Platinum-based chemotherapy promotes antigen presenting potential in monocytes of patients with high-grade serous ovarian carcinoma

Abstract

Ovarian cancer (OC) is the most lethal gynecologic malignancy worldwide. The major clinical challenge includes the asymptomatic state of the disease, making diagnosis possible only at advanced stages. Another OC complication is the high relapse rate and poor prognosis following the standard first-line treatment with platinum-based chemotherapy. At present, numerous clinical trials are being conducted focusing on immunotherapy in OC; nevertheless, there are still no FDA-approved indications. Personalized decision regarding the immunotherapy, including immune checkpoint blockade and immune cell-based immunotherapies, can depend on the effective antigen presentation required for the cytotoxic immune response. The major aim of our study was to uncover tumor-specific transcriptional and epigenetic changes in peripheral blood monocytes in patients with high-grade serous ovarian cancer (HGSOC). Another key point was to elucidate how chemotherapy can reprogram monocytes and how that relates to changes in other immune subpopulations in the blood. To this end, we performed single-cell RNA sequencing of peripheral blood mononuclear cells (PBMCs) from patients with HGSOC who underwent neoadjuvant chemotherapeutic treatment (NACT) and in treatment-naïve patients. Monocyte cluster was significantly affected by tumor-derived factors as well as by chemotherapeutic treatment. Bioinformatical analysis revealed three distinct monocyte subpopulations within PBMCs based on feature gene expression - CD14.Mn.S100A8.9hi, CD14.Mn.MHC2hi and CD16.Mn subsets. The intriguing result was that NACT induced antigen presentation in monocytes by the transcriptional upregulation of MHC class II molecules, but not by epigenetic changes. Increased MHC class II gene expression was a feature observed across all three monocyte subpopulations after chemotherapy. Our data also demonstrated that chemotherapy inhibited interferon-dependent signaling pathways, but activated some TGFb-related genes. Our results can enable personalized decision regarding the necessity to systemically re-educate immune cells to prime ovarian cancer to respond to anti-cancer therapy or to improve personalized prescription of existing immunotherapy in either combination with chemotherapy or a monotherapy regimen.

Keywords: antigen presentation; chemotherapy; methylation; monocyte; ovarian cancer; single cell sequencing; transcriptome.

Figure 1

Single cell RNAseq analysis of PBMCs in high-grade serous ovarian carcinoma revealed tumor-specific programming of monocytes (A) Overall study design. (B) Distribution of PBMC cell clusters and individual samples in UMAP 2D dimensions. (C) Dot plot discriminates marker genes for each cluster. (D) Volcano plots demonstrate p-value and log2fold-change value for DEGs in each monocyte subpopulation (CD14.Mn.S100A8.9hi; CD14.Mn.MHC2hi; CD16.Mn) versus healthy control (|L2FC|>0.58, FDR<0.05). (E) Bar plots with GSEA results indicate top deregulated pathways in each monocyte subpopulation (CD14.Mn.S100A8.9hi; CD14.Mn.MHC2hi; CD16.Mn) versus healthy control (FDR<0.1). (F) Localization of SIGLEC1 and MS4A4A gene expression among individual cells on UMAP. DnN – healthy control (donor group), OvN – treatment-naive group, OvY – NAC-treated group.

Figure 2

Chemotherapy-induced transcriptional programming in cancer monocytes (A) Augur demonstrates top-affected cell types in naive tumor PBMCs (upside) and in NAC-treated PBMCs (underside) (indicated by Augur score). (B) NAC-related gene signature obtained using UCell tool. (C) Abundance plot demonstrated cell content in studied cohorts. *Significant difference in cell abundance in NAC vs. Dn (FDR<0.1). (D) Volcano plots demonstrate p-value and log2fold-change value for DEGs in each monocyte subpopulation (CD14.Mn.S100A8.9hi; CD14.Mn.MHC2hi; CD16.Mn) of NAC-treated monocytes versus healthy control (|L2FC|>0.58, FDR<0.05). (E) Bar plots with GSEA results indicate top deregulated pathways in each monocyte subpopulation (CD14.Mn.S100A8.9hi; CD14.Mn.MHC2hi; CD16.Mn) of NAC-treated monocytes versus healthy control (FDR<0.1). DnN – healthy control (donor group), OvN – treatment-naive group, OvY – NAC-treated group.

Figure 3

Chemotherapy induces activation of antigen presentation in peripheral blood monocytes. (A) NAC-related gene signature (MHC2 signature) obtained using UCell tool. (B) CD8-associated gene signature obtained using UCell tool (C) Factor loadings per each group (constructed using LIANA with Tensor-cell2cell) (Upside). Top 15 ligand-receptor pairs associated with factor 1, factor 2 and factor 5 (Underside). (D) Module enrichment with GSEA in individual comparison by monocyte subpopulations (CD14.Mn.S100A8.9hi; CD14.Mn.MHC2hi; CD16.Mn) (FDR<0.05). NS, not significant. OvNvsDnN – treatment-naive vs. healthy control; OvYvsDn – NAC-treated vs. healthy control; OvYvsOvN – NAC-treated vs. treatment-naive. (E) DEGs (FDR<0.1) reveled in distinct module-associated groups.

Figure 4

Methylation level in CpG sites in CD14+ monocytes of treated vs. untreated samples (A) PCA plot demonstrated analyzed samples (OvY – NAC-treated monocytes; OvN – treatment naive monocytes). (B) Heatmap with hierarchical clustering of OvY vs. OvN samples using top 10 000 variable CpG cites. (C) Volcano plot with differentially methylated CpG sites near promoter region (|L2FC|>1, FDR<0.25). (D) Functional annotation of genes with differentially methylated CpG sites near promoter region performed with Enrichr (FDR<0.1).