ABSTRACT: Challenges in the precision diagnosis and treatment of nasopharyngeal carcinoma (NPC) persist, primarily due to the lack of a multi-omics-based molecular typing system and effective targeted therapeutic strategies. In this study, we integrated proteomic and phosphoproteomic analyses of NPC tissues and normal nasopharyngeal tissues to systematically identify key dysregulated protein networks, which are enriched in genes associated with cell cycle, cilium function, immune response, metabolism, mTOR signaling, and responses to virus infection. Based on these proteomic characteristics, we classified NPC into two molecular subtypes, subtype 1 and 2, which exhibit significant clinical heterogeneity. Moreover, we identify protein signatures for both normal and cancerous tissues, as well as for the two molecular subtypes, and developed diagnostic and prognostic models. Through computational drug repurposing and experimental validation, we identified the pan-histone deacetylase inhibitor Panobinostat as a potent anti-tumor agent for NPC in both in vitro and in vivo models. Our findings revealed that Panobinostat inhibits MYC signaling, thereby disrupting the transcriptional activation of the ATR-CHK1 axis. This reduction in ATR-CHK1 activity leads to impaired homologous recombination repair efficiency, ultimately leading to DNA damage accumulation. Notably, we demonstrated a synergistic therapeutic effect when Panobinostat was combined with radiotherapy, resulting in enhanced cytotoxicity against NPC cells. In summary, we established a multi-omics molecular typing system for NPC, developed diagnostic and prognostic marker panels, and highlighted the therapeutic potential of Panobinostat in combination with radiotherapy. These findings lay a solid foundation for precision diagnosis, prognostic stratification, and personalized treatment strategies in NPC.
