Harnessing ZIKV NS2A RNA for alleviating acute hepatitis and cytokine release storm by targeting translation machinery

Abstract

Background and Aims:

Hyperactivated inflammatory responses induced by cytokine release syndrome are the primary causes of tissue damage and even death. The translation process is precisely regulated to control the production of proinflammatory cytokines. However, it is largely unknown whether targeting translation can effectively limit the hyperactivated inflammatory responses during acute hepatitis and graft-versus-host disease.

Approach and Results:

By using in vitro translation and cellular overexpression systems, we have found that the nonstructural protein gene NS2A of Zika virus functions as RNA molecules to suppress the translation of both ectopic genes and endogenous proinflammatory cytokines. Mechanistically, results from RNA pulldown and co-immunoprecipitation assays have demonstrated that NS2A RNA interacts with the translation initiation factor eIF2α to disrupt the dynamic balance of the eIF2/eIF2B complex and translation initiation, which is the rate-limiting step of translation. In the acetaminophen-induced, lipopolysaccharide/D-galactosamine-induced, viral infection-induced acute hepatitis, and graft-versus-host disease mouse models, mice with myeloid cell-specific knock-in of NS2A show decreased levels of serum proinflammatory cytokines and reduced tissue damage.

Conclusions:

Zika virus NS2A dampens the production of proinflammatory cytokines and alleviates inflammatory injuries by interfering translation process as RNA molecules, which suggests that NS2A RNA is potentially used to treat numerous acute inflammatory diseases characterized by cytokine release syndrome.

INTRODUCTION

Inflammatory responses are usually elicited to defend against invading pathogens or to clear damaged cells. Nevertheless, hyperactivated inflammatory responses have consistently been associated with cytokine release syndrome (CRS) and other disorders with elevated cytokine levels, ultimately leading to fatal organ damage.¹ Numerous clinical conditions are characterized by an overactive immune system and high levels of proinflammatory cytokines, including COVID-19, sepsis, acute hepatitis, and graft-versus-host disease (GVHD).^1–3 Therefore, limiting the release of excessive proinflammatory cytokines is crucial for balancing host immune responses and protecting patients.

In clinical conditions involving hyperactivated inflammation, the monocytes and macrophages are the major cell types that produce proinflammatory cytokines such as IL-6, TNF-α, and IL-1β.^4,5 Drug administration-induced and viral infection-induced acute hepatitis involve the activation of hepatic resident and inflammatory macrophages, resulting in the elevation of proinflammatory cytokines over the anti-inflammatory cytokines.^6,7 This process exacerbates hepatocyte damage and liver injury.⁸ As a typical clinical disorder that provokes CRS, GVHD, caused by allogeneic hematopoietic cell transplantation with severe symptoms, is associated with poor prognosis and low survival rates.³ Moreover, previous studies have indicated that macrophage infiltration and inflammation induction contribute to direct cytotoxicity and promote the development of GVHD.^9,10

Currently, corticosteroid and anti-cytokine-blocking antibody treatments are the 2 primary clinical therapies for acute inflammatory disorders. Corticosteroids, especially glucocorticoids, function as anti-inflammatory agents and effectively suppress the expression of proinflammatory genes by targeting specific transcription factors such as NF-κB in inflammation signaling pathways.¹¹ The monoclonal antibody tocilizumab targeting the IL-6 receptor and anakinra antagonizing the IL-1 receptor have been reported to be effective in treating severe cases of rheumatoid arthritis and systemic lupus erythematosus, respectively.^12,13 Furthermore, a recent study suggests that the translation process may serve as an effective therapeutic target for reducing the production of proinflammatory factor IL-6.¹⁴ This is based on the translational control of key regulators involved in inflammation.^14–16 The study revealed that metoprolol was able to target the translation of IL-6 mRNA to reduce its synthesis by downregulating the translation elongation mediated by the eEF2K-eEF2 axis.¹⁴ This strategy has been shown to alleviate CRS effectively, suggesting its potential application in ameliorating clinical inflammatory conditions by targeting the translation process.

Nonstructural protein encoded by the NS2A gene of Zika virus (ZIKV) has been shown to degrade adherens junction proteins that connect neural stem cells to impair cortical neurogenesis.¹⁷ NS2A also promotes the degradation of the primary nucleocytoplasmic transporter KPNA2 through chaperone-mediated autophagy, thereby facilitating ZIKV infection.¹⁸ Moreover, NS2A inhibits the activation of TBK1 and IRF3 to antagonize ectopic RIG-I and MDA5-mediated production of interferon-β (IFN-β).^19,20 However, it is unknown the underlying mechanism for the decrease in ectopic gene expression caused by NS2A and whether the downregulation of antiviral immunity by NS2A is a result of reduced protein levels of ectopic signals rather than direct suppression of the antiviral pathway.

Here, we present evidence that ZIKV NS2A extensively inhibits ectopic protein expression at the translational level. Mechanistically, NS2A targets cellular translation machinery as an RNA molecule, occupying the translation factor eukaryotic translation initiation factor 2 (eIF2)α to interfere with the translation process when an abundance of translation factors are active. Myeloid cell-specific introduction of the NS2A cassette into the mice or transplantation of bone marrow cells overexpressing NS2A significantly alleviates acute hepatitis and CRS of GVHD in mice, providing potential therapeutic strategies for alleviating hyperactivated inflammation conditions.

METHODS

Mice

NS2A conditional knock-in mice were constructed by Cyagen Biosciences (Suzhou, China) by inserting a cassette of the promoter CAG-driven NS2A open reading frame flanking the loxP and STOP (ploy A) sequence into the Rosa26 locus of the C57BL/6N mouse genome. Knock-in mice that specifically expressed NS2A in myeloid cells (NS2A^mKI) were generated by breeding loxP-STOP-flanked NS2A mice with LysM-Cre mice. The genotyping primers were 5′-GGCAACGTGCTGGTTATTGTG-3′ (NS2A-F) and 5′-CTCCAGTGTTCATTTCCGCGAAG-3′ (NS2A-R). The PCR products for NS2A^floxp and NS2A^mKI mice were 1293 bp and 416 bp, respectively. All mice were maintained in a specific pathogen-free environment at the Suzhou Institute of Systems Medicine (ISM) under a controlled temperature (25°C) and a 12-hour day-night cycle.

Surface sensing of translation assay

The surface sensing of translation (SUnSET) assay was performed as described.²¹ Briefly, HEK293T cells were transfected with plasmids expressing a nonstructured protein from ZIKV, with or without green fluorescent protein (GFP). After 24 hours, the medium was replaced with fresh medium, and the cells were incubated with puromycin (0.5 µg/mL) for 30 minutes before being harvested with cold PBS. Cells were lysed with NP-40 lysis buffer supplemented with a protease inhibitor for 30 minutes on ice. Proteins were subjected to SDS-PAGE and blotted with an anti-puro antibody (clone 12D10). For in vivo assay to detect protein synthesis in peritoneal macrophages, WT and NS2A^mKI mice were intraperitoneally injected with 2 mL of thioglycolate for 72 hours, followed by injection of 0.04 μmol/g puromycin.²² After 30 minutes, peritoneal macrophages were harvested to detect the incorporation efficiency of puromycin using immunoblotting.

Mouse models for hepatitis and CRS

All animal experiments were conducted following the US National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Animal Service Center of ISM (AUP no. ISM-IACUC-0011-R). The experiments were performed in compliance with the ARRIVE (Animals in Research: Reporting In Vivo Experiments) guidelines. Acute hepatitis and acute liver failure were induced using either acetaminophen (APAP) or lipopolysaccharide (LPS)/D-galactosamine (D-GalN). Concanavalin A was employed to facilitate T-cell-mediated hepatitis, while the murine hepatitis virus (MHV)-A59 strain was used to model fulminant hepatitis. Additionally, GVHD was implemented to investigate CRS.

Statistical analysis

The data represent the mean of at least 3 independent experiments, and the error bars represent the SD of the mean. Statistical significance was determined by unpaired two-tailed Student's t-test using GraphPad Prism 7 software. Survival data in vivo were analyzed using the Log-rank (Mantel-Cox) test. p<0.05 was considered as statistically significant difference.

RESULTS

ZIKV NS2A RNA inhibits translation independent of NS2A protein

To assess the effect of ZIKV NS proteins on translation in mammalian cells, we used the SUnSET assay by measuring puromycin incorporation efficiency into elongating polypeptide chains.²¹ Among all ZIKV NS proteins, only NS2A significantly resulted in translational inhibition of exogenous GFP (Figure 1A and Supplemental Figure S1A, http://links.lww.com/HEP/I690). The fluorescence observation, flow cytometry analysis, and western blotting results further confirmed that NS2A inhibited overexpressed GFP in comparison to the empty vector but had no effect on GFP transcription or stably expressed GFP (Figure 1B, C and Supplemental Figure S1B, http://links.lww.com/HEP/I690). The synthesis of nascent proteins was also assessed using ³⁵S labeling and O-propargyl-puromycin incorporation assays. The results indicated that NS2A did not influence the overall translation of the endogenous proteins (Figure 1D, E and Supplemental Figure S1C, http://links.lww.com/HEP/I690). NS2A exhibited a specific inhibitory effect on the translation of exogenous proteins in cells overexpressing GFP or mCherry (Figure 1D and F, G). Compared with other flaviviruses, only ZIKV NS2A showed a significant inhibitory impact on GFP expression, thereby indicating the specificity of ZIKV NS2A in disrupting the translation process within mammalian cells (Supplemental Figure S1D, http://links.lww.com/HEP/I690).

FIGURE 1.

Inhibition of translation by NS2A RNA from ZIKV. (A) SUnSET assay and GFP detection by immunoblot analysis in HEK293T cells co-overexpressed with GFP and nonstructured proteins from ZIKV or EV for 24 hours, α-tubulin was shown as a loading control. (B) GFP fluorescence detection by fluorescent microscope and MFI calculation by flow cytometry in NS2A and EV overexpressed HEK293T cells. (C) NS2A and NS2A mutant (NS2Amut) plasmids were transfected into HEK293T cells for 24 hours to detect the GFP protein level by immunoblot analysis and mRNA level by RT-qPCR, * represents the expected protein band of NS2A. (D) 50 µCi/mL of [³⁵S] methionine was incorporated into HEK293T cells that overexpressed NS2A or control vectors, with or without GFP, for subsequent SDS-PAGE and autoradiography. (E) Flow cytometry analysis of the MFI of fluorescently labeled O-propargyl-puromycin in HEK293T cells. (F and G) Flow cytometry analysis of the MFI of overexpressed mCherry (F) and the incorporation of O-propargyl-puromycin (G) in HEK293T cells. (H) 5 pmol of in vitro transcribed Luc and NS2A RNA molecules were transfected into GFP-overexpressing HEK293T cells for 12 hours to detect GFP expression by immunoblot analysis. (I) HEK293T cells were co-transfected with HA-mCherry plasmid, pSIF-H1-NS2A plasmid, or EV for 24 hours to detect mCherry expression by immunoblot analysis. (J) The working model of in vitro coupled transcription and translation system with RRL. (K) 100 ng of luciferase DNA template and 200 ng of GFP, NS2A, or NS2Amut RNA molecules were added in the RRL system to detect the relative activity (left) and mRNA level (right) of luciferase with GFP as the standard. (L) In total, 100 ng of luciferase DNA and 0, 0.2, 1 μg of NS2A RNA molecules were added to the RRL system, and the group without luciferase DNA was used as a NTC to calculate the relative activity of luciferase and levels of transcribed luciferase mRNA and NS2A RNA. Data of (B, right), (C, right), (E-G), and (K, L) are shown as the mean ± SD from 3 independent experiments. *p<0.05 and ***p<0.001 by unpaired Student's t-test. Data of (A), (B, left), (C, left), (D), and (H, I) are representative results from 3 independent experiments, scale bar, 100 μm. Abbreviations: EV, empty vector; GFP, green fluorescent protein; HA, hemagglutinin; MFI, mean fluorescence intensity; NTC, negative control; OPP, O-propargyl-puromycin; RRL, rabbit reticulocyte lysate; RT-qPCR, real-time quantitative PCR; SSC-A, side scatter-area; ZIKV, Zika virus.

To exclude the possibility that NS2A targets specific promoters, we constructed multiple GFP plasmids driven by different type II promoters, including CMV, EH, LTR, EF1α, CAG, and UBC, and found that NS2A significantly inhibited all GFP expression (Supplemental Figure S1E, http://links.lww.com/HEP/I690). Additionally, inhibitors targeting the proteasome or lysosome were unable to restore NS2A-mediated inhibition of the GFP protein (Supplemental Figure S1F, http://links.lww.com/HEP/I690), excluding the main post-translational protein regulatory pathways. These results suggest that NS2A suppresses protein expression mainly at the translational level rather than at the transcriptional, post-transcriptional, and post-translational levels.

Furthermore, we confirmed that the overexpression of NS2A did not affect cell proliferation and viability (Supplemental Figure S1G, H, http://links.lww.com/HEP/I690). RNA-seq analysis also indicated that NS2A did not alter cellular housekeeping gene expression or normal physiological functions (Supplemental Figure S1I, http://links.lww.com/HEP/I690). Although previous studies have successfully expressed NS2A protein in the form of fusion expression or tag labeling in mammalian cells,^17,18 we could not detect a specific protein band with an anticipated molecular weight (~25 kDa) when the hemagglutinin tag was fused at the 3′-terminal of the NS2A open reading frame (Figure 1C). Interestingly, when the start codon ATG was mutated to ATCG to disrupt the NS2A open reading frame (Supplemental Figure S1J, http://links.lww.com/HEP/I690), it still exhibited the ability to suppress GFP expression (Figure 1C). Therefore, we speculated that NS2A inhibited ectopic GFP expression independently of its protein.

Next, we used the nontranslated NS2A RNA, the in vitro transcribed or H1 RNA polymerase III promoter-driven RNA transcripts of NS2A to assess GFP expression. Our findings indicated that both forms of NS2A RNA molecules significantly reduced the expression level of exogenous proteins (Figure 1H, I), which suggested that NS2A RNA molecules, but not the NS2A protein, play important roles in the suppression of protein translation. Subsequently, we further validated the function of NS2A in an in vitro transcription and translation system using luciferase DNA molecules as a template (Figure 1J). Consistent with the suppression of GFP expression by NS2A, NS2A and NS2A-mutant RNAs inhibited luciferase expression at the protein level rather than at the mRNA level, and the inhibitory effect of NS2A was dose dependent (Figure 1K, L).

NS2A RNA hijacks cellular translation machinery to inhibit protein translation

Protein translation relies on multiple intracellular initiation, elongation, and termination factors. To investigate the potential translation factors that bind to NS2A RNA, we conducted RNA pulldown and mass spectrometry assays using biotin-labeled in vitro transcripts of NS2A (Supplemental Figure S2A, http://links.lww.com/HEP/I690 and Supplemental Table S1, http://links.lww.com/HEP/I691). The western blotting results verified the association between NS2A RNA and the translation initiation factors eIF2, eIF4E, and the elongation factor eEF1A1 (Figure 2A and Supplemental Figure S2B, http://links.lww.com/HEP/I690). We subsequently confirmed the interaction between endogenous eIF2α and NS2A RNA through an RNA immunoprecipitation (RIP)-quantitative polymerase chain reaction (qPCR) assay in HEK293T cells overexpressing NS2A (Figure 2B). It has been reported that eIF2α interacts with -2 and -3 nucleotides of mRNA to facilitate recognition of the AUG codon.²³ Herein, we performed RIP-qPCR to assess the occupancy of GFP mRNA on eIF2α and revealed that NS2A hindered the ability of eIF2α to recognize GFP mRNA, resulting in a decrease in translation efficiency (Figure 2B).

FIGURE 2.

NS2A RNA inhibits the translation initiation process. (A) RNA pulldown using 30 pmol of biotinylated GFP or NS2A RNA to detect bound proteins from cell lysates of iBMM and HEK293T by immunoblot analysis, with GAPDH as a loading control. (B) RNA immunoprecipitation-quantitative PCR analysis for the interaction of eIF2α with NS2A or GFP RNA in HEK293T cells with or without overexpressing GFP plasmid. (C) RNA pulldown assay with biotin-NS2A RNA and eIF2α truncations fused with GFP protein. (D) RNA pulldown using truncations of biotin-NS2A RNAs to detect eIF2α from cell lysates of HEK293T. (E) Immunoblot analysis for GFP expression in HEK293T co-transfected with GFP and NS2A truncations or control plasmids. (F) Co-immunoprecipitation assay using anit-eIF2α antibody and isotype anti-IgG in HEK293T cells co-transfected with GFP and NS4B or NS2A plasmids. (G) Proximity ligation assay detection and positive signals analysis for endogenous interaction of eIF2α and eIF2Bβ in WT and NS2A knock-in bone marrow-derived macrophages. (H) eIF2α and eIF2Bβ complex was detected using co-immunoprecipitation assay in GFP and NS4B or NS2A overexpressed HEK293T cells treated with 200 nM ISRIB for 24 hours or not. (I) The working model of NS2A RNA hijacking translation initiation process. Data of (A), (C), and (D–H) are representative results from 3 independent experiments, scale bar, 25 μm. Data of (B), (G, right) are shown as the mean ± SD from 3 independent experiments. **p<0.01 and ***p<0.001 by unpaired Student's t-test. Abbreviations: CTD, C-terminal domain; eIF2, eukaryotic translation initiation factor 2; EV, empty vector; GFP, green fluorescent protein; iBMM, immortalized bone marrow-derived macrophage; IP, immunoprecipitation; ISRIB, integrated stress response inhibitor; NTD, N-terminal domain; WT, wild-type.

To determine the specific sites of interaction between eIF2α and NS2A, we generated truncated and GFP-fused proteins of eIF2α and NS2A RNA fragments. RNA pulldown assay results indicated that the C-terminal domain (CTD, aa 187-303) of eIF2α interacted with the 5′-terminal 312 nucleotides of NS2A RNA (Figure 2C-D). NS2A truncation with the 5′-terminal 312 nucleotides demonstrated a significant capacity to suppress the expression of GFP to the same extent as the full-length gene (Figure 2E). Previous studies have shown that the CTD (aa 188-280) and N-terminal domain (NTD, aa 1-183) of eIF2α interact with each other in physiological states.²⁴ This intramolecular interaction impedes the binding surface of eIF2α-NTD on eIF2B, a guanine nucleotide exchange factor (GEF) responsible for activating eIF2. The deletion of the CTD or the phosphorylation of the NTD, which destabilizes the interaction, can lead to direct or indirect exposure of the binding surface. This exposure further prevents nucleotide exchange and translation initiation.^25,26 Thus, we hypothesized that NS2A RNA, which binds to eIF2α CTD, disrupts the intramolecular interaction between CTD and NTD to control the regular translation process. The in situ proximity ligation assay and co-immunoprecipitation results indicated a more constrained eIF2/eIF2B inhibitory complex that restricted GEF activity in NS2A-overexpressing HEK293T cells and bone marrow-derived macrophages compared to the control groups (Figure 2F, G). Notably, NS2A did not affect the phosphorylation of eIF2α or the upstream kinase protein kinase R (Supplemental Figure S2C, http://links.lww.com/HEP/I690). Integrated stress response inhibitor (ISRIB) is a compound that inhibits the integrated stress response by destabilizing the robust interaction between eIF2α and eIF2B to antagonize the inhibitory effect of eIF2α on the GEF activity of eIF2B. Our results also revealed that ISRIB reversed this constrained association and translation inhibition of GFP protein induced by NS2A (Figure 2H). These results suggest that NS2A impedes translation initiation by sequestering eIF2B in the eIF2 complex through its interaction with the CTD of eIF2α (Figure 2I).

In addition to cap-dependent translation in eukaryotic cells, numerous viruses have evolved a cap-independent mechanism for initiating translation by targeting the internal ribosome entry site (IRES) within the 5′ untranslated region (5′ UTR) of viral RNA. However, viral IRESs use a limited or absent set of eIFs to initiate translation.^27,28 To investigate the reliance of NS2A RNA on eIFs, we used in vitro transcripts that encode luciferase downstream of the cap-dependent SV40 promoter or viral IRESs. Compared with the control group, NS2A exhibited the capacity to suppress translation in different systems (Supplemental Figure S2D, http://links.lww.com/HEP/I690). As the intergenic cricket paralysis virus IRES does not require any eIFs for translation initiation, we speculated that NS2A RNA inhibits translation by simultaneously targeting both translation initiation and elongation.

Taken together, we found that NS2A RNA occupies the cellular translation machinery to disrupt mRNA scanning and block the translation of overexpressed proteins.

NS2A suppresses the inflammatory response at the translational level

In the innate immune cells, the transcription factor NF-κB is activated in response to external stimuli or pathogens, leading to the production of proinflammatory factors and chemokines. Myeloid differentiation primary response protein 88 (MyD88) and Toll-interleukin-1 receptor (TIR)‍-domain-containing adaptor-inducing IFN-‍β (TRIF) serve as key downstream adaptors of Toll-like receptors signaling pathways to activate NF-κB.²⁹ MyD88, Toll-interleukin-1 receptor (TIR)‍-domain-containing adaptor-inducing IFN-‍β (TRIF), and the receptor-interacting kinase 1 could activate the NF-κB-responsive promoter in the luciferase reporter system. Overexpression of the NS2A significantly inhibited the activation of the NF-κB promoter induced by these signals (Figure 3A). Additionally, the mRNA levels of the NF-κB downstream inflammatory genes TNF-α and CXCL10 were also reduced in the HEK293T cells overexpressing NS2A and proinflammatory signals (Figure 3B). Furthermore, we tested the protein levels of these overexpressed signals and found that they were markedly inhibited in the cells overexpressing NS2A (Figure 3C). These results suggest that NS2A suppresses the activity of NF-κB by restricting the production of excessive upstream signals.

FIGURE 3.

NS2A suppresses inflammatory response and proinflammatory factor translation in a dose-dependent manner. (A and B) HEK293T cells were co-transfected with luciferase reporter plasmid of NF-κB promoter, plasmid of NS2A or EV, and plasmid encoding stimulating factors MyD88/TRIF/RIP1 for 24 hours to detect luciferase activities (A) and TNFα and CXCL10 mRNA levels (B). (C) Immunoblot analysis for overexpressed stimulating factors MyD88/TRIF/RIP1 with or without co-transfection with NS2A plasmid, β-actin was shown as a loading control. (D) The construction strategy of Firefly and Renilla luciferase reporters using 5′ and 3′ UTR of indicated proinflammatory factors. (E) Analysis of the relative luciferase activities in HEK293T cells transfected with Firefly or Renilla reporters, along with NS4B or NS2A plasmid. (F) The luciferase activity percentage was determined in HEK293T cells through co-transfection with reporter plasmids in different ratios together with NS2A or NS4B plasmids, using NS4B groups as the reference point at 100%. Data of (A, B) and (E) are shown as the mean ± SD from 3 independent experiments. *p<0.05, **p<0.01, and ***p<0.001 by unpaired Student's t-test. Data of (C) are representative results from 3 independent experiments. Abbreviations: EV, empty vector; GFP, green fluorescent protein; HA, hemagglutinin; LUC, luciferase; MCP-1, monocyte chemoattractant protein-1; MyD88, myeloid differentiation primary response protein 88; RIP1, receptor-interacting kinase 1; TRIF, Toll-interleukin-1 receptor (TIR)domain-containing adaptor-inducing IFN-β; UTR, untranslated region.

The eukaryotic 5′ UTR and 3′ UTR play vital roles in translation initiation and regulation of protein expression. To investigate the effect of NS2A on the translational control of proinflammatory factors, we constructed translational reporter systems containing specific 5′ UTR and 3′ UTR, driven by the SV40 promoter (Figure 3D). In contrast to the NS4B group, NS2A notably inhibited both Firefly and Renilla luciferase activities regulated by the UTRs of IL-6, monocyte chemoattractant protein-1, and TNF-α, which demonstrated that NS2A targets the translation process of proinflammatory cytokines (Figure 3E). We also assessed the inhibitory capability of NS2A on different levels of luciferase reporters and found that NS2A exhibited increased inhibitory potency with a higher proportion of specific luciferase reporters (Figure 3F).

Together, these results suggest that NS2A potentially controls the highly induced inflammatory response at the translational level.

NS2A alleviates proinflammatory activity in the activated primary macrophages

To further explore the role of NS2A in the elevated inflammatory response in primary macrophages and in vivo, we generated NS2A knock-in mice that specifically targeted myeloid cells (NS2A^mKI) using the CRISPR-Cas9 system. In this model, NS2A was integrated into the intron of the Rosa26 gene and expressed under the control of the CAG promoter (Supplemental Figure S3A, http://links.lww.com/HEP/I690). Genotyping results confirmed NS2A expression in myeloid cells upon deletion of the stop sequence using Cre recombinase (Supplemental Figure S3B, http://links.lww.com/HEP/I690), and the offspring of heterozygous NS2A myeloid knock-in mice followed Mendel law of inheritance (Supplemental Figure S3C, http://links.lww.com/HEP/I690). The absolute copy number of NS2A RNA from NS2A^mKI mice was measured to assess the high abundance of NS2A in primary macrophages (Supplemental Figure S3D, http://links.lww.com/HEP/I690). Weight measurements of NS2A myeloid and whole-body (E2A-Cre) knock-in mice, as well as matched control mice from 4 to 20 weeks of age, indicated that NS2A knock-in did not affect normal mouse growth and development (Supplemental Figure S3E, http://links.lww.com/HEP/I690). Next, we conducted a global RNA-seq analysis using bone marrow-derived macrophages from knock-in and control mice. Gene Set Enrichment Analysis results indicated that there were no significant differences in the expression of housekeeping genes, macrophage activation, and cytokine production (Supplemental Figure S4A, http://links.lww.com/HEP/I690). Therefore, the NS2A myeloid knock-in mouse model is suitable for studying the role of NS2A in vivo.

Upon stimulation with LPS to induce an inflammatory response, a reduction of secretory proinflammatory factors was observed in NS2A knock-in cells, suggesting the inhibitory role of NS2A in primary macrophages (Figure 4A). Moreover, we checked the mRNA levels of Il6, Tnfα, Mcp1, and the activity of inflammatory signaling and found that NS2A did not affect the relative upstream pathway at early time points (Figure 4B and Supplemental Figure S4B, http://links.lww.com/HEP/I690). APAP-induced acute hepatitis would recruit immune cells to the inflammatory sites. Liver tissue suspensions from APAP model mice were prepared to test their ability to recruit NS2A^mKI and control macrophages through the release of chemokines or cytokines. NS2A^mKI macrophages exhibited a reduction of inflammatory responses and cell migration triggered by APAP (Figure 4C). Thioglycolate-induced peritonitis also resulted in the decreased recruitment of macrophages and neutrophils in NS2A^mKI mice (Figure 4D). Furthermore, macrophage proliferation was diminished due to the introduction of NS2A (Figure 4E). The SUnSET assay conducted in thioglycolate-treated mice indicated downregulation of translation efficiency in NS2A^mKI mice, as evidenced by the incorporation of puromycin in vivo (Figure 4F). This effect might be attributed to the intrinsic interaction between eIF2α and NS2A RNA, as confirmed by RIP-qPCR in peritoneal macrophages (Figure 4G). Additionally, ISRIB treatment partially reversed the suppressive effect of NS2A on proinflammatory cytokine production, which is consistent with the effect of the eIF2/eIF2B inhibitory complex on the expression of exogenous GFP protein, indicating the involvement of eIF2α-mediated translation initiation (Supplemental Figure S4C, http://links.lww.com/HEP/I690).

FIGURE 4.

NS2A inhibits the proinflammatory activities of primary macrophages. (A) Peritoneal macrophage isolated from WT and NS2A knock-in (NS2A^mKI) mice were stimulated with LPS (100 ng/mL) for indicated time points. NS2A RNA level was detected by RT-qPCR, and proinflammatory factors of IL-6, TNF-α, and MCP-1 released into cell supernatant were analyzed by ELISA. (B) RT-qPCR analysis of Il6, Tnfα, and Mcp1 mRNA levels in LPS-stimulated peritoneal macrophages from WT and NS2A^mKI mice. (C) Transwell assay was performed with liver suspension from APAP-treated mice to recruit WT and NS2A^mKI bone marrow-derived macrophages, followed by staining with 1% crystal violet. (D) Flow cytometry analysis for cell numbers of macrophage and neutrophil from the peritoneal cavity of thioglycolate-treated WT and NS2A^mKI mice. (E) Proliferation of bone marrow-derived macrophages from WT and NS2A^mKI mice with or without LPS stimulation was evaluated by Ki-67 staining for flow cytometry analysis. (F) SUnSET assay for translation efficiency in thioglycolate-recruited peritoneal cells from WT and NS2A^mKI mice. (G) RIP-qPCR analysis for the interaction of eIF2α with NS2A RNA in NS2A^mKI macrophage with or without LPS stimulation for 4 hours. Data of (A, B), (D), (E, right), and (G) are shown as the mean ± SD from 3 independent experiments. *p<0.05, **p<0.01, and ***p<0.001 by unpaired Student's t-test. Data of (C) are representative images from 3 independent experiments, scale bar, 100 μm. Abbreviations: APAP, acetaminophen; LPS, lipopolysaccharide; MCP-1, monocyte chemoattractant protein-1; RT-qPCR, real-time quantitative PCR; RIP-qPCR, RNA immunoprecipitation-quantitative PCR; WT, wild-type.

Taken together, these results suggest that NS2A inhibits the inflammatory responses of primary macrophages and chemotaxis of immune cells, which is possibly due to the limited translation efficiency.

NS2A protects mice from acute liver inflammatory injuries

To explore the function of NS2A in vivo, an acute liver inflammation model was employed utilizing both wild-type (WT) and NS2A^mKI mice. APAP overdose results in excessive glutathione consumption in the liver, leading to mitochondrial oxidative stress and liver cell necrosis, which triggers an acute inflammatory response.² APAP was administered intraperitoneally to WT and NS2A^mKI mice for 24 hours (Figure 5A). NS2A^mKI mice exhibited attenuated liver injury, as indicated by lower levels of serum ALT and AST (Figure 5B). In addition, analysis of circulating proinflammatory cytokines in the serum revealed significantly lower levels of IL-6, TNF-α, monocyte chemoattractant protein-1, and IL-1β in NS2A^mKI mice than in WT control mice (Figure 5C). Histopathological examination of the liver showed that NS2A^mKI mice displayed fewer infiltrated inflammatory cells, less liver necrosis, and sinusoidal congestion, as well as lower Suzuki’s histological scores (Figure 5D). Immunohistochemical staining for myeloperoxidase release, which indicated the infiltration of neutrophil cells in the liver, and flow cytometry analysis results demonstrated that NS2A knock-in restrained hepatic neutrophil infiltration and protected mice from severe liver damage induced by APAP (Figure 5E, F).

FIGURE 5.

NS2A protects mice from liver inflammatory injuries induced by APAP. (A) WT (n=6) and NS2A^mKI (n=6) mice were injected with APAP i.p. with a dose of 300 mg/kg for 24 hours, and the same volume of saline was injected into the control group (n=3). (B and C) The serum from WT and NS2A^mKI mice in modeling and control groups were collected to detect the enzymatic activities of ALT and AST (B) and the number of proinflammatory factors IL-6, TNF-α, MCP-1, and IL-1β by ELISA (C). (D) Liver histopathological sections from WT and NS2A^mKI mice in APAP modeling and control groups were analyzed by H&E staining, and the Suzuki's tissue score in the modeling group was analyzed. The white arrows represent the vacuolization of hepatocytes, and the black triangles represent congestion in the liver sinusoids. Scale bar, “—” represents 400 μm, “----” represents 100 μm. (E) Liver histopathological sections from WT and NS2A^mKI mice were analyzed by MPO staining and the number of MPO-positive cells in the modeling group was counted under high magnification (HPF, ×400). Scale bar, “---” represents 400 μm, “----” represents 200 μm. (F) Liver nonparenchymal cells were isolated to analyze the proportion of neutrophils in CD45-positive cells by flow analysis. Data are shown as the mean ± SD and representative of 3 independent experiments, and one dot represents a mouse, *p <0.05, **p<0.01, and ***p<0.001 by unpaired Student's t-test. Abbreviations: APAP, acetaminophen; H&E, hematoxylin and eosin; HPF, high-power field; MCP-1, monocyte chemoattractant protein-1; MPO, myeloperoxidase; WT, wild-type.

In contrast to hepatocyte necrosis induced by APAP, the combination of LPS and D-GalN led to acute liver failure resulting from TNF-mediated hepatocyte apoptosis in a septic shock model.³⁰ We established a mouse model using LPS/D-GalN and observed that NS2A^mKI mice also alleviated liver damage, as indicated by lower levels of ALT and AST (Supplemental Figure S5A, B, http://links.lww.com/HEP/I690). The levels of circulating inflammatory factors IL-6, TNF-α, and monocyte chemoattractant protein-1 were repressed in NS2A^mKI mice, and the histopathological analysis and myeloperoxidase staining results illustrated the role of NS2A in inhibiting inflammation response (Supplemental Figure S5C–E, http://links.lww.com/HEP/I690). Furthermore, the TdT-mediated dUTP Nick-End Labeling assay was employed to assess hepatocyte apoptosis, and NS2A^mKI mice demonstrated a lower number of apoptotic hepatocytes, suggesting that NS2A alleviated acute liver failure induced by LPS/D-GalN (Supplemental Figure S5F, http://links.lww.com/HEP/I690).

To explore the specificity of NS2A toward macrophages in NS2A^mKI mice, we conducted a T-cell-mediated hepatitis model induced by concanavalin A and assessed the extent of liver injury and inflammation. These results indicated that NS2A^mKI mice did not influence the expression of proinflammatory factors in T cells or the progression of T-cell-mediated hepatitis (Supplemental Figure S6, http://links.lww.com/HEP/I690).

Taken together, NS2A inhibits inflammatory responses in the context of acute liver inflammation and confers protection against liver injury in mice.

NS2A alleviates fulminant hepatitis during viral infection

The coronavirus MHV causes severe liver damage and fulminant viral hepatitis (FH), characterized by extensive hepatocyte necrosis.³¹ Previous studies have indicated that FH progression is linked to inflammatory responses and immune cell activation.³² Here, we investigated the potential role of NS2A in FH by infecting NS2A^mKI mice and WT littermates with MHV-A59 (Figure 6A). The levels of ALT, AST, and the released proinflammatory cytokines were dramatically decreased in the virus-infected NS2A^mKI mice (Figure 6B, C). Noticeably, the expression of antiviral genes, such as IFN-β and MX1, as well as the MHV genes, was suppressed in the livers of NS2A^mKI mice (Figure 6D). Histological analysis using hematoxylin and eosin, myeloperoxidase, and TdT-mediated dUTP nick-end labeling staining showed reduced liver necrosis, fibrosis, neutrophil infiltration, and apoptosis in NS2A^mKI mice (Figure 6E–G). Therefore, these results demonstrate that NS2A attenuates MHV infection-induced FH and protects mice from further viral infection.

FIGURE 6.

NS2A alleviates MHV-induced fulminant hepatitis. (A) 5×10⁵ pfu/mouse of MHV-A59 were injected i.p. into WT and NS2A^mKI mice for 72 hours. (B and C) The serum from WT and NS2A^mKI mice in the MHV infection group and control group (saline) was taken to detect the enzymatic activities of ALT and AST (B) and levels of proinflammatory factors IL-6, TNF-α, and MCP-1 (C). (D) RT-qPCR analysis of the mRNA levels of Ifnb1, Mx1, MHV, and NS2A in liver tissues. (E and F) The liver histopathological sections were prepared for H&E staining (E) and MPO staining (F), and the number of MPO-positive cells in the infection group was counted under high magnification (HPF, ×400), scale bar, “—” represents 400 μm, “----” represents 200 μm. (G) TUNEL staining of liver histopathological sections and the number of TUNEL-positive cells per field in the infection group were counted by scale bar, 100 μm. Data are shown as the mean ± SD and representative of 3 independent experiments, and one dot represents a mouse, *p <0.05, **p<0.01, and ***p<0.001 by unpaired Student's t-test. Abbreviations: H&E, hematoxylin and eosin; HPF, high-power field; MHV, murine hepatitis virus; MCP-1, monocyte chemoattractant protein-1; MPO, myeloperoxidase; RT-qPCR, real-time quantitative PCR; TUNEL, TdT-mediated dUTP nick-end labeling; WT, wild-type.

NS2A attenuates CRS in the GVHD mouse model

To explore the potential application of NS2A in treating CRS by utilizing its ability to suppress the production of proinflammatory factors, we conducted a study in which bone marrow cells from NS2A^mKI mice and WT littermates were transplanted into irradiated CD45.1 mice and subsequently challenged with APAP after 8 weeks. The results of serum ALT and AST levels and pathological staining indicated that the bone marrow harboring the NS2A cassette could mitigate liver inflammatory injuries (Figure 7A, B).

FIGURE 7.

NS2A attenuates cytokine release syndrome in the graft-versus-host disease mouse model. (A and B) Bone marrow cells isolated from WT and NS2A^mKI mice were injected into the tail vein of irritated CD45.1 mice and challenged with APAP after 8 weeks to detect the ALT and AST activities (A) and liver pathology (B). (C) Survival analysis of BALB/c recipients that received BM cells from WT and NS2A^mKI C57BL/6 mice with or without CD3⁺ T cells. (D) Proinflammatory factors were analyzed by ELSA in the intestines of recipients who received BM and CD3⁺ T cells. (E) Histology analysis of small intestine, colon, and liver from recipients received BM with or without CD3⁺ T cells by H&E staining. Data (A) and (D) are shown as the mean ± SD and representative of 3 independent experiments, and one dot represents a mouse, *p <0.05, **p<0.01, and ***p<0.001 by unpaired Student’s t-test. Data (C) are evaluated by the Log-rank (Mantel-Cox) test, *p<0.05. Data of (B) and (E) are representative images with scale bars representing 100 μm and 50 μm, respectively. Abbreviations: APAP, acetaminophen; BALB, Bagg's albino; BM, bone marrow; BMT, bone marrow transplant; H&E, hematoxylin and eosin; MCP-1, monocyte chemoattractant protein-1; WT, wild type.

GVHD-triggered CRS is characterized by increased levels of proinflammatory cytokines. To establish the GVHD model, we transplanted bone marrow and splenic CD3⁺ T cells from WT and NS2A^mKI mice into irradiated Bagg's albino/c mice. The WT group experienced significantly higher mortality due to GVHD than the NS2A knock-in group (Figure 7C).

Our findings demonstrated that NS2A^mKI mice exhibited lower levels of proinflammatory factors expressed in the intestines, the primary organ affected in the GVHD model (Figure 7D). Additionally, organs such as the liver, colon, and small intestine also showed less severe damage than the WT group (Figure 7E). Therefore, NS2A exhibits an efficient capability to control cytokine storm and inflammation-induced organ damage.

DISCUSSION

In this study, we demonstrate that NS2A potently suppresses the synthesis of exogenous and substantially induced proteins at the translational levels. NS2A RNA molecules can directly occupy the translation machinery independently of the NS2A protein. In various acute inflammation models, we observed a significant reduction in tissue damage and the inflammatory response in mice with NS2A conditional knock-in in myeloid cells.

NS2A is crucial for the replication and pathogenicity of ZIKV as a nonstructured protein.^17,33 In this study, we explored the beneficial functions of NS2A by harnessing NS2A RNA to inhibit hyperactivated inflammation. Manipulation of viruses or viral elements is a well-established approach in translational and clinical research. ZIKV, owing to its neurotropic properties, has been exploited to develop oncolytic viruses for glioblastoma treatment.^34,35 Our study further expands the therapeutic possibilities of ZIKV genes or their derivatives.

We have demonstrated that NS2A effectively alleviates inflammatory injuries in murine models of hepatitis and GVHD. Given the extensive anti-inflammatory properties of NS2A, NS2A may have potential applications in targeting additional disease models involving increased proinflammatory cytokines beyond hepatitis and GVHD. Compared with conventional chemical and antibody-based anti-inflammatory drugs, NS2A RNA functions as a nucleic acid drug with reduced acquisition costs and enhanced specificity. Consequently, our future studies will concentrate on developing NS2A RNA delivery systems, such as lipid nanoparticles and adeno-associated viruses,^36,37 and exploring their efficacy in treating a broader spectrum of acute inflammatory disorders.

Previous studies have shown that NS2A is involved in the degradation of the adherens junction complex between neural stem cells and the nucleocytoplasmic transport protein KPNA2 at the post-translational level.^17,18 Additionally, NS2A has been shown to decrease the expression of exogenous proteins RIG-I, MDA5, and signaling molecules such as TBK1, thereby inhibiting the transcriptional activity of IFN-I and NF-κB and the corresponding immune response.^19,20 However, our study indicated that NS2A also plays an important role in regulating the translation of exogenous or excessively induced proteins while having no impact on the normal expression of intracellular endogenous proteins in the resting state. These results suggest that NS2A inhibits the translation process unless the cells require a substantial amount of translation machinery.

During the initial stage of protein translation, a variety of eukaryotic translation initiation factors, including eIF2α, facilitate the assembly of the 40S and 60S ribosomal subunits onto the mRNA and position the ribosome and initiating methionine tRNA in proximity.³⁸ Upon invasion of the virus into a host cell, the cell initiates mechanisms such as protein kinase R-eIF2α pathways to hinder the early stage of protein translation, thereby restricting the synthesis of viral proteins and the generation of viral particles.³⁹ Nevertheless, a variety of viral RNA elements can modulate the translation process within host cells, either by assisting the virus in completing the expression of viral proteins or hindering the translation of host proteins to evade the immune response.⁴⁰ For instance, certain viruses produce small RNAs, including Adenovirus "virus-associated" RNA I, Epstein-Barr virus encoded RNA, and trans-activation response element RNA from human immunodeficiency virus type 1, to inhibit activation of the protein kinase R-eIF2α axis and facilitate the normal synthesis of viral proteins.^41–43 Additionally, the stable downstream loop (DLP) structure found in the coding region of certain alphavirus subgenomic mRNAs can bind to the 40S ribosome subunit and the ES6S region of the 18S rRNA, stalling the ribosome and enhancing the translation of viral proteins in an eIF2-independent manner.⁴⁴ Moreover, the NS1 of SARS-CoV-2 blocks the mRNA-binding channel of host cells by interacting with components of the 40S ribosome subunit, which dampens global host protein translation and Interferon-β production.^45,46 In this study it has shown that the NS2A gene of ZIKV, which is known to encode proteins, can act as a noncoding RNA before translation. This noncoding RNA has been shown to disrupt the translation process within mammalian cells, contrary to the conventional understanding that RNA with protein-coding potential exclusively operates at the protein level, as per the central dogma.

The investigation into the mechanism by which NS2A interacts with the host translation machinery via RNA molecules revealed that in vitro transcribed NS2A RNA can bind several initiation factors and elongation factors associated with translation. Despite the inability of in vitro transcribed NS2A RNA to encode proteins, its interaction with translation factors may competitively impede the translation process of mRNA molecules within host cells. Our investigation has demonstrated the strongest binding affinity between eIF2α and NS2A RNA, as evidenced by RNA pulldown and RIP-qPCR assays. However, it remains to be determined whether NS2A functions through other translation factors based on our RNA pulldown and mass spectrometry results. Additionally, in the luciferase reporter system initiated by viral IRES elements, which are independent of eIF2α and other specific translation initiation factors, NS2A is capable of inhibiting luciferase activity. This suggests a potential mechanism by which NS2A suppresses the translation process independently of eIF2α.

ISRIB is recognized as an inhibitor of the integrated stress response that involves the phosphorylation of eIF2α and the blocking of GEF activity of eIF2B, which is induced by various stresses.⁴⁷ However, NS2A overexpression did not affect the phosphorylation of p-eIF2α, as well as XBP1 splicing or the mRNA levels of BIP and CHOP, the markers of the integrated stress response, especially endoplasmic reticulum stress. Therefore, it was postulated that NS2A overexpression does not induce a typical integrated stress response in vivo and in vitro. Given the inhibitory effect of NS2A RNA on the in vitro translation system, it is plausible to infer that NS2A works by directly targeting the translation machinery.

Our study demonstrates that NS2A hinders protein translation in the form of RNA at the cellular level. In primary macrophages, NS2A knock-in leads to a significant decrease in the expression of proinflammatory cytokines but does not affect transcription and signaling activation during the early stages. Furthermore, translation reporters of proinflammatory cytokines mediated by the 5′ and 3′ UTR are inhibited by NS2A, indicating its inhibitory effect on the translation process. However, the targeting specificity of NS2A during inflammatory responses in primary macrophages and in vivo inflammation models remains to be determined.

In conclusion, our study expands the understanding of the impact of viral protein-encoding genes as noncoding RNA molecules on the host translation process. We have also validated the potential capacity of the RNA cassette to control inflammatory responses. These findings add to the knowledge of translational regulation in the context of inflammation and provide strategies for limiting CRS in clinical situations.

Supplementary Material

hep-82-110-s001.pdf ^{(2.1MB, pdf)}

hep-82-110-s002.xlsx ^{(67.5KB, xlsx)}