Based on our previously reported DeepAVC framework, which achieved highly accurate phenotype- and target-guided prediction of broad-spectrum antiviral compounds, we identified NVP-BVU972 (NVP) as a novel candidate and conducted detailed experimental validation of its antiviral efficacy in RAW 264.7 cells. RAW 264.7 cells were concurrently treated for 12 hours with NVP at concentrations of 25, 50, or 100 µM, or DMSO control, and infected at a multiplicity of infection (MOI) of 0.1 with vesicular stomatitis virus (VSV), herpes simplex virus type 1 (HSV-1), mouse hepatitis virus (MHV), or encephalomyocarditis virus (EMCV). RT-qPCR analysis demonstrated a dose-dependent reduction in viral mRNA levels upon NVP treatment compared with DMSO controls, with maximum suppression observed at 100 µM ( Figure 1A ).

To confirm that these antiviral effects extended to human cells, we treated HeLa, HT29, and HT1080 cell lines with 50 μM NVP and concurrently infected with VSV, HSV-1, or EMCV. RT-qPCR quantification indicated a significant 50–70% decrease in viral mRNA levels in NVP-treated cells relative to controls ( Figures 1B-D ). Additionally, flow cytometry analysis at 12 hours post-infection with GFP-expressing VSV (VSV-GFP, MOI=0.1) showed substantially fewer GFP-positive cells in the NVP-treated groups, confirming effective inhibition of viral replication ( Figures 1E, F ).

We further assessed the broad-spectrum antiviral capability of NVP using both DNA viruses (HSV-1 and Vaccinia virus [VACV]) and RNA viruses (VSV) expressing GFP. Fluorescence microscopy ( Figure 1G ) and flow cytometry ( Figure 1H ) consistently revealed significant decreases in GFP expression upon NVP treatment, demonstrating robust suppression of viral replication regardless of virus type.

Dose-dependent antiviral activity was further verified by western blot analysis of GFP protein expression in RAW 264.7 cells infected with HSV, VACV, or VSV ( Figure 1I ). Clear antiviral effects emerged at an NVP concentration of 50 μM, with complete viral suppression achieved at 100 μM.

Encouraged by these compelling in vitro findings, we next evaluated the protective effect of NVP in vivo. Six-week-old C57BL/6J mice were intraperitoneally infected with MHV, VSV, HSV-1, or EMCV, followed by subsequent treatment with NVP or vehicle control. Survival analysis over a 7-day period revealed significantly improved survival rates in mice receiving NVP treatment compared to untreated controls ( Figure 1J ). Thus, at 24 hours post-infection, RT-qPCR analyses of blood, liver, and spleen tissues from a subset of mice indicated markedly reduced viral levels across all tested viruses in the NVP-treated groups ( Figures 1K-N and Supplementary Figure S1A ), further confirming potent antiviral activity in vivo.

Together, these results demonstrated that NVP exerted potent, dose-dependent antiviral effects across multiple cell lines and provided broad-spectrum protection against both RNA and DNA viruses in vivo.

To determine whether NVP possessed anti-inflammatory activity during viral infections, we first evaluated the expression of pro-inflammatory cytokines at the mRNA level in vitro. RAW264.7 macrophages were infected with VSV, HSV-1, EMCV or MHV, and simultaneously treated with 100 µM NVP or DMSO control ( Figure 2A ). Additionally, three human cell lines, including HeLa, HT29 and HT1080 were similarly infected with VSV, HSV-1, or EMCV under identical treatment conditions ( Figures 2B-D ). RT-qPCR analysis at 12 hours post-infection indicated that NVP treatment strongly suppressed virus-induced transcription of pro-inflammatory cytokines Il6, Tnfα, and Il1β compared to the DMSO controls, approaching baseline expression levels in RAW264.7 cells across all viral challenges. Consistent and robust inhibition of these cytokines was also observed across all tested human cell lines, demonstrating that NVP exerted a broad and cell-type-independent anti-inflammatory effect. Additionally, UV−inactivated VSV and poly(I:C) assays showed that NVP directly blocked NF−κB−driven cytokine induction independent of viral replication ( Supplementary Figure S2A, B ).

To further investigate whether NVP-mediated suppression of virus-triggered inflammation could be translated in vivo, we infected Six-week-old C57BL/6J mice intraperitoneally with VSV, HSV-1, EMCV, or MHV ( Figures 2E-H ). At 24 hours post-infection, RT-qPCR analysis of pro-inflammatory cytokine mRNA levels in blood, liver, and spleen tissues revealed that NVP administration significantly attenuated the induction of Il6, Tnfα, and Il1β compared to PBS-treated control mice, reflecting systemic suppression of viral inflammation.

Histopathological examination of liver sections using hematoxylin and eosin (H&E) staining further validated the protective effects of NVP in vivo. Vehicle-treated, virus-infected mice exhibited pronounced inflammatory cell infiltration, severe hepatocyte damage, and disrupted liver architecture. By contrast, liver tissues from NVP-treated mice displayed significantly reduced inflammatory infiltration, preserved tissue architecture, and minimal hepatocellular injury, consistent with the observed molecular findings ( Figures 2I-L ).

Collectively, these comprehensive in vitro and in vivo findings established that NVP effectively reduced virus-induced inflammatory responses through suppression of inflammatory cytokine expression, inhibition of NF-κB signaling pathway activation, and alleviation of virus-mediated tissue injury, highlighting its therapeutic potential as a dual antiviral and anti-inflammatory compound.

Building upon our observations of antiviral and virus-induced anti-inflammatory activity, we next evaluated the direct anti-inflammatory effects of NVP in a classic LPS-induced inflammatory model and assessed subacute safety.

First, we treated RAW 264.7 macrophages with 100 µM NVP for 3 hours in the presence of 2 μg/mL lipopolysaccharide from E. coli O55:B5. RT-qPCR analysis showed that LPS significantly induced mRNA expression of multiple inflammatory genes, including Il1β, Il6, Tnfα, Acod1, Ccl3, Lta, Ltb, Nos2, and Ptgs2 ( Figures 3A-C ). NVP pretreatment substantially reduced these increases, bringing mRNA levels close to baseline. Importantly, NVP treatment significantly reduced basal Il6 mRNA levels, whereas Il1β remained unchanged and Tnfα exhibited only a slight, non−significant decrease ( Supplementary Figure S1C ), demonstrating that NVP did not directly upregulate pro−inflammatory cytokines under resting conditions.

LPS triggered a Toll-like receptor 4 (TLR4)-mediated signaling cascade that activated the IκB kinase (IKK) complex. IKK then phosphorylated the inhibitory protein IκBα, marking it for degradation and thereby releasing the NF-κB p65 subunit to translocate into the nucleus and induce pro-inflammatory gene expression. To explore the underlying mechanism, we performed Western blot analyses of key proteins involved in the NF-κB signaling pathway ( Figure 3D ). LPS stimulation markedly increased phosphorylation of IKKα/β, IκBα, and P65, and significantly induced COX-2 and iNOS protein expression. Treatment with NVP fully blocked these changes without affecting total protein levels of IκBα and P65, indicating that NVP exerted its anti-inflammatory effects by suppressing NF-κB pathway activation.

Next, we examined the anti-inflammatory effects of NVP in vivo. Six-week-old C57BL/6J mice received the pretreatment of a single dose of NVP (20 mg/kg, i.p.) followed by a lethal intraperitoneal injection of LPS (20 mg/kg) one hour later. All vehicle-treated mice died within 48 hours, while nearly 50% of the NVP-treated mice survived ( Figure 3E ). To further assess systemic inflammation, we harvested blood, heart, liver, spleen, lung, and kidney tissues 12 hours post-LPS administration and measured inflammatory cytokine mRNA levels. NVP-treated mice showed significantly decreased levels of Il1β, Il6, and Tnfα mRNAs across all tissues compared to controls, demonstrating effective suppression of systemic inflammation ( Figure 3F-K ).

Lastly, we assessed the subacute toxicity of NVP. Mice received daily intraperitoneal injections of NVP (20 mg/kg) for 14 consecutive days. After the treatment period, no significant differences in organ weight ( Figure 3L ), organ appearance ( Figure 3M ), or histopathological structure (heart, liver) ( Figure 3N ) were observed between NVP-treated and control groups. These results confirmed a favorable safety and tolerability profile. In addition, NVP exhibited no significant cytotoxicity in cultured cells as assessed by the CCK-8 assay in RAW264.7, HeLa, HT29, HT1080 cell lines ( Supplementary Figure S1B ).

Collectively, our findings demonstrated that NVP effectively suppressed LPS-induced inflammation by inhibiting NF-κB activation and cytokine production both in vitro and in vivo, while maintaining excellent safety upon repeated dosing.

To understand how NVP modulated host responses, RAW264.7 macrophages were treated with 50 µM NVP or DMSO for 12 hours in two independent biological replicates, then we performed bulk RNA sequencing on each sample. Differential expression analysis identified 1402 upregulated and 1706 downregulated genes upon NVP treatment. A heatmap of these differentially expressed genes (DEGs) showed marked suppression of proinflammatory transcripts after NVP treatment ( Figure 4A ). Volcano plot visualization confirmed these cytokine genes among the most significantly downregulated targets ( Figure 4B ). KEGG pathway analysis of all DEGs revealed enrichment in “Autophagy” “Epstein-Barr virus infection” and “MAPK signaling pathway” as the top affected pathways. Gene Ontology enrichment highlighted “cellular response to virus” “heterochromatin organization” “negative regulation of inflammatory response” and “N-methyltransferase activity” ( Figure 4C ). These data indicated that NVP broadly suppressed inflammatory gene programs at the transcriptional level. Notably, the appearance of “histone modifying activity” in GO terms suggested that NVP might alter histone marks to repress inflammatory programs.

To determine whether these transcriptional changes correlated with chromatin accessibility, we carried out ATAC sequencing on cells treated with NVP (50 µM) or DMSO for 12 hours, as in our bulk RNA-seq experiments. Compared to DMSO treatment, NVP induced a net loss of approximately 1696 accessible peaks, with only one peak gained ( Figures 4D, E ). To link these chromatin changes to gene expression, we intersected the list of genes with significantly decreased ATAC peaks and the downregulated transcripts from RNA-seq. This overlap produced 243 genes ( Figure 4F ). GO and KEGG enrichment of this intersected gene set again emphasized “chromatin sliencing complex” “methylation” “viral myocarditis” and “ATP-dependent chromatin remodeling” ( Figure 4G ). Together, these findings suggested that NVP repressed inflammatory gene expression by reducing chromatin accessibility.

Guided by the GO enrichment of histone modifying activity, we next measured global levels of two key histone modifications, H3K9me3 (repressive) and H3K27ac (activating) by Western blot ( Figure 4H ). The result showed that H3K9me3 was increased strongly, whereas H3K27ac levels remained unchanged. This selective enhancement of H3K9me3 supported the hypothesis that NVP promoted deposition of repressive methylation marks at inflammatory gene loci, leading to reduced chromatin accessibility and transcriptional repression. To further confirm that NVP increased H3K9me3 at specific loci, we performed CUT&Tag profiling for this mark in NVP- and DMSO-treated cells for 12 hours, as in our bulk RNA-seq experiments ( Figure 4I, J ). Differential peak calling revealed 2060 loci with elevated H3K9me3 signal under NVP treatment. Intersecting these NVP-induced H3K9me3 peaks with the RNA-seq downregulated genes, 150 genes emerged as candidates whose promoters or enhancers gained repressive methylation concomitant with decreased expression ( Figure 4K ). KEGG analysis of these 150 genes again enriched “Viral life cycle” “MAPK signaling pathway” and “Inflammatory bowel disease” confirming that NVP specifically targeted both antiviral and inflammatory pathways at the epigenetic level ( Figure 4L ).

To test whether H3K9me3 deposition was required for the antiviral activity of NVP, we pretreated cells with Chaetocin, an inhibitor of H3K9 methyltransferases, before NVP exposure and viral infection. Western blots demonstrated that Chaetocin effectively prevented the NVP-induced increase in H3K9me3 ( Figure 4M ). Functionally, flow cytometry showed that VSV-GFP infected cells were significantly higher NVP and Chaetocin co-treated groups compared to NVP alone ( Figure 4N ), indicating that blockade of H3K9 methylation diminished the antiviral efficacy of NVP. Finally, we then assessed the impact of H3K9me3 inhibition on the performance of NVP in vivo ( Figure 4O ). In the VSV infection model and LPS-induced inflammation model, survival analysis demonstrated that the addition of Chaetocin abolished the antiviral and anti-inflammation effects of NVP. Addtionally, Chaetocin co-treatment specifically reversed anti-inflammatory actions of NVP, and Chaetocin itself did not non-specifically enhance viral replication or cytokine induction ( Supplementary Figure S2C-F ).

Collectively, NVP induced H3K9me3 deposition at key proinflammatory and pro-viral gene loci, leading to chromatin compaction, reduced accessibility, and transcriptional repression. Inhibition of H3K9 methylation by Chaetocin reversed these epigenetic changes and undermined NVP’ s antiviral and anti-inflammatory efficacy both in vitro and in vivo, confirming that H3K9me3-mediated chromatin remodeling was essential for the dual therapeutic actions of NVP.

Rabbit anti-COX2 antibodies were purchased from Wanleibio. Rabbit antibodies against iNOS (catalog no. 80517-1-RR) and Beta Tubulin (catalog no. 10094-1-AP), along with mouse antibodies targeting the DYKDDDDK tag (catalog no. 66008-4-Ig), were procured from Proteintech. Additional antibodies, including rabbit phospho-NF-κB p65 (Ser536) (93H1) (catalog no. 3033), NF-κB p65 (D14E12) XP (catalog no. 8242), phospho-IκBα (Ser32) (14D4) (catalog no. 2859), IκBα (44D4) (catalog no. 4812) and phospho-IKKα (Ser176)/IKKβ (Ser177) (C84E11) (catalog no. 2078), were obtained from Cell Signaling Technology. Secondary detection utilized HRP-conjugated Affinipure Goat Anti-Rabbit IgG (H+L) (catalog no. SA00001-2) and HRP-conjugated Affinipure Goat Anti-Mouse IgG (H+L) (catalog no. SA00001-1), also from Proteintech. Lipopolysaccharides (LPS) (catalog no. HY-D1056) and NVP-BVU972 (catalog no. HY-15456) were sourced from MedChemExpress. Dimethyl sulfoxide (DMSO) (catalog no. D8371) was purchased from Solarbio.

The RAW264.7 (RRID: CVCL_0493), HeLa (RRID: CVCL_0030), HT1080 (RRID: CVCL_0316), HT29 (RRID: CVCL_0320), 17-Cl1 (RRID: CVCL_VT75) and Vero (RRID: CVCL_0059) cell lines were sourced from the American Type Culture Collection (ATCC). Cells were authenticated by STR profiling and routinely tested negative for mycoplasma contamination using Mycolor One-Step Mycoplasma Detector (Vazyme, D201-01). RAW264.7, HT1080, 17Cl-1 and Vero were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (03.1002C, EallBio), while HeLa and HT29 were maintained in RPMI 1640 (03.4001C, EallBio). All culture media were supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin for optimal cell growth and maintenance.

Cells at 70-80% confluence were infected with the following viruses at the indicated multiplicity of infection (MOI), vesicular stomatitis virus (VSV, MOI = 0.1), herpes simplex virus 1 (HSV-1, F strain, MOI = 0.5), encephalomyocarditis virus (EMCV, MOI = 0.1), mouse hepatitis virus (MHV, A59 strain, MOI = 0.1) and GFP-expressing viruses, including VSV (MOI=0.1), HSV (MOI=0.1) and vaccinia virus (VACV, MOI=0.1).

Virus stocks were prepared as follows, VSV Indiana strain and VSV-GFP, generously provided by J. Rose (Yale University), were propagated in Vero cells. HSV-1 strain 17, VSV-GFP, HSV-GFP and VACV-GFP were gifts from Zhengfan Jiang (Peking University), and were cultured in the same cell line for viral amplification. MHV-A59 strain (ATCC VR-764) and EMCV (ATCC VR-129B) were obtained from ATCC and propagated in 17Cl-1 and Vero cells, respectively.

Wild-type (WT) C57BL/6J mice were purchased from the Department of Laboratory Animal Science, Peking University Health Science Center. All animal care and procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals by the Chinese Association for Laboratory Animal Science. The protocols were approved by the Animal Care Committee of Peking University Health Science Center (permit number: LA 2016240). Mice were bred and housed under specific pathogen-free conditions at the Laboratory Animal Center of Peking University. Only mice aged 6 to 8 weeks were used in the experiments.

Age- and sex- matched C57BL/6J littermates were used for all in vivo experiments. Six-week-old mice were infected with different viruses at a dose of 1×10⁸ plaque-forming units (PFU) per mouse via intraperitoneal injection (IP), and immediately received the first IP dose of NVP (20 mg/kg) or vehicle control, with subsequent treatments administered every other day. Survivals were monitored daily.

We used C57BL/6J mice because they were a well-characterized, immunocompetent inbred strain widely employed in inflammatory and infection models, including LPS-induced sepsis and viral challenge.

This study employed lipopolysaccharide (LPS)-induced sepsis models to investigate systemic inflammatory responses characteristic of sepsis. Male C57BL/6J mice (6–8 weeks old; 20–25 g) were randomized into two groups (n=12 per group): LPS+PBS only and LPS+NVP. A solution of LPS was prepared in sterile saline at a concentration of 20 mg/kg body weight for the WT C57BL/6J mice and administered intraperitoneally using sterile insulin syringes. After LPS injection, a subset of mice received an intraperitoneal injection of NVP. At the same time, the control group was injected with an equivalent volume of PBS. Survival was monitored for 48 hours, blood and lung tissues were subsequently collected at 12 hours post-LPS for downstream analyses.

The tissues were quickly placed in cold saline solution and rinsed after they were collected, then fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin prior to sectioning at 5 mm, and sections were stained with hematoxylin and eosin.

The total RNA was extracted using TRIzol Reagent (TIANGEN, A0123A01) following the manufacturer’s protocol. Complementary DNA (cDNA) was synthesized from total RNA using HiScript II RT SuperMix (Vazyme, R223-01). mRNA expression levels of viral and immune related genes were analyzed by quantitative reverse transcription PCR (RT-qPCR) using Taq Pro Universal SYBR qPCR Master Mix (Vazyme). The relative expression levels of target genes were normalized to the housekeeping gene Actb, and data were analyzed using the comparative Ct (2^−ΔΔCt) method. The primer sequences used for qPCR analysis in the study can be found in Supplementary Table S1 .

The cells were harvested and lysed using ice-cold RAPI buffer (MedChemExpress, HY-K1001), which was supplemented with protease inhibitors for 30 min. Cell lysates were mixed with protein loading buffer and denatured at 100°C for 5 min, followed by centrifugation at 12,000 rpm for 10 min at 4°C. Protein samples were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a nitrocellulose membrane (Beyotime Biotechnology, FFN08) by electroblotting. Membranes were blocked with 5% non-fat milk at room temperature for 1 h and subsequently immunoblotted with indicated primary antibodies (1:1000) at 4°C overnight, followed by incubation with HRP-conjugated secondary antibodies (1:10000) at 37°C for 1 h. Protein signals were detected by using an enhanced ECL system (EallBio, 07.10009-50) according to the manufacturer’s instructions.

Target cells were seeded into a 96-well plate at a density of 2×10³ cells per well in 100 μL of complete medium. Cells were treated with increasing concentrations of NVP-BVU972 (MedChemExpress, HY-15456) and incubated at 37°C in a humidified 5% CO² incubator for 48 hours. And then, we added 10 μL of CCK-8 reagent (YEASEN, 40203ES60) to each well and gently shook the plate to mix. After incubation at 37°C for 1 hour, we measured the absorbance at 450 nm using a microplate reader.

Confluent monolayers of Vero cells were plated in 24-well plates and infected with 10-fold serial dilutions of virus for 1 hours at 37°C to allow adsorption. The inoculum was then removed, and cells were washed twice with PBS before being overlaid with DMEM containing 0.5% carboxymethyl cellulose (CMC) (Sigma, 419338). Plates were incubated at 37°C for 48 hours to allow plaques to form. The cells were then fixed with 4% paraformaldehyde and stained with 1% crystal violet to visualize plaques. Viral titer was calculated as PFU/mL based on the plaque counts, dilution factor, and inoculum volume.

Cells were infected with GFP-tagged viruses at MOI of 0.1 for 12 hours. After treatment with NVP, cells were visualized under a Nikon fluorescence microscope equipped with a 488 nm excitation filter and a 510 nm emission filter. Images were captured at a magnification of 100×.

After infection, cells were washed twice with phosphate-buffered saline (PBS) to remove residual virus, followed by trypsinization to generate a single-cell suspension. The cells were collected by centrifugation at 1600 rpm for 5 minutes, and the supernatant was discarded. The resulting cell pellets were resuspended in PBS and transferred to flow cytometry tubes for subsequent analysis. To evaluate the infection efficiency, the percentage of GFP-positive cells was quantified using a flow cytometer (BD Biosciences).

Total RNA was extracted with the TIANGEN A0123A01 high-throughput kit. Quality control, library preparation, sequencing and downstream analysis were performed by Suzhou GENEWIZ following their standard protocols. Gene counts were generated by featureCounts v2.0.0 and normalized to FPKM. Differential expression was assessed in DESeq2 v1.38.3 using |log₂FC|>2 and p<0.05.

We performed ATAC-seq using the Hyperactive ATAC-Seq Library Prep Kit (TD711, Vazyme Biotech). Tn5 transposase, preloaded with sequencing adapters, was incubated with isolated nuclei to insert adapters into open chromatin. After PCR amplification with indexed primers, libraries were sequenced by GENEWIZ (Suzhou, China), yielding a genome-wide map of chromatin accessibility.

Cells were captured on ConA-coated magnetic beads and permeabilized with digitonin. After incubation with a primary antibody and a Protein A/G–linked secondary antibody fused to Tn5 transposase, Mg²⁺ activation triggered targeted DNA cleavage and simultaneous adapter insertion. The resulting fragments were PCR-amplified using the Transgen CUT&Tag Library Prep Kit (Vazyme, KP172) and sequenced by GENEWIZ (Suzhou, China).

Raw ATAC-seq and CUT&Tag reads (paired-end) from GENEWIZ were first trimmed with Trim-Galore v0.6.4, then aligned to the mouse mm10 genome using Bowtie2 v2.3.5.1 (–very-sensitive -X 2000) and piped through SAMtools v1.10 to produce sorted, indexed BAM files. Peaks were called for each sample with MACS3 v3.0.0a5 (default settings), yielding per-sample BED files (56).

For ATAC-seq, individual peak sets were merged across all samples with Bedtools v2.31.1, and genome-wide coverage tracks were generated in RPKM units using deepTools v3.3.2 (bamCoverage –binSize 100 –normalizeUsing RPKM –effectiveGenomeSize 2864785220 –ignoreForNormalization chrM –extendReads). CUT&Tag signal within called peaks was normalized similarly, and treatment vs. input ratios were computed with bamCompare (–binSize 145 –normalizeUsing RPKM –effectiveGenomeSize 2864785220 –ignoreForNormalization chrM –extendReads –scaleFactorsMethod None) (56).

Differential accessibility/enrichment analyses were performed in R using the csaw package v1.38.0, and peaks were annotated with ChIPseeker v1.34.1. Final track visualization was carried out in IGV v2.17.4.

Statistical analyses were conducted using Student’s t-test for two-group comparisons (indicated by square brackets) and one-way ANOVA for multi-group experiments. Survival was evaluated by Kaplan–Meier analysis. Data are shown as mean ± SEM (n≥3). N.S.,not significant, p > 0.05, *p < 0.05, ** p < 0.01, ***p < 0.001, ****p < 0.0001.