Characterization of the immune cell landscape in neuroblastoma : for the discovery of novel therapeutics

The most common extracranial and deadliest tumor in infants, neuroblastoma (NB), is both clinically and genetically heterogenous in nature. NB patients with low-risk disease have good prognosis with over 90% of children surviving. In contrast, patients with high-risk disease presenting with bone/bone marrow metastasis at diagnosis have roughly a 40% chance of survival. Over the past decades, immune cells have been extensively studied in NB, however, conflicting results remain. In addition, studies have largely focused on a single cell type and metastatic tumors have hardly been included. Anti-GD2 immunotherapy is one element of the current standard of care but comes with severe toxicities leading to patients dropping out of treatment. Other immunotherapy approaches, such as T-cell-based approaches, are currently making their way into the standard of care treatment regimen. To further improve immunotherapy for these patients, it is necessary we understand the immune cell composition and their function in NB primary tumors and bone metastasis. To this end, we aimed to unravel the immune cell landscape in NB primary tumors (paper I) and bone/bone marrow metastasis (BMets) (paper II). Furthermore, we sought to elucidate alpha beta (aß) and gamma delta (yo) T-cell clonal expansion, and phenotypic and functional properties of T cells infiltrating NB primary tumors (paper III).

We employed single-cell RNA sequencing (scRNAseq) and VDJ sequencing (scVDJseq) techniques, flow cytometry, immunohistochemistry (IHC) and bioinformatic tools. For scRNAseq, in paper I, 19 primary tumor samples derived from 17 patients were included, paper II included 7 non-metastatic and 8 metastatic bone/bone marrow samples, and for paper III, 11 primary tumor samples plus 4 matched blood samples were used for scRNA/VDJseq. To validate scRNAseq findings, IHC and flow cytometry were performed. Bioinformatic tools were used to extract T-cell clonotypes from scRNAseq data. Flow cytometry was also used for functional testing of yoTCR reactivity against a panel of NB cell lines.

Paper I and II revealed a complex interactive immune cell network in NB primary tumors and BMets. Most immune cell subpopulations correlated with improved prognosis in NB primary tumors except for Macro-1. The main myeloid populations in primary tumors were macrophages and monocytes, whereas in BMets a higher number of neutrophils were detected. Furthermore, a significant reduction in B cells was found in BMets compared to non-metastatic samples (nonMets). There was an increase in the fraction of CD56bright NK cells in BMets whereas in NB primary tumors, CD56dim NK cells were increased in frequency in intermediate- and high-risk disease compared to low-risk. CTLs were significantly increased in numbers in BMets compared to nonMets and had high expression of cytotoxicity- related genes in both BMets and NB primary tumors. The gene signature score for Treg activity was unchanged comparing risk groups in primary tumors whereas this was significantly increased in BMets compared to nonMets. Paper III revealed a heterogeneous T-cell landscape in NB primary tumors and matched blood samples including an interesting 'repair' population. We detected a significant increase in percentage of V81+ and Vy9-V81-V82- in primary tumor compared to blood samples. Drastic clonal expansion was found for both aß and yo T cells suggesting antigen recognition in the tumors. And, expanded Vy9V83, Vy8V83 and Vy5V81 clonotypes cloned into Jurkat cells were reactive towards one or more NB cell lines measured by CD69 upregulation. This hints to their respective ligands existing on NB cells.

Taken together, we mapped the immune cell landscape in NB primary tumors and BMets showing differences in cell numbers and phenotypic properties. Generally, immune cells were correlated to improved prognosis in NB primary tumors. Clonal expansion of aß and yo T cells, and yoTCR reactivity suggests tumor antigen recognition. This thesis outlines a deeper comprehension of the NB tumor microenvironment in both primary tumors and BMets. It provides novel insights into T-cell clonality and phenotypes, and functions as a basis for the discovery of novel immunotherapeutic targets.

List of scientific papers

I. Verhoeven BM, Mei S, Olsen TK, Gustafsson K, Valind A, Lindström A, Gisselsson D, Fard SS, Hagerling C, Kharchenko PV, Kogner P, Johnsen JI, Baryawno N. The immune cell atlas of human neuroblastoma. Cell Rep Med. 2022 Jun 21;3(6):100657. Epub 2022 Jun 9. https://doi.org/10.1016/j.xcrm.2022.100657

II. Mei S, Alchahin AM, Embaie BT, Gavriliuc IM, Verhoeven BM, Zhao T, Li X, Jeffries NE, Pepich A, Sarkar H, Olsen TK, Wickström M, Stenman J, Reina- Bedoya O, Kharchenko PV, Saylor PJ, Johnsen JI, Sykes DB, Kogner P, Baryawno N. Single-cell analyses of metastatic bone marrow in human neuroblastoma reveals microenvironmental remodeling and metastatic signature. JCI Insight. 2024 Feb 15;9(6):e173337. https://doi.org/10.1172/jci.insight.173337

III. Verhoeven BM, Ye K, Stenman J, Bedoya Reina O, Kreslavskiy T, Baryawno N. Single-cell clonal composition of T-cell repertoires in human neuroblastoma. [Manuscript]