Hyperinflammatory repolarisation of ovarian cancer patient macrophages by anti-tumour IgE antibody, MOv18, restricts an immunosuppressive macrophage:Treg cell interaction

Abstract

Ovarian cancer is the most lethal gynaecological cancer and treatment options remain limited. In a recent first-in-class Phase I trial, the monoclonal IgE antibody MOv18, specific for the tumour-associated antigen Folate Receptor-α, was well-tolerated and preliminary anti-tumoural activity observed. Pre-clinical studies identified macrophages as mediators of tumour restriction and pro-inflammatory activation by IgE. However, the mechanisms of IgE-mediated modulation of macrophages and downstream tumour immunity in human cancer remain unclear. Here we study macrophages from patients with epithelial ovarian cancers naive to IgE therapy. High-dimensional flow cytometry and RNA-seq demonstrate immunosuppressive, FcεR-expressing macrophage phenotypes. Ex vivo co-cultures and RNA-seq interaction analyses reveal immunosuppressive associations between patient-derived macrophages and regulatory T (Treg) cells. MOv18 IgE-engaged patient-derived macrophages undergo pro-inflammatory repolarisation ex vivo and display induction of a hyperinflammatory, T cell-stimulatory subset. IgE reverses macrophage-promoted Treg cell induction to increase CD8+ T cell expansion, a signature associated with improved patient prognosis. On-treatment tumours from the MOv18 IgE Phase I trial show evidence of this IgE-driven immune signature, with increased CD68+ and CD3+ cell infiltration. We demonstrate that IgE induces hyperinflammatory repolarised states of patient-derived macrophages to inhibit Treg cell immunosuppression. These processes may collectively promote immune activation in ovarian cancer patients receiving IgE therapy.

Fig. 1. In vitro-derived human macrophage subsets express FcεRs, bind MOv18 IgE and exhibit a net pro-inflammatory phenotypic shift following IgE stimulation.

Evaluation of in vitro-derived human macrophage subsets FcεR expression, MOv18 IgE binding and phenotypes following FcεR:IgE cross-linking by polyclonal anti-IgE antibodies. a Heatmap displaying the scaled expression of surface markers (MFI) (flow cytometry) (n = 9 CD86, CD80, CD40, CD206, CD204, CD163, MerTK; n = 8 FcεRI; n = 7 CD23; n = 3 CD200R) and secreted factors (pg/ml) in cell culture supernatants (Luminex) (n = 5). b tSNE plot visualising the in vitro-derived macrophage subsets based on surface marker expression (flow cytometry) (n = 2). c Comparison of FcεRI (n = 9) and CD23 (n = 7) expression, with representative flow cytometry plots for M2a macrophages. d Assessment of MOv18 IgE binding (n = 3), with representative flow cytometry plots for M1 and M2a macrophages. e Comparison of surface marker expression following FcεR:IgE cross-linking (anti-IgE: cross-linking endogenous IgE only; anti-IgE + MOv18 IgE: cross-linking endogenous IgE and exogenously applied MOv18 IgE) (flow cytometry) (n = 9 CD163, CD204, MerTK; n = 6 FcεRI; n = 5 CD40, CD80, CD86). f Composite and faceted tSNE plots visualising the in vitro-derived macrophage subsets based on surface marker expression following FcεR:IgE cross-linking (flow cytometry) (n = 2). g Following detection of IgE-mediated repolarisation of M2a and M2c, the effect of FcεR:IgE cross-linking on the secretome of M2a and M2c, as well as M1, was assessed by Luminex measurement of secreted factors in macrophage culture supernatants. Heatmap displaying the scaled expression of surface markers (MFI) (flow cytometry) (n = 9 CD86, CD80, CD40, CD206, CD204, CD163, MerTK; n = 8 FcεRI; n = 7 CD23; n = 5 CD39; n = 3 CD200R) and secreted factors (pg/ml) in cell culture supernatants (Luminex) (n = 8) following FcεR:IgE cross-linking. Data shown as mean ± SEM. Statistical significance was calculated using a repeated measures 1-way ANOVA with Tukey’s post hoc test (c, e); *Padj <0.05, **Padj <0.01, ***Padj <0.001 and ****Padj <0.0001. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 2. The ovarian tumour microenvironment (TME) is enriched for alternatively-activated macrophage phenotypes, which associate with poor patient survival, IL10 signaling and FCER1A/FcεRI expression.

Bulk (TCGA-OV; n = 378) and single-cell (GSE165897; n = 6) RNA-seq and flow cytometric evaluations of the TME of ovarian cancer patients. a Heatmap displaying the scaled abundance of immune cells in TCGA-OV primary tumours as determined by the CIBERSORTx deconvolution package and a pie chart representing mean immune cell abundance across patients. b Kaplan–Meier plots (Survival and Survminer packages) stratifying TCGA-OV patients by high and low levels (quartiles) of tumour immune cell abundance (Immune Score; ConsensusTME deconvolution package) and tumour CD68 and CD163 expression. c Left: Following patient stratification by CD163 tumour expression (n = 190), differentially expressed genes (DEGs) were determined using the DESeq2 package and enrichment of genes sets evaluated within the human Reactome database (v2023.1) by the fgsea package. Graph displaying the top 20 upregulated Reactome pathways ordered by Normalised Enrichment Score, with Interleukin 10 Signaling highlighted. Right: Heatmap displaying the scaled expression (TPM: transcripts per million) of genes in the Interleukin 10 Signaling pathway (n = 34 genes). d Boxplot comparison of FCER1A expression (log₁₀(TPM + 0.001)) between CD163^high and CD163^low tumours (n = 190)(P < 0.0001). e Top: UMAP visualising the annotated immune cell types in GSE165897 (metastatic peritoneal tumours; n = 7142 cells) following unsupervised clustering using the Seurat package and a pie chart displaying mean cell type abundance across patients. Bottom: Heatmap displaying the scaled expression (TPM) of lineage marker genes used for cell type annotation and FCER1A. f UMAP visualising the annotated monocyte and macrophage subsets in GSE165897 (n = 1374 cells) following unsupervised clustering. g Flow cytometry evaluation of TAM FcεRI expression (n = 14) (top) and TAM (n = 11), dendritic cell (DC) (n = 8) and mast cell (n = 4) frequencies in patient ascites (bottom). Data shown as median (centre line), interquartile range (IQR) (box) and range within 1.5 x IQR (whiskers) (d) and mean ± SEM (g). Statistical significance was calculated using a Flemington-Harrington weighted log rank test (b), Wald test with Benjamini–Hochberg correctionh (c - DEGs), permutation testing (c - gene sets), a two-tailed Wilcoxon signed rank test (d) and a mixed effects analysis with Tukey’s post hoc test (g); *P/Padj <0.05, **P/Padj <0.01 and ****P/Padj <0.0001. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 3. Ovarian cancer patient ascites-conditioned macrophages (MAsc) and ascites tumour-associated macrophages (TAMs) exhibit strong phenotypic similarity to IL-10 polarised macrophages.

Evaluation of MAsc and ascites TAM phenotypes by flow cytometric assessment of surface marker expression and single-cell RNA-seq pseudotime differentiation trajectory analysis of macrophage subsets in metastatic ovarian tumours. a Schematic of the experimental workflow for maturing monocytes from healthy volunteer leukocyte cones into MAsc. b Comparison of marker expression between unpolarised M0 and MAsc, with representative flow cytometry plots (n = 19) (P = 0.0005 (CD163); 0.025 (MerTK); 0.0071 (CD80); 0.0307 (PD-L1); 0.0377 (HLA-DR)). c Heatmap displaying the scaled expression of markers (MFI) between MAsc (n = 19) and the in vitro-derived macrophage subsets M1, M2a and M2c (n = 6). d UMAP visualising MAsc (n = 19) and the in vitro-derived macrophage subsets M1, M2a and M2c (n = 6), with scaled expression of selected markers superimposed. e Comparison of the expression of markers across MAsc, ascites TAMs and the in vitro-derived macrophage subsets M1, M2a-d by heatmap (scaled % positive cells) and bar charts (MAsc, n = 19; TAMs, n = 10; in vitro-derived macrophage subsets, n = 9 CD86, CD80, CD40, CD206, CD163, MerTK, FcεRI, n = 7 CD23). f Single-cell RNA-seq pseudotime (Slingshot package) analysis (GSE165897; n = 4 metastatic peritoneal tumours), displaying the differentiation trajectories of monocyte and macrophage subsets (n = 1374). UMAP visualising the subsets with the pseudotime minimum spanning tree superimposed. g DEGs between the subsets were determined using the Seurat package (two-tailed Wilcoxon signed-rank test), and the top 100 (upregulated and downregulated) used for gene over-representation analysis via g:Profiler (human Reactome (v2023.1) and Gene Ontology Biological Processes (v2023.1) databases) to investigate enrichment of gene sets. Scores for the specific statistically significant pathways between subsets visualised by UMAP on a per-cell basis and dot plot on a per-subset basis. Data are shown as mean ± SEM. Statistical significance was calculated using a paired two-tailed t-test (b) and a 1-way ANOVA with Tukey’s post hoc test (e); (b) *P < 0.05, **P < 0.01 and ***P < 0.001, (e) *Padj <0.05. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 4. Ovarian cancer patient ascites-conditioned macrophages (MAsc) and ascites tumour-associated macrophages (TAMs) exhibit an IgE-mediated pro-inflammatory repolarisation and TAMs display MOv18 IgE-induced killing of ovarian cancer cells.

Evaluation of MAsc and ascites TAM MOv18 IgE binding and phenotypes following FcεR:IgE cross-linking and MOv18 IgE-specific killing of ovarian cancer cells by TAMs. a Schematic of the experimental workflow for assays FcεR:IgE cross-linking MAsc and TAM by polyclonal anti-IgE antibodies. Contains images created in BioRender. Karagiannis, S. (2025) https://BioRender.com/c06z015. b Assessment of MOv18 IgE binding to MAsc (n = 4) and TAMs (n = 2), with representative flow cytometry plots for TAMs. c Evaluation of surface marker expression on unpolarised M0 and MAsc following FcεR:IgE cross-linking by polyclonal anti-IgE antibodies, with representative flow cytometry plots (anti-IgE: cross-linking endogenous IgE only; anti-IgE + MOv18 IgE: cross-linking endogenous IgE and exogenously applied MOv18 IgE) (n = 19). d Heatmap displaying the scaled expression of surface markers (MFI) (flow cytometry) and secreted factors (pg/ml) (Luminex) in cell culture supernatants, following MAsc FcεR:IgE cross-linking by polyclonal anti-IgE antibodies (n = 19). e tSNE plot visualising MAsc following FcεR:IgE cross-linking (n = 19) and the in vitro-derived macrophage subsets M1, M2a and M2c (n = 6), based on surface marker expression (flow cytometry). MDS plot visualising the same experiment with each dot representing an individual sample. f Evaluation of surface marker expression on unpolarised M0 and MAsc following co-culture with FRα-expressing IGROV1 ovarian cancer cells, in the presence of MOv18 IgE, with representative flow cytometry plots (n = 6). g Evaluation of surface marker expression on TAMs following FcεR:IgE cross-linking by polyclonal anti-IgE antibodies, via heatmap (scaled MFI) and bar charts, with representative flow cytometry plots (n = 13). h Evaluation of IGROV1 cell killing (Padj=0.0054) and TAM CD163 expression (Padj = 0.03), following co-culture between TAMs and IGROV1 cells in the presence of MOv18 IgE or isotype control (anti-NIP IgE), with representative flow cytometry plots (n = 5). Data are shown as mean ± SEM. Statistical significance was calculated using a repeated measures 1-way ANOVA with Tukey’s post hoc test (a, f, g, h left panel) and a Friedman test with Dunn’s post hoc test (h right panel); *Padj < 0.05, **Padj < 0.01, ***Padj < 0.001 and ****Padj < 0.0001. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 5. IgE mediates a repolarisation of patient macrophage subsets away from immunosuppression towards an IgE-induced hyperinflammatory subset.

Subset level evaluation of ovarian cancer patient ascites-conditioned macrophages (MAsc) (n = 19) and ascites TAMs (n = 7) by flow cytometric analysis of surface markers following FcεR:IgE cross-linking by polyclonal anti-IgE antibodies. a tSNE plot visualising unpolarised M0 and MAsc with the 11 subsets generated by the FLOWSOM algorithm superimposed. b Heatmap displaying the scaled expression of markers (MFI) in each of the subsets and the mean proportion (%) of each of the subsets across the samples. c Comparison of the proportions (%) of each of the subsets per sample. d Comparison of the proportions (%) of differentially enriched subsets. e tSNE plot visualising TAMs with the 13 subsets generated by the FLOWSOM algorithm superimposed. f Heatmap displaying the scaled expression of markers (MFI) in each of the subsets and the mean proportion (%) of each of the subsets across the samples. g Comparison of the proportions (%) of differentially enriched subsets. Data shown as mean ± SEM. Statistical significance was calculated using a repeated measures 1-way ANOVA with Tukey’s post hoc test; *Padj <0.05, **Padj <0.01, ***Padj <0.001 and ****Padj <0.0001. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 6. An association between immunosuppressive macrophages and regulatory T (Treg) cells can be reversed by IgE-mediated macrophage repolarisation.

Evaluation of the association between macrophages and Treg cells in bulk (TCGA-OV; n = 378 primary ovarian tumours) and single-cell (GSE165897; n = 4 peritoneal metastatic ovarian tumours) RNA-seq datasets and ex vivo co-cultures following FcεR:IgE cross-linking (flow cytometry). a Evaluation of cell:cell interactions between immune cell types in GSE165897 using the Liana package. Heatmap representing the frequency of statistically significant sent and received interactions (Padj <0.01) between cell types. b Scatter graph displaying the Spearman’s Rank Correlation between TCGA-OV tumour expression of CD163 and FOXP3 (R is Spearman’s Rank Correlation Coefficient). c TCGA-OV patients were stratified by high and low levels (quartiles) of CD163 tumour expression (n = 190) and differentially expressed genes (DEGs) were determined using the DESeq2 package. Gene set enrichment was evaluated within the human Gene Ontology Biological Processes (v2023.1) database by the fgsea package. Heatmap displaying the scaled expression (TPM: transcripts per million) of genes for the Positive Regulation of Treg Differentiation pathway (n = 15 genes). d Schematic of the experimental workflow for IL-10 polarised M2c and ovarian cancer patient ascites-conditioned macrophage (MAsc) co-cultures with naïve CD4+ T cells following FcεR:IgE cross-linking. e M2c were co-cultured with allogeneic naive CD4+ T cells; comparison of the proportion of Treg and Teff cells of total CD4+ cells and the Treg:Teff cell ratio, with representative flow cytometry plots (n = 6 T cell monocultures, n = 13 co-cultures). f Comparison of the percentage of Treg and Teff cells expressing TGF-β (P = 0.0075 (Teff); 0.0129 (Treg)), IL-10 (P = 0.0384 (Teff); 0.0159 (Treg)) and TNF, with representative flow cytometry plots for Treg cells (n = 17). Data are shown as mean ± SEM. Statistical significance was calculated using permutation testing (a, c - gene sets), a Wald test with Benjamini–Hochberg correction (c - DEGs), a mixed effects analysis with Tukey’s post hoc test (e) and a paired two-tailed t-test (f); *P/Padj < 0.05, **P/Padj < 0.01, ***P/Padj < 0.001 and **** P/Padj < 0.0001. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 7. IgE-mediated hyperinflammatory repolarisation of patient macrophages reverses macrophage-promoted Treg cell activity to increase CD8+ T cell expansion.

Evaluation of effect of FcεR:IgE cross-linking in ex vivo co-cultures on: ovarian cancer patient ascites-conditioned macrophage (MAsc)-mediated Treg cell induction from allogeneic naïve CD4+ T cells and Treg cell and MAsc cytokine expression; suppressive function of Treg cells co-cultured with M2c macrophages. a Unpolarised M0 or MAsc were co-cultured with allogeneic naïve CD4+ T cells; comparison of the proportion of Treg and Teff cells of total CD4+ cells and Treg:Teff cell ratio, with representative flow cytometry plots (n = 10 T cell monocultures; n = 19 co-cultures). b Comparison of the percentage of Treg and Teff cells expressing TGF-β, IL-10 and TNF, with representative flow cytometry plots for Treg cells (n = 19). c Comparison of MAsc expression of TGF-β, with a representative flow cytometry plot (n = 12). d Following naïve CD4+ T cell co-cultures with IL-10-polarised M2c macrophages, purified CD4+ CD25+ T cells (Treg and Teff cells) were co-cultured with allogeneic peripheral blood mononuclear cells (PBMCs); comparison of the percentage relative proliferation of CD4+ and CD8+ T cells in PBMCs, with representative flow cytometry histograms (PBMC:CD25+ T cell ratio of 8) (n = 8), and IL-2 concentration (pg/ml) in the cell culture supernatants (ELISA) (n = 13). e Kaplan–Meier plot stratifying TCGA-OV patients by high and low levels (tertiles) of CD163 and CD8B expression (n = 378). Data are shown as mean ± SEM. Statistical significance was calculated using a mixed effects analysis with Tukey’s post hoc test (a), a repeated measures 1-way ANOVA with Tukey’s post hoc test (b, c), a paired two-tailed t-test (d) and a Flemington-Harrington weighted log rank test (e); *P/Padj < 0.05, **P/Padj < 0.01, ***P/Padj < 0.001 and ****P/Padj < 0.0001. Source data and exact P/Padj values are provided as a Source Data file.

Fig. 8. Following MOv18 IgE treatment, both tumours from rats and patient tumour biopsies in the Phase I trial display similar IgE-immune activation signatures to those observed in ex vivo functional assays.

Transcriptomic (microarray) evaluations of MOv18 IgE and phosphate-buffered saline (PBS)-treated tumours from a syngeneic lung metastasis rat model (n = 2) and immunohistochemistry (IHC) characterisation of matched pre- and on-treatment tumour biopsies from the Phase I trial of MOv18 IgE (n = 2). a Boxplot comparisons of gene expression (log₂) between tumours from MOv18 IgE- (n = 2) and PBS-treated (n = 2) rats. b Differentially expressed genes (DEGs) between MOv18 IgE and PBS-treated rat tumours were determined using the limma package and enrichment of genes sets evaluated within the human Gene Ontology Biological Processes (v2023.1) database by the fgsea package. Graph displays selected upregulated immune pathways and the top 10 downregulated pathways ordered by Normalised Enrichment Score. c Heatmaps displaying scaled expression of genes from selected upregulated and downregulated pathways: Positive Regulation of αβ T cell Proliferation (n = 16 genes); Natural Killer (NK) cell-mediated Immunity (n = 39 genes); Monocyte Chemotaxis (n = 24 genes); Positive Regulation of TNF-mediated signaling (n = 6 genes); Positive Regulation of IL12 production (n = 24 genes); Phagocytosis (n = 106 genes); Macrophage Activation (n = 46 genes); Cell Killing (n = 72 genes); DNA-templated DNA Replication (n = 85 genes); Cell Cycle Checkpoint Signaling (n = 71 genes). d Immunohistochemical evaluation of the proportion of immune cell markers identified intratumourally in pre- and on-treatment tumour biopsies from the Phase I trial of MOv18 IgE (n = 2). Representative images of CD68 expression in pre- and on-treatment biopsies; stained images shown on the left, with the pixel classifiers used to quantify tumour (pink) and CD68 expression (green) superimposed on the right. Data shown as median (centre line), IQR (box) and range (whiskers) (a). Statistical significance was calculated using permutation testing (b). Source data and exact P/Padj values are provided as a Source Data file.