MYCN Drives a Tumor Immunosuppressive Environment Which Impacts Survival in Neuroblastoma

Abstract

A wide range of malignancies presents MYCN amplification (MNA) or dysregulation. MYCN is associated with poor prognosis and its over-expression leads to several dysregulations including metabolic reprogramming, mitochondria alteration, and cancer stem cell phenotype. Some hints suggest that MYCN overexpression leads to cancer immune-escape. However, this relationship presents various open questions. Our work investigated in details the relationship of MYCN with the immune system, finding a correlated immune-suppressive phenotype in neuroblastoma (NB) and different cancers where MYCN is up-regulated. We found a downregulated Th1-lymphocytes/M1-Macrophages axis and upregulated Th2-lymphocytes/M2-macrophages in MNA NB patients. Moreover, we unveiled a complex immune network orchestrated by N-Myc and we identified 16 genes modules associated to MNA NB. We also identified a MYCN-associated immune signature that has a prognostic value in NB and recapitulates clinical features. Our signature also discriminates patients with poor survival in non-MNA NB patients where MYCN expression is not discriminative. Finally, we showed that targeted inhibition of MYCN by BGA002 (anti-MYCN antigene PNA) is able to restore NK sensibility in MYCN-expressing NB cells. Overall, our study unveils a MYCN-driven immune network in NB and shows a therapeutic option to restore sensibility to immune cells.

Keywords: MYCN; MYCN blocking; anti-MYCN antigene PNA; immune network; immune signature; immune system; neuroblastoma.

Figure 1

MYCN but not MYC anticorrelates with immune pathways in cancers. (A) Bar plot represents Gene Ontology enriched terms in the MYCN negative correlated genes in two neuroblastoma (NB) datasets (left panel: E-MTAB-1781, right panel: TARGET). Bar length represents NES absolute value while color intensity represents -log10 FDR. (B) Pathway enrichment for five select immune pathways (GO terms) in different cancer datasets. Symbol size and color intensity indicate—log10 FDR and NES. GO terms enriched in MYCN (left panel) and MYC (right panel) correlated genes. (C) Color intensity indicated the mean of T-helper subset relative abundance in MYCN amplified (MNA) and non-MNA patient expression profiles plotted as heatmap. (D) Macrophage relative abundance (left panel: M1 population, right panel: M2 population) in MNA and non-MNA patient gene expression profiles. Each symbol represents an individual patient (MNA = 122, non-MNA = 580), the middle line represents the median, the first and third quartiles are indicated as box limits, whiskers represents 1.5 box lengths, extreme values are indicated as single dots. Wilcoxon matched pair test; **P < 0.01; ****P < 0.0001.

Figure 2

MYCN up-regulation impacts a wide immune interaction network. (A, B) Circular plots. Immune subpopulations are specified outside the circle, outer circle represents cell types, inner circle represents activation status. Connecting lines indicate connection between two subpopulations and are proportional to the number of connections. (A) Immune network in non-(MYCN amplification) MNA patients. (B) Immune network in MNA patients. (C) Immune system gene module network in neuroblastoma (NB). Edges size is proportional to Pearson correlation coefficients, correlation is indicated in gray and negative correlation in red. Modules with no connections are not shown, module size is proportional to the number of genes within. Pie chart colors correspond to immune cell types, the size of the slices corresponds to the number of the genes. (D) Heatmap of MNA and non-MNA patient gene expression profiles (MNA = 122, non-MNA = 580) in NB1 cohort (E-MTAB-1781). Color intensity is proportional to z-score of the average eigengenes for each gene module. Main pathway enrichment for each module is listed on the left, full list is present in the Supplementary Tables. (E) heatmap representing the normalize relative abundance of regulons in NB1 cohort (E-MTAB-1781). Hierarchical clustering is conducted on the row and the columns using the Euclidean distance. Clinical data are on top.

Figure 3

MYCN effect on immune system has a prognostic impact. (A–E) Analyses conducted on E-MTAB-1781 dataset. (A) Uniform Manifold Approximation and Projection (UMAP) projection of MYCN amplification (MNA) and non-MNA patient gene expression profiles (PGEP). (B) Violin plots represent normalized MYCN immune score in MNA and non-MNA PGEP. (C) Kaplan–Meier plots for the probability of overall survival over time for patients associated with MYCN immune score (high enriched, n = 114; medium enriched, n = 311; low enriched, n = 277). Associated P value is shown in the middle of the plot (log-rank test). (D) Normalized MYCN immune score in different International Neuroblastoma Staging System Committee (INSS) classification stages. (E) Ki-67 log2 expression in patients associated with MYCN immune. Wilcoxon matched pair test; ****P < 0.0001.

Figure 4

BGA002 blocks CD276 and restores neuroblastoma (NB) susceptibility to natural killer (NK) cells. (A) Pathway enrichment for four select immune pathways [Gene Ontology (GO) terms] associated to NK cells. (NB1: E-MTAB-1781, NB2: TARGET NB). Symbol size and color intensity indicate—log10 FDR and NES. GO terms enriched in MYCN (left) anti-correlated genes and GO terms enriched in non-MYCN amplification (MNA) (right) patients. (B) mRNA expression inhibition of different genes (MYCN, CD276, CD274, HMGA1, PVR) in NB cell lines measured through real-time PCR after 12 h of treatment with BGA002 2.5 µM (black is the control, red the treatment, n = 4 biological replicates for each cell line). Wilcoxon matched pair test; *P < 0.05, **P < 0.01, where not shown is not significant (P > 0.05). (C) Kelly-luc cell line (MNA NB cell line transfected with luciferase) viability after treatment with BGA002 2.5 µM and NK co-culture (five independent experiments). Wilcoxon matched pair test; **P < 0.01.