Multinucleated giant cells are hallmarks of ovarian aging with unique immune and degradation-associated molecular signatures

Abstract

The ovary is one of the first organs to exhibit signs of aging, characterized by reduced tissue function, chronic inflammation, and fibrosis. Multinucleated giant cells (MNGCs), formed by macrophage fusion, typically occur in chronic immune pathologies, including infectious and non-infectious granulomas and the foreign body response, but are also observed in the aging ovary. The function and consequence of ovarian MNGCs remain unknown as their biological activity is highly context-dependent, and their large size has limited their isolation and analysis through technologies such as single-cell RNA sequencing. In this study, we define ovarian MNGCs through a deep analysis of their presence across age and species using advanced imaging technologies as well as their unique transcriptome using laser capture microdissection. MNGCs form complex interconnected networks that increase with age in both mouse and nonhuman primate ovaries. MNGCs are characterized by high Gpnmb expression, a putative marker of ovarian and non-ovarian MNGCs. Pathway analysis highlighted functions in apoptotic cell clearance, lipid metabolism, proteolysis, immune processes, and increased oxidative phosphorylation and antioxidant activity. Thus, MNGCs have signatures related to degradative processes, immune function, and high metabolic activity. These processes were enriched in MNGCs compared to primary ovarian macrophages, suggesting discrete functionality. MNGCs express CD4 and colocalize with T-cells, which were enriched in regions of MNGCs, indicative of a close interaction between these immune cell types. These findings implicate MNGCs in modulation of the ovarian immune landscape during aging given their high penetrance and unique molecular signature that supports degradative and immune functions.

Fig 1. Multinucleated giant cells (MNGCs) are age-associated in mammalian ovaries.

(A) Analysis workflow of the study. (B) Representative mouse and nonhuman primate (NHP) ovarian histological sections (top and middle panels) and Trainable Weka Segmentation (ImageJ) of MNGCs (bottom panel) across age. (C) Quantification of MNGC area in mouse ovaries across age. RY; reproductively young 6-12 wks. (D) Quantification of MNGC area in rhesus macaque ovaries across age. (E) F4/80 immunolabeling of mouse ovarian MNGCs. (F) CD68 immunolabeling of NHP MNGCs. Z-stack of autofluorescent MNGCs in cleared mouse ovaries at macro (G) and micro (H) scales, with white arrows in the last panel of (H) highlighting putative nuclei. Scale bars for B are = 200 µm mouse top and middle, 2,000 µm for NHP top panel, 200 µm for NHP middle panel. Scale bars = 100 µm for E, 200 µm for G, and 50 µm for H. For C, data is presented as means ± SEM with 4-7 mice of each age. One-way ANOVA with Tukey’s multiple comparison post-test (*, p < 0.05) was used for statistical analysis of C while a simple linear regression was used to determine the goodness to fit (R²) and the likelihood of the slope being non-zero (P) in D. Portions of panel 1A were created in BioRender. Lab, D. (2025) https://BioRender.com/mtmex43. The data underlying the graphs can be accessed at https://doi.org/10.5061/dryad.kh18932j4.

Fig 2. Transcriptomic analysis of ovarian MNGCs.

(A) Schematic of laser capture microdissection (LCM) of ovarian MNGCs for transcriptomic analysis. (B) Representative images depicting MNGC identification by serial H&E-stained section (i) and autofluorescence (iv-vi), and capture for LCM (ii-iii). (C) Table of the top 50 MNGC expressed genes categorized by biological processes to which they relate. (D) Pathway analysis of top 1,000 expressed MNGC genes. (E) Localization of Gpnmb RNA (RNAscope) and protein (IHC) in mouse (12m) ovarian sections. (F) Localization of GPNMB protein in rhesus macaque (13yr) ovarian tissue section. For C, (*) denotes genes that fall into more than one process. Scale bars = 150 µm for B; 200 µm for E; 1,000 µm for F. Transcriptomic analysis was performed on 4 biological replicates. Portions of panel 2A were created in BioRender. Lab, D. (2025) https://BioRender.com/2pi739t. The data underlying the graphs can be accessed at https://doi.org/10.5061/dryad.kh18932j4.

Fig 3. Comparison of ovarian MNGCs and primary ovarian macrophages.

(A) Schematic of isolation procedure for ovarian macrophages by immunomagnetic pull-down. (B) Confirmation of macrophage identity of F4/80-pull down cells by phagocytic capacity. (C) Principal component analysis (PCA) of primary ovarian macrophage (Mac) and MNGC transcriptomes. (D) Expression profiles of macrophage marker genes in primary macrophages, MNGCs, and non-MNGC stromal samples. (E) Volcano plot of differentially expressed genes (DEGs) between MNGCs and primary ovarian macrophages (p < 0.05 and Log₂FC≥2). (F) Pathway analysis of upregulated DEGs in MNGCs compared to primary macrophages. (G) Pathway analysis of downregulated DEGs in MNGCs compared to primary macrophages. (H) Heat maps of MNGC vs. primary macrophage DEGs for select biological processes. Transcriptomic analysis was performed on 4 biological replicates for each sample type. Panel 3A was created in BioRender. Lab, D. (2025) https://BioRender.com/x0ry468. The data underlying the graphs can be accessed at https://doi.org/10.5061/dryad.kh18932j4.

Fig 4. Assessment of gene signatures unique to MNGCs.

(A) Pathway analysis of genes unique to ovarian macrophages from young mice. (B) Pathway analysis of genes unique to ovarian MNGCs. (C) Top 15 expressed unique MNGCs genes. (D) Heat map of expression profiles of T-cell marker genes. (E) Immunohistochemical analysis of macrophage (F4/80) and T-cell (CD3ε, CD4, CD8a) marker protein expression in ovarian MNGCs. Bottom two panels show positive area (red) after DAB deconvolution and thresholding. (F) CD4 positive area of MNGCs from 3-4 individual MNGCs in four biological samples (#1-4). Quantification of CD3ε (G) and CD4 (H) positivity of MNGCs, stroma, and corpora lutea. (I) A schematic overview of our findings on ovarian MNGC functionality and interaction with T-cell populations. For F-H, data represents means ± SEM. For G and H, data were analyzed by one-way ANOVA with Tukey’s multiple comparison post-test (****, p < 0.0001). All transcriptomic and histological analyses were performed on 4 biological replicates. The data underlying the graphs can be accessed at https://doi.org/10.5061/dryad.kh18932j4.