β-hydroxybutyrylation and O-GlcNAc modifications of STAT1 modulate antiviral defense in aging

Abstract

Aging changes the protein activity status to affect the body’s functions. However, how aging regulates protein posttranslational modifications (PTMs) to modulate the antiviral defense ability of the body remains unclear. Here, we found that aging promotes STAT1 β-hydroxybutyrylation (Kbhb) at Lys592, which inhibits the interaction between STAT1 and type-I interferon (IFN-I) receptor 2 (IFNAR2), thereby attenuating IFN-I-mediated antiviral defense activity. Additionally, we discovered that a small molecule from a plant source, hydroxy camptothecine, can effectively reduce the level of STAT1 Kbhb, thus increasing antiviral defense ability in vivo. Further studies revealed that STAT1 O-GlcNAc modifications at Thr699 block CBP-induced STAT1 Kbhb. Importantly, fructose can improve IFN-I antiviral defense activity by orchestrating STAT1 O-GlcNAc and Kbhb modifications. This study reveals the significance of the switch between STAT1 Kbhb and O-GlcNAc modifications in regulating IFN-I antiviral immunity during aging and provides potential strategies to improve the body’s antiviral defense ability in elderly individuals.

Introduction

Understanding and mitigating the process of aging is an enormous challenge. A key initial step in the field of aging research was the observation in 1939 that restriction of caloric intake in mice and rats increased lifespan [1]. Genetic and environmental factors, as well as metabolic changes, play crucial roles in regulating the aging process [2]. In addition, the aging process affects numerous physiological functions, such as dysregulated nutrient sensing, altered intercellular communication, and epigenetic alterations [3]. Notably, aging is closely related to certain specific immune responses. Aging results in decreased production of lymphocytes, which are basal components of the adaptive immune system [4, 5]. In addition, age-related changes in different types of immune cells, including macrophages, neutrophils, natural killer (NK) cells and dendritic cells (DCs), have been reported [6–9]. However, how aging impairs antiviral innate immune defense remains poorly understood.

Protein posttranslational modifications (PTMs) modulate the activity status of cellular proteins, thus regulating various physiological and pathological processes. Recent studies have demonstrated a strong correlation between PTMs and age-related diseases, including neurodegenerative diseases (such as Alzheimer’s disease), metabolic diseases (such as type 2 diabetes), and cardiovascular diseases (such as coronary artery disease) [10]. For example, the increase in the phosphorylation and ubiquitination levels of tau protein has been well documented in Alzheimer’s disease [11]. Moreover, SUMOylation has been implicated in the processes of aging and neurodegeneration, suggesting that targeting the SUMO pathway may hold promise as a therapeutic strategy for neurodegenerative disorders [12]. In addition, arginine methylation decreases in stem cells during aging [13]. Recently, acylation modifications have been increasingly demonstrated to be critical for cellular biological activities. However, how specific acylation modifications are involved in aging and aging-related pathological processes is largely unknown.

Acylation modifications have been reported to play roles in various biological processes, including gene transcription, signal transduction, and metabolism [14]. To date, Zhao et al. identified a variety of acylation modifications, such as lactylation, succinylation, and β-hydroxybutyrylation (Kbhb), via mass spectrometry [15–17]. These different acylation modifications serve distinct functions: p53 lactylation promotes tumor development [18]; SIRT5-regulated succinylation is a functional modification with the potential to impact mitochondrial metabolism [19]; and histone Kbhb affects metabolic gene expression [16]. Recent studies have revealed that SIRT3 serves as an eraser of histone (H3K9) Kbhb [20]. In addition, p300 can catalyze histone Kbhb, while histone deacetylase 1 (HDAC1) and HDAC2 remove Kbhb enzymatically [21]. Despite the advances in research on nuclear histone Kbhb, how the Kbhb modifications of other cellular proteins regulate their biological functions needs to be further explored.

Here, we revealed a significant increase in STAT1 Kbhb modifications at Lys592, which inhibits STAT1 activation and IFN-I signaling, in aging mice. By screening a library of small molecule compounds, we discovered that hydroxy camptothecine (10-HCPT) can effectively inhibit STAT1 Kbhb to increase the IFN antiviral activity of aging mice. Furthermore, the region of STAT1 that binds to CBP is modified by O-GlcNAc, which inhibits CBP-STAT1 binding and therefore restricts CBP-induced STAT1 Kbhb. Moreover, we revealed that fructose can orchestrate STAT1 O-GlcNAc and Kbhb modifications and ultimately improve IFN-I antiviral immune defense in aging mice.

Results

Aging promotes the Kbhb modification of STAT1

To explore the effects of acylation modifications in aging on antiviral immune defense in the body, tissue samples from young and aging mice were used to determine the levels of different acylation modifications via an acylation modification detection antibody kit. Compared with the other acylation modifications analyzed in this study, only the Kbhb modification had significantly increased levels in the spleen tissues of aging mice (Fig. 1A and Supplementary Fig. 1a). Furthermore, the increase in Kbhb modification during aging was confirmed in the kidney and liver tissues of aging mice compared with those of young mice (Fig. 1B, C). Next, we used a specific anti-Kbhb antibody to pull down the potential proteins undergoing Kbhb modifications from the tissues of aging mice and then utilized mass spectrometry to identify these proteins. Among the top ten proteins identified, STAT1, a central signaling protein in the IFN-I signaling pathway, was found to potentially harbor Kbhb modifications (Fig. 1D). Moreover, STAT2 could also harbor Kbhb modifications (Supplementary Fig. 1b). Further analysis by immunoblotting confirmed that, compared with those in young mice, STAT1 proteins in aging mice presented higher levels of Kbhb modifications (Fig. 1E), whereas the level of STAT2 Kbhb modifications remained unchanged (Fig. 1F). Collectively, these observations demonstrated that aging promotes STAT1 Kbhb modifications.

Fig. 1.

Aging promotes the Kbhb modification of STAT1. A Immunoblotting (IB) analysis of acylation modifications, including pan-Khib, Kglu, Kmal, Kpr, Kbu, Kcr, Ksucc, Kla, Kbhb, Kbz and Ace, in spleen tissues from young or aging mice (n = 3 for each group) was performed via an Acylation Antibody Kit. (Ace acetyllysine, Kglu glutarylation, Kma malonylation, Kpr propionylation, Kbu butyrylation, Kcr crotonylation, Ksucc succinylation, Kla lactylation, Kbhb β-hydroxybutyrylation, Kbz benzoylation, Khib 2-hydroxyisobutyrylation). B, C Western blot analysis of pan-Kbhb levels in kidney (B) and liver (C) tissues from young or aging mice. The relative levels of pan-Kbhb in the aging group are shown as the fold change relative to those in the young group after quantification of the immunoblotting signals via densitometric analyses. D Whole-cell lysates (WCLs) extracted from the spleen of aging mice were subjected to immunoprecipitation (IP) using a pan-Kbhb antibody. Mass spectrometry was used to analyze the potential proteins with Kbhb modifications. The “b” and “y” symbols indicate the MS-identified fragment ions from the N-termini (b) and C-termini (y) of the peptides after fragmentation. The presented diagrams provide the MS/MS spectra of the identified STAT1 peptides. E, F IP-IB analysis of STAT1 (E) or STAT2 (F) Kbhb in spleen tissues from young or aging mice (n = 3). The data are representative of three independent experiments (A–C, E, F)

STAT1 harbors Kbhb modifications at Lys592 regulated by CBP and SIRT3

Given that STAT1 Kbhb modifications are upregulated during aging, we further identified the key enzymes responsible for writing or erasing the Kbhb modification of STAT1. The results showed that CBP but not p300 upregulated STAT1 Kbhb modifications (Fig. 2A). Consistently, the specific CBP inhibitor A-458 dramatically reduced the Kbhb modification of STAT1 (Supplementary Fig. 2a). In addition, SIRT3, but not SIRT1 or SIRT2, can downregulate STAT1 Kbhb modifications (Fig. 2B and Supplementary Fig. 2b). Furthermore, the overexpression of SIRT3 reduced STAT1 Kbhb in a dose-dependent manner (Fig. 2C and Supplementary Fig. 2c). Conversely, SIRT3 deficiency markedly increased STAT1 Kbhb levels (Fig. 2D). Moreover, the specific SIRT3 inhibitor 3-TYP markedly upregulated STAT1 Kbhb modifications in different types of cells, including HEK293T cells and macrophages (Fig. 2E and Supplementary Fig. 2d), suggesting that the enzyme activity of SIRT3 is required for the regulation of STAT1 Kbhb.

Fig. 2.

STAT1 harbors Kbhb modifications at Lys592 regulated by CBP and SIRT3. A IP-IB analysis of STAT1 Kbhb in HEK293T cells cotransfected with control vectors or HA-CBP or HA-p300, together with or without Myc-STAT1. B IP-IB analysis of STAT1 Kbhb in HEK293T cells transfected with control vectors (CON) or Flag-SIRT1, Flag-SIRT2 or Flag-SIRT3. C IP-IB analysis of STAT1 Kbhb in HEK293T cells transfected with increasing amounts of Flag-SIRT3. D IP-IB analysis of STAT1 Kbhb in Sirt3^+/+ (WT) or Sirt3^−/− (KO) HEK293T cells transfected with Myc-STAT1. E IP-IB analysis of STAT1 Kbhb in HEK293T cells transfected with Myc-STAT1 and then treated with 3-TYP (50 µM, 24 h). F HEK293T cells were transfected with Myc-STAT1. Myc-STAT1 proteins were immunoprecipitated with an anti-Myc antibody, and then mass spectrometry was used to analyze the Kbhb modification site. G IP-IB analysis of STAT1 Kbhb in HEK293T cells transfected with Myc-STAT1 (WT or K592R). H IP-IB analysis of STAT1 Kbhb in Stat1^−/− HEK293T cells cotransfected with Myc-STAT1 (WT or K592R) and HA-CBP. I IP-IB analysis of STAT1 Kbhb in Stat1^−/− HEK293T cells transfected with Myc-STAT1 (WT or K592R) and then treated with 3-TYP (50 µM, 24 h). The data are representative of three independent experiments (A–E, G–I). N.S. not significant, **p < 0.01, ***p < 0.001 (two-tailed unpaired Student’s t test)

Next, we investigated which residue of STAT1 harbors the Kbhb modification. Through mass spectrometry analysis, we identified Lys592 (K592) as the residue of STAT1 that underwent Kbhb modification (Fig. 2F). Consistently, the STAT1 K592 mutation dramatically reduced the Kbhb modification of STAT1 (Fig. 2G and Supplementary Fig. 2e). Moreover, mutating STAT1 K592 abolished both CBP-mediated (Figs. 2H) and 3-TYP-mediated (Fig. 2I) upregulation of STAT1 Kbhb. Consistently, the K592 mutation blocked the decrease in STAT1 Kbhb caused by SIRT3 overexpression (Supplementary Fig. 2f). These results suggest that STAT1 K592 is the major residue of STAT1 whose Kbhb modification is regulated by CBP and SIRT3. Taken together, these findings demonstrated that STAT1 proteins harbor Kbhb modifications at Lys592, which are regulated by CBP and SIRT3.

STAT1 Lys592 Kbhb attenuates STAT1-IFNAR2 binding and inhibits IFN-I signaling

We next explored whether and how STAT1 Kbhb affects STAT1 activation and IFN-I signaling. We found that SIRT3 overexpression promoted IFN-I-induced STAT1 Tyr701 phosphorylation, which is a marker of STAT1 activation (Supplementary Fig. 2g), whereas SIRT3 knockout largely inhibited STAT1 Tyr701 phosphorylation (Supplementary Fig. 2h). Moreover, the IFN-I-induced expression of representative ISGs, including Ifit1, Isg15 and Viperin, was significantly increased by SIRT3 overexpression (Supplementary Fig. 2i). In addition, both the inhibition of SIRT3 by 3-TYP and the knockout of SIRT3 largely inhibited IFN-I-induced ISG expression (Supplementary Fig. 2j, k). Consistent with the above findings, the STAT1 K592 mutation strongly promoted IFN-induced STAT1 activation (Fig. 3A). Moreover, mutating K592 of STAT1 enhanced IFN-I-induced ISG expression (Fig. 3B and Supplementary Fig. 3a) and promoted IFN-I antiviral activity (Fig. 3C and Supplementary Fig. 3b).

Fig. 3.

STAT1 Lys592 Kbhb attenuates STAT1-IFNAR2 binding and inhibits IFN-I signaling. A Western blot analysis of STAT1 Tyr701 phosphorylation (pY701) in Stat1^−/− HEK293T cells transfected with Myc-STAT1 (WT or K592R) and then treated with IFNα (1000 IU/ml) for 30 min. B RT‒qPCR was used to analyze the expression of representative ISGs (Ifit1 and Viperin) in Stat1^−/− HEK293T cells transfected with Myc-STAT1 (WT or K592R) and then stimulated with IFNα (1000 IU/ml) for 4 h. Western blotting was used to detect Myc-STAT1 protein levels. C Fluorescence microscopy of VSV-GFP in Stat1^−/− HEK293T cells transfected with Myc-STAT1 (WT or K592R) and then treated with IFNα (60 IU/ml) for 20 h, followed by infection with VSV-GFP (MOI = 0.5) for 24 h. Scale: 100 µm. D IP-IB analysis of the interaction between Myc-STAT1 (WT or K592R) and Flag-IFNAR1 (left) or HA-IFNAR2 (right) in HEK293T cells. E IP-IB analysis of the interaction of the STAT1 domains with IFNAR2 in HEK293T cells cotransfected with Myc-STAT1 (WT or mutants) and HA-IFNAR2. F In vitro binding assay to analyze the interaction between HA-IFNAR2 and Flag-STAT1 (WT or K592R), which were immunoprecipitated from HEK293T cells transfected with either HA-IFNAR2 or Flag-STAT1 (WT or K592R). G IP-IB analysis of the interaction between STAT1 and IFNAR2 in Sirt3^+/+ or Sirt3^−/− HEK293T cells cotransfected with Myc-STAT1 and HA-IFNAR2. The data are representative of three independent experiments (A, D–G) or are shown as the means and s.d.s of three biological replicates (B). N.S. not significant; ***p < 0.001 (two-tailed unpaired Student’s t test)

To explore the underlying mechanism by which STAT1 Kbhb modifications affect its Tyr701 phosphorylation and activation, an immunoprecipitation experiment was conducted. First, we found that mutation of K592 of STAT1 did not affect the interaction between STAT1 and STAT2 (Supplementary Fig. 3c). However, compared with wild-type (WT) STAT1, the STAT1-K592R mutant exhibited increased interaction with IFNAR2 but not with IFNAR1 (Fig. 3D). Consistently, the binding of STAT1-K592R to endogenous JAK1, which constitutively interacts with IFNAR2, was also upregulated (Supplementary Fig. 3d). We subsequently identified the specific region of STAT1 that interacts with IFNAR2. The results revealed that the amino acid 577--703 region of STAT1 is responsible for IFNAR2 binding (Fig. 3E). We noted that the K592 site at which STAT1 undergoes Kbhb modification is located in this region. To clarify the effect of the STAT1 Kbhb modification on the STAT1-IFNAR2 interaction, an in vitro binding assay was performed. The results revealed stronger binding between IFNAR2 and STAT1-K592R than between IFNAR2 and STAT1-WT (Fig. 3F). Moreover, SIRT3 knockout resulted in elevated levels of STAT1 Kbhb modification and decreased binding of STAT1 to IFNAR2 (Fig. 3G). Taken together, these findings suggest that STAT1 Kbhb at Lys592 inhibits its binding to the IFN-I receptor IFNAR2, thereby impeding STAT1 activation.

Hydroxy camptothecine downregulates STAT1 Kbhb and improves the IFN-I antiviral activity of aging mice

Given that STAT1 Kbhb modifications inhibit IFN-I signaling, we sought to identify potential drugs to lower the levels of STAT1 Kbhb. To this end, a drug library containing 214 small molecules from plant sources, all of which have been approved for clinical treatment, was employed. Interestingly, hydroxy camptothecine (10-HCPT) dramatically reduced the Kbhb levels of STAT1 in cells (Fig. 4A). However, 10-HCPT did not strongly affect the pan-Kbhb levels of total proteins (Supplementary Fig. 3e) or the SIRT3 and CBP protein levels (Supplementary Fig. 3f) in cells. Furthermore, among several small molecule compounds that have the potential to lower the level of STAT1 Kbhb, 10-HCPT had the strongest effect on the upregulation of IFN-I-induced ISG expression (Fig. 4B and Supplementary Fig. 3g, h). Next, we examined the effects of different concentrations of 10-HCPT on STAT1 Kbhb levels. We found that 10-HCPT at concentrations ranging from 1 µM to 10 µM could effectively downregulate STAT1 Kbhb levels (Supplementary Fig. 3i), and the IC50 of 10-HCPT against STAT1 Kbhb was approximately 2.128 µM (Supplementary Fig. 3j). Within this concentration range, 10-HCPT decreased STAT1 Kbhb levels in a dose-dependent manner (Fig. 4C). Consistent with the downregulation of STAT1 Kbhb, 10-HCPT promoted the interaction between STAT1 and IFNAR2 (Supplementary Fig. 3k) and increased IFN-induced Tyr701 phosphorylation (Supplementary Fig. 3l). Moreover, 10-HCPT treatment enhanced the antiviral ability of the cells (Supplementary Fig. 3m).

Fig. 4.

Hydroxy camptothecine decreases the level of STAT1 Kbhb and improves the IFN antiviral activity of aging mice. A drug library containing 214 clinically approved small molecules from plant sources was employed to screen for potential small molecules that can inhibit STAT1 Kbhb in HEK293T cells transfected with Myc-STAT1. IP-IB was used to detect STAT1 Kbhb levels, and the relative levels of STAT1 Kbhb are shown as the fold change relative to those in the control group (vehicle) after quantification by densitometric analysis. B RT‒qPCR analysis of representative ISGs (Ifit1) in HT1080 cells treated with several compounds (#21: bergenin; #27: formononetin; #29: gossypol-acetic acid; #51: taxifolin; #72: berbamine; #110: hydroxy camptothecine) for 24 h and then stimulated with IFNα (1000 IU/ml) for 4 h. C IP-IB analysis of STAT1 Kbhb in HEK293T cells transfected with Myc-STAT1 and then treated with increasing amounts of hydroxy camptothecine (5 or 10 µM) for 24 h. D Diagram of mouse experiments in which aging mice were first administered hydroxy camptothecine (3 mg/kg body weight) for 3 days and then injected with VSV viruses (1 × 10⁸ PFU per gram of mouse body weight) for 24 h. Western blot or RT‒qPCR was used to analyze the levels of STAT1 Kbhb, ISGs and viral RNAs from various tissues. E IP-IB analysis of STAT1 Kbhb in spleen tissues from aging mice (n = 3) treated with hydroxy camptothecine (3 mg/kg body weight) for 3 days. F IP-IB analysis of the interaction between STAT1 and IFNAR2 in spleen tissues from an aging mouse treated as described in (E). G RT‒qPCR analysis of the mRNA levels of representative ISGs (Viperin, Ifit1 and Isg15) in liver tissues from aging mice (n = 6) treated with hydroxy camptothecine (3 mg/kg body weight) for 3 days. H RT‒qPCR analysis of representative ISG (viperin) mRNA in kidney and spleen tissues from aging mice (n = 6) treated as described in (G). I RT‒qPCR analysis of VSV viral RNA in spleen, liver, lung, and kidney tissues from aging mice (n = 6) treated with hydroxy camptothecine (3 mg/kg body weight) for 3 days and then infected with VSV (1 × 10⁸ PFU per gram body weight mouse, i.p.) for 24 h. J IB analysis of VSV-G proteins in the spleen, lung, and kidney tissues of aging mice (n = 6) treated as described in (I). The data are representative of three independent experiments (C, E, F, J) or are shown as the means and s.d.s of three biological replicates (B). The graphs show the means ± s.e.m. for six individual mice (G–I). N.S. not significant, ***p < 0.001 (two-tailed unpaired Student’s t test)

Next, we determined whether 10-HCPT could improve the in vivo IFN-I antiviral immune defense ability of aging mice. To this end, aging mice were administered 10-HCPT for 3 days and subsequently infected with VSV viruses (Fig. 4D). The results showed that 10-HCPT administration greatly decreased STAT1 Kbhb modifications in aging mice (Fig. 4E). In line with the downregulation of STAT1 Kbhb, 10-HCPT administration promoted the interaction between STAT1 and IFNAR2 or JAK1 in the spleen tissues of aging mice (Fig. 4F and Supplementary Fig. 3n). Consistently, 10-HCPT administration increased the expression of representative ISGs in various tissues of aging mice challenged with VSV viruses (Fig. 4G, H and Supplementary Fig. 3o, p). As a consequence, 10-HCPT administration inhibited virus infection, as shown by reduced levels of viral RNAs and virus-encoded proteins in various tissues of aging mice (Fig. 4I, J). Thus, we revealed that hydroxy camptothecine can inhibit STAT1 Kbhb, thus enhancing in vivo IFN-I signaling and the antiviral defense ability of aging mice.

STAT1 O-GlcNAc modification at Thr699 inhibits CBP binding to STAT1

Given that CBP interacts with STAT1 to induce STAT1 Kbhb, we further explored strategies to inhibit CBP binding to STAT1 to reduce STAT1 Kbhb. To this end, we first determined the region of STAT1 that interacts with CBP. The results revealed that the amino acid 658–703 region of STAT1 is responsible for CBP binding (Fig. 5A and Supplementary Fig. 4a). We next hypothesized that certain PTMs in the amino acid 658–703 region of STAT1 could block CBP binding. Thus, we used mass spectrometry to identify potential PTMs in this region. The results revealed three PTMs in this specific region (Fig. 5B), including Lys665 acetylation (Supplementary Fig. 4b), Lys673 ubiquitination (Supplementary Fig. 4c), and Thr699 O-GlcNAc (Fig. 5C and Supplementary Fig. 4d). Furthermore, we demonstrated that among these three residues containing PTMs, mutation of STAT1 Thr699 (T699A) resulted in increased interaction between STAT1 and CBP (Fig. 5D), suggesting that STAT1 Thr699 O-GlcNAc modifications could inhibit CBP binding to STAT1.

Fig. 5.

The domain of STAT1 that interacts with CBP harbors the O-GlcNAc modification at Thr699. A diagram of the interaction between STAT1 and CBP (left) and IP-IB analysis of the interaction of the STAT1 domains with CBP in HEK293T cells cotransfected with HA-CBP and Myc-STAT1 (WT or mutants) (right). B Diagram of three posttranslational modifications of STAT1, including K665 acetylation, K673 ubiquitination and T699 O-GlcNAc. C HEK293T cells were transfected with Myc-STAT1. Myc-STAT1 proteins were immunoprecipitated with an anti-Myc antibody, and then mass spectrometry was used to analyze the O-GlcNAc modification site T699. D IP-IB analysis of the interaction between CBP and STAT1 (WT or mutants) in HEK293T cells cotransfected with HA-CBP and Myc-STAT1 (WT, K665R, K673R or T699A). E IP-IB analysis of the O-GlcNAc level of STAT1 (WT or T699A) in HEK293T cells transfected with Flag-STAT1 (WT or T699A). F, G IP-IB analysis of the O-GlcNAc level of STAT1 (F) or STAT2 (G) in spleen tissues from young or aging mice (n = 3). The data are representative of three independent experiments (A, D–G). N.S. not significant, *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed unpaired Student’s t test)

Consistent with the above findings, mutation of STAT1 Thr699 abolished STAT1 O-GlcNAc modifications (Fig. 5E), whereas mutation of STAT1 Lys592, which was demonstrated to undergo Kbhb modification in our aforementioned studies, did not affect STAT1 O-GlcNAc modifications (Supplementary Fig. 4e). Next, we analyzed in vivo STAT1 O-GlcNAc modifications in the tissues of aging mice. The results revealed that STAT1 O-GlcNAc modifications were attenuated in the tissues of aging mice compared with those of young mice (Fig. 5F). However, the O-GlcNAc levels of both STAT2 and JAK1 remained relatively stable in aging mice (Fig. 5G and Supplementary Fig. 4f). Additionally, the STAT1 Kbhb levels in macrophages from aging mice were greater than those in macrophages from young mice, whereas the STAT1 O-GlcNAc levels in macrophages from aging mice were lower than those in macrophages from young mice (Supplementary Fig. 4g). Taken together, these findings demonstrated that STAT1 O-GlcNAc modifications at the Thr699 residue inhibit CBP binding to STAT1 and that aging reduces the levels of STAT1 O-GlcNAc modifications.

OGT induces STAT1 O-GlcNAc at Thr699 to inhibit STAT1 Kbhb

O-GlcNAc modification has been implicated in various biological functions, such as chromatin remodeling, mitochondrial function, protein stability and transcription [22]. Uridine diphosphate-GlcNAc (UDP-GlcNAc) serves as a key metabolite in the hexosamine biosynthesis pathway (HBP). O-GlcNAc transferase (OGT) utilizes UDP-GlcNAc to attach O-GlcNAc moieties to proteins, whereas O-GlcNAcase (OGA) removes these modifications [23]. We further found that OGT overexpression increased the level of STAT1 O-GlcNAc in a dose-dependent manner (Fig. 6A). In contrast, OGT knockdown inhibited STAT1 O-GlcNAc (Fig. 6B). In addition, mutation of the O-GlcNAc transferase active site (Lys908) of OGT inhibited OGT-induced STAT1 O-GlcNAc (Fig. 6C). OGT-induced STAT1 O-GlcNAc was abolished when STAT1 Thr699 was mutated (Fig. 6D). These findings demonstrated that OGT induces STAT1 O-GlcNAc modifications at Thr699.

Fig. 6.

STAT1 O-GlcNAc at Thr699 by OGT inhibits CBP binding and STAT1 Kbhb. A IP-IB analysis of the O-GlcNAc level of STAT1 in HEK293T cells transfected with increasing amounts of Myc-OGT. B IP-IB analysis of the O-GlcNAc level of STAT1 in HEK293T cells transfected with either control shRNAs (−) or two shRNAs against OGT (shOGT: #1, #2). C IP-IB analysis of the O-GlcNAc level of STAT1 in HEK293T cells transfected with OGT wild-type (WT) or its enzyme inactive mutant (K908A). D IP-IB analysis of the O-GlcNAc level of STAT1 in Stat1^−/− HEK293T cells cotransfected with Flag-STAT1 (WT or T699A) and Myc-OGT. E IP-IB analysis of Flag-STAT1 Kbhb in HEK293T cells transfected with Flag-STAT1 (WT or T699A). F IP-IB analysis of Flag-STAT1 Kbhb in HEK293T cells cotransfected with Flag-STAT1 and increasing amounts of Myc-OGT. G IP-IB analysis of the CBP-STAT1 interaction and STAT1 Kbhb in HEK293T cells cotransfected with HA-CBP, Myc-STAT1 and increasing amounts of OGT. H Diagram of the competitive inhibition of STAT1 Kbhb by STAT1. CBP-induced STAT1 Kbhb at K592 inhibits the STAT1-IFNAR2 interaction and IFN-I signaling, whereas STAT1 O-GlcNAc at T699 induced by OGT inhibits the CBP-STAT1 interaction, thereby improving IFN-I antiviral immunity. The data are representative of three independent experiments (A–G)

We further determined whether STAT1 O-GlcNAc at Thr699 inhibits STAT1 Kbhb modifications. The results revealed that mutation of STAT1 Thr699 (T699A) led to stronger Kbhb modification of STAT1 (Fig. 6E) but did not obviously affect the other acylation modifications of STAT1 observed in this study (Supplementary Fig. 5a). Notably, OGT overexpression increased STAT1 O-GlcNAc modifications and simultaneously decreased STAT1 Kbhb modifications in a dose-dependent manner (Fig. 6F). Consistently, OGT overexpression inhibited the STAT1-CBP interaction (Fig. 6G). Taken together, these findings suggest that CBP induces STAT1 Kbhb at Lys592, which blocks STAT1 binding to IFNAR2 and therefore inhibits IFN-I signaling, whereas OGT induces STAT1 O-GlcNAc at Thr699, which inhibits CBP binding to STAT1, thus restricting CBP-induced STAT1 Kbhb modifications (Fig. 6H).

STAT1 Thr699 O-GlcNAc promotes IFN-I signaling and IFN-I antiviral immune activity

Further studies demonstrated that OGT overexpression facilitated IFN-I-induced STAT1 Tyr701 phosphorylation (Supplementary Fig. 5b, c), which was dependent on the enzyme activity of OGT (Supplementary Fig. 5d). Conversely, OGT knockdown markedly inhibited STAT1 Tyr701 phosphorylation (Supplementary Fig. 5e). Furthermore, OGT promoted IFN-I-induced ISG expression in various cell lines (Supplementary Fig. 6a, b), whereas OGT deficiency inhibited the IFN-I-induced expression of ISGs (Supplementary Fig. 6c, d). As a consequence, OGT overexpression promoted IFN-I-mediated antiviral activity (Supplementary Fig. 6e), which was dependent on the enzyme activity of OGT (Supplementary Fig. 6f), whereas OGT knockdown attenuated IFN-I-mediated antiviral activity (Supplementary Fig. 6g, h).

Next, we found that mutation of STAT1 at Thr699 inhibited the interaction between STAT1 and IFNAR2 (Fig. 7A). Consistently, OGT knockdown reduced STAT1 O-GlcNAc and inhibited the STAT1-IFNAR2 interaction (Fig. 7B). Conversely, OGT overexpression facilitated the STAT1-IFNAR2 interaction (Supplementary Fig. 7a). In line with the reduced STAT1-IFNAR2 interaction, mutation of STAT1 Thr699 dramatically inhibited IFN-I-induced STAT1 Tyr701 phosphorylation (Fig. 7C and Supplementary Fig. 7b–e). Next, we investigated the effects of STAT1 Thr699 O-GlcNAc on IFN-I-induced ISG expression and IFN-I antiviral activity. The results revealed that mutation of STAT1 at Thr699 significantly inhibited the IFN-I-induced expression of ISGs (Fig. 7D, E and Supplementary Fig. 7f). Consistently, mutation of STAT1 Thr699 attenuated IFN-I-mediated antiviral immune activity (Fig. 7F and Supplementary Fig. 7g). Collectively, these findings demonstrate that OGT-induced STAT1 O-GlcNAc modifications at Thr699 facilitate IFN-I signaling and antiviral immune defense activity.

Fig. 7.

STAT1 O-GlcNAc modifications at Thr699 enhance IFN-I signaling. A IP-IB analysis of the interaction between HA-IFNAR2 and Myc-STAT1 in HEK293T cells cotransfected with HA-IFNAR2 and Myc-STAT1 (WT or T699A). B IP-IB analysis of the interaction between HA-IFNAR2 and Myc-STAT1 in HEK293T cells cotransfected with HA-IFNAR2 and Myc-STAT1, together with shOGT (#1, #2). C Western blot analysis of STAT1-pY701 in Stat1^−/− HEK293T cells transfected with Flag-STAT1 (WT or T699A) and then treated with IFNα (1000 or 3000 IU/ml) for 30 min. D, E RT‒qPCR analysis of representative ISGs (Ifit1 and Viperin) in U3A (D) or Stat1^−/− HEK293T cells (E) transfected with Myc-STAT1 (WT or T699A) and then stimulated with IFNα (1000 IU/ml) for 4 h. F Fluorescence microscopy of VSV-GFP in Stat1^−/− HEK293T cells transfected with Myc-STAT1 (WT or T699A) and then treated with IFNα (60 IU/ml) for 20 h, followed by infection with VSV-GFP (MOI = 0.5) for 24 h. Scale: 100 µm. The data are representative of three independent experiments (A–C) or are shown as the means and s.d.s of three biological replicates (D, E). **p < 0.01, ***p < 0.001 (two-tailed unpaired Student’s t test)

Fructose orchestrates STAT1 O-GlcNAc and Kbhb and improves the in vivo IFN-I antiviral defense ability of aging mice

Given that STAT1 O-GlcNAc modifications can rescue the decrease in IFN-I antiviral immune activity caused by aging-related STAT1 Kbhb, we speculated that increasing the level of STAT1 O-GlcNAc could increase IFN-I antiviral defense during aging. Fructose metabolism is able to provide the metabolite UDP-GlcNAc to promote O-GlcNAc modification (Fig. 8A). Thus, we first observed the effect of fructose on STAT1 O-GlcNAc. The concentration of fructose used in the study was 5 or 10 mM, since the published studies usually used 5–25 mM fructose [24–26]. We found that fructose increased STAT1 O-GlcNAc modifications and correspondingly downregulated STAT1 Kbhb modifications (Fig. 8B and Supplementary Fig. 8a). Importantly, mutation of STAT1 at Thr699 blocked fructose-mediated upregulation of STAT1 O-GlcNAc (Fig. 8C), and mutation of STAT1 at K592 abolished fructose-mediated downregulation of STAT1 Kbhb (Fig. 8D). In addition, a specific inhibitor of OGT (OSMI-4) inhibited the effect of fructose on the regulation of STAT1 O-GlcNAc and Kbhb modifications (Supplementary Fig. 8b). Next, we found that fructose promoted the interaction between STAT1 and IFNAR2 (Fig. 8E) and enhanced IFN-induced STAT1 Tyr701 phosphorylation (Fig. 8F). Further analysis demonstrated that fructose significantly upregulated the transcription of ISGs induced by IFN-I (Fig. 8G and Supplementary Fig. 8c, d) but was inhibited by OSMI-4 (Supplementary Fig. 8e), thereby increasing IFN-I antiviral activity (Supplementary Fig. 8f). Moreover, mutation of STAT1 Thr699 inhibited the fructose-mediated increase in ISG expression induced by IFN-I (Fig. 8H).

Fig. 8.

Fructose orchestrates STAT1 O-GlcNAc and Kbhb to improve in vivo IFN-I antiviral immunity in aging mice. A diagram of the process of O-GlcNAc modification stimulated by glucose and fructose. B IP-IB analysis of the levels of both O-GlcNAc and Kbhb of STAT1 in HT1080 cells treated with fructose (5 or 10 mM) for 24 h. C IP-IB analysis of STAT1 O-GlcNAc in HEK293T cells transfected with Flag-STAT1 (WT or T699A) and then treated with fructose (10 mM) for 24 h. D IP-IB analysis of STAT1 Kbhb in HEK293T cells transfected with Flag-STAT1 (WT or K592R) and then treated with fructose (10 mM) for 24 h. E IP-IB analysis of the interaction between HA-IFNAR2 and STAT1 in HEK293T cells transfected with HA-IFNAR2 and then treated with fructose (10 mM, 24 h). F Western blot analysis of STAT1-pY701 in HT1080 cells treated with fructose (10 mM, 24 h) and then stimulated with IFNα (1000 IU/ml, 30 min). G RT‒qPCR analysis of representative ISG (Ifit1 and Isg54) mRNAs in HT1080 cells treated with fructose (10 mM, 24 h) and then stimulated with IFNα (1000 IU/ml, 4 h). H RT‒qPCR analysis of representative ISGs (Ifit1 and Isg54) in Stat1^−/− HEK293T cells transfected with Myc-STAT1 (WT or T699A) and then treated with fructose (10 mM, 24 h), followed by IFNα treatment (1000 IU/ml, 4 h). I Diagram of mouse experiments in which aging mice were fed 30% fructose-water for 1 week and then infected with VSV viruses (1 × 10⁸ PFU per gram body mouse, i.p.) for 24 h. J IP-IB analysis of both O-GlcNAc and Kbhb levels of STAT1 in liver tissues from aging mice (n = 3) fed 30% fructose-water for 1 week. K RT‒qPCR analysis of representative ISG (Ifit1) mRNA in liver, spleen, and kidney tissues from aging mice (n = 6) fed 30% fructose-water for 1 week. L RT‒qPCR analysis of VSV viral RNA in liver, spleen, and kidney tissues from aging mice (n = 6) fed 30% fructose-water for 1 week and then infected with VSV (1 × 10⁸ PFU per gram body weight mouse, i.p.) for 24 h. M IB analysis of VSV-G proteins in liver and kidney tissues treated as described in (J). The data are representative of three independent experiments (B–F). The graphs show the means ± s.e.m. for six individual mice (K, L). N.S. not significant, *p < 0.05, **p < 0.01, ***p < 0.001 (two-tailed unpaired Student’s t test)

To investigate the in vivo effects of fructose on STAT1 modifications and IFN-I antiviral immune activity, aged mice were fed 30% fructose in drinking water (30% fructose-water) for one week according to the literature (Fig. 8I). We observed that fructose feeding upregulated STAT1 O-GlcNAc and downregulated STAT1 Kbhb in the liver tissues of aging mice (Fig. 8J). Furthermore, fructose feeding significantly promoted the expression of ISGs in the liver, spleen, kidney, heart and lung tissues of aging mice challenged with viruses (Fig. 8K and Supplementary Fig. 8g–i). Consistent with increased ISG expression, fructose feeding improved the antiviral defense ability of aging mice, as shown by decreased VSV viral RNA levels and virus-encoded proteins in various tissues of mice fed fructose (Fig. 8L, M and Supplementary Fig. 8j). Taken together, these findings demonstrated that fructose can orchestrate STAT1 O-GlcNAc and Kbhb modifications and, in turn, enhance the IFN-I-mediated antiviral defense ability of the body during aging.

Discussion

Aging changes cell metabolism and leads to abnormal production of metabolites, which could affect PTMs of cellular proteins. Recent studies have explored the changes in PTMs of nuclear histones during aging [27]. For example, it has been reported that H3K4me3 levels are upregulated with age in mouse hematopoietic stem cells (HSCs) [28]. In addition, several studies have demonstrated correlations between aging and histone phosphorylation and ubiquitination [29, 30]. Nevertheless, whether and how aging regulates PTMs of cytoplasmic signaling proteins remain largely unexplored. In this study, we found that aging regulates both the Kbhb and O-GlcNAc modifications of STAT1, which is a central signaling protein of the IFN-I pathway. Importantly, this study revealed that the upregulation of STAT1 Kbhb in aging blocks STAT1 binding to the IFN-I receptor IFNAR2 and therefore attenuates IFN-I antiviral immune activity in aging mice, whereas STAT1 O-GlcNAc results in a switch from a “inhibitory” signaling modification (Kbhb) to a “active” signaling modification (O-GlcNAc) of STAT1. Moreover, we identified fructose as a rescuer that promotes STAT1 O-GlcNAc to restrict STAT1 Kbhb and ultimately enhances the IFN-I antiviral defense ability of aging mice.

Recently, PTM-based therapeutics have been considered an attractive strategy for treating aging-related diseases. For example, kinase inhibitors have been evaluated in some aging-associated diseases, such as AD [31] and diabetes [32]. Therefore, it is imperative to explore potential small molecule compounds that target various PTMs for the treatment of clinical diseases. In this study, we screened a drug library containing 214 small molecules from plant sources and identified hydroxy camptothecine as a potent small molecule compound that inhibits STAT1 Kbhb modifications. Interestingly, the protein levels of CBP and SIRT3 were not significantly affected by hydroxy camptothecine. Thus, we speculated that the hydroxy camptothecin may affect the enzyme activity of either SIRT3 or CBP to regulate STAT1 Kbhb modifications. In addition, it is also possible that hydroxy camptothecin may regulate the interaction between STAT1 and CBP. Given that we have demonstrated that STAT1 Kbhb is upregulated during aging and inhibits IFN-I antiviral immune activity, we believe that downregulation of STAT1 Kbhb could enhance the antiviral defense ability of aging mice. Further in vivo observations proved that hydroxy camptothecine administration lowered in vivo STAT1 Kbhb levels and improved the IFN-I antiviral defense activity of aging mice. These observations not only confirmed the role of STAT1 Kbhb in regulating IFN-I signaling but also provided a potential strategy for enhancing antiviral immunity in the body during aging.

Despite the potential of hydroxy camptothecin in improving the IFN-I antiviral defense activity of the body during aging, the use of hydroxy camptothecine could have unreported side effects, which may restrict its usage. Thus, we further explored strategies to inhibit CBP-induced STAT1 Kbhb. We believe that certain PTMs located at the protein‒protein binding region could inhibit the interaction between two proteins, which could be an effective strategy to control the activity of the corresponding proteins. Thus, we further identified the STAT1-CBP binding region and reported that there is an O-GlcNAc modification located at Thr699 of this binding region of STAT1. We demonstrated that STAT1 O-GlcNAc at Thr699 can inhibit CBP binding to STAT1, thus inhibiting CBP-induced Kbhb modifications of STAT1. On the basis of these findings, we further explored the effect of fructose on STAT1 O-GlcNAc, since fructose could provide the metabolite UDP-GlcNAc to increase O-GlcNAc modification. Further in vitro and in vivo results demonstrated that fructose can effectively promote STAT1 O-GlcNAc, inhibit STAT1 Kbhb, and ultimately enhance the IFN-I antiviral defense ability of aging mice. Thus, this study suggests that an appropriate high-fructose diet during a virus epidemic could benefit the antiviral defense ability of the body in elderly individuals.

This study revealed crosstalk between STAT1 O-GlcNAc and Kbhb. Recently, researchers have paid increasing attention to the interplay between two different PTMs of cellular proteins. For example, the glycosylation of IRF5 at S430 promotes its K63-linked ubiquitination [33]. MAVS O-GlcNAc modifications promote K63 ubiquitination and subsequent IFN production [34]. In addition, O-GlcNAc modification inhibits IRS1 activity by increasing its phosphorylation at Ser307 and Ser632/635 [35]. However, the interplay between the O-GlcNAc modification and the Kbhb modification remains unknown. In this study, we demonstrated that STAT1 O-GlcNAc inhibits STAT1 Kbhb. Mechanistically, OGT-induced STAT1 O-GlcNAc at Thr699 blocks CBP binding to STAT1, thus inhibiting STAT1 Kbhb modifications at Lys592 induced by CBP. The crosstalk between STAT1 O-GlcNAc and Kbhb determines cellular IFN-I antiviral immune activity. In summary, this study reveals the significance of the switch between STAT1 Kbhb and O-GlcNAc modifications in regulating IFN-I antiviral immune activity during aging and provides potential strategies to increase the antiviral defense ability of the body in elderly individuals.

Materials and methods

Mice

Wild-type (WT) C57BL/6J and aging mice were purchased from Shanghai JiHui Company. All the mice were maintained under specific-pathogen-free (SPF) conditions in the animal facility of Soochow University. In this study, 6~8-week-old or 16~18-month-old mice were used in the corresponding experiments. The animal care and use protocol adhered to the National Regulations for the Administration of Affairs Concerning Experimental Animals. All animal experiments received ethical approval from the Ethics Committee of Soochow University (ethical approval numbers: 202303A0093, 202310A0182, 202407A0863 and 202410A0241) and were carried out in accordance with the Laboratory Animal Management Regulations with the approval of the Scientific Investigation Board of Soochow University, Suzhou.

Animal experiments

For mouse experiments with hydroxy camptothecine, hydroxy camptothecine (3 mg/kg body weight) was prepared in 5% DMSO, 40% PEG300 and 5% Tween 80 and then diluted with ddH₂O. Hydroxy camptothecine and vehicle were intraperitoneally (i.p.) injected into the mice once a day. After 3 days, the mice were given i.p. injections of VSV (1 × 10⁸ PFU per gram body weight mouse). Twenty-four hours after infection, mouse heart, liver, spleen, lung and kidney tissues were harvested. RT‒qPCR or western blotting was subsequently used for the analysis of vesicular stomatitis virus (VSV) viral RNAs and virus-encoded proteins.

Cell culture and reagents

HEK293T, HT1080 and RAW264.7 cells were obtained from ATCC. U3A cells were kindly provided by Guo-Qiang Chen (Shanghai Jiaotong University, China). The cells were cultured at 37 °C under 5% CO₂ in DMEM (Gibco) supplemented with 10% FBS (GIBCO, Life Technologies), 100 unit/ml penicillin, and 100 µg/ml streptomycin. Recombinant human IFNɑ was obtained from PBL Interferon Source. The SIRT3 inhibitor 3-TYP was purchased from Selleck. A-458, OSMI-4 and Thiamet G were purchased from Selleck. Flag peptides (F3290), puromycin and other chemicals were purchased from Sigma.

Plasmids and transfection

The HA-CBP and HA-p300 plasmids were gifts from Dr. Jin Liu (The Affiliated Infectious Diseases Hospital of Soochow University). Myc-OGT, Flag-SIRT1 and Flag-SIRT2 were gifts from Dr. Lin Hu (Soochow University). Flag-STAT1 was generated via PCR amplification from pIND-STAT1-V5 from Dr. Steven Johnson. Flag-SIRT3 was purchased from Vigene Biosciences. HA-IFNAR2 was a gift from Dr. Serge Y. Fuchs (University of Pennsylvania). All plasmids were confirmed by DNA sequencing. All the mutations were generated via the QuickChange Lightning site-Directed Mutagenesis Kit (Stratagene, 210518). Transient transfections of different cell lines were carried out via LongTrans or GenePORTER2 (Ucallm, TF/07; Genlantis, T202015).

Mass spectrometry (MS) analysis

HEK293T cells were transfected with Myc-STAT1. Forty-eight hours after transfection, the cells were harvested with Nonidet P-40 lysis buffer [150 mM NaCl, 1% NP-40, Tris-HCl (20 mM, pH 7.4), PMSF (50 mg/mL), 0.5 mM EDTA and protease inhibitor mixtures]. An anti-Myc antibody was used to pull down Myc-STAT1. Mass spectrometry analyses were performed as follows: SDS‒PAGE gels were minimally stained with Coomassie brilliant blue and then cut into 1 × 1 mm gel blocks, followed by digestion with trypsin. Next, the resulting tryptic peptides were purified via the C18 Zip Tip. The peptides were subsequently analyzed via an Orbitrap Elite hybrid mass spectrometer (Thermo Fisher) coupled with a Dionex LC. Next, MS/MS spectra were collected for the selected precursor ions within a 0.02 Da mass isolation window. Afterwards, the spectral data were searched via Proteome Discoverer 1.4 against the UniProt protein database. Peptide spectrum matches (PSMs) for STAT1 or STAT2 were obtained after a database search.

Immunoblotting (IB) and immunoprecipitation (IP)

The cells were harvested via lysis buffer containing 1% Nonidet P-40 (NP-40), 150 mM NaCl, Tris-HCl (20 mM, pH 7.4), 0.5 mM EDTA, PMSF (50 µg/ml) and protease inhibitor mixtures (Sigma). Cellular proteins from whole-cell lysates (WCLs) were first subjected to SDS‒PAGE and then transferred to PVDF membranes (Millipore). After blocking with 5% nonfat milk or 5% BSA (for the antibodies targeting phosphorylated proteins) for 1 h, the membranes were incubated with the corresponding primary antibodies overnight, followed by incubation with the secondary antibodies (Bioworld or Abbkine) for 1 h. All immunoreactive bands were visualized with SuperSignal West Dura Extended kits (Thermo Scientific).

Immunoprecipitation was carried out using specific antibodies at 4 °C. Protein G agarose beads (Millipore, #16--266) were incubated with the samples on a rotor at 4 °C. After washing five times with lysis buffer, the immunoprecipitates were eluted by heating at 95 °C with loading buffer containing β-mercaptoethanol for 10 min. The samples were then analyzed by SDS‒PAGE and subsequent immunoblotting.

The antibodies with the indicated dilutions were as follows: Acylation Antibody Sampler Kit Plus (PTM BIO, PTM-6680, 1:1000), Anti-β-Hydroxybutyryllysine (PTM BIO, PTM-1201RM, 1:1000), Anti-pY701 (STAT1) (Cell Signaling Technology, 9167S, 1:1000), Anti-Flag (Sigma, F7425, 1:5000), Anti-HA (Abcam, ab9110, 1:3000), Anti-STAT1 (Cell Signaling Technology, 9172S, 1:1000), Anti-STAT2 (Cell Signaling Technology, 72604S, 1:1000), Anti-O-GlcNAc (Cell Signaling Technology, 9875S, 1:1000), Anti-Myc (Abmart, M20002H, 1:2000), Anti-VSV-G (abcam, ab1874, 1:2000), Anti-OGT (Proteintech, 11576-2-AP, 1:1000), Anti-β-Actin (Proteintech, 66009-1-Ig, 1:1000), Anti-Tubulin (Proteintech, 66031-1-Ig, 1:3000), Anti-IFNAR2 (Affinity, DF2535, 1:1000), Anti-STAT1 (Santa Cruz, sc-464, 1:1000) and Anti-SIRT3 (Affinity, AF5135, 1:1000), Anti-CBP (Affinity, AF5487, 1:1000).

RNA isolation and quantitative real-time PCR

Total RNA was isolated from different cells or mouse tissues via TRIzol reagent (Invitrogen). Briefly, cDNA was synthesized via the RevertAid First Strand cDNA Synthesis Kit (abm, G490). RT‒qPCR was performed via a StepOne Plus real-time PCR system (Applied Bioscience). The results from three independent experiments are presented as the average mean ± standard deviation (s.d.). The sequences of primers used were as follows:

human Ifit1:

Forward: 5’- GCCTTGCTGAAGTGTGGAGGAA -3’

Reverse: 5’- ATCCAGGCGATAGGCAGAGATC -3’;

human Isg15:

Forward: 5’-CTCTGAGCATCCTGGTGAGGAA-3’

Reverse: 5’-AAGGTCAGCCAGAACAGGTCGT-3’;

human Isg54:

Forward: 5’-GGAGCAGATTCTGAGGCTTTGC-3’

Reverse: 5’-GGATGAGGCTTCCAGACTCCAA-3’;

human Viperin:

Forward: 5’-CCAGTGCAACTACAAATGCGGC-3’

Reverse: 5’-CGGTCTTGAAGAAATGGCTCTCC-3’;

VSV:

Forward: 5’-ACGGCGTACTTCCAGATGG-3’

Reverse: 5’-CTCGGTTCAAGATCCAGGT-3’;

mouse Ifit1:

Forward: 5’-TACAGGCTGGAGTGTGCTGAGA-3’

Reverse: 5’-CTCCACTTTCAGAGCCTTCGCA-3’;

mouse Rsad2 (Viperin):

Forward: 5’-CAGGCTGGTTTGGAGAAGATCAAC-3’

Reverse: 5’-TACTCCCCATAGTCCTTGAACCATC-3’;

Mouse Isg15:

Forward: 5’-CATCCTGGTGAGGAACGAAAGG-3’

Reverse: 5’-CTCAGCCAGAACTGGTCTTCGT-3’;

β-actin:

Forward: 5’-ACCAACTGGGACGACATGGAGAAA-3’

Reverse: 5’-ATAGCACAGCCTGGATAGCAACG-3’.

GAPDH:

Forward: 5’-CATCACTGCCACCCAGAAGACTG-3’

Reverse: 5’-ATGCCAGTGAGCTTCCCGTTCAG-3’.

CRISPR–Cas9-mediated genome editing

The lenti-CRISPRv2 vector was a gift from Dr. Fangfang Zhou (Soochow University, China). For gene knockout, small guide RNAs were first cloned and inserted into the lenti-CRISPRv2 vector and then transfected into HEK293T cells. Forty-eight hours after transfection, the cells were cultured under puromycin (1.5 µg/ml) selection for 2 weeks, after which they were identified by immunoblotting analysis. After that, the cells were transferred to 96-well plates and cultured for further experiments.

The guide RNA sequences used were as follows:

human Sirt3 (sgRNA): 5′-CTCTACACGCAGAACATCGA-3′;

human Stat1 (sgRNA1): 5′-CTGTGGTCTGAAGTCTAGAA-3′;

human Stat1 (sgRNA2): 5′-TGACGAGGTGTCTCGGATAG-3′.

Screening of the small-molecule compound library

The drug library with 214 clinically approved small molecule compounds from plant sources was purchased from Selleck Chemicals (L2000--Z659310). All drugs were diluted to 10 mM in DMSO. The cells were plated in 6-well plates and treated with vehicle (Ctrl) or each compound (5 μM) for 24 h. After that, the cells were harvested. IP-IB was then used to detect STAT1 Kbhb levels, and the relative levels of STAT1 Kbhb are shown as the fold change relative to those in the control group (vehicle) after quantification via densitometric analysis of the immunoblotting signals.

In vitro binding assay

HEK293T cells were transfected with the corresponding constructs. Flag-STAT1 (WT and K592R) proteins were immunoprecipitated with M2 (Flag) agarose and then eluted with the Flag peptides. HA-IFNAR2 proteins were pulled down from HEK293T cells transfected with HA-IFNAR2 by the HA antibody and protein G beads. Then, the Flag-STAT1 (WT or K592R) eluates were mixed and vibrated with the HA-IFNAR2 immunoprecipitates containing protein G beads at 4 °C. The products of this reaction were separated via SDS‒PAGE and analyzed by immunoblotting with anti-Flag or anti-HA antibodies.

Viruses and viral infection in vitro

Vesicular stomatitis virus (VSV) and VSV-GFP were described previously [36, 37]. The cells were first transfected with the corresponding plasmids and then treated with IFNα (60 IU/ml) overnight. After being washed twice, the cells were infected with either VSV-GFP or VSV at a multiplicity of infection (MOI) of 0.5 or 1.0 in serum-free medium for 2 h for virus entry. The infection medium was subsequently removed by washing twice with 1x PBS. After that, the cells were cultured with fresh medium (10% FBS) for 24 h. Then, the cells were analyzed by immunofluorescence, RT‒qPCR, or western blotting.

Viral infection in vivo

The procedures for in vivo viral infection were described previously [38, 39]. Briefly, aging mice were given i.p. injections of VSV (1 × 10⁸ PFU per gram body weight mouse). Twenty-four hours after infection, mouse tissues, including lung, kidney, liver, spleen and heart tissues, were harvested. RT‒qPCR was subsequently used to analyze vesicular stomatitis virus (VSV) RNA levels. In addition, western blotting was carried out using whole-cell lysates from mouse spleen, lung, liver and kidney tissues.

Isolation of peritoneal macrophages

Peritoneal macrophages were harvested from the mice 4 days after the injection of thioglycollate (BD) and cultured in DMEM supplemented with 5% FBS. Then, the cells were treated with different inhibitors for further analysis.

Immunofluorescence microscopy

The cells infected with VSV-GFP were imaged under an upright fluorescence microscope at a magnification of ×200. The data were analyzed via ImageJ software.

Statistics

Two-tailed unpaired Student’s t tests were used to compare the significance of differences between different groups. All differences were considered statistically significant at p < 0.05. P values are indicated by the asterisks in the corresponding figures as follows: *p < 0.05, **p < 0.01 and ***p < 0.001.

Author contributions

YZ, QW, WT and ZZ performed the experiments. WH, RZ and QZ assisted with the mouse experiments, tissue processing and analysis. HZ and YZ designed the experiments, analyzed the data and wrote the paper. YM and YY helped analyze the data. HZ, JW, YZ and QW discussed the manuscript. HZ was responsible for research supervision, coordination, and strategy.

Data availability

All the data generated or analyzed during this study are included in Figs. 1–8 and Supplementary Figs. 1–8. The raw data containing uncropped images of all gels and blots are included in the Supplementary Information. Additional datasets that support the findings of this study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41423-025-01266-x.