Comparative Study of Single-Cell and Bulk RNA Sequencing Data from Metastatic Bone Marrow Neuroblastoma Samples

Neuroblastoma is characterized by frequent involvement of bone marrow (BM) as a site of cell dissemination and spread. In this study, single-cell RNA sequencing (scRNA-seq) was used to analyze the cellular heterogeneity of a subset of metastatic BM samples collected at initial diagnosis. Comparison of the single-cell data with bulk RNA sequencing further refined the analysis. An enrichment of regulatory T cells relative to a healthy control and activation of the CD24, CD47, and CD200 "don't eat me" signals were documented. Computational analyses highlighted communication between neuroblastoma and myeloid cells via the amyloid precursor protein (APP) and midkine (MK) signaling networks. Within neuroblastoma cells, mutually exclusive adrenergic and transitory cell states were identified, and ten sub-clusters were denoted. In addition, common and unique tumor cell antigens were investigated. CNTFR and CHRNA3, as high-ranking candidates, were validated, confirming their strong selectivity for neuroblastoma cells. Taken together, these findings support the existence of a significant tumor-dependent modulation of the BM ecosystem, which should be considered when introducing immunotherapy. Furthermore, they highlight the potential to investigate new antigens at the single-cell resolution.