. Three separate experiments are represented by the data (mean ± SEM). *P < 0.05, or **P < 0.01.

Although pI73R blocks the nuclear translocation of overexpressed IRF3 [27], it remains unclear whether the same effect occurs with endogenous IRF3. To investigate this, pCMV-Myc-I73R was transfected into the cells, followed by poly(dA:dT) stimulation. The subcellular localization of endogenous IRF3 was examined (Figure 2E). In the absence of stimulation, endogenous IRF3 was primarily localized in the cytoplasm. After stimulation, 35% of the IRF3 migrated to the nucleus in control cells. In contrast, only 4% of IRF3 was observed in the nuclei of cells expressing pI73R (Figure 2F). These findings demonstrate that ASFV pI73R suppresses the nuclear translocation of endogenous IRF3.

ASFV pI73R suppresses IFN-β activation by targeting TBK1 or downstream molecules

The promoter activity was evaluated to explore the molecules potentially targeted by pI73R. Results showed that pI73R markedly suppressed the activation of IFN-β (Figure 3) and ISRE (Figure S3A-D) promoters enhanced by these elements in a dose-dependent manner. Western blotting revealed that ASFV pI73R had no effect on the protein expression of cGAS (Figure 3A) and STING (Figure 3B) but suppressed the protein expression of TBK1 (Figure 3C) and IRF3 (Figure 3D). TBK1 and IRF3 phosphorylation was also reduced when TBK1 (Figure S4E) or IRF3 (Figure S4F) was combined with pI73R. Therefore, we speculated that ASFV pI73R targeted TBK1 or downstream molecules to block IFN-β production.

Figure 3.

ASFV pI73R suppresses the IFN-β activation by targeting TBK1 or downstream molecules. (A-D) HEK293 T cells were co-transfected with pCMV-HA-cGAS (A), pCMV-HA-STING (B), pCMV-HA-TBK1 (C), or pCMV-HA-IRF3 (D), together with IFN-β reporter, Renilla-TK reporter, and increasing amounts (0, 100, 250, or 500 ng) of pCMV-Myc-I73R. Luciferase activity was measured 24 h after transfection. Firefly luciferase activity was normalized to Renilla luciferase activity to account for transfection efficiency. Western blotting was used to identify protein expression. Three separate experiments are represented by the data (mean ± SEM). *P < 0.05, or **P < 0.01.

ASFV pI73R promotes nuclear retention of TBK1 and IRF3 mRNAs by binding to their mRNAs

Interactions between pI73R and cGAS, STING, TBK1, and IRF3 were evaluated using a glutathione-transferase (GST) pulldown assay. The results showed that pI73R did not pull down these proteins (Figure. S4A). Silver staining after the co-immunoprecipitation (CO-IP) assay revealed no significant differences between the control and experimental groups (Figures S4B).

Given that ASFV pI73R is a nucleic acid-binding protein with an affinity for both RNA and DNA, and exhibits a preference for host cellular RNAs with high GC content [23], an RIP assay was performed to verify the association between pI73R and TBK1 or IRF3 mRNAs. The results showed that pI73R specifically bound to endogenous TBK1 and IRF3 mRNAs, while control GAPDH and IKKβ mRNAs showed no significant enrichment (Figure 4A).

← Figure 4.

ASFV pI73R promotes the nuclear retention of TBK1 and IRF3 mRNAs by binding to their mRNAs. (A) HEK293 T cells were transfected with pCMV-Myc-I73R. Cell lysates were subjected to RNA immunoprecipitation (RIP) analysis. Protein expression was identified by western blotting. The mRNA levels of TBK1, IRF3, GAPDH, or IKKβ were quantified by qRT-PCR. (B-I) HEK293 T cells were transfected with pCMV-Myc-I73R or empty vector. The subcellular localization of TBK1, IRF3, GAPDH, or IKKβ mRNA was analyzed by fluorescence in situ hybridization (FISH). Laser confocal microscopy was used to visualize subcellular localization of TBK1 (B), IRF3 (D), GAPDH (F), or IKKβ (H) mRNA (red), pI73R (green), and nuclei (blue). Bars, 50 μm. The nucleus-to-cytoplasm fluorescence intensity ratios for TBK1 (C), IRF3 (E), GAPDH (G), or IKKβ (I) mRNA were calculated from 20 cells per condition. (J-K) HEK293 T cells were transfected with pCMV-HA-TBK1 (J) or pCMV-HA-IRF3 (K), alone or together with pCMV-Myc-I73R. Expression of TBK1, IRF3, pI73R, and β-actin was detected by western blotting. Three separate experiments are represented by the data (mean ± SEM). ***P < 0.001.

Given that pI73R can bind to TBK1 and IRF3 mRNAs, FISH was used to label TBK1 and IRF3 mRNA to determine whether pI73R affects mRNA localization. In the vector group, TBK1 and IRF3 mRNAs were predominantly localized in the cytoplasm. The aggregation of TBK1 (Figure 4B) and IRF3 (Figure 4D) mRNAs was observed in the nucleus when pI73R was overexpressed. Quantitative analysis of the nucleus-to-cytoplasmic fluorescence intensity ratio confirmed the statistical significance of nuclear retention (Figure 4C and E). In contrast, the localization of control GAPDH (Figure 4F and G) and IKKβ (Figure 4H and I) mRNAs was unaffected. Consistent with the mRNA retention, overexpression of pI73R reduced TBK1 (Figure 4J) and IRF3 (Figure 4K) protein expression. These findings show that pI73R affects the nuclear export of TBK1 and IRF3 mRNAs by binding to their mRNAs, promoting their nuclear retention, and reducing TBK1 and IRF3 protein levels.

To identify the key domains of pI73R that contribute to its immunosuppressive function, four mutant forms of pI73R were generated (Figures S5A). Except for the weak expression of I73R-4, the other proteins were expressed at levels comparable to those of the full-length protein (Figure S5B). We then assessed the functional impact of these truncations on cGAS-STING-mediated IFN-β promoter activity. Consistent with previous results, the intact I73R-EGFP protein significantly inhibited the activation of IFN-β promoter, but all mutants lost their inhibitory capacity (Figure S5). Furthermore, RIP assays demonstrated that the truncation diminished the ability of pI73R to bind TBK1 and IRF3 mRNAs (Figure S6). Notably, the I73R-2 mutant completely lost its ability to bind to TBK1 mRNA while retaining partial binding to IRF3 mRNA (Figure S6C). These results supports the speculation that pI73R attenuates IFN-β production by binding to TBK1 and IRF3 mRNAs, and underscore the critical dependence of this mechanism on an intact pI73R structure.

ASFV pI73R inhibits IFN-β activation during viral infection by suppressing TBK1 and IRF3

To investigate the physiological kinetics of pI73R expression and its correlation with the host response, we analysed endogenous pI73R expression at multiple time points post-ASFV infection. Results showed that pI73R was detectable as early as 4 hpi. Its transcriptional (Figure S7A) and protein (Figure S7C) expression kinetics mirrored those of the known early gene CP204L (p30), confirming its classification as an early expressed gene. Concurrent monitoring of the host IFN-β response revealed a sharp increase in IFN-β mRNA levels following infection, peaking at 8 hpi before gradually declining (Figure S7B). Notably, this peak in innate immune activation coincided with the initial accumulation phase of pI73R. We also examined the expression of TBK1 and IRF3. Although ASFV infection did not significantly alter total mRNA levels, protein levels began to decrease from 4 hpi and were markedly reduced by 24 hpi (Figure S7C). Collectively, the early expression of pI73R is temporally synchronized with the onset of the host IFN-β response and the subsequent decline of TBK1/IRF3 proteins, supporting its pivotal role in promptly suppressing the initial antiviral defence.

To explore the role of pI73R in IFN-β induction during ASFV infection, PAMs were infected with ASFV China/LN/2018/1 48 h after transfection with siI73R. Compared to siNC, siI73R downregulated I73R transcription (Figure 5A) and protein levels (Figure 5F), as well as decreased the transcription (Figure 5B) and protein (Figure 5F) levels of B646L. Meanwhile, after siI73R transfection, IFN-β (Figure 5C), ISG54 (Figure 5D), and ISG56 (Figure 5E) mRNA levels were upregulated. Compared to the downregulated expression of TBK1 and IRF3 during ASFV infection, TBK1 and IRF3 protein expression increased upon siI73R transfection (Figure 5F). These date reveals that pI73R downregulates IFN-β production by suppressing TBK1 and IRF3 and thus inhibits the viral replication.

Figure 5.

ASFV pI73R inhibits IFN-β activation during viral infection by suppressing TBK1 and IRF3. (A-F) Primary porcine alveolar macrophages (PAMs) were transfected with siRNA targeting the I73R gene or a negative control siRNA for 12 h and subsequently infected with ASFV China/LN/2018/1 at a MOI of 0.1 for 48 h. (A-E) Cells were harvested, and mRNA levels of I73R (A), B646L (B), IFN-β (C), ISG54 (D), and ISG56 (E) were analyzed by qRT-PCR. (F) Cell lysates were analyzed by western blotting to detect the expression of TBK1, IRF3, ASFV p72, ASFV pI73R, and GAPDH. Three separate experiments are represented by the data (mean ± SEM). *P < 0.05, or **P < 0.01.

Virtual screening and identification of small molecules targeting pI73R

pI73R forms stable dimers [25], and three key residues (Asn44, Tyr48, and Trp68) [25] located at the dimer interface have been identified. Based on this knowledge, we reasoned that compounds that bind to this interface may disrupt the dimerization of pI73R or its interaction with host factors. To test this hypothesis, we defined a putative binding pocket centred on these three residues for structure-based virtual screening to identify small molecules with complementary steric and chemical features.

In total, 127,694 compounds from the ZINC natural library were docked to pI73R (Figure 6A). Results were ranked according to the predicted binding energies, which spanned from −10.6 to −2.4 kcal/mol (Figure 6B). Compounds with a binding energy stronger than −9 kcal/mol were considered to possess a high affinity potential. To enhance binding specificity and reduce potential off-target effects, compounds exhibiting binding energies more favourable than −10 kcal/mol and forming hydrogen bonds with key residues of pI73R were prioritized for screening. Finally, eight purchasable compounds were obtained.

Figure 6.

Virtual screening and identification of small molecules targeting pI73R. (A) Structure-based virtual screening workflow diagram. (B) Distribution of binding energy for compounds in libraries. The amount of compounds with predicted binding energies within specified ranges is shown by bars. (C) 2D chemical structure of STL527159. (D) The interactions between STL527159 and pI73R were analyzed by PyMOL and LigPlus. The blue folded structure represents the three-dimensional conformation of pI73R (left). The 3D (middle) and 2D (right) conformations of the docking result. The green stick structure denotes STL527159, while the blue stick structure indicates the amino acid residues of pI73R bound to STL527159. The red lines represent hydrogen bonds. (E) Surface plasmon resonance evaluation of affinity between STL527159 and pI73R. His-tagged pI73R was captured on NTA chip, and the chip surface was exposed to a 2-fold gradient concentrations of STL527159 from 50 μM to 1.5625 μM. The real-time binding curves are presented.

Viral inhibition experiments were performed to assess the antiviral effects of these candidates. The compounds STK924213, STL527159, STK923882, STL534844, and STK923899 exhibited varying degrees of antiviral activity (Figure S8). Among these, STL527159 demonstrated the most potent inhibitory effect (Figure S8). Therefore, STL527159 was selected for the subsequent experiments.

The structure of STL527159 is shown in Figure 6C. To characterize the binding mode, STL527159 was docked into the defined pocket of pI73R. Protein–ligand interactions were analysed using PyMOL and LigPlus. The analysis showed that STL527159 formed strong hydrogen bonds with the pI73R amino acid residues Tyr48 and Lys59 (Figure 6D). Furthermore, STL527159 engaged in hydrophobic interactions with several residues of each monomer of the pI73R dimer, including Ser51, Lys56, and Trp68 (Figure 6D). To assess its ability to bind to pI73R, His-pI73R was purified (Figure S9) and immobilized on an NTA chip for SPR detection. Results indicated that STL527159 presents strong binding to pI73R, exhibiting a KD value of 3.4 μM (Figure 6E).

STL527159 inhibits pI73R-RNA binding in HEK293 T cells

A CCK-assay was employed to assess the impact of STL527159 on cell HEK293 T cell activity. Results showed that after serial dilutions, STL527159 concentrations exceeding 12.5 μM exhibited significant negative effect on cell viability (Figure 7A). Thus we selected 12.5 μM of STL527159 for subsequent experiments.

Figure 7.

STL527159 inhibits pI73R-RNA binding in HEK293 T cells. (A) Cytotoxicity of STL527159 toward HEK293 T cells was assessed using the CCK-8 assay. HEK293 T cells (2 × 10⁵ cells/well) were seeded in 96-well plates and cultured overnight. Cells were treated with the indicated concentrations of DMSO or STL527159 for 24 h, followed by incubation with CCK-8 reagent for 1 h. Cell viability was determined by measuring the OD₄₅₀ value and normalized to that of cells cultured without STL527159 (set as 100%). (B-D) HEK293 T cells were co-transfected with pCMV-HA-cGAS, pCMV-HA-STING, pCMV-Myc-I73R or empty vector, together with IFN-β reporter and Renilla-TK reporter. Cells were then treated with DMSO or STL527159 (12.5 μM) for 24 h. (B) Cell lysates were analyzed for luciferase activity. (C) IFN-β protein levels in the culture supernatant were measured by ELISA. (D) IFN-β mRNA levels were determined by qRT-PCR. (E-F) HEK293 T cells were transfected with pCMV-Myc-I73R and treated with DMSO (E) or STL527159 (12.5 μM) (F) for 24 h. Cells were harvested for RIP. TBK1 and IRF3 mRNA levels were quantified by qRT-PCR. Western blotting was used to identify protein expression. Three separate experiments are represented by the data (mean ± SEM).*P < 0.05, **P < 0.01 or ***P < 0.001.

To assess the effect of STL527159 on pI73R-mediated inhibition of the cGAS-STING pathway, HEK293 T cells were co-transfected with cGAS and STING expression plasmids, together with either pCMV-Myc-I73R or empty vector, and then treated with DMSO or STL527159. Luciferase reporter assays revealed that pI73R potently inhibited cGAS-STING-induced IFN-β promoter activity by approximately 35.7-fold in DMSO-treated cells. This inhibition was substantially alleviated by STL527159 treatment, which reduced the fold inhibition to 6.6 (Figure 7B). At the protein level, ELISA confirmed that pI73R suppressed IFN-β secretion (3.1-fold inhibition in the DMSO group), and STL527159 treatment partially restored IFN-β levels (1.6-fold inhibition in its presence) (Figure 7C). Consistent with promoter activity and ELISA results, STL527159 also attenuated the inhibition of pI73R on IFN-β mRNA expression (Figure 7D). Overall, these results show that STL527159 significantly weakened the inhibitory capacity of pI73R.

To determine whether STL527159 directly interfered with the RNA-binding function of pI73R, we performed RIP assays. HEK293 T cells expressing Myc-tagged pI73R were treated with either DMSO or STL527159. Immunoprecipitation was performed using an anti-Myc antibody, with a control IgG antibody used to assess non-specific background. In DMSO-treated cells, pI73R specifically bound to endogenous TBK1 and IRF3 mRNAs, showing enrichments that were 3.6-fold and 6.3-fold higher, respectively, than that in the IgG control (Figure 7E). In contrast, STL527159 treatment markedly reduced binding capacity. The enrichment of TBK1 mRNA dropped only 1.4-fold compared to that of the IgG control, whereas binding to IRF3 mRNA was completely abolished (Figure 7F). These results confirm that STL527159 directly targets pI73R and inhibits its ability to bind specific host mRNAs, providing a mechanistic explanation for its capacity to restore IFN-β production.

STL527159 reverses nuclear retention of TBK1 and IRF3 mRNAs in PAMs

We first evaluated the non-toxic concentration of STL527159 in PAMs using the CCK-8 assay. Results revealed that STL527159 concentrations exceeding 12.5 μM exhibited significant negative effect on cell viability (Figure 8A). The 50% cytotoxic concentration (CC₅₀) is 31.0μM (Figure S10A).

Figure 8.

STL527159 reverses nuclear retention of TBK1 and IRF3 mRNAs in PAMs. (A) Cytotoxicity of STL527159 toward PAMs was assessed by CCK-8. PAMs (2 × 10⁵ cells/well) were seeded in 96-well plates and cultured overnight. Cells were treated with the indicated concentrations of DMSO or STL527159 for 72 h, followed by incubation with CCK-8 reagent for 1 h. Cell viability was determined by measuring the OD₄₅₀ value and normalized to that of cells cultured without STL527159 (set as 100%). (B–F) PAMs were infected with ASFV at a MOI of 1 for 2 h. The medium was then replaced with fresh medium containing 10 μM STL527159 or DMSO. At 24 h post-infection, cells were harvested for nuclear-cytoplasmic fractionation. The mRNA levels of U6 (B), GAPDH (C), TBK1 (D), IRF3 (E), and IKKβ (F) in the cytoplasmic and nuclear fractions were quantified by qRT-PCR. The percentage of transcript abundance in each fraction was calculated. Three separate experiments are represented by the data (mean ± SEM). ***P < 0.001.

To directly examine whether STL527159 counteracts the nuclear mRNA retention function of pI73R during infection, we performed subcellular fractionation of ASFV-infected PAMs treated with STL527159. The purity of the cytoplasmic and nuclear fractions was validated by the distinct localization of the reference genes U6 (nuclear) (Figure 8B) and GAPDH (cytoplasmic) (Figure 8C), which remained unchanged across the experimental groups. Analysis of the target mRNAs revealed a striking pattern. In the mock group, TBK1 and IRF3 mRNAs levels were predominantly in the cytoplasm (∼80% and ∼83%, respectively). ASFV infection alone caused significant redistribution, increasing the nuclear fractions to approximately 50% and 33%, respectively (Figure 8D and E). Treatment with STL527159 substantially reversed this effect, reducing the nuclear retention of TBK1 and IRF3 mRNAs to levels close to those in the mock group (∼27% and ∼11% nuclear retention, respectively). In contrast, the subcellular distribution of IKKβ mRNA, a non-target transcript in this pathway, was not significantly altered by either ASFV infection or STL527159 treatment (Figure 8F). These results demonstrate that STL527159 specifically alleviates the nuclear retention of TBK1 and IRF3 mRNAs during ASFV infection without affecting general mRNA localization.

STL527159 inhibits ASFV infection and restores IFN-β production during viral replication

To assess the antiviral impact of STL527159 on ASFV replication in PAMs, we measured the viral gene copy number, viral titre, and viral protein expression. qPCR results revealed that STL527159 inhibited ASFV B646L expression in a dose-dependent manner (Figure 9A). The 50% inhibitory concentration (IC₅₀) is approximately 1.1 μM (Figure S10B).

← Figure 9.

STL527159 inhibits ASFV infection during viral replication. (A-E) PAMs were infected with ASFV (MOI = 0.1) and treated with several doses of STL527159 (12.5, 10, 7.5, 5, 2.5 and 0 μM) (A) Supernatant was collected at 72 hpi, and ASFV B646L genome numbers were measured via qPCR using a standard curve. (B) Viral titres were measured by HAD assay using combined supernatants and lysates collected at 72 hpi. (C) The HAD₅₀ of ASFV in PAMs exposed to various doses of STL527159 at 72 hpi. (D) ASFV pI73R, p30, and p72 proteins expression at 72 hpi were detected by western blotting after treating with different STL527159 concentrations. (E) Immunofluorescence was performed to assess the dose-dependent antiviral effects of STL527159 at 72 hpi. p30 (green) and nuclear staining (blue) were examined under a fluorescence microscope Bars, 100 μm. (F) Diagram of the virucidal assay. STL527159 (10 μM) was incubated with ASFV (MOI = 0.1), diluted 20-fold, and then added to PAMs. Viral inactivation was examined at 72 hpi by qPCR (G) and western blotting (H). (I) Diagram of time-of-addition experiment. STL527159 (10 μM) and ASFV (MOI = 0.1) were added to PAMs at various time points. The antiviral effects of STL527159 at different stages of infection is presented as viral genome copies (J) and western blotting (K). Three separate experiments are represented by the data (mean ± SEM). ***P < 0.001.

In the HAD assay, STL527159 exhibited dose-dependent antiviral activity across the entire non-toxic concentration range tested (0-12.5 µM), inhibiting both the HAD phenomenon (Figure 9B) and viral titres. The antiviral effect was pronounced at 12.5 µM, where the viral titre was reduced by approximately 3 logs, from 10⁶ to 10³ HAD⁵⁰/mL (Figure 9C). Western blotting analysis revealed that STL527159 treatment at 2.5, 5, 7.5, 10, and 12.5 µM significantly inhibited the expression of viral proteins p30, p72 and pI73R (Figure 9D). Indirect immunofluorescence results showed a dose-dependent decrease in the fluorescence intensity of p30 following STL527159 treatment (Figure 9E). In summary, these results indicate that STL527159 potently inhibits the progression of ASFV in vitro infection.

We investigated whether STL527159 directly inactivates ASFV. ASFV was incubated with STL527159, diluted, and added to the PAMs (Figure 9F). qPCR and western blotting were performed at 72 h post infection (hpi). The results revealed that pre-incubation with STL527159 did not affect viral copy number (Figure 9G) or p72 protein expression (Figure 9H), suggesting that STL527159 does not directly inactivate ASFV.

To clarify the phase of the ASFV replication cycle, STL527159 exerts its effects, a time-of-addition assay, was performed. PAMs were infected with ASFV at a MOI of 0.1. The compound STL527159 (10 µM) or DMSO control was added at distinct time points relative to infection, defining four treatment groups (Figure 9I). In the pretreatment group, cells were pretreated with STL527159 for 1 h. The compound was then removed, followed by virus adsorption for 1 h at 4°C. In during group, STL527159 and virus were added simultaneously for 1 h at 37°C. In the post-1 group, virus adsorption was performed for 1 h at 4°C, followed by compound addition. The post-2 group was identical to post-1 except that virus adsorption occurred at 37°C. The antiviral effect was assessed by quantifying the viral genomic DNA (Figure 9J) and expression of the viral protein p72 (Figure 9K) at 72 hpi. STL527159 exhibited significant antiviral activity in all treatment phases. Pronounced inhibition of ASFV replication was observed when the compound was present during the adsorption phase (during) or immediately after adsorption (post-1 and post-2). Notably, the inhibitory efficacy of STL527159 against both viral DNA replication and p72 protein expression in the pre-treatment group was consistently lower than that observed in the during – and post-treatment groups. These results indicated that STL527159 exerts its antiviral activity during the post-entry and replication stages of ASFV infection.

To directly link the compound's effect to pI73R targeting and immune restoration, we examined the impact of STL527159 on IFN-β and its downstream signalling in ASFV-infected PAMs. The results showed that STL527159 treatment significantly downregulated viral gene transcription while markedly elevating IFN-β mRNA levels in infected cells (Figure S11A). Notably, although the total mRNA levels of TBK1 and IRF3 were not significantly altered, their protein levels increased substantially after treatment (Figure S11A and B). These findings suggest that STL527159 alleviates nuclear retention of TBK1 and IRF3 mRNAs, thereby restoring IFN-β production.

Discussion

We confirmed that pI73R bound to TBK1 and IRF3 mRNAs, thereby promoting their nuclear retention and attenuating IFN-β activation. This is the first study to describe a novel mechanism by which ASFV utilizes its RNA-binding protein to regulate IFN induction. Additionally, we used virtual screening and SPR assays to identify small molecules that target pI73R. For the first time, we report that the compound STL527159, obtained from the ZINC natural product database, effectively blocked pI73R-RNA binding by interacting with pI73R and significantly inhibiting ASFV infection in PAMs. This study provides a novel perspective for drug development based on ASFV protein structures.

ASFV encodes multiple proteins that antagonize the host innate immune response by targeting various nodes of the cGAS-STING signalling pathway to facilitate infection. In this study, we discovered that ASFV pI73R targeted TBK1 or downstream molecules to block IFN-β production (Figure 3). We observed that pI73R significantly downregulated protein expression and phosphorylation of TBK1 and IRF3 (Figure 2). Consistent with previous reports [27], it effectively blocked the nuclear translocation of IRF3 (Figure 2). However, pI73R did not interact directly with the core components of this pathway (cGAS, STING, TBK1, and IRF3) (Figure S4). Studies have suggested that viruses employ various mechanisms to reprogramme the host RNA metabolism and block cellular protein synthesis, thus facilitating viral protein expression and replication [23,33,34]. For instance, influenza A virus NS1 blocks the expression of interferon by preventing host mRNA nuclear export of host mRNA [33]. Previous studies reported that ASFV pI73R preferentially binds to high-GC-content host mRNA during ASFV replication, inhibiting host antiviral protein synthesis and altering cellular RNA metabolism [23,26]. We hypothesized that it may regulate IFN-β production by binding to mRNAs. RIP and RNA-FISH analysis showed that pI73R could bind to TBK1 and IRF3 mRNAs and promote their nuclear retention (Figure 4), thereby suppressing IFN-β production. Further analysis revealed that the GC content of IRF3 mRNA was as high as 62%, while TBK1 mRNA also contained some GC content regions, which may explain why pI73R bound more strongly to IRF3 than to TBK1 mRNA. This is in contrast to previously reported ASFV proteins involved in direct degradation or inhibition of host signalling proteins [35–38]. Thus, pI73R represents a novel escape tactic that operates during mRNA export, making it a promising drug target.

Although pI73R possesses an intrinsic affinity for GC-rich RNA [23], our experimental data indicated that the binding and nuclear retention of TBK1 and IRF3 mRNAs were specific relative to control transcripts (Figure 4). This specificity may arise from the unique sequence features or secondary structures within these immune-related mRNAs, such as GC-enriched regions or specific stem-loop motifs, which could provide high-affinity binding sites for pI73R. Further elucidation of these determinants requires transcriptome and structural analyses. Moreover, nuclear retention induced by pI73R interferes with the host mRNA export machinery. Canonical mRNA export depends on NXF1 (TAP)/NXT1 heterodimeric export receptors and adaptor proteins, including ALYREF [39,40]. Although the present study did not directly demonstrate an interaction between pI73R and this pathway, previous research has indicated that pI73R-mediated suppression of CRNKL1 may disrupt host mRNA splicing and/or nuclear export, thereby reducing the cytoplasmic availability of specific host factors [26]. Thus, pI73R could impede the proper assembly or function of the NXF1 complex through competitive binding to target mRNAs, disruption of adaptor protein activity, or direct interaction with the export receptor, leading to the selective sequestration of key immune signalling mRNAs. This aligns with the common strategy of multiple viruses to hijack host RNA export pathways to facilitate viral replication [41–43].

Our results showed that pI73R was detectable as early as 4 hpi, which is consistent with previous reports [23,24]. Notably, its accumulation phase overlaps significantly with the onset of the host's endogenous IFN-β response, which peaks around 8 h (Figure S7). This temporal correlation suggests that pI73R is pre-positioned to initiate the clogging of the mRNA export channel, just as the host initiates the transcription of antiviral genes. Consequently, the virus may exploit this time to evade the immune system before the host establishes a fully effective antiviral state. In this study, knockdown of I73R was shown to significantly enhance the host IFN-β response and reduce viral replication (Figure 5). Although our attempts to generate an I73R deletion mutant of ASFV China/LN/2018/1 were unsuccessful, this implied that I73R may be crucial for viral replication under our experimental conditions. Previous studies have indicated that this gene is not essential for viral replication, but plays a significant role during in vitro infections [23]. These divergent observations may stem from variations among different ASFV isolates of the same genotype. Furthermore, during long-term adaptation or selection, viruses may develop compensatory mechanisms. Similar to PCNA from other species, which can partially compensate for the function of ASFV pE301R [44], the function of pI73R may be supplemented by other ASFV proteins or structurally related host factors, such as transcription factors of the FOX family [26].

Recognizing pI73R as a promising drug target, we performed structure-based virtual screening to identify compounds that target its dimerization interface. The available structure revealed that this interface is critical for the RNA-binding activity of pI73R, with key residues directly involved in nucleic acid interactions [25]. Therefore, targeting this interface is designed to competitively or allosterically disrupt the binding of pI73R to its RNA substrates rather than non-specifically inhibiting dimerization. From this screen, the candidate compound STL527159 was identified. STL527159 effectively blocks the ability of pI73R to bind target mRNAs, thereby rescuing IFN-β production (Figure 6–8, Figure S11). This suggests that STL527159 likely binds to pI73R and sterically hinders its RNA-binding domain, leading to the release of sequestered TBK1 and IRF3 mRNAs and the subsequent promotion of IFN-β generation. The restored IFN-β response subsequently activates interferon-stimulated genes, establishing an antiviral state that effectively suppresses viral replication [45]. This mechanism explains the significant inhibitory effects observed during viral entry and replication (Figure 9).

In conclusion, we present evidence that ASFV pI73R promotes nuclear retention and attenuates IFN-β production by binding to TBK1 and IRF3 mRNAs. Knockdown of I73R promotes the transcription of ISGs and increases TBK1 and IRF3 protein expression. Unlike reported mechanisms of inhibition of IFN-β by other ASFV proteins, pI73R inhibits IFN-β through its RNA binding function. We also identified STL527159 as a lead compound capable of disrupting pI73R-RNA interaction and significantly hampering ASFV infection in vitro. These findings provide new perspectives on the immune evasion strategies employed by ASFV and offer a novel theoretical foundation for the development of antiviral drugs and novel vaccine candidates.

Funding Statement

This work was supported by Key research and development program (Modern Agriculture) project of Jiangsu Province (grant numbers BE2021331 and BE2020398), the 111 Project (grant number D18007), the Priority Academic Program Development of Jiangsu Higher Education Institutions (grant number PAPD).

Disclosure statement

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

Data availability statement

All data needed to evaluate the conclusions in the paper are present in the paper and the provided supplementary information.

Supplemental Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/22221751.2026.2640703.