Macrophage multinucleated giant cells – hallmarks of the aging ovary

Abstract

Inflammaging is a state of chronic, low grade inflammation associated with biological aging which contributes to age-related diseases. Recently, an age-associated increase in inflammation has been documented in the mammalian ovary, which is accompanied by a shift in the immune cell profile. In this Point of View article, we consider a unique population of macrophage-derived multinucleated giant cells, found in reproductively old mouse ovaries, as potential markers or functional drivers of inflammation in ovarian aging.

The immune system undergoes profound changes over the course of a human lifetime, including deterioration of the adaptive immune system and an increase in innate immune activity. These changes are manifested in the form of increased numbers of natural killer cells and pro-inflammatory cytokines likely released by macrophages (Weiskopf et al., 2009). The increasing levels of innate immune activity can become pathogenic and lead to an acceleration of normal biological aging. This chronic, sterile, low-grade inflammation in the absence of overt infection is termed “inflammaging” and is being explored in the context of age-dependent diseases such as atherosclerosis, Type II diabetes, and cancer (Franceschi and Campisi, 2014).

Inflammaging in the ovary is just beginning to be explored but may have important implications for fertility and overall health. The female reproductive system is the first to age in humans and is characterized by a decline in both the number and the quality of oocytes (Broekmans et al., 2009). Inflammation and fibrosis occur in ovaries of mice with advanced reproductive age, which could negatively affect oocyte development and lead to additional sequelae of reduced gamete quality, including meiotic nondisjunction and aneuploidy (Zhang et al., 2020). The aging ovarian environment is characterized by an increase in collagen I and III as well as increased expression of the proinflammatory cytokines interleukin 1β (IL-1β) and tumor necrosis factor alpha (TNFα) (Briley et al., 2016). IL-1β and TNFα are commonly secreted during acute inflammatory responses by innate immune cells including macrophages. There is also a dramatic age-dependent increase in the production and secretion of the pleotropic cytokine, interleukin-6 (IL-6) (Briley et al., 2016). When secreted by macrophages, IL-6 has pro-inflammatory effects, and serum levels of IL-6 are a marker of inflammaging in large cohort studies (De Martinis et al., 2005). These findings implicate an important role of immune cells in reproductive aging, specifically with regards to macrophages. A unique population of ovarian macrophage-derived multinucleated giant cells (MNGCs) may serve as important beacons and drivers of ovarian aging (Figure 1).

Figure 1.

Ovarian Multinucleated Giant Cells. (A) Mouse ovarian tissue section stained with hematoxylin and eosin (H&E) highlights clusters of MNGCs (light brown staining, black arrows). (B) An electron micrograph of an ovarian tissue section at 3000X showing three stromal fibroblasts (asterisks highlighting nuclei) abutting a MNGC containing vacuoles, inclusion bodies, and a foamy appearance. (C) Schematic illustrating the potential causes and consequences of ovarian MNGCs. In panel (i), MNGCs are visible in a cluster (black arrow) in an ovarian section from a reproductively old mouse stained with an antibody against α-SMA. Note the close proximity of these MNGCs to a growing follicle. The scale bar is 200 μm. In panel (ii), a single MNGC is shown containing 5 nuclei visible in the same plane of an H&E stained histological section (left) from a reproductively old mouse. This cell is autofluorescent likely due to lipofuscin accumulation (right). The scale bar is 10 μm.

Macrophages are mononuclear, phagocytic cells developmentally derived from yolk-sac macrophages and fetal liver monocytes or postnatally derived from bone marrow monocytes. Upon activation, macrophages can be categorized into the M1 (classically activated) and M2 (alternatively activated) subgroups based on phenotype. M1 macrophages traditionally play a critical role in the acute stages of inflammation and respond to the presence of damage or danger associated molecular pattern (DAMP) molecules such as extracellular matrix fragments (e.g. low molecular weight hyaluronan), and normally intracellular molecules (e.g. nuclear and mitochondrial DNA) and pathogen associated molecular patterns (PAMPs) (e.g. lipopolysaccharide). Cytokines such as interferon-γ (IFN-γ) can also result in M1 macrophage polarization (Ivashkiv, 2018). M1 macrophages contribute to the inflammatory response by producing pro-inflammatory cytokines IL-6 and TNFα, the same cytokines reported to be abundant in aged ovaries (Briley et al., 2016). By responding to and secreting pro-inflammatory cytokines, M1 macrophages contribute to a positive feedback inflammation loop. On the other hand, M2 macrophages are active during the later stages of infection or response to tissue injury when they play key roles in inflammation resolution and wound healing, which could result in fibrosis if injury persists. These macrophages secrete anti-inflammatory cytokines such as interleukin-4 and interleukin-13 (Hesketh et al., 2017). Although these macrophage groupings are useful for high level classification, it may be more accurate and biologically relevant to consider macrophage phenotypes on a spectrum rather than in discrete groups because macrophages change in response to the environment (Stout et al., 2005).

In the ovary, macrophages have well-developed roles in multiple biological processes, including folliculogenesis, ovulation, and corpus luteum formation and regression (Wu et al., 2004). With advanced reproductive age, there is not a change in the overall number of macrophages as assessed by the pan-macrophage marker (F4/80) using an immunofluorescence-based histological approach. However, there is a unique population of multinucleated macrophage giant cells (MNGCs) which are highly penetrant in reproductively old mice across strains but are absent in ovaries from reproductively young mice (Briley et al., 2016) (Figure 1A). The data on MNGC are limited, and more research is needed to fully elucidate the role of these cells in the ovary and ovarian aging. However, we posit that the presence of MNGCs in the aging ovary is likely indicative of a failure of normal physiology given the role of similar giant cells in other tissues. Generally, MNGCs are specialized phagocytic cells that result from the fusion of macrophages, usually in response to infection or a foreign body, and are commonly found at sites of chronic inflammation (Milde et al., 2015). For example, MNGCs are found in granulomas, a collection of immune cells that form to effectively “wall off” areas of inflammation from healthy tissue (Timmermans et al., 2016). Macrophages and MNGCs in granulomas are activated by IFN-γ and can secrete proinflammatory cytokines such as interleukin 12 (IL-12), IL-1β, and TNFα to amplify the inflammatory response. MNGCs are also found in autoimmune conditions, including rheumatoid arthritis, autoimmune myocarditis, and autoimmune enteropathy. Another subtype of MNGCs, foreign body giant cells (FBGCs), are commonly found at the site of surgical implants and may be specialized to ingest large targets. It is thought that FBGCs form when mononuclear M2 macrophages are not able to efficiently phagocytose their target due to size (Milde et al., 2015). As the degradation of an implant is impossible for one macrophage, FBGCs form to enhance degradational capacity. There are also examples of MNGCs involved in normal physiologic functions, such as osteoclasts. Osteoclasts are a specialized subtype of bone macrophage which form MNGC to absorb bone tissue and are integral to both bone maintenance and repair (Boyce et al., 2009).

Given their penetrance with age, MNGCs likely play a central role in ovarian aging, but a more comprehensive understanding of this unique population is necessary. Based on initial characterization, these cells stain positively for F4/80, a cell surface marker that is reliably expressed by murine monocyte-derived macrophages, but staining is variable at the cell periphery and is sometimes reduced or absent altogether (Briley et al., 2016). This decreased staining may reflect fusogenic activity between macrophages of different lineages. These cells also stain light brown with H&E (Figure 1A, C), appear foamy in electron micrographs (Figure 1B), and are autofluorescent (Figure 1C) due to lipofuscin accumulation (Briley et al., 2016). The ratio of M1/M2 macrophages in the ovary changes with age, resulting in more M2 macrophage markers in aged ovaries (Zhang et al., 2020). This is coupled with a reduction in classic M1 macrophage markers, such as iNOS. It is tempting to speculate that the increase in M2 macrophage markers is due to MNGCs migrating into the ovary because their appearance correlates with increased expression of M2-associated markers such as Arginase 1 and resistin-like molecule alpha (Retnla). If MNGC are in fact M2 macrophages, these cells may have a role in fibrosis or fibrosis resolution, but further experiments are needed to validate this possibility.

In addition to further characterizing the unique population of MNGCs in the aging ovary and their function, a major unanswered question is what triggers their presence (Figure 1C). MNGCs may appear in the ovary in response to the continued accumulation of debris in the ovary over time. In fact, this type of debris-clearance behavior is observed in osteoclasts (Boyce et al., 2009). Throughout their reproductive lifespan, mammals including mice and human, will undergo hundreds of ovulatory cycles including follicular rupture, wound healing, repair of the ovarian surface epithelium, and formation and regression of corpora lutea. Repeated ovulation resulting in cycles of injury, and incomplete repair may lead to extracellular matrix (ECM) accumulation and eventual organ dysfunction. Moreover, if clearance mechanisms are also impaired, cellular debris may accumulate over time and ultimately trigger an inflammatory response.

The cumulative number of ovulations, however, cannot be the sole explanation for reproductive aging phenotypes because women who are on oral contraceptives, and therefore experience ovulation suppression, undergo menopause at similar ages as women who are not (Vries, 2001). Another contributing factor to reproductive aging phenotypes and the presence of MNGCs may be the significant amount of cell death that occurs in the ovary under normal physiologic conditions. Primordial follicles in the ovary are continuously activated, but only a fraction of the total number of follicles ultimately reach the large antral follicles stage and ultimately release an egg at the time of ovulation. The rest of the follicles undergo atresia before the oocyte is fully matured, and this process can occur at any follicle stage (Kim and Tilly, 2004). This atresia results in a large amount of cellular debris which must be cleared. As more and more follicles are activated and become atretic over the course of a reproductive lifespan, the ability of the ovary to effectively clear this debris may be reduced. Thus, the presence of MNGCs may represent the need for enhanced degradational capacity in response to this accumulation and ultimately contribute to a positive inflammatory feedback loop via cytokine secretion (Figure 1C). The accumulation of cellular debris may reach a critical threshold and serve as a biological timer of ovarian aging.

Changes in the ovarian ECM may also play a critical role in generating the unique aging innate immune response. For example, hyaluronan is a ubiquitous glycosaminoglycan that is found in the ovarian stroma (Mara et al., 2020). Low molecular weight hyaluronan fragments induce an inflammatory response in ovarian stromal cells, in vitro, that partially recapitulates the response observed in the aging ovary (Briley et al., 2016; Rowley et al., 2020) In addition, collagens in general can be extremely long-lived proteins which can be post-translationally modified over time, and ultimately these new protein forms may become immunogenic (Toyama and Hetzer, 2013). Collagen I and III increase in the aging ovarian stroma, and it will be interesting to determine whether these ovarian ECM components are long-lived, accumulate damage and post-translational modifications, or become cross-linked with aging. Thus, MNGCs may exist in the ovary in an attempt to clear collagen-dense fibrotic foci, and these regions may play a nucleating role in ovarian inflammation and adaptive immune responses to highly modified collagen (Figure 1C). Interestingly, ovarian MNGCs are commonly found in close proximity to cells that stain positive for α-smooth muscle actin (α-SMA) (Figure 1C) (Briley et al., 2016). α-SMA is indicative of fibroblast activation, which is associated with increased collagen secretion. Increasing collagen in the aging ovary may have an endocrine origin because mice lacking aromatase, a key enzyme in the estrogen biosynthetic pathway, have ovaries with increased collagen, significant follicle loss, and a large influx of macrophages (Britt et al., 2000). Thus, falling levels of estrogen characteristic of reproductive aging may regulate collagen deposition in the ovary and the presence of MNGCs (Figure 1C).

The reasons proposed here for why MNGCs are found in the aging ovary may not be mutually exclusive, and future studies are warranted to decipher whether these cells are a cause or consequence of reproductive aging. Such understanding will enable the use of MNGCs and their secretome as potential biomarkers of reproductive aging. Alternatively, their selective clearance may serve as a novel therapeutic target to promote reproductive longevity.