Aging and sex: Impact on microglia phagocytosis

Abstract Microglia dysfunction and activation are important hallmarks of the aging brain and are concomitant with age‐related neurodegeneration and cognitive decline. Age‐associated changes in microglia migration and phagocytic capacity result in maladaptive responses, chronic neuroinflammation, and worsened outcomes in neurodegenerative disorders. Given the sex bias in the incidence, prevalence, and therapy response of most neurological disorders, we have here examined whether the phagocytic activity of aged microglia is different in males and females. With this aim, the phagocytosis activity of male and female cells was compared in an in vitro aged microglia model and in microglia isolated from adult (5‐month‐old) or aged (18‐month‐old) mice. In both models, the phagocytosis of neural debris increased with aging in male and female cells and was higher in aged female microglia than in aged male cells. However, female aged microglia lost its ability to adapt its phagocytic activity to inflammatory conditions. These findings suggest that microglia phagocytosis of neural debris may represent a previously unexplored neuroprotective characteristic of aged microglia that may contribute to the generation of sex differences in the manifestation of neurodegenerative diseases.

. The aged microglial phenotype is characterized by reduced migration and phagocytosis activity, as well as by exacerbated inflammatory responses (microglia priming) and deficits in chemotactic functions (Damani et al., 2011;Koellhoffer, McCullough, & Ritzel, 2017). Overall, age-related alterations in the expression of receptors implicated in innate immunity and phagocytosis, and the inability to mount a normal response to injury or inflammation, limit the capacity of microglia to cope with pathogens or neurodegeneration and contribute to an increased susceptibility and neurodegeneration (Liang, Domon, Hosur, Wang, & Hajishengallis, 2009).
Although there are robust sex differences in the epidemiology, clinical features, and pathophysiology of many neurological disorders (The Lancet, 2019), little attention has been paid to the sex differences in microglia function with aging. In this study, we aimed to determine the relevance of sex in the aged phenotype of microglia, analyzing two key functional responses of these cells: migration and phagocytosis. With this aim, we studied microglia isolated from adult or aged mouse brains (from 5-and 18-month-old animals, respectively) and in an experimental aging model in vitro after isolation of microglia from neonatal mouse brains.

| Microglia phagocytosis is affected by aging in a sex-specific way
As previously mentioned, microglia migration, motility (Damani et al., 2011;Hefendehl et al., 2014), and phagocytic activity are impaired with aging (Damani et al., 2011;Koellhoffer et al., 2017). Some studies had found sex differences in the phagocytic activity of microglia in early developmental stages or even in adulthood (Hanamsagar & Bilbo, 2016;Villa et al., 2018;Yanguas-Casás et al., 2018); however, possible sex differences in microglia phagocytosis had not been assessed in aged brains. Therefore, we decided to focus our studies on the phagocytic responses of microglial cells directly purified from adult and aged mouse brains. For this, we performed three different engulfment assays to evaluate: (a) nonspecific phagocytosis (quantifying fluorescent bead intake); (b) pathogen-specific phagocytosis (measuring Escherichia coli bioparticles uptake); and (c) neural debris phagocytosis (analyzing the intake of Cy TM 3-labeled neural debris) (Figure 1), using IFN-γ as a pro-phagocytic stimulus (Yanguas-Casás et al., 2018). Then, we measured the amount of internalized particles in actively engulfing microglial cells in the different experimental conditions. Male and female microglia purified from adult (5 months) brains showed similar internalization of uncoated beads (nonspecific phagocytosis), E. coli bioparticles (pathogen-specific phagocytosis), and neural debris under basal conditions ( Figure 1). The comparison of the results obtained with microglia purified from adult and aged brains revealed that aging had significant effects on basal microglia phagocytosis of neural debris. Thus, both male and female microglia isolated from aged brain (18 months) significantly increased the internalization of neural debris compared to microglia isolated from adult brains (Figure 1i). IFN-γ treatment increased nonspecific, pathogen-specific, and neural debris intake in both male and female microglia purified from adult brains (Figure 1c,f,i). However, adult female microglia showed a F I G U R E 1 Phagocytosis of microglia purified from adult (5 months) or aged (18 months) mouse brain. Representative images of microglia: nonspecific bead intake (a, b), pathogen-specific (d, e), and neural debris (g, h) phagocytosis. (c, f, i) Amount of internalized particles per cell. +++ p < .001 effect of time measured by two-way ANOVA followed by Bonferroni post hoc test; $$$ p < .001 sex differences, **p < .01, ***p < .001 effect of IFN-γ treatment measured by one-way ANOVA followed by Tukey's post hoc test. Dots show mean ± SEM. Blue: male; Purple: female; Dark: IFN-γ treatment much higher increase in bead internalization than adult male microglia ( Figure 1c). This sex difference was lost in microglia isolated from aged brains, in which bead phagocytosis was irresponsive to IFN-γ stimulation in both sexes ( Figure 1c). In addition, aged male microglia did not increase the internalization of E. coli bioparticles upon IFN-γ stimulation ( Figure 1f). Furthermore, IFN-γ stimulation was ineffective to increase neural debris internalization in microglia isolated from aged female brains. Thus, aging affects microglia phagocytosis in response to an inflammatory challenge in a sex-specific way.

| Perinatal male and female microglia acquire a senescent-like phenotype after 16 days in vitro
Previous studies had described an experimental model to reproduce irresponsive/senescent microglia in vitro (Caldeira et al., 2014); however, they did not evaluate the relevance of sex in the senescence process. For this reason, we characterized the senescent phenotype in microglia obtained separately from male and female mouse brains in this in vitro model.
We first determined β-galactosidase activity, which is increased in the senescence phenotype, at 2 and 16 days in vitro (DIV) in male and female microglial cells. There was a time-dependent increase in the senescent phenotype regardless of the sex (Figure 2a,b). We also found decreased miRNA-124a, miRNA-146a, and miRNA-155 expression, a characteristic of aged microglia, at 16 DIV in both sexes We next evaluated the mRNA expression of other senescence markers such as Beclin-1, which plays a central role in autophagosome formation, TLR2 and TLR4, which are associated with microglia activation, and interleukin (IL)-1β, IL-6, and tumor necrosis factor-α (TNF-α), whose expression by microglia has been shown to be altered with physiological aging and in disease (Caldeira et al., 2014;Koellhoffer et al., 2017). We found decreased Beclin-1, These results show that both male and female microglia acquire a senescent phenotype when kept in culture over 16 DIV, being the effect more pronounced in male than in female cells.

| IFN-γ induces a sex-specific inflammatory response in primary microglia that is altered in the in vitro aging model
We next stimulated microglial cells at 2 and 16 DIV, using IFN-γ as a pro-inflammatory stimulus. The levels of this cytokine are increased in the aged brain, and converging evidences point to its involvement in different mechanisms of aging (Monteiro, Roque, Marques, Correia-Neves, & Cerqueira, 2017).
As observed in the previous experiment, female microglia showed increased basal levels of IL-6 mRNA compared to male microglia at 2 DIV. However, by 16 DIV the basal mRNA levels of IL-6 were reduced and reached male values ( Figure 3b). In contrast, the basal levels of IL-1β mRNA expression were increased in male microglia by 16 DIV over basal female levels ( Figure 3a).
The effect of IFN-γ was also different in male and female cells.

| Sex differences in microglia motility disappear after 16 DIV
In a healthy brain, microglial cells constantly monitor their immediate surroundings by extension and retraction of their motile processes, allowing homeostasis maintenance and fine-tuning of neuronal activity. These cells also have the potential to move their soma, which allows a fast response for many pathophysiological processes (Kettenmann et al., 2011;Nimmerjahn, Kirchhoff, & Helmchen, 2005). Recent studies report that aged microglia show impaired migration and decreased motility and are unable to respond to several chemotactic stimuli in mice (Damani et al., 2011;Hefendehl et al., 2014). To test microglia motility in our aging model in vitro, we first analyzed the mRNA expression of migration-related genes. Therefore, microglia aged in vitro display a limited motility regardless of the sex, hence losing their physiological sex differences, and also lose the ability to respond to IFN-γ stimulation.

| In vitro aging alters the expression of phagocytosis receptors in microglia
As a first step to determine the effects of in vitro aging on microglia phagocytosis, we analyzed the mRNA expression of several genes

| Microglia aged in vitro show similar alterations in phagocytosis activity as microglia purified from aged brains
To investigate whether microglia phagocytosis was affected by sex in our aging model in vitro, we performed the same engulfment assays we had previously tested in microglia isolated from adult and  While those studies suggest that microglia are involved in the generation of sex differences in neurodevelopmental and psychiatric disorders and in the neurodegenerative response after acute traumatic brain injury or stroke in young animals, our findings have implications for the possible role of microglia in the generation of sex differences in the response of the aging brain to neurodegenerative conditions.

| D ISCUSS I ON
Reproducing microglia aging in vitro using long-term murine cultures provides a useful tool to study senescent microglia, given the limitations of isolating degenerating microglia from the aged brains for experimentation, as only the more resistant microglia will survive the isolation procedure, while susceptible microglia are lost in this process (Njie et al., 2012). Microglia isolated from newborn and adult brains maintain sex-specific features when in culture, such as postnatal sex differences in the phagocytic and migratory activity (Villa et al., 2018;Yanguas-Casás et al., 2018). Our findings also indicate that microglia derived from newborn male or female mice brains acquire a similar phenotype when cultured during 16 DIV.
This phenotype is characterized by increased β-galactosidase activity, decreased expression of specific miRNAs and mRNA levels of Beclin-1, TLR2, and TLR4, and reduced motility and mRNA expression of motility-related genes (Caldeira et al., 2014;Sieber, Claus, Witte, & Frahm, 2011). Furthermore, microglia maintained for 16 DIV show an altered inflammatory response when exposed to IFN-γ stimulation.
Although all these modifications are compatible with the phenotype of senescent microglia (Caldeira et al., 2014;Scheiblich, Trombly, Ramirez, & Heneka, 2020;Sieber et al., 2011), additional methods and molecular markers would be necessary to confirm cellular senescence in 16 DIV microglia cultures. Moreover, it is important to note that the molecular mechanisms that determine aging process of microglia in vitro may differ from those that cause microglia dysfunction in the aged brain (Stojiljkovic et al., 2019). For instance, the expression of Gal3 and MHCII in phagocytic cells has been shown to increase with age (Shobin et al., 2017). Here, we find that the mRNA expression of Gal3 decays in the aged phenotype of microglia, while MHCII mRNA expression levels remained unaffected. The expression of these markers is also linked to microglia priming (Holtman et al., 2015); therefore, these differences may be due to a different priming state of microglial cells aged in vitro or to the infiltration of macrophages and the expression of these markers by other cell types besides microglia in the aged brain. However, with independence of the differences in the triggering mechanisms, the observed functional characteristics of microglia aged in vitro are reminiscent of the functional changes that occur in microglia in vivo with the aging process, such as impaired inflammatory response (Scheiblich et al., 2020), decreased process motility, soma movement, and cellular migration and recruitment in the injured tissue compared to young microglia (Damani et al., 2011;Hefendehl et al., 2014).
In this study, our principal aim was to analyze the influence of aging and sex on microglia phagocytosis, a functional response of these cells that are involved in the regulation of brain tissue homeostasis under physiological and pathological conditions. Our findings Our results are in agreement with previous findings of sex differences in the activity of phagocytosis of developing microglia in vitro and in vivo (Weinhard et al., 2018;Yanguas-Casás et al., 2018) and However, further studies are necessary to unveil the precise mechanism and determine the functional consequences of this increase in phagocytosis activity by aging microglia. Nevertheless, as neuronal vulnerability increases with age, increased phagocytosis of neural debris might contribute to maintain homeostasis in the aged brain. Indeed, clearance of cellular debris from the parenchyma is essential to avoid further degeneration and exacerbated inflammatory responses in the brain under pathological conditions (Neumann et al., 2009). Alternatively, it may be hypothesized that the increased phagocytosis activity of neural debris by aged microglia could be part of the cell priming process associated with microglia senescence and may, therefore, contribute to brain deterioration with aging.
Another important observation in our study is that the phagocytosis of neural debris was impaired only in aged female microglia under inflammatory conditions elicited by IFN-γ stimulation. This effect of aging may be associated with the marked decrease in P2RY6 mRNA expression detected in aged female microglia, since P2RY6 mediates UDP-evoked phagocytosis of debris from damaged neurons (Koizumi et al., 2007). We can only speculate on the possible consequences of the impaired phagocytosis of neural debris by aging female microglia under inflammatory conditions. However, considering that brain aging is associated with increased neuroinflammation, our finding may implicate a decreased efficiency of microglia to maintain homeostasis in the aged female brain compared to the aged male brain or, on the contrary, a better control of the inflammatory stimulation of microglia phagocytosis by female cells to avoid further damage. In either case, the different phagocytic response of male and female microglia to inflammation may contribute to the generation of the well-characterized sex differences in the incidence of neurodegenerative diseases with aging (Loke et al., 2015;The Lancet, 2019). Even though IFN-γ is a factor aged microglia are physiologically exposed to (Monteiro et al., 2017), we cannot rule out that other inflammatory stimuli in vivo may elicit a different response or affect the one we are describing.
Finally, it should be emphasized that the sex-specific alterations in the phagocytosis activity of aged microglia are associated with modifications in the inflammatory response that are also different between male and female cells and occur in parallel with impaired cell motility, a characteristic of microglia in several neurological diseases (O'Connor, Borsig, & Heikenwalder, 2015;Raoul et al., 2010). The combination of all these circumstances most probably will magnify the consequences of sex dimorphic microglia changes in the aged brain.

| Reagents
DNase I, papain, and dispase II were purchased from Sigma-Aldrich.

| Microglia purification from adult mouse brain
5-and 18-month-old C57BL/6 male and female mice (4-5 animals per experimental group) were anesthetized with pentobarbital (Dolethal, 50 mg/kg body weight, intraperitoneal) and perfused transcardially with 0.9% saline. Adult or aged microglia were isolated as previously described (Lee & Tansey, 2013). Brain tissue was minced and digested at 37°C for 30 min with gentle shaking in a buffer containing papain, dispase II, and DNase I followed by mechanical dissociation. After neutralization of the reaction, cells were centrifuged and filtered by a 40-µm mesh. Microglia fraction was obtained in a 30%-70% SIP Percoll gradient by centrifugation of the cells at 500g and 18°C for 30 min, with no brake. After centrifugation, myelin was discarded and 6ml of the interphase was collected and mixed with 40 ml RPMI. The cells were centrifuged at 500g and 18°C for 7 min with regular brake, and the pellet was resuspended on 2ml RPMI.
After counting the cells, they were centrifuged at 168g for 10 min and resuspended in 180μL MACS buffer. Afterward, 10 μl/10 7 cells of CD11b microbeads were added to the mix. After washing, magnetically labeled cells were collected using MACS column system (Miltenyi Biotec).

| Microglia cultures from newborn mouse brain
Primary cultures of microglial cells were obtained from newborn (P0) to 2-day-old (P2) CD1 mouse forebrains. Pups were sexed via measurement of anogenital distance, and a separate cohort of animals was used for each experiment. Homogenized forebrains from male or female pups were grown separately in DMEM supplemented with 10% FBS, 10% HS, and P/S (DMEM 10:10:1) in 75-cm 2 flasks, coated with poly-l-lysine (10 μg/ml) as described previously (Mecha et al., 2011). Briefly, after reaching confluence, cells were shaken at 230 rpm for 3 hr at 37°C. Detached cells were centrifuged at 168g for 10 min. To avoid the estrogenic effects of phenol red, purified microglia were plated in warm antibiotic-and phenol red-free RPMI 1640 supplemented with 0.1% FBS. All the subsequent procedures were carried out using this medium.

| In vitro aging model
An age-like phenotype was induced in microglia cultures as previously described (Caldeira et al., 2014), with minor modifications.
After microglia purification from newborn brains, the cells were seeded on 6-well plates coated with poly-l-lysine (10 μg/ml) at a density of 100,000 cells/cm 2 for PCR analysis, 25,000 cells/cm 2 for senescence assays, or 50,000 cells/cm 2 for phagocytosis assays. Cells were maintained for 2, 10, or 16 days at 37º C and 5% CO 2 in RPMI medium containing 0.5% FBS. The cells were incubated for at least 12 hr in antibiotic-free serum-free RPMI prior to IFN-γ (20 ng/ml) treatment. Senescence, gene expression, phagocytic capacity, and motility of the cells were evaluated at the three time points.

| Cell senescence
Microglia senescence was evaluated using the Senescence Cells Histochemical Staining Kit (Sigma) according to the manufacturer's protocol. Microglia were seeded at a density of 21,000 cells/cm 2 and kept at 37°C in fresh RPMI with 0.5% FBS and P/S for 2 or 16 days.
At the selected times, cells were washed with PBS, fixed for 7 min at room temperature, and stained for 2 hr at 37°C. Images for quantification of β-galactosidase (senescent)-positive cells were acquired using a 10× lens by phase contrast in a Leica DMI6000 microscope.

| RNA purification of microglial cells and qPCR
To study microglial mRNA expression by quantitative PCR, microglial cells were seeded at a density of 100,000 cells/cm 2 and lysed 24h after treatment with IFN-γ at each time point (2 or 16 DIV). Total

RNA was extracted using an Illustra RNAspin Mini RNA Isolation
Kit (GE Healthcare) to assess the mRNA expression levels of the set of genes listed in Table 1 relative concentration should be between −0.1 and 0.1). The C t was determined for each target gene in duplicate. ΔC t was calculated by the difference between the C t of each target gene and the C t of an artificial BestKeeper reference gene based on the C t values of two independent reference genes: RPL13A and RPS29 calculated using the BestKeeper © software (http://gene-quant ifica tion.com/bestk eeper.html), which helps determine stable housekeeping genes, differentially regulated target genes, and sample integrity.

| MicroRNA purification
MicroRNAs were purified following manufacturer's protocol of Isolation Kit mirVana ™ . After acid phenol:chloroform extraction, miRNAs were isolated after washing with different ethanol concentrations, and collected with RNase-free water. Less than 5 ng/ µl miRNA were used in the retrotranscription/amplification phase.

| Neuron debris production and labeling
Neurons were obtained from the brains of mouse embryos as previously described (Hilgenberg & Smith, 2007). Mouse embryos were sexed prior to brain extraction, and brains were minced and digested with a trypsin solution at 37°C for 15 min, and after mechanical disaggregation, the cells were resuspended in neurobasal medium containing B27, ampicillin, and GlutaMAX TM . After pelleting the cells by centrifugation at 168g for 5 min, neurons were resuspended in sodium carbonate buffer (0.1 M NaHCO 3 -Na 2 CO 3 , pH 9.3) for optimal labeling and sonicated in a bar sonicator. Neuron debris was labeled using Cy TM 3 Mono-Reactive Dye Pack (Amersham Biosciences) according to the manufacturer's specifications and kept stored at 4°C until use.

| Phagocytosis assays
To determine the microglia phagocytosis, cells were seeded on 10-mm-diameter glass coverslips coated with poly-l-lysine 10 μg/ ml for in vitro microglia cultures or 50 μg/ml for adult mouse brain-derived microglia at a density of 50,000 or 25,000 cells/ cm 2 , respectively. After 24-hr incubation in serum-free RPMI or IFN-γ (20 ng/ml) treatment, the cells were washed twice with warm RPMI medium and the phagocytosis reagents were added for 1 hr.

| Time-lapse acquisition of microglia motility
To study microglial motility, male and female microglial cells were seeded at a density of 25,000 cells/cm 2 in Multiwell 6-well plastic plates (FALCON, Corning Incorporated-Life Sciences) coated with poly-l-lysine (10 µg/ml).
Microglia motility was analyzed 2 or 16 days after microglia seeding, in cells obtained from the same cell culture. Phase-contrast images of three fields per well were acquired every 2 min for 3 hr with a 203 0.70 DRY Leica DMI6000B lens in a time-lapse Leica AF 6500-7000 microscope and analyzed using Fiji software. Microglia motility was evaluated using two parameters: total area covered by the cells during the acquisition (expressed as the area in mm 2 covered by a single cell during a 3-hr period) and the wandering of the cells (displacement, measured as the area in µm 2 a single cell moves around in 1 min). Changes in cell shape and brightness did not allow a reliable automatic tracking, and therefore, this possibility was discarded. To determine the total area covered by each microglial cell, acquired images were stabilized and cropped, and the series of images was stacked with the StackReg plug-in. The area covered by each cell was measured in the projection in µm 2 . Total path distance wandered by each cell was determined by tracking the trajectory of each cell body during the acquisition period with the Manual Tracking plug-in.

| Statistical analysis
The data in bar graphs are expressed as the mean ± SEM, and the data in scattered plots are expressed as the median ± range of a representative replicate, being the cells from the same replicate at the three time points analyzed when in vitro. GraphPad Prism software version 5.0 for Windows and SPSS 22 software (IBM Corporation) was used for the statistical analysis. Normality of the data was assessed with the Kolmogorov-Smirnov test, to satisfy the assumption of normality for the analysis of variance (ANOVA). Whenever normality was not achieved, statistical significance was determined with nonparametric tests (Kruskal-Wallis and post hoc pairwise comparisons with Mann-Whitney U test). One-way ANOVA was used for comparison of multiple samples, followed by a Tukey's post hoc test to determine the statistical significance. Interactions between sex and treatment, sex and time, and time and treatment were determined using the two-way ANOVA interaction model, with post hoc Bonferroni's comparisons. Statistical significance was set at p < .05 in all cases.

ACK N OWLED G M ENTS
We thank Ms. Elisa Baides Rosell for excellent technical assistance. The study was supported by a grant from Agencia Estatal de Investigación (AEI), co-funded by Fondo Europeo de Desarrollo Regional (FEDER): BFU2017-82754-R and by CIBERFES.

CO N FLI C T O F I NTE R E S T
The authors of the manuscript declare no conflict of interest. They certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or nonfinancial interest (such as personal or professional relationships, affiliations, knowledge, or beliefs) in the subject matter or materials discussed in this manuscript.

AUTH O R CO NTR I B UTI O N S
NYC conceived and designed the study, and analyzed and interpreted the data. ACC and NYC conducted the experimental procedures. LMGS and MAA advised on the experimental design and contributed materials and animals. All authors contributed to manuscript writing and revision and approved the final version.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.