Cross‑kingdom antiviral action of Scutellaria baicalensis Georgi and Chrysanthemum indicum L. miRNAs: A novel strategy against duck viral hepatitis Authors: Mengxin Hu, Zhipeng Chen, Kexin Wang, Xu Liu, Shuoyan Guan, Xiang Fu, Bolin Gu, Xinyu Liu, Zefei Yan, Jiaguo Liu Published: 2026 ABSTRACT Duck viral hepatitis (DVH) requires novel therapeutic strategies. This study investigated antiviral miRNAs in separate decoctions of Scutellaria baicalensis Georgi and Chrysanthemum indicum L. High-throughput sequencing identified 14 highly expressed miRNAs, which were chemically synthesized for functional validation. CCK8 assay quantified cytoprotection in duck embryonic hepatocytes (DEHs) and RT-qPCR measured DHAV-1 gene expression, enabling the selection of the most effective miRNAs. Two candidates, novel-miR2 (from S. baicalensis) and novel-miR10 (from C. indicum), were selected for detailed study. RT-qPCR along with fluorescent probes assessed miRNA concentrations, stability, and their effects on DEHs biological characteristics. In vivo evaluations included mortality rates, liver histopathology/function, and markers of inflammation/apoptosis. Both miRNAs exhibited differential concentrations in fresh versus dried herbs and demonstrated good thermal stability. In vitro, miRNAs reduced DHAV-1 replication, apoptosis, necrosis, and ROS, while enhancing cell viability. Oral administration improved survival rates in infected ducklings, lowered serum liver enzymes, and alleviated hepatic damage. Specifically, novel-miR2 predominantly reduced apoptosis, whereas novel-miR10 alleviated inflammation. Collectively, novel-miR2 and novel-miR10 suppressed DHAV-1-induced hepatitis, exerting hepatoprotective effects. These findings provided new evidence for clarifying the active constituents of herbal decoctions and their miRNA‑mediated cross‑kingdom antiviral mechanism. Introduction Duck viral hepatitis (DVH) is a severe acute disease in ducklings, characterized by rapid spread and over 95% mortality, posing a critical threat to poultry production(Ou et al., 2022). Clinical signs include high fever, convulsions, and opisthotonos, with hepatitis and hemorrhage as the primary pathological manifestations. Currently, no highly effective treatment exists for this WOAH-notifiable disease. In China, duck hepatitis A virus type 1 (DHAV-1) is the predominant causative agent(Zhang et al., 2020). DHAV-1, an enveloped spherical virus, belongs to the genus Avihepatovirus (family Picornaviridae)(Yugo et al., 2016). Its genome is a positive-sense, single-stranded RNA (∼7.6 – 8.0 kb) featuring a 5’UTR, an open reading frame (ORF), a 3’UTR, and a poly-A tail. Translation produces a polyprotein that is cleaved into structural proteins (VP0, VP3, VP1) and multiple non-structural proteins. The latter include 2A (with subtypes 2A1, 2A2, 2A3), 2B, 2C, 3A, 3B, 3C, and 3D(Jiang et al., 2024; Zhang et al., 2024). Among these, proteins 2A2, 2C, 3A, and 3D interact with host PI4KB to enhance viral replication. Protein 2C also possesses NTPase activity(Li et al., 2021; Li et al., 2024). Separately, protein 2B disrupts cellular homeostasis, promotes NF-κB/NLRP3-mediated inflammation, and induces autophagy(Liu et al., 2021; Mao et al., 2024). Protein 3C, moreover, antagonizes type I interferon signaling and facilitates viral replication(Lai et al., 2021; Zhang et al., 2024). Scutellaria baicalensis Georgi and Chrysanthemum indicum L. are widely distributed traditional Chinese medicinal herbs(Wang et al., 2025b; Xiang et al., 2021). Both possess bitter taste and cold nature, serving to clear liver heat and detoxify. From S. baicalensis, 126 small molecules and 6 polysaccharides have been isolated. Its small molecules fall into four structural categories, which comprise free flavonoids, flavonoid glycosides, phenethyl glycosides, and other compounds(Wang et al., 2018). These constituents demonstrate diverse pharmacological activities, including anti-inflammatory, antibacterial, antiviral, antioxidant, and anticancer effects(Song et al., 2020; Yu et al., 2023; Zhang et al., 2025). From C. indicum, over 190 chemical constituents have been identified, such as flavonoids, terpenoids, phenylpropanoids, polysaccharides, and trace elements(Shao et al., 2020). Extracts or compounds from this herb exhibit anti-inflammatory, antioxidant, antimicrobial, anticancer, immunomodulatory, and hepatoprotective properties(Kim et al., 2012; Zhang et al., 2023a; Zhang et al., 2022). Our prior work indicated that active components like baicalin and polysaccharides from both herbs could mitigate DHAV-1-induced DVH(Chen et al., 2018b; Hu et al., 2025; Qiu et al., 2022). Beyond these chemicals, their decoctions also contain biological components including miRNAs. Therefore, this study aimed to investigate the potential therapeutic role of these herbal miRNAs against DVH. Traditional antiviral research on herbal medicines has primarily focused on small molecules and polysaccharides, which typically act through direct viral inhibition or immune modulation(Paul et al., 2021; Prajapati et al., 2022; Rizwan et al., 2022). In contrast, plant-derived miRNAs have emerged as novel bioactive agents, offering fresh insights into pharmacological principles. These miRNAs exhibit distinct characteristics compared to classical constituents like flavonoids or polysaccharides. Certain plant miRNAs can be absorbed by animals and demonstrate notable stability in circulation, enabling cross-kingdom gene regulation(Li et al., 2023). As key regulatory molecules, their sequences are highly conserved(Chen and Rechavi, 2021). Pioneering work in 2012 first identified abundant plant miRNAs in human and animal serum. Specifically, rice-derived MIR168a was shown to enter the murine bloodstream upon ingestion and target hepatic genes to influence cholesterol metabolism(Zhang et al., 2012). This finding rapidly stimulated extensive research. Moreover, miRNAs possess precise targeting capabilities. Subsequent studies revealed that honeysuckle MIR2911 specifically inhibits influenza virus in animal models(Zhou et al., 2015). Furthermore, miRNAs from dietary plants and herbal medicines have been detected in animal blood, tissues, and milk, displaying remarkable stability(Chen et al., 2021; Chin et al., 2016; Lukasik et al., 2017). Additionally, miRNA sources are clearly traceable and amenable to synthetic validation. High-throughput sequencing allows precise identification, while chemical synthesis enables rapid functional testing and optimization, thus supporting quality control and standardized production. By comparison, polysaccharide or flavonoid extracts are often complex mixtures, complicating the identification of active monomers and structure-activity analysis. However, therapeutic application of miRNAs faces considerable challenges. Most encounter stability and delivery issues, being vulnerable to RNase degradation and exhibiting low oral bioavailability, which necessitates protective formulations or delivery systems(Kara et al., 2022). Furthermore, miRNA levels in herbs vary significantly with growth conditions, harvest time, and processing methods, while stringent extraction and storage requirements hinder their standardization as active ingredients(Wei et al., 2015). Current research primarily focuses on model plants and food crops. For traditional medicinal plants, key aspects are not yet systematically characterized. These include the miRNAome of their decoctions, along with pharmacokinetics and antiviral mechanisms of these miRNAs. Therefore, this study examined S. baicalensis and C. indicum, herbs traditionally used against viral hepatitis. We investigated whether specific miRNAs enriched in their decoctions could exert direct antiviral activity within duck hosts to combat DHAV-1 infection. This work aimed to identify candidate miRNAs and provide a foundation for developing nucleic acid-based antiviral drugs from traditional Chinese medicine. Materials and methods Ethics statement All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) and the Nanjing Agricultural University IACUC (Approval No. NJAU.No20241121238). To minimize distress, ducklings received intramuscular inoculations and were then humanely euthanized with diethyl ether at 4, 8, or 54 hpi (hours post injection) or upon exhibiting opisthotonos or convulsions. miRNA sequencing of S. baicalensis and C. indicum decoctions Sample Preparation and miRNA Quality Control. Three 100-g portions of each fresh herb were boiled vigorously in 1 L of deionized water, then simmered for 30 minutes. After brief cooling, the decoctions were filtered through gauze. The filtrate was concentrated by boiling to a final volume of 100 mL. The three S. baicalensis decoctions were labeled SD-1 to SD-3, and the three C. indicum decoctions were labeled CID-1 to CID-3. All samples were submitted to Biomarker Technologies (Beijing, China) for miRNA extraction, quality assessment, and sequencing library construction. Library Construction and Sequencing. Small RNAs were enzymatically ligated with 3’ and 5’ adapters, then reverse-transcribed into cDNA. Subsequent PCR amplification with IndexX Primers simultaneously enriched the library and added sample indices. Target fragments were size-selected and excised using PAGE. Following gel extraction and purification, the resulting libraries were subjected to quality control via Qsep-400. Finally, qualified libraries were sequenced on an Illumina NovaSeq 6000 platform. Data Quality Control and Analysis. Raw sequencing data underwent base calling to generate FASTQ files. These raw reads were then processed to remove adapter sequences and low-quality bases. Reads shorter than 18 nt or longer than 30 nt were discarded, along with those containing ≥10% ambiguous bases (N). The remaining high-quality sequences constituted the clean reads. Clean reads were sequentially aligned against the Silva, GtRNAdb, Rfam, and Repbase databases using Bowtie to filter out non-coding RNAs (rRNA, tRNA, snRNA, snoRNA) and repetitive sequences. The resulting unannotated reads were subsequently mapped to the reference genome with Bowtie to obtain mapped reads. miRNA Identification. For known miRNAs, mapped reads were aligned to mature miRNA sequences in miRBase (v22), extending 2-nt upstream and 5-nt downstream. Reads with one or fewer mismatches to these reference regions were classified as known miRNAs. Since miRNAs primarily originate from intergenic regions, introns, or antisense strands and are processed from hairpin precursors by Dicer/DCL enzymes, unmapped sequences were analyzed as potential novel miRNAs using miRDeep2. This software first identified potential genomic precursor loci. It then employed a Bayesian model to score candidates based on their read distribution (consistent with mature, star, and loop sequences), and on precursor folding energy calculated by RNAfold and randfold. Predicted novel miRNAs were assigned names in the format “novel-miR” followed by sequential number. miRNA Quantification. miRNA expression was normalized using the transcripts per million (TPM) method. The calculation followed the formula: TPM = (Read count × 1,000,000) / Mapped Reads, where Read count denotes reads aligned to specific miRNA and Mapped Reads represents the total mapped reads for all miRNAs. Subsequently, miRNAs with TPM values >10,000 were chemically synthesized (Sangon, Shanghai, China) for further experiments. DEHs preparation and viral challenge A 0.20% trypsin solution was prepared in sterile D-Hank’s buffer. Complete medium comprised 89% DMEM (Yuanpei, Shanghai, China), 10% heat-inactivated neonatal bovine serum (NBS; Sijiqing, Zhejiang, China), and 1% penicillin‑streptomycin (Beyotime, Shanghai, China). Maintenance medium contained 98% DMEM, 1% heat-inactivated serum, and 1% penicillin‑streptomycin. Duck embryonic hepatocytes (DEHs) were isolated from 14‑day‑old embryos (Chaoyang Hatchery, Anhui, China). Following aseptic dissection, hepatic tissue was excised, rinsed with D‑Hank’s buffer, and digested with 0.2% trypsin at 37°C. The dispersed cells were resuspended in complete medium at optimal density and seeded into culture dishes. After 24 h of incubation at 37°C with 5% CO₂, the medium was removed. Cells were infected with DHAV‑1 (20 TCID₅₀; LQ2 strain, Shandong Poultry Research Institute, GenBank: KF826748.1) for 2 h(Ma et al., 2015). The inoculum was then replaced with fresh maintenance medium, and cultures were incubated for an additional 24 h to generate DHAV‑1‑infected DEHs for subsequent experiments. Cell viability assay DEHs were seeded into 96‑well plate at 5 × 10⁵ cells/mL in complete medium and incubated for 24 h. Cells were assigned to four groups: uninfected cells control (CC), DHAV‑1‑infected viral control (VC), miRNA‑treated infected cells, and negative control treated with ncRNA. In particular, the ncRNA was a scrambled non-targeting sequence provided by Sangon Biotech, and it showed no homology to either the test miRNA or the viral genome. Chemically synthesized miRNAs (Tab. 1) were complexed with RNATransMate (Sangon) in DMEM to generate transfection mixtures at concentrations of 2.5, 5, 10, 20, and 30 nmol/L. Two hours post‑infection, treatment and ncRNA groups received maintenance medium containing the respective miRNA complexes, whereas the VC group received maintenance medium alone. After another 24 h, the medium was exchanged for CCK‑8 working solution (Abbkine, Wuhan, China) and incubated for 1 h. Absorbance was measured with microplate reader, and cell viability was calculated as: (ODTreatment -ODBlank)/(ODControl -ODBlank) × 100%. The miRNA concentration yielding high viability was selected for further experiments.Table 1Primer sequences.Table 1 dummy alt textGenesForward Primer (5’-3’)Reverse Primer (5’-3’)β-actinCTTTCTTGGGTATGGAGTCCTGTGATTTTCATCGTGCTGGGTDHAV-1GGACCTCCAAGTTCCAGACAGGAGTGTGCTACCATTGCTGnovel-miR2CAGTAGCGTTCTGATCAAGGTTTCAATCCAGTGCAGGGTCCGAGGnovel-miR10ACGACAATTCCCGGCAACGGATCCAGTGCAGGGTCCGAGGBcl-2GGATGCCTTCGTGGAGTTGTATGGGATGCAAGCTCCCACCAGAACCCaspase-3TGAGGCAGACAGTGGACCAGATGGCATTCCGCCAGGAGTAATAGCCIL-6AAGCATCTGGCAACGACGATAAGGTGTGAGGAGGGATTTCTGGGTAGCIL-10CAACCTGCTGCTGAGCCTGAAGCGCCTTGTAGATGCCGTTCTCG Table 1Primer sequences.Table 1 dummy alt textGenesForward Primer (5’-3’)Reverse Primer (5’-3’)β-actinCTTTCTTGGGTATGGAGTCCTGTGATTTTCATCGTGCTGGGTDHAV-1GGACCTCCAAGTTCCAGACAGGAGTGTGCTACCATTGCTGnovel-miR2CAGTAGCGTTCTGATCAAGGTTTCAATCCAGTGCAGGGTCCGAGGnovel-miR10ACGACAATTCCCGGCAACGGATCCAGTGCAGGGTCCGAGGBcl-2GGATGCCTTCGTGGAGTTGTATGGGATGCAAGCTCCCACCAGAACCCaspase-3TGAGGCAGACAGTGGACCAGATGGCATTCCGCCAGGAGTAATAGCCIL-6AAGCATCTGGCAACGACGATAAGGTGTGAGGAGGGATTTCTGGGTAGCIL-10CAACCTGCTGCTGAGCCTGAAGCGCCTTGTAGATGCCGTTCTCG Primer sequences. Reverse-transcription quantitative PCR (RT‑qPCR) for DHAV-1 gene expression DEHs were seeded in 6‑well plates at 1 × 10⁶ cells/mL and incubated for 24 h. Cells were grouped as described in Section Cell Viability Assay and treated with miRNA transfection complexes at 2.5, 5, or 10 nmol/L. Total RNA was extracted using TRIzol reagent (Vazyme, Nanjing, China), quantified by NanoDrop spectrophotometer (Thermo, Waltham, USA) and reverse-transcribed using first-strand cDNA synthesis kit (Vazyme). SYBR Green-based qPCR reactions were performed on 7500 Real-Time System (ABI, Waltham, USA). Gene expression quantification employed the 2−ΔΔCT method. Primer sequences were shown in Table 1. Based on cell viability and DHAV‑1 gene expression, novel‑miR2 (from S. baicalensis) and novel‑miR10 (from C. indicum) were selected for further study and submitted to miRBase for formal naming. miRNA target gene prediction The complete DHAV‑1 genome sequence was retrieved from NCBI (Gene ID: 5076728). Potential miRNA‑binding sites were predicted using the RNAhybrid algorithm. Predictions followed two stringent criteria: 1) the minimum free energy (MFE) of binding must be ≤‑20 kcal/mol; 2) the seed region of the miRNA must align perfectly with the target sequence. Apoptosis and necrosis assay DEHs were seeded in 6‑well plates at 1 × 10⁶ cells/mL and cultured for 24 h. Following the grouping scheme in Section Cell Viability Assay, cells were assigned to control (CC), virus control (VC), ncRNA group, and two treatment groups (novel‑miR2 and novel‑miR10). For staining, cells were incubated with 5 μL Hoechst 33342 and 5 μL propidium iodide in 1 mL of buffer at 4°C for 30 min (Beyotime). After one wash with PBS, samples were observed under fluorescence microscope. Apoptotic and necrotic cell populations were quantified based on fluorescence intensity using ImageJ software. Intracellular reactive oxygen species (ROS) detection Cells were grouped and treated following Section Apoptosis and Necrosis Assay. The maintenance medium was removed, and cells were incubated with 10 μM DCFH‑DA (1 mL/well) at 37°C for 20 min. After three washes with DMEM to remove extracellular probe, intracellular DCF fluorescence was immediately visualized using fluorescence microscopy. Relative ROS levels were quantified from the images with ImageJ software. Animals experimental design and sampling Pharmacokinetic Assessment of miRNA. 60 one‑day‑old ducklings (Chaoyang Hatchery) were acclimated for seven days with warm water and commercial feed. They were then randomly divided into two equal groups of thirty. Prior to administration, five ducklings per group were sampled under anesthesia to collect baseline plasma and liver specimens. Following oral gavage of novel‑miR2 or novel‑miR10 (100 pmol per duckling), five animals from each group were sampled at 0.5, 1, 3, 6, and 12 hours post‑administration. miRNA Efficacy Assessment for DVH. 150 one-day-old ducklings (Chaoyang Hatchery) were housed and acclimatized for seven days, then randomly allocated into five groups: blank control (BC), virus control (VC), negative control (ncRNA), and two miRNA treatment groups (novel‑miR2 and novel‑miR10). All ducklings except the BC group received intramuscular injection of 0.2 mL DHAV‑1 (100 TCID₅₀). One hour post injection (hpi), all groups were provided normal drinking water. Additionally, the VC and BC groups received sterile water via oral gavage, while the ncRNA and miRNA groups were gavaged with 100 pmol per duckling of the respective miRNA. This gavage procedure was repeated daily for three consecutive days. The BC group was maintained separately under identical conditions and injected with 0.2 mL sterile saline one hour before the first gavage. At 4 hpi (early phase), 8 hpi (acute phase), and 54 hpi (stable phase) of DVH, five ducklings per group were anesthetized for serum and liver collection. Survival analysis and hepatic pathology 150 one-day-old ducklings were grouped according to section miRNA Efficacy Assessment for DVH. Survival was monitored every 12 hours until mortality ceased. All deceased ducklings were necropsied, and hepatic lesions were assessed by gross examination. A lesion severity score (0–5) was assigned. The scoring criteria were: 0, no lesions; 1, discoloration without hemorrhage; 2, <20 petechiae; 3, >20 petechiae; 4, diffuse petechiae with <50% ecchymosis; 5, >50% ecchymotic coverage(Wang et al., 2025a). Hepatic histopathological evaluation Liver specimens fixed in paraformaldehyde were paraffin‑embedded and sectioned. The resulting sections were stained with hematoxylin and eosin (H&E) to visualize cellular and nuclear morphology for assessment of hepatic histopathological alterations. Determination of serum hepatic function parameters Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (ALP) were quantified in ducklings using automated biochemistry analyzer (Beckman Coulter, California, USA). Determination of IL-10, IFN-γ, and DHAV-1 levels in liver tissues Liver tissues (0.1 g) were homogenized in 1 mL PBS at 4°C. Following homogenization, samples were centrifuged to collect the supernatant. Concentrations of IL‑10, IFN‑γ, and DHAV‑1 in liver homogenates were then determined using specific ELISA kits according to the manufacturer’s instructions. RT‑qPCR analysis of apoptosis- and inflammation-related gene expression Total RNA was extracted from cells or tissues using TRIzol reagent. Subsequent procedures for cDNA synthesis and quantitative PCR followed the protocol detailed in Section 2.5. All primer sequences were listed in Table 2.Table 2The sequences of miRNA with TPM >10,000 in decoctions of S. baicalensis and C. indicum.Table 2 dummy alt textSpeciesmiRNAsSense (5’-3’)ncRNAUUCUCCGAACGUGUCACGUTTS. baicalensisnovel-miR22UUUGGAUUGAAGGGAGCUCUS. baicalensisnta-miR319bUUGGACUGAAGGGAGCUCCCUS. baicalensisnovel-miR2UUCUGAUCAAGGUUUCAAUUGGACUS. baicalensissly-miR319c-3pUUGGACUGAAGGGAGCUCCUUS. baicalensisnovel-miR16CGAAUCCGGCAGGUCGGACCAS. baicalensisnovel-miR26ACCUAGCUCUGAUACCAAUUUAS. baicalensisnovel-miR41CGAACGAUCGGCUCAUCUUUGAGCAS. baicalensisnovel-miR14CUUGUGUGAGACUGUCUCAUGGAC. indicumnovel-miR10UUCCCGGCAACGGAACCAC. indicumnovel-miR34CUUCGUAGAUCUCUGUCCAGC. indicumnovel-miR71ACGGUCUUUUAGAGCUUCUUAC. indicumnovel-miR73CGUCGUAAUUUGUUGCUGAC. indicumnovel-miR111GUCCCACGGAAGGCGCCAC. indicumnovel-miR47UUAUGAAGGUAGUCUAACUCA Table 2The sequences of miRNA with TPM >10,000 in decoctions of S. baicalensis and C. indicum.Table 2 dummy alt textSpeciesmiRNAsSense (5’-3’)ncRNAUUCUCCGAACGUGUCACGUTTS. baicalensisnovel-miR22UUUGGAUUGAAGGGAGCUCUS. baicalensisnta-miR319bUUGGACUGAAGGGAGCUCCCUS. baicalensisnovel-miR2UUCUGAUCAAGGUUUCAAUUGGACUS. baicalensissly-miR319c-3pUUGGACUGAAGGGAGCUCCUUS. baicalensisnovel-miR16CGAAUCCGGCAGGUCGGACCAS. baicalensisnovel-miR26ACCUAGCUCUGAUACCAAUUUAS. baicalensisnovel-miR41CGAACGAUCGGCUCAUCUUUGAGCAS. baicalensisnovel-miR14CUUGUGUGAGACUGUCUCAUGGAC. indicumnovel-miR10UUCCCGGCAACGGAACCAC. indicumnovel-miR34CUUCGUAGAUCUCUGUCCAGC. indicumnovel-miR71ACGGUCUUUUAGAGCUUCUUAC. indicumnovel-miR73CGUCGUAAUUUGUUGCUGAC. indicumnovel-miR111GUCCCACGGAAGGCGCCAC. indicumnovel-miR47UUAUGAAGGUAGUCUAACUCA The sequences of miRNA with TPM >10,000 in decoctions of S. baicalensis and C. indicum. Absolute quantitative RT‑qPCR for miRNA concentration miRNA Extraction from S. baicalensis and C. indicum Samples. Three 1‑g portions of each dried herb were ground into powder under liquid nitrogen and set aside for direct RNA extraction. Separately, decoctions were prepared from both fresh and dried herbs. Three 100‑g portions of each fresh herb and three 100‑g portions of each dried herb were individually boiled in 1 L of deionized water. Each mixture was simmered for 30 min, cooled, filtered, and the filtrate concentrated to 100 mL. For RNA extraction, 100 mg of each reserved powder or 100 μL of each decoction was lysed in 1 mL RNA‑Solv Reagent at room temperature for 3 min. miRNA was subsequently purified according to the manufacturer’s protocol (Omegabiotek, Norcross, USA). miRNA Extraction from Plasma and Liver Samples. For plasma miRNA extraction, 200 µL of plasma was mixed with 900 µL MZA lysis buffer by inversion for 30 s and incubated at room temperature for 5 min. miRNA was then isolated using the miRcute Serum/Plasma miRNA Isolation Kit according to the manufacturer’s instructions (TIANGEN, Beijing, China). For liver miRNA extraction, 50 mg of tissue was homogenized in 1 mL RNA‑Solv Reagent and incubated for 3 min at room temperature. Subsequent miRNA purification was performed following the corresponding kit protocol (Omegabiotek). Heat and RNase A Stability Assays. The stability of chemically synthesized novel‑miR2 and novel‑miR10 was assessed under heat and nuclease challenge. For the heat assay, miRNAs were heated to 100°C and held at 80°C. For the nuclease assay, parallel samples were treated with 1 µg/mL RNase A. Aliquots from both assays were collected at 1 h and 12 h time points, stored at -80°C, and subsequently used for miRNA quantification. RT‑qPCR for Novel‑miR2 and Novel‑miR10 Concentration. Standard curves for novel‑miR2 and novel‑miR10 were constructed via eight‑step, ten‑fold serial dilutions of synthetic standards. miRNA samples extracted in sections 2.15.1–2.15.3 were quantified using NanoDrop spectrophotometer and then reverse‑transcribed with stem‑loop miRNA cDNA synthesis kit (Vazyme). Subsequent SYBR Green qPCR was performed on Gentier 96E/96R PCR Analysis System (TIANLONG, Xi’an, China). miRNA concentrations in experimental samples were calculated by interpolation from the corresponding standard curve. All primer sequences were provided in Table 2. Statistical analysis Data were presented as mean ± SEM. One‑way ANOVA followed by GraphPad Prism 9 was employed for comparisons among multiple groups, with P < 0.05 considered statistically significant. Survival curves were analyzed using the Kaplan‑Meier method. Results Identification and screening of anti-DHAV-1 miRNAs from herbal decoctions High-throughput sequencing revealed multiple miRNAs with TPM > 10,000 in S. baicalensis and C. indicum decoctions, highlighting those exceeding TPM 100,000 in red (Fig. 1A, B). The miRNA sequences were shown in Table 2. These highly expressed miRNAs were chemically synthesized and screened for anti-DHAV-1 activity in DEHs. Among miRNAs derived from S. baicalensis, novel-miR2, novel-miR16, and novel-miR26 exhibited the strongest cytoprotective effects at 2.5 and 5 nmol/L, as evidenced by maximal cell viability (Fig. 1C). Similarly, novel-miR10 and novel-miR34 from C. indicum conferred the greatest protection at identical concentrations (Fig. 1D). Dose-response experiments demonstrated that both novel-miR2 and novel-miR10 suppressed DHAV-1 gene expression (Fig. 1E, F). Notably, 2.5 nmol/L novel-miR2 failed to reduce viral RNA levels compared to the VC group. Therefore, the 5 nmol/L concentration was selected for both miRNAs in subsequent experiments. The negative control ncRNA showed no antiviral activity throughout the screening process.Fig. 1Highly expressed miRNAs in herbal decoctions and their anti-DHAV-1 activity in DEHs. (A, B) Expression levels of miRNAs with TPM >10,000 in decoctions of (A) S. baicalensis and (B) C. indicum. (C, D) Cell viability after DHAV‑1 infection, measured by CCK‑8 assay in DEHs treated with chemically synthesized miRNAs corresponding to the highly expressed sequences from (C) S. baicalensis or (D) C. indicum. (E, F) DHAV‑1 gene expression in DEHs treated with synthetic miRNAs from (E) S. baicalensis and (F) C. indicum, as determined by RT‑qPCR. Data represent mean ± SEM, n = 3. Distinct letters (a-e) indicate statistically significant differences (p < 0.05) among groups.Fig 1 dummy alt text Highly expressed miRNAs in herbal decoctions and their anti-DHAV-1 activity in DEHs. (A, B) Expression levels of miRNAs with TPM >10,000 in decoctions of (A) S. baicalensis and (B) C. indicum. (C, D) Cell viability after DHAV‑1 infection, measured by CCK‑8 assay in DEHs treated with chemically synthesized miRNAs corresponding to the highly expressed sequences from (C) S. baicalensis or (D) C. indicum. (E, F) DHAV‑1 gene expression in DEHs treated with synthetic miRNAs from (E) S. baicalensis and (F) C. indicum, as determined by RT‑qPCR. Data represent mean ± SEM, n = 3. Distinct letters (a-e) indicate statistically significant differences (p < 0.05) among groups. Physicochemical stability and predicted viral targets of candidate miRNAs To determine the feasibility of oral administration, the stability of these miRNAs was evaluated under various conditions. Their potential mechanism of action was then explored through target prediction. While novel‑miR2 and novel‑miR10 concentrations were comparable between dried powders and their respective decoctions, both were significantly lower than in decoctions from fresh herbs (Fig. 2A). Synthesized miRNAs were challenged by heating or RNase A treatment. Heat exposure for 12 h reduced concentrations by approximately half. Conversely, after 12 h RNase A digestion, only 0.60 nmol/L novel‑miR2 and 0.05 nmol/L novel‑miR10 remained detectable (Fig. 2B). Bioinformatic analysis predicted that each miRNA possesses three target sites within the DHAV‑1 genome, located on the 2A2, 2B, 2C, and 3C genes (Fig. 2C). Prediction scores for their precursor structures were low, with no complementary star strands identified, likely reflecting RNA degradation during high‑temperature decoction. Nonetheless, the enrichment of both miRNAs in decoctions indicated considerable intrinsic stability (Fig. 2D).Fig. 2Analysis of novel‑miR2 and novel‑miR10 abundance, stability, and target prediction. (A) Quantification of novel‑miR2 or novel‑miR10 in S. baicalensis or C. indicum samples by RT‑qPCR. (B) Stability assessment of synthesized miRNAs under high‑temperature or RNase A treatment, measured by RT‑qPCR. (C) Prediction of DHAV‑1 target sites for novel‑miR2 and novel‑miR10 using the RNAhybrid algorithm. (D) Predicted precursor structures and sequences for both miRNAs from miRDeep2 analysis. Data represent mean ± SEM, n = 3. Distinct letters (a-b) indicate statistically significant differences (p < 0.05) among groups.Fig 2 dummy alt text Analysis of novel‑miR2 and novel‑miR10 abundance, stability, and target prediction. (A) Quantification of novel‑miR2 or novel‑miR10 in S. baicalensis or C. indicum samples by RT‑qPCR. (B) Stability assessment of synthesized miRNAs under high‑temperature or RNase A treatment, measured by RT‑qPCR. (C) Prediction of DHAV‑1 target sites for novel‑miR2 and novel‑miR10 using the RNAhybrid algorithm. (D) Predicted precursor structures and sequences for both miRNAs from miRDeep2 analysis. Data represent mean ± SEM, n = 3. Distinct letters (a-b) indicate statistically significant differences (p < 0.05) among groups. Both miRNAs attenuate DHAV-1-induced cytopathology Whether novel-miR2 and novel-miR10 could protect DEHs from DHAV-1-induced cellular damage was next investigated, following confirmation of their stability and predicted viral targets. Hoechst 33342 and PI staining distinguished viable (faint blue), apoptotic (bright blue), and necrotic (bright red) cells. Quantitative analysis revealed that novel-miR2 primarily reduced DHAV-1-induced apoptosis, whereas novel-miR10 alleviated necrosis (Fig. 3A, B). Consistent with these observations, viral infection downregulated the anti-apoptotic marker Bcl-2 and upregulated Caspase-3 expression relative to CC group. Both miRNAs reversed these trends, with novel-miR2 exhibiting more pronounced effects on Caspase-3 gene expression (Fig. 3C). Additionally, the elevated levels of IL-6 and IL-10 following infection were reduced by miRNA treatment; novel-miR10 showed greater suppression of IL-10 specifically (Fig. 3C). Intracellular ROS levels, reflected by green fluorescence intensity, increased after DHAV-1 infection. Treatment with either novel-miR2 or novel-miR10 attenuated ROS accumulation in infected DEHs (Fig. 3D, E).Fig. 3novel‑miR2 and novel‑miR10 mitigate DHAV‑1‑induced cellular injury. (A) Apoptosis and necrosis in DEHs detected by Hoechst 33342 and PI staining (scale bar: 10 µm). (B) Quantification of average fluorescence intensity for red and blue channels. (C) RT‑qPCR showed relative mRNA expression of apoptosis- and inflammation-related genes in DEHs across groups. (D) DCFH-DA assay measured ROS levels in DEHs (scale bar: 50 μm). (E) Quantification of average fluorescence intensity for green channels. Data represent mean ± SEM, n = 3. Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each indicator.Fig 3 dummy alt text novel‑miR2 and novel‑miR10 mitigate DHAV‑1‑induced cellular injury. (A) Apoptosis and necrosis in DEHs detected by Hoechst 33342 and PI staining (scale bar: 10 µm). (B) Quantification of average fluorescence intensity for red and blue channels. (C) RT‑qPCR showed relative mRNA expression of apoptosis- and inflammation-related genes in DEHs across groups. (D) DCFH-DA assay measured ROS levels in DEHs (scale bar: 50 μm). (E) Quantification of average fluorescence intensity for green channels. Data represent mean ± SEM, n = 3. Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each indicator. Plasma and liver pharmacokinetic profiles of both miRNAs To determine whether orally administered miRNAs could reach liver in vivo, their pharmacokinetic profiles were examined in ducklings following oral gavage. Pharmacokinetic profiling following oral administration revealed that plasma concentrations of both miRNAs rose rapidly between 0.5 and 3 h, peaked at 3 h, declined sharply until 6 h, and then tapered gradually (Fig. 4A, B). Hepatic concentrations exhibited distinct profiles. novel-miR2 levels increased rapidly until 0.5 h, slowed from 0.5 to 3 h, and surged again until 6 h; novel-miR10 accumulated steadily until 3 h followed by a slower increase. Both miRNAs reached peak liver concentrations at 6 h before declining rapidly (Fig. 4C, D).Fig. 4Pharmacokinetic profiles of novel-miR2 and novel-miR10 in duckling plasma and liver. (A, B) Plasma concentration-time curves of (A) novel-miR2 and (B) novel-miR10 following oral administration, quantified by RT-qPCR. (C, D) Hepatic accumulation kinetics of (C) novel-miR2 and (D) novel-miR10 over time. Data represent mean ± SEM, n = 5. Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each time point.Fig 4 dummy alt text Pharmacokinetic profiles of novel-miR2 and novel-miR10 in duckling plasma and liver. (A, B) Plasma concentration-time curves of (A) novel-miR2 and (B) novel-miR10 following oral administration, quantified by RT-qPCR. (C, D) Hepatic accumulation kinetics of (C) novel-miR2 and (D) novel-miR10 over time. Data represent mean ± SEM, n = 5. Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each time point. Both miRNAs alleviate DHAV-1-induced liver injury Based on their favorable hepatic accumulation, the therapeutic efficacy of novel-miR2 and novel-miR10 was next evaluated in a duckling model of DHAV-1 infection. Survival analysis demonstrated that the VC and ncRNA groups exhibited acute mortality peaks at 12 and 24 hpi (Fig. 5A). In contrast, deaths in the miRNA‑treated groups were distributed gradually throughout the observation period. At 12 hpi, survival rates were 40.0% (VC), 36.7% (ncRNA), 83.3% (novel‑miR2), and 80.0% (novel‑miR10). Only 3.3% of VC group ducklings survived at 36 hpi, with no subsequent deaths. Final survival rates reached 6.7% (ncRNA), 36.7% (novel‑miR2), and 26.7% (novel‑miR10). Both miRNA groups showed higher overall survival than the VC group, but no statistical difference was observed between the two miRNA treatments.Fig. 5Novel-miR2 and novel-miR10 alleviate DHAV-1-induced mortality and hepatopathology in ducklings. (A) Survival curves of ducklings in each group. (B) Hepatic gross pathology alterations (red arrows: surface abnormalities; background grid: 5 × 5 cm). (C) Serum levels of liver enzymes (ALT, AST, ALP). (D) Liver histopathological features (red arrows: vacuolar degeneration; yellow boxes: inflammatory infiltration; orange arrows: lipid droplets; yellow arrows: inflammatory cells; green arrows: acidophilic lesions; black arrows: erythrocytes; blue arrows: fibrinoid necrosis; scale bar: 50 μm). Data represent mean ± SEM, n = 30 (A) or 5 (C). Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each time point.Fig 5 dummy alt text Novel-miR2 and novel-miR10 alleviate DHAV-1-induced mortality and hepatopathology in ducklings. (A) Survival curves of ducklings in each group. (B) Hepatic gross pathology alterations (red arrows: surface abnormalities; background grid: 5 × 5 cm). (C) Serum levels of liver enzymes (ALT, AST, ALP). (D) Liver histopathological features (red arrows: vacuolar degeneration; yellow boxes: inflammatory infiltration; orange arrows: lipid droplets; yellow arrows: inflammatory cells; green arrows: acidophilic lesions; black arrows: erythrocytes; blue arrows: fibrinoid necrosis; scale bar: 50 μm). Data represent mean ± SEM, n = 30 (A) or 5 (C). Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each time point. Gross liver examination revealed normal morphology in BC livers throughout (Fig. 5B). At 4 hpi, a few hemorrhagic foci were present on VC, ncRNA, and novel‑miR2 livers, but absent in the novel‑miR10 group. By 8 hpi, VC and ncRNA livers showed severe hemorrhage and rib impressions, while miRNA groups exhibited only minor spots. At 54 hpi, VC livers appeared green with extensive hemorrhage, whereas treated groups had fewer hemorrhagic foci. Notably, novel‑miR10 livers retained a near‑normal red coloration similar to BC controls. Serum levels of ALT, AST, and ALP were elevated in VC and ncRNA groups compared to BC controls at all time points (Fig. 5C). At 4 hpi, novel‑miR10 reduced both AST and ALT, while novel‑miR2 lowered AST relative to the VC group. By 8 and 54 hpi, all three enzyme levels were lower in miRNA‑treated groups. HE‑stained liver sections revealed distinct pathological progression (Fig. 5D). At 4 hpi, all groups except BC displayed vacuolar degeneration with inflammatory infiltration. Vacuolation was milder in the novel‑miR10 group than in the VC group. By 8 hpi, VC and ncRNA groups exhibited hepatic congestion, extensive eosinophilic lesions, and apoptotic bodies, alongside inflammatory infiltration. In contrast, miRNA‑treated groups still showed vacuolation and inflammation, though lesions had worsened relative to 4 hpi. At 54 hpi, VC livers presented fibrinoid degeneration, fragmented necrosis, and massive inflammatory infiltration with complete loss of normal architecture. Treatment groups primarily maintained vacuolation and inflammation; notably, the novel‑miR10 group showed reduced severity compared to its 8 hpi state. Table 3 summarized macroscopic liver lesion scores in ducklings. Except for the BC group, all groups displayed liver injury to varying degrees. The VC and ncRNA groups contained the highest numbers of ducklings (n=26 each) with severe scores of 4 or 5. In contrast, the novel‑miR2 group had the fewest such cases (n=17). Mean lesion scores decreased in the following order: VC group > ncRNA group > novel‑miR10 group > novel‑miR2 group > BC group.Table 3Statistics of liver lesions of ducklings.Table 3 dummy alt textGroupSample number of lesion scores in liver / ducklingsAverage score(mean ± SEM)0 point1 point2 point3 point4 point5 pointBC30000000.00 ± 0.00VC02025214.43 ± 0.20ncRNA00228184.40 ± 0.16novel-miR20436983.47 ± 0.25novel-miR10031411113.87 ± 0.23 Table 3Statistics of liver lesions of ducklings.Table 3 dummy alt textGroupSample number of lesion scores in liver / ducklingsAverage score(mean ± SEM)0 point1 point2 point3 point4 point5 pointBC30000000.00 ± 0.00VC02025214.43 ± 0.20ncRNA00228184.40 ± 0.16novel-miR20436983.47 ± 0.25novel-miR10031411113.87 ± 0.23 Statistics of liver lesions of ducklings. Both miRNAs modulate DHAV-1-induced immune and apoptotic responses To explore hepatoprotective mechanisms, hepatic viral load, cytokines, and apoptotic markers were analyzed following miRNA treatment. ELISA demonstrated that both novel‑miR2 and novel‑miR10 reversed DHAV‑1‑induced elevations in viral antigen, IFN‑γ, and IL‑10 at 8 and 54 hpi (Fig. 6A, B). Corresponding changes were observed at the transcriptional level (Fig. 6C, D). At 8 hpi, VC and ncRNA groups showed increased expression of IL‑6, IL‑10, and Caspase-3 genes alongside decreased Bcl‑2 versus the BC group. Both miRNAs reduced IL‑6, IL‑10, and Caspase-3 transcription compared to the VC group; however, only novel‑miR2 elevated Bcl‑2 levels. By 54 hpi, both treatments lowered IL‑10 and Caspase-3 while raising Bcl‑2 relative to the VC group. Notably, novel‑miR2 suppressed Caspase-3 more effectively than novel‑miR10 at this late stage.Fig. 6Novel-miR2 and novel-miR10 modulate viral load, inflammatory cytokines, and apoptosis-related gene expression in DHAV-1-infected duckling livers. (A, B) Hepatic levels of (A) DHAV‑1, (B) IL‑10 and IFN‑γ antigens measured by ELISA. (C, D) Relative mRNA expression of (C) inflammation- and (D) apoptosis-related genes assessed via RT‑qPCR. Data represent mean ± SEM, n = 5. Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each time point.Fig 6 dummy alt text Novel-miR2 and novel-miR10 modulate viral load, inflammatory cytokines, and apoptosis-related gene expression in DHAV-1-infected duckling livers. (A, B) Hepatic levels of (A) DHAV‑1, (B) IL‑10 and IFN‑γ antigens measured by ELISA. (C, D) Relative mRNA expression of (C) inflammation- and (D) apoptosis-related genes assessed via RT‑qPCR. Data represent mean ± SEM, n = 5. Distinct letters (a-d) indicate statistically significant differences (p < 0.05) among groups, with comparisons made separately for each time point. Discussion This study provides the first evidence that specific miRNAs from traditional herbal decoctions of S. baicalensis and C. indicum possess direct anti‑DHAV‑1 activity. Two novel miRNAs, novel‑miR2 and novel‑miR10, were identified through a high‑throughput screening strategy that prioritized the most abundant miRNAs in the decoctions. Both molecules exhibited in vitro cytoprotection and in vivo hepatoprotection, improving survival and reducing liver pathology in infected ducklings. These findings reveal that novel‑miR2 and novel‑miR10 possess antiviral activity, adding to the growing list of bioactive constituents in these herbs that could explain their traditional use against DVH(Chen et al., 2018a; Qiu et al., 2022). One of the main challenges facing therapeutic miRNAs is their susceptibility to degradation(Li et al., 2022). In this study, novel‑miR2 and novel‑miR10 remained detectable after high‑temperature decoction. Both miRNAs showed heat resistance but RNase sensitivity in vitro (Fig. 2A, B), aligning with canonical RNA properties. Pharmacokinetic analysis indicated that orally administered miRNAs entered circulation and accumulated in the liver, with hepatic peak levels following plasma peaks (Fig. 4). The in vivo stability of such plant miRNAs may be maintained through 3’-terminal 2’‑O‑methylation(Li et al., 2023). Alternatively, association with protective carriers like exosome‑like vesicles or protein complexes could also contribute to their stability(Xu et al., 2024). This absorption pattern suggested uptake in the upper gastrointestinal tract, which aligned with prior dietary miRNA reports(Chen et al., 2021). Although the exact absorption mechanism remains unclear, the observed in vivo efficacy supported sufficient miRNA delivery to target organs. Bioinformatic analysis predicted binding sites for both miRNAs on the DHAV‑1 genome, specifically within the 2A2, 2B, 2C, and 3C genes (Fig. 2C). Furthermore, both miRNAs suppressed viral gene expression (Fig. 1E, F) and reduced hepatic viral load (Fig. 6A). This raised the possibility of direct viral targeting, similar to honeysuckle MIR2911 which directly inhibits influenza A virus(Zhou et al., 2015). However, additional host‑directed effects cannot be excluded. Although the precise mechanisms of plant miRNA absorption and transport remain unclear, accumulating functional evidence strongly supports their cross‑kingdom regulatory roles. For example, ginger‑derived exosomal osa‑miR164d targeted macrophage TAB1 mRNA, initiating downstream cascades(Yan et al., 2024). Documented functions of plant miRNAs span nutrition, immunity, and disease control. Ginseng miRNA156 enhanced metabolic clearance and muscle glycogen storage in mice(Wang et al., 2024). Plant miRNAs also modulated dopamine metabolism in giant pandas, influencing food preference(Tan et al., 2025). In immunity, Centella asiatica exosomal miRNAs alleviated inflammation, rebalanced gut microbiota, and boosted immune cell activity(Shi et al., 2025). Therapeutically, various plant miRNAs have been shown to combat viral infections and inhibit tumor growth(Kalarikkal and Sundaram, 2021; Zhang et al., 2023b). Correspondingly, in this study, novel‑miR2 and novel‑miR10 provided additional hepatoprotection by reducing apoptosis, necrosis, inflammation, and oxidative stress, ultimately improving liver function and attenuating pathological lesions (Figs. 3, 5, and 6). Additionally, the two miRNAs exhibited functional divergence: novel‑miR2 more potently mitigated apoptosis, whereas novel‑miR10 excelled at suppressing inflammation and ameliorating gross liver lesions, likely due to distinct target gene networks. Although this study had certain limitations regarding dose optimization and mechanistic characterization, current findings carry significant implications. They revealed that traditional herbal decoctions contain not only small molecules and polysaccharides but also functionally active miRNAs that can survive the decoction process and exert cross‑kingdom antiviral effects(Chen et al., 2018b; Hu et al., 2025; Qiu et al., 2022). For veterinary practice, these miRNAs could be developed as orally administrable, sequence‑specific antiviral agents against DHAV‑1. Future work should focus on (i) validating direct viral targeting using reporter systems, (ii) identifying host targets via RNA‑seq and RIP‑seq, (iii) optimizing delivery formulations to enhance stability, (iv) conducting systematic dose‑response studies to determine the optimal therapeutic window, and (v) evaluating safety and efficacy in large‑scale field trials. In conclusion, this study demonstrated that novel‑miR2 and novel‑miR10 from S. baicalensis and C. indicum decoctions alleviate DHAV‑1‑induced hepatitis by suppressing viral replication, apoptosis, and inflammation. These findings expanded the understanding of herbal pharmacology and provide a promising strategy for developing miRNA‑based therapeutics against poultry viral diseases. Conclusion This study successfully identified two thermostable plant miRNAs, novel-miR2 and novel-miR10, from the decoctions of S. baicalensis and C. indicum. Both miRNAs exhibited anti-DHAV-1 activity in vitro and in vivo. They protected DEHs by inhibiting viral replication, reducing apoptosis, necrosis, and oxidative stress. Oral administration of these miRNAs improved survival rates, alleviated hepatic damage, and modulated inflammatory responses in infected ducklings. The results elucidated that plant-derived miRNAs constitute a novel material basis for the traditional use of these herbs against viral hepatitis, highlighting a cross-kingdom regulatory mechanism. This discovery opened new avenues for exploring nucleic acid-based active components in traditional medicine and developing novel antiviral strategies for poultry diseases. Funding statement The work was supported by 10.13039/501100001809National Natural Science Foundation of China (Grant No. 32473074). Author Contribution Mengxin Hu: Conceived the core hypothesis and designed the experimental workflow. She led the investigative process, performed all cell-based functional assays, and took primary responsibility for the systematic analysis and interpretation of the raw experimental data. She also composed the initial draft of the manuscript. Zhipeng Chen: Established the computational framework. He conducted the bioinformatics analysis of high-throughput sequencing data, including miRNA identification, quantification, and target gene prediction, and interpreted the sequencing results. Kexin Wang: Served as the statistical guarantor for the sequencing data. She designed the data normalization strategy and conducted the comparative statistical evaluations that confirmed the reproducibility of our findings. Xu Liu, Shuoyan Guan, Xiang Fu, Bolin Gu, Xinyu Liu, and Zefei Yan: Collectively participated in the establishment and troubleshooting of the animal models and molecular assays (including RT-qPCR and ELISA). Beyond technical execution, they were actively engaged in the daily critical discussions regarding data quality control and contributed to the preliminary interpretation of the phenotypic results. Jiaguo Liu: Conceptualized the overarching therapeutic strategy and provided the key mechanistic insights that connected the in-vitro data to the in-vivo effects. He directed the scientific decision-making throughout the study and critically revised the manuscript for intellectual content to ensure the arguments were logically sound. Disclosures The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper