Identification of AURKA as a Key Target in Hepatocellular Carcinoma: Integrated Bioinformatics Analysis and Evaluation of Natural Compounds via Molecular Docking

Abstract

Objective: This study aimed to identify a central molecular target associated with hepatocellular carcinoma (HCC) using bioinformatic methods and to comparatively evaluate the binding potential of reference inhibitors and natural compounds against this target by molecular docking.Materials and Methods: In the GSE76427 microarray dataset (Illumina HumanHT-12 V4.0), 115 HCC tumor tissues were compared with 52 non-tumor liver tissues, and differentially expressed probes were determined with the limma package (adjusted p < 0.05 and |log2 fold change| ≥ 1). A STRING protein–protein interaction (PPI) network was constructed from the up-regulated genes, nodes were ranked by degree, and Aurora kinase A (AURKA) was selected as the docking target based on expression robustness, biological relevance, structural data quality, and druggability. Two reference inhibitors (Sorafenib and Alisertib) and six natural compounds (Curcumin, Quercetin, Luteolin, Kaempferol, Apigenin, Resveratrol) were docked into the 4JBQ crystal structure of AURKA using AutoDock Vina with a shared grid box and identical parameters.Results: A total of 362 probes were differentially expressed (95 up-regulated, 267 down-regulated); down-regulated genes were associated with liver-specific metabolic and detoxification functions, whereas up-regulated genes were related to cell cycle and mitosis. The STRING network comprised 69 nodes and 422 edges (PPI enrichment p < 1.0 × 10⁻¹⁶), and nearly all of the most central nodes were mitotic regulators, with AURKB (degree = 33) and AURKA (degree = 32) ranking first and second. In docking, the highest affinities were obtained for Sorafenib (−9.537 kcal/mol) and Alisertib (−9.066 kcal/mol); among the natural compounds, Curcumin (−8.414 kcal/mol) yielded an energy close to that of the reference inhibitors. Interaction analysis showed that sorafenib formed a hydrogen bond with Ala273 through its urea bridge, a π-cation interaction with the catalytic Lys162, and a halogen bond with Glu181, exhibiting an ATP-competitive binding mode that extended toward the DFG region (Asp274).Conclusion: This integrated approach combining transcriptomic analysis, network biology, and molecular docking highlights AURKA as a central HCC-associated hub gene and computationally prioritizes natural polyphenols, particularly curcumin, as potential AURKA-binding candidates. These docking-based predictions are hypothesis-generating and require experimental validation before any conclusions regarding biological or therapeutic efficacy can be drawn.