The cellular signaling crosstalk between memory B cells and tumor cells in nasopharyngeal carcinoma cannot be overlooked: Their involvement in tumor progression and treatment strategy is significant

Background: Nasopharyngeal carcinoma (NPC) refers to a cancerous tumor that develops in the upper and side walls of the nasopharyngeal cavity. Typically, individuals are often diagnosed with the disease when it has already progressed significantly, and those with advanced NPC tend to have an unfavorable outlook in terms of response rate to targeted treatments and overall clinical survival. Various molecular mechanisms, including Myeloid-derived suppressor cells and factors like PD-L1, have been explored to enhance the outcome of NPC. However, there are still challenges to be addressed in terms of identifying symptoms at an early stage, making precise predictions about the chances of cancer returning and spreading, and devising successful approaches for treatment. The activation of B cells and their corresponding pathways holds potential for developing enhanced immune therapeutic strategies. Nevertheless, the comprehensive understanding of the intricate association between B cells and NPC tumor cells remains incomplete. Hence, this study employed single-cell multi-omics analysis to investigate the molecular biomarkers and prognostic factors linked to B cell subpopulations in human NPC while examining the underlying mechanisms. Materials and Methods: The Gene Expression Omnibus database provided tumor and blood samples obtained from patients diagnosed with NPC. Subsequently, we analyzed these single-cell data. Following the assessment of NPC sample quality, we employed the R package 'Harmony' to mitigate batch discrepancies using PCA outcomes. The analysis of Gene Ontology, Gene Set Enrichment Analysis, and Kyoto Encyclopedia of Genes and Genomes was used to examine differentially expressed genes in B cell subpopulations of NPC tumors. The pseudo-temporal trajectories of B cells in NPC were studied using the Monocle and Slingshot software tools. In addition, the CellChat package was utilized to predict the incidence of intercellular communication between different subpopulations of B cells and cancerous cells. Furthermore, we utilized univariate Cox regression, LASSO, and multivariate Cox regression analysis to construct prognostic models. The immune cell infiltration was evaluated in tumor tissues using ESTIMATE, CIBERSORT, and xCell. Furthermore, the infercnv was employed to assess the extent of copy number variation in NPC cells. To forecast the potential reaction of particular tumor samples to chemotherapy, the R package called 'pRRophetic' was utilized.