nature communications Single-cell transcriptomics and epigenomics unravel the role of monocytes in neuroblastoma bone marrow metastasis

nature communications Single-cell transcriptomics and epigenomics unravel the role of monocytes in neuroblastoma bone marrow metastasis

Metastasis is the major cause of cancer-related deaths. Neuroblastoma, a childhood tumor has been molecularly defined at the primary cancer site, however, the bone marrow as the metastatic niche of neuroblastoma is poorly characterized. Here we perform single-cell transcriptomic and epigenomic profiling of bone marrow aspirates from eleven subjects spanning three major neuroblastoma subtypes and compare these to five age-matched and metastasis-free bone marrow, followed by in-depth single cell analyses of tissue diversity and cell-cell interactions, as well as functional validation. We show that cellular plasticity of neuroblastoma tumor cells is conserved upon metastasis and tumor cell type composition is neuroblastoma subtype-dependent. Neuroblastoma cells signal to the bone marrow microenvironment, rewiring via macrophage migration inhibitory factor and midkine signaling specifically monocytes, which exhibit M one and M two features, are marked by activation of pro- and anti-inflammatory programs, and express tumor-promoting factors, reminiscent of tumor-associated macrophages. The interactions and pathways characterized in our study provide the basis for therapeutic approaches that target tumor-to-microenvironment interactions.

Neuroblastoma accounts for fifteen percent of childhood cancer-related deaths, where greater than ninety percent of metastatic stage stage M neuroblastoma tumors disseminate to the bone marrow, which acts as a site for disease relapse and progression. Genetic neuroblastoma tumor heterogeneity and plasticity have been suggested to contribute to differentiation or metastasis and relapse, serving as intrinsic oncogenic drivers. Main genetic factors involved in disease onset and progression include amplification of MYCN, mutation of TP fifty-three, amplification or mutation of ALK and other Ras MAPK pathway genes, and dysregulation of telomere maintenance via rearrangements of TERT or alternative lengthening of telomeres, which is often associated with mutated or truncated ATRX. However, recent whole-genome sequencing studies have identified a scarcity of recurrent somatic alterations, but show that a subgroup of metastatic neuroblastoma is rather defined by large segmental chromosomal aberrations (herein referred to as sporadic).

Numerous studies in recent years have focused in defining cell types and lineage trajectories of the developing adrenal medulla, aiming to uncover the cell of origin in neuroblastoma, which arises from neural crest-derived sympatho-adrenal progenitor cells at different stages during embryonic development of the sympathetic nervous system. Although the majority of tumor cells in primary neuroblastoma resemble healthy sympathoblasts, especially in untreated and low-risk tumors, some tumor cells in pretreated and high-risk cases have shown enriched signatures of chromaffin cells and their progenitors, i.e., Schwann cell precursors, but also of mesenchymal and neural crest-like cells, suggesting that their abundance and differentiation state is associated with prognosis. Immune cells within the tumor microenvironment have been shown to carry either tumor promoting or suppressing activities. Studies of primary neuroblastoma tumors found increased levels of T-, NK, and dendritic cells in the tumor microenvironment of low-risk neuroblastoma compared to high-risk neuroblastoma. Previous studies applying tissue imaging and bulk transcriptomics link myelocytes with inflammatory signatures. A comprehensive characterization of the immune microenvironment and crosstalk with tumor cells at the metastatic site remains elusive.

Recently, the use of single-cell technologies have emerged as powerful tools to comprehensively characterize cellular states within healthy and diseased tissues. These approaches have been applied to characterize the tumor heterogeneity, along with the tumor microenvironment, of primary neuroblastoma tumors, and the composition of adult human bone marrow in normal and disease settings has been investigated, e.g., in leukemia and bone metastases in prostate cancer. However, such approaches have yet to be deployed across different neuroblastoma subgroups, i.e., MYCN amplification, ATRX mutation, and sporadic at the bone marrow, the metastatic niche of neuroblastoma. Here, we apply single-cell ATAC-sequencing and single-cell RNA-sequencing across consensus neuroblastoma subgroups in tandem with proteomics and functional assays to: (i) study differences in cellular plasticity across neuroblastoma subtypes in metastatic and primary tumors, (ii) investigate interactions between tumor cells and the bone marrow microenvironment, and (iii) unravel metastasis-induced alterations in the bone marrow.

We find that neuroblastoma subgroups determine cell type composition, and that tumor phenotype is conserved upon metastasis. Neuroblastoma cells primarily interact with myeloid cells, which present with M one and M two features, indicated by aberrant pro- and anti-inflammatory core TF regulatory loops, pro-differentiation, and reduction of cell cycle genes as well as expression of tumor-promoting factors. Collectively, these data provide insights into the molecular and cellular architecture of neuroblastoma across all subgroups, as well as with the potential to inform future studies aimed at improving patient outcomes.

Results

The single-cell atlas of neuroblastoma bone marrow metastasis We integrated genome-wide single-cell RNA-sequencing and single-cell ATAC-sequencing profiling in bone marrow samples originating from benign tumors without bone marrow metastases, herein defined as controls, and eleven samples across metastatic neuroblastoma subtypes: MYCN amplification, ATRX mutation, and sporadic (lacking either alteration). Each tumor sample was molecularly and cytogenetically characterized, substantiating subtype classification. Single-cell RNA-sequencing yielded a total of eighty thousand seven hundred eighty-nine single cells with a median of one thousand two hundred seventy-eight genes per cell. Following integration and clustering of the single-cell RNA-sequencing data, cells were classified using five reference datasets. This yielded seven major cell types, comprised of various types of immune cells: T-cells, NK-cells, B-cells, myeloid cells, and plasmacytoid dendritic cells, followed by erythroid cells and stem cells, which were supported by expression of canonical metastatic tumor cells. Neuroblastoma subtypes are grouped into MYCN amplification and non-MYCN amplification tumors. Correlation analysis of pseudobulk gene expression demarcates the M tumors from A and S tumors. Source data are provided as a Source Data file.

marker genes. Additionally, we identified a cluster of NB cells, which was (i) classified as neurons, (ii) expressed key NB markers, and (iii) was absent in control samples. An unspecified cell cluster consisting of twenty-two cells was defined as "other" and was excluded from further analysis. To further confirm the demarcation between microenvironment and tumor cells, we used two complementary strategies. First, we calculated the tumor infiltration rates based on the evaluation of tumor markers GD two and L one CAM by flow cytometry, which were in concordance with the scRNA-seq data assignments. Second, we inferred copy number variants based on scRNA-seq data, which were present in NB cells and absent in non-NB cells. The scRNA-seq-based CNV calls were validated using bulk tumor profiling by SNP-arrays, further corroborating our tumor cell assignment.