Integrated multi-omics study identifies ursolic acid as a novel therapeutic agent targeting the TNF-α/TAK1/IKKβ/NF-κB axis in hepatic sinusoidal obstruction syndrome

Abstract Background Hepatic sinusoidal obstruction syndrome (SOS) is a lethal complication of hematopoietic stem cell transplantation (HSCT) with limited targeted therapies. The busulfan-cyclophosphamide (BUCY) conditioning regimen induces endothelial injury, thereby initiating a thromboinflammatory cascade. While TNF-α/NF-κB signaling is implicated in this process, its precise role and therapeutic targeting in SOS remain undefined. Ursolic acid (UA), a natural pentacyclic triterpenoid, possesses anti-inflammatory properties, yet its efficacy and mechanism in SOS are unknown. Methods Multi-omics integration: Network pharmacology identified shared drug (BUCY, UA) and post-HSCT disease (SOS) targets and enriched pathways (supported by GEO/ArrayExpress analysis). Molecular docking and dynamics (MD) simulations (GROMACS) assessed binding affinities and stability. Subsequently, surface plasmon resonance (SPR) quantified UA-TNF-α binding. In vitro: Primary rat HSECs were treated with BUCY ± UA, NF-κB was overexpressed (OE-p65 plasmid) or inhibited (BAY 11-7082). Apoptosis (Annexin V/PI flow cytometry), ROS (DCFH-DA), mRNA (qPCR), and protein (WB) levels of TNF-α/NF-κB pathway components and antioxidants were measured. In vivo: SOS was induced in male SD rats with BUCY. Rats received prophylactic UA, AAV-shRNA-p65 (liver-specific NF-κB knockdown), or vehicle. Endpoints included hepatic sinusoidal injuries (histology), serum biomarkers (ELISA), liver ROS, liver function, apoptosis markers, and target pathway analysis. Results Target identification and binding: network pharmacology and transcriptomics identified TNF-α signaling as the top shared pathway (KEGG q=3.2x10⁻⁵). MD simulations revealed stable binding of BUCY metabolites (e.g., acrolein, ΔG=-8.2 kcal/mol) and UA to TNF-α (Tyr59/Trp107, ΔG=-9.3 kcal/mol). SPR confirmed high-affinity UA-TNF-α binding (Kd=0.32 ±0.05μM). UA suppressed TAK1/IKKβ phosphorylation by 82.3% (p<0.001). In Vitro: BUCY significantly induced HSEC apoptosis (↑3.3-fold, p<0.001) and ROS (↑3.8-fold, p<0.001), activating TNF-α, p-TAK1, p-IKKβ, p-IκBα, p65, and downstream cytokines (ICAM-1, VCAM-1, IL-1β, IL-6, MMP9). UA co-treatment attenuated apoptosis (↓70%, p<0.001), ROS (↓3.1-fold, p<0.001), and NF-κB pathway activation. OE-p65 mimicked BUCY damage; UA or NF-κB inhibitor reversed it. In Vivo: BUCY induced SOS (85% incidence), elevated serum TNF-α (↑4.1-fold), ALT/AST (↑4.3-fold), liver ROS, and caused severe HSEC ultrastructural damage (SEM/TEM: loss of fenestrae, basement membrane disruption). UA prophylaxis significantly reduced SOS incidence (35%, p<0.01), serum TNF-α (↓74.2%), ALT/AST (↓87.4%, p<0.01), ROS, and restored sinusoidal architecture, comparable to AAV-shRNA-p65. UA and shRNA-p65 both suppressed hepatic p65, ICAM-1, VCAM-1, IL-1β, IL-6, MMP9, and apoptosis markers (TUNEL↓, BAX↓, Caspase-3↓, FASL↓, Bcl-2↑). UA uniquely enhanced antioxidants (SOD1↑3.2-fold, CAT↑) early, mitigating a biphasic NF-κB response: early compensatory antioxidant upregulation (SOD1↑) transitioning to exhaustion (T-AOC↓63%, p<0.001) in sustained injury. UA outperformed shRNA-p65 in systemic cytokine suppression and apoptosis reduction (Caspase-3↓79.3% vs ↓52.1%, p<0.001). Conclusion TNF-α/NF-κB signaling critically drives BUCY-induced HSEC injury and SOS. UA, a first-in-class natural TNF-α inhibitor, binds TNF-α with high affinity, thereby blocking downstream TAK1/IKKβ/NF-κB activation. This mechanism suppresses endothelial inflammation, oxidative stress, and apoptosis. Significantly, UA's efficacy in preventing SOS and protecting HSECs equals or surpasses NF-κB knockdown, demonstrating its polypharmacology. Collectively, the study redefines SOS pathogenesis as “TNF-α-initiated thromboinflammation” and strongly supports UA's clinical translation for preventing SOS/endothelial complications in HSCT.