Copper chelation redirects neutrophil function to enhance anti-GD2 antibody therapy in neuroblastoma

Abstract

Anti-disialoganglioside (GD2) antibody therapy has provided clinical benefit to patients with neuroblastoma however efficacy is likely impaired by the immunosuppressive tumor microenvironment. We have previously defined a link between intratumoral copper levels and immune evasion. Here, we report that adjuvant copper chelation potentiates anti-GD2 antibody therapy to confer durable tumor control in immunocompetent models of neuroblastoma. Mechanistic studies reveal copper chelation creates an immune-primed tumor microenvironment through enhanced infiltration and activity of Fc-receptor-bearing cells, specifically neutrophils which are emerging as key effectors of antibody therapy. Moreover, we report copper sequestration by neuroblastoma attenuates neutrophil function which can be successfully reversed using copper chelation to increase pro-inflammatory effector functions. Importantly, we repurpose the clinically approved copper chelating agent Cuprior as a non-toxic, efficacious immunomodulatory strategy. Collectively, our findings provide evidence for the clinical testing of Cuprior as an adjuvant to enhance the activity of anti-GD2 antibody therapy and improve outcomes for patients with neuroblastoma.

Fig. 1. Copper chelation potentiates antitumor activity of anti-GD2 immunotherapy.

a Experimental design and dosing strategy. Schematic created in BioRender. Vittorio, O. (2024). BioRender.com/c53h577. b Individual tumor kinetics of Th-MYCN mice. c Kaplan–Meier survival curves of Th-MYCN mice presented in (b). Statistical pairwise comparisons were calculated using a two-tailed Mantel–Cox log-rank test with p values displayed in the figure. For (b) and (c), data are n = 11 (Saline + IgG2a) or n = 10 (all other groups) biological replicates, one independent experiment. d Representative images of merged OPAL multiplex immunofluorescence spectra depicting the tumoral distribution of NCR1⁺ natural killer cells (red), CD8⁺ cytotoxic T cells (yellow), CD11b⁺ myeloid (white) and DAPI nuclei stain (blue) in Th-MYCN neuroblastoma tumor tissue 14 days post-treatment. Scale bar, 100 µm. e Immune cell quantification of (d) as positive counts per 1000 nuclei. Significance was calculated using a two-tailed Mann–Whitney U test with p values displayed in figure. For (e), Data are mean ± SEM n = 4 (Saline + IgG2a) or n = 3 (all other groups) biological replicates with a minimum of two technical replicates, one independent experiment. Abbreviations: IP intraperitoneal, p.o. orally. Source data are provided as a Source Data file.

Fig. 2. Copper chelation promotes an immune-permissive tumor microenvironment.

a Experimental design and dosing strategy of Th-MYCN model with peripheral blood and tumors obtained after 1 week of treatment. Schematic created in BioRender. Vittorio, O. (2024). BioRender.com/c53h577. b Cytokine levels in sera and tumoral lysates obtained from control and TEPA-treated mice. Serum data (excluding TGF-β) are presented as mean ± SEM, n = 4 (both groups) biological replicates, one independent experiment. Tumor microenvironment (TME) data (excluding TGF-β) are presented as mean ± SEM, n = 10 (Control) and n = 6 (TEPA) biological replicates, one independent experiment. TGF-β serum and TME data are presented as mean ± SEM, n = 7 (both groups) biological replicates, two independent experiments. Significance was calculated using a two-tailed Mann–Whitney U test with p-values displayed in the figure. c Flow cytometric analysis of myeloid subset frequencies from tumors after 1 week of treatment. Data are presented as mean ± SEM, n = 3 (both groups) biological replicates, one independent experiment. Significance was calculated using a two-tailed t-test with Welch’s correction with p-value displayed in the figure. d Representative flow cytometry plots of CD11b⁺Ly6G⁺ neutrophil frequency of tumors plotted in (c). e Count and percentage of circulating neutrophils obtained from control and TEPA-treated mice. Data are presented as mean ± SEM, n = 4 (both groups) biological replicates, one independent experiment. Significance was calculated using a two-tailed Mann–Whitney U test with p-values displayed in the figure. Abbreviations: n.d. no data available, p.o. orally, TME tumor microenvironment. Source data are provided as a Source Data file.

Fig. 3. Copper chelation reinvigorates antitumor immunity via pro-inflammatory signaling.

a Representative images of a tissue microarray containing control and TEPA-treated Th-MYCN tumor cores (n = 10/treatment, biological replicates in technical duplicate, total n = 20 cores; one independent experiment) and resected after 0, 3 or 7 days for NanoString GeoMx Digital Spatial Profiling, stained with fluorescently conjugated antibodies to PanCK (red), smooth muscle actin (yellow), CD45 (green) with DAPI nuclei stain (blue). Scale bar, 300 µm. One independent experiment. b Sankey diagram of Th-MYCN tumor cores depicting distribution of CD45-infiltrated low/high regions of interest, plotted against treatment group and duration. c Volcano plot of genes upregulated and downregulated in TEPA-treated high versus low regions of interest. False discovery rate (FDR) was adjusted using the two-tailed Benjamini–Hochberg procedure. Thresholds: p < 0.1; |log₂FC | > 0.5. d Gene set enrichment analysis for TEPA-treated high-infiltrated tumor regions compared to low-infiltrated regions presented as a bar plot. e Network representation of selected pathways in (d) displaying differentially expressed genes as branches. The magnitude of change is reported as log₂FC using colored nodes. Abbreviations: FC fold change, PanCK pancytokeratin.

Fig. 4. The neuroblastoma tumor microenvironment is sensitive to copper chelation therapy and promotes neutrophil infiltration.

a Experimental design and tumor processing workflow for single-cell RNA sequencing. Schematic created in BioRender. Vittorio, O. (2024). BioRender.com/j88j488. b Uniform manifold approximation and projection (UMAP) representation of integrated samples in the tumoral compartment (13,544 cells), colored by treatment group. c Violin plots of gene expression levels associated with intracellular copper levels (Mt1, Mt2) and neuroblastoma oncogene Mycn, split by treatment group. Significance was calculated using two-tailed differential expression analysis using the MAST algorithm after batch correction with p-values displayed in figure. Horizontal line indicates data median. d Gene set enrichment analysis plot for HALLMARK_MYC_TARGETS_V1 using the fgsea package. e Split UMAP representation of immune cell compartment (12,127 cells) according to treatment arm and colored by annotated immune subsets. f Bar plot of the proportion of immune cell subsets shown in (e). g Dot plot of gene expression markers used to classify the immune subsets defined in (e).

Fig. 5. Neutrophils supersede tumorigenic signaling to drive reinvigoration of antitumor immunity.

a Overrepresentation analysis of pathways relatively enriched in TEPA-treated immune cell clusters compared to control, presented as nodes with associated genes as branches using single-cell RNA sequencing. b Circle plot of the aggregated cell-cell communication networks in control and TEPA-treated single-cell samples. Edge width is proportional to the number of ligand-receptor interactions between cell types. c Scatter plots comparing the outgoing and incoming interaction strengths between control and TEPA-treated samples.

Fig. 6. Copper chelation facilitates infiltration and N1-polarization of neutrophils via copper mobilization to exert an anti-tumor response.

a Heatmap comparing the average expression of genes associated with copper metabolism across treatment arms within immune cell clusters. b Heatmap of log₂FC in expression between treatment arms for genes associated with migration and extravasation within the neutrophils cluster. Heatmap of log₂FC in expression for genes associated with (c) N1 anti-tumorigenic or (d) N2 pro-tumorigenic neutrophil phenotypes from control vs. TEPA-treated Th-MYCN tumors within the neutrophils cluster. Data presented in (a–d) were obtained from single-cell RNA sequencing with relevant cell values averaged and scaled. e Gene set enrichment analysis shows top pathways relatively enriched in TEPA-treated neutrophils. The “N1_ANTI_TUMOR NEUTROPHILS” signature was constructed using the N1-associated genes listed in (b). f IncuCyte cell imaging of neuroblastoma cell line SK-N-BE(2)-C transfected with a plasmid encoding tGFP-tagged MT1X protein following 24 h of TEPA treatment (10× objective). Representative image obtained from one independent experiment. Scale bar, 100 µm. g Concentration of copper in conditioned media before and after 30 min incubation with naive neutrophils isolated from healthy donors. Data are presented as mean ± SEM, n = 4/condition, biological replicates (healthy donors), three independent experiments. Significance was calculated using a two-tailed paired t-test with p value displayed in the figure. h qRT-PCR analysis for the expression of genes in human neutrophils associated with intracellular copper (MT1X), migration (S100A8) and pro-inflammatory activation (ISG15) obtained after 30 min incubation in conditioned media as per (e). Data are presented as mean, n = 2 biological replicates (healthy donors), one independent experiment. i Transwell migration assay of neutrophils towards untreated or TEPA-treated SK-N-BE(2)-C cells. Migrated neutrophils were counted using flow cytometry and percentage transmigration was calculated relative to input cells. Data are presented in a violin plot, n = 2 biological replicates (healthy donors) in triplicate, two independent experiments. Data minima and maxima values are as indicated, the median (solid line), and the first and third quartiles (dotted horizontal lines). Significance was calculated using an ordinary one-way ANOVA with Tukey’s post-hoc test with p-value displayed in the figure. j Antibody-dependent cytotoxicity assay against the Kelly neuroblastoma cell line using neutrophils isolated from healthy donors in the presence of anti-GD2 antibody (1 µg/ml) with caspase 3/7 activity quantified after 8 h. Data are presented as mean ± SEM, n = 3/condition, biological replicates (healthy donors) in triplicate, one independent experiment. Significance was calculated using a two-tailed Mann–Whitney U test with p-value displayed in figure. Abbreviations: FC fold change, tGFP turbo green fluorescent protein. Source data are provided as a Source Data file.

Fig. 7. Copper chelating agent TETA synergizes with anti-GD2 therapy to mediate antitumor activity in the syngeneic NXS2 model of neuroblastoma.

a Tumor growth kinetics in a syngeneic model of neuroblastoma involving the subcutaneous inoculation of A/J mice with NXS2 cells. Animals commenced treatment 1 week after inoculation (black arrow) and were treated by oral gavage with saline (control) or TETA (400 mg/kg/day) for 7 days before blood and tumor collection. Data are presented as mean ± SEM, n = 6 (Control) and n = 7 (TETA) biological replicates, two independent experiments. b Flow cytometric analysis of neutrophil frequencies in NXS2 tumors after 1 week of TETA treatment. Data are presented as mean ± SEM, n = 4 (both groups) biological replicates, one independent experiment. Significance was calculated using a two-tailed Mann–Whitney U test with p-value displayed in the figure. c Experimental design of the syngeneic NXS2 → A/J preclinical model and immunocombination dosing strategy. Schematic created in BioRender. Vittorio, O. (2024). BioRender.com/c53h577. For (d–f), arrows indicate the treatment period. d Relative weight change in tumor-bearing mice measured from date of inoculation. e Tumor growth kinetics of individual tumors measured from date of inoculation. f Kaplan–Meier survival curves of tumor-bearing mice measured from date of inoculation. Statistical pairwise comparisons were calculated using a two-tailed Mantel–Cox log-rank test with p-values displayed in figure. For (d–f), data are presented as mean ± SEM, n = 8 (all groups) biological replicates, one independent experiment. g Representative images of merged OPAL multiplex immunofluorescence spectra depicting the tumoral distribution of NCR1⁺ natural killer cells (red), CD8⁺ cytotoxic T cells (yellow), CD11b⁺ myeloid (white) and DAPI nuclei stain (blue) in NXS2 neuroblastoma tumor tissue 14 days post-treatment. Scale bar, 100 µm. h Immune cell quantification of (g) as positive counts per 1000 nuclei. Data presented as mean ± SEM, n = 3 (Saline + IgG2a); n = 4, (Saline + anti-GD2, TETA + anti-GD2), n = 5 (TETA + IgG2a) biological replicates with three technical replicates, one independent experiment. Significance was calculated using a two-tailed Mann–Whitney U test with p values displayed in the figure. Abbreviations: IP intraperitoneal, p.o. oral gavage, TF tumor-free. Source data are provided as a Source Data file.