Aging is the underlying cause of neurodegenerative diseases such as Alzheimer's disease (AD) and Parkinson's disease (PD) [1-3]. Neurodegenerative diseases are common in the elderly population, and it is rare to find an elderly brain without disease [2]. The pathogenesis of neurodegenerative diseases is a complex interplay of genetic, environmental, and lifestyle factors [3]. Amyloid beta (Aβ) and tau accumulation, which are pathological hallmarks of AD that are associated with memory loss, gradually increase during the aging process in healthy adults leading to chronic inflammation in the brain [4,5]. Notably, women are more susceptible to AD compared to men due to the enzyme ubiquitin-specific peptidase 11, encoded in an X chromosome gene, which prevents tau degradation and promotes its abnormal aggregation [6]. However, the precise mechanisms underlying the association between aging and the striking vulnerability of women to AD remain elusive.

Aging is associated with a decline in blood-brain barrier (BBB) integrity, with initial BBB leakage occurring in the hippocampus, a region of the brain involved in memory and learning [7]. BBB leakage leads to increased levels of inflammatory mediators in the brain, which can damage neurons [8,9]. Microglia are resident immune cells in the brain, which play a protective role in defending the brain from damage [10,11]. Microglia function is often assessed by examining their morphology [11]. However, aging-induced microglial hyperactivation can produce proinflammatory cytokines and reactive oxygen species, which can damage neurons [12].

Galectin-3 modulates inflammatory response in the nervous system and is implicated in the pathogenesis of various neurological disorders, including AD [13-15]. Galectin-3 is expressed in immune cells, such as T and B cells, and is particularly upregulated in activated microglia during neuroinflammation [14,16,17]. Galectin-3 expression in microglia during neuroinflammation promotes inflammatory responses and Aβ oligomerization–induced phagocytosis [18]. High levels of galectin-3 have detrimental effects on hippocampal memory function [15,18,19]; however, the role of galectin-3 in the aged hippocampus is still not fully understood.

Here, we used 6-, 12-, and 24-month-old female mice to investigate the changes in neurodegeneration, BBB breakdown, and neuroinflammation-related proteins in the hippocampus. We found that microglial activation and galectin-3 levels increased with age. In addition, we found upregulation of galectin-3 in lipopolysaccharide (LPS)-treated primary mouse microglial cells. These findings suggest that galectin-3 may play an important role in microglial activation and neuroinflammation during brain aging.

Animals

Female C57BL/6J mice were purchased from Central Lab Animal Inc and maintained in the animal facility at Gyeongsang National University. All experiments were conducted in accordance with the protocol approved by The Animal Care Committee for Animal Research of Gyeongsang National University (GNU-190701-M0033). In addition, in vitro study was approved by the Institutional Animal Care and Use Committee of Gyeongsang National University (No. GNU-231109-M0212). Mice were divided into three groups: 6 months old (n = 6), 12 months old (n = 10), and 24 months old (n = 10). All mice were housed in virus‐free facilities on a 12 h light/12 h dark cycle.

Primary microglia culture

Whole brains from mouse pups at postnatal day 1 were obtained. After removal of the meninges, the brains were washed with Dulbecco's Modified Eagle's Medium (DMEM, GenDEPOT) three times. The brains were transferred to 0.25% Trypsin-ethylenediamine tetraacetic acid (EDTA) solution followed by 10 min of gentle agitation. DMEM/F12 complete medium was used to stop trypsinization. The brains were washed three times in this medium again. A single-cell suspension was obtained by trituration. Cell debris and aggregates were removed by passing the single-cell suspension through a 100 μm nylon mesh. The final single-cell suspension was thus achieved and cultured in DMEM/F12 (Gibco) supplemented with fetal bovine serum (Gibco) and penicillin-streptomycin (Gibco) on T-flask for 10 days. The medium was changed every 3 days. The mixed glial cell population was separated into an astrocyte-rich fraction and a microglia-rich fraction using the CD11b MicroBeats UltraPure mouse kit (Miltenyi Biotec), and the pour-off fraction containing microglia was separately cultured. Primary microglia cells were cultured in 6-well plates (1 × 10⁶ cells/well). After starvation induced, cells were treated with LPS of 500 ng/ml for 12 h and then harvested.

Tissue preparations

Mice were anesthetized with Zoletil (20 mg/kg, Virbac Laboratories) and Rompun (5 mg/kg, Bayer Korea). Following their extraction from the left ventricle, blood was centrifuged. Heparinized saline and 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) were used to perfuse the brain. The brains were fixed for 12 h at 4°C and then were successively submersed in 0.1 M PBS containing 15%, 20%, and 30% sucrose. The brains were embedded in a Tissue-Tek embedding medium (OCT compound, Sakura Finetek Inc.) and frozen in liquid nitrogen. Finally, the frozen brains were sliced into 40-μm coronal sections.

Cresyl violet staining

To determine neuronal neurodegeneration in aged mice, the brain sections were stained with 0.5% Cresyl violet (Thermo Fisher Scientific). The images were captured with a BX51 light microscope (Olympus). Photomicrographs of the CA1 and dentate gyrus (DG) were taken at 40× magnification. The thickness of pyramidal and granule cell layers of the hippocampus in CA1 and DG subregions were measured with ImageJ (Version 1.52a, National Institutes of Health) respectively.

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay

TUNEL was conducted on frozen tissue sections according to the manufacturer’s protocol using an in situ cell death detection kit (Roche Molecular Biochemicals). The sections were examined using an FV3000 fluorescent microscope (Olympus), and digital pictures were taken.

Enzyme-linked immunosorbent assay (ELISA)

Serum galectin-3 levels (n = 7 mice per group) were measured using a mouse galectin-3 (Abcam) ELISA kit according to the manufacturer’s protocols.

ProteoStat staining assay

To assay the aggresomes in the hippocampal CA1 region, we employed a ProteoStat aggresome detection kit (Enzo Life Sciences Inc.) in accordance with the manufacturer’s instructions. Free-floating brain slices were permeabilized with 0.1 M PBS containing 0.5% Triton X-100 and 3 mM EDTA for 30 min. ProteoStat dye was then added at a 1:2,000 dilution, and Hoechst dye for nuclear staining was added at a 1:1,000 dilution for 30 min at room temperature (RT). A confocal laser microscope model FV3000 (Olympus) was used to capture the images. The fluorescence intensity in each selected region was measured with ImageJ (Version 1.52a, National Institutes of Health).

Western blot analysis

The brain (n = 3–4 mice per group) was quickly removed from the mouse skull, and hippocampi were dissected and frozen. Hippocampal and cell lysates were homogenized in tissue protein extraction reagent (Thermo Fisher Scientific), incubated on ice for 30 min, and sonicated twice for 3 min. After centrifuging the hippocampal lysates for 30 min at 12,000 rpm at 4°C, the supernatants were transferred to new vials and kept at −80°C. Protein concentrations were measured using a bicinchoninic acid protein assay reagent kit (Thermo Fisher Scientific). Then, proteins (10–25 μg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis on 8%–12% acrylamide gels and transferred onto polyvinylidene fluoride membranes (Bio-Rad). Proteins were probed with primary antibodies; phosphorylated (p)Tau (ab108387, 1:1,000, Abcam), Tau (sc-32274, 1:1,000, Santa Cruz Biotechnology), zonula occludens-1 (ZO-1, sc-33725, 1:1,000, Santa Cruz Biotechnology), lipocalin-2 (LCN2, AF3508, 1:1,000, R&D Systems), glial fibrillary acidic protein (GFAP, G3738, 1:5,000, Sigma-Aldrich), interleukin-1β (IL-1β, sc-32294, 1:1,000, Santa Cruz Biotechnology), ionized calcium-binding adapter molecule-1 (Iba-1, #016-20001, 1:1,000, Wako), triggering receptor expressed on myeloid cells-2 (TREM-2, sc-373828, 1:1,000, Santa Cruz Biotechnology), and galectin-3 (sc-23938, 1:2,000, Santa Cruz Biotechnology). An enhancer (Thermo Fisher Scientific) was used to detect immune-reactive antigens, and anti-β-actin (A5441, 1:100,000, Sigma-Aldrich) antibody served as a loading control to normalize the signal intensity of each target protein. The target band density was determined using the Multi-Gauge V 3.0 image analysis program (Fujifilm).

Immunohistochemistry

After washing with cold 0.1 M PBS, free-floating brain slices (n = 3–4 mice per group) were blocked with 5% donkey serum for 1 h at RT and then incubated with primary antibody against anti-Iba-1 (#019-19741, 1:500, Wako) or anti-galectin-3 (1:200), overnight at 4°C, followed by incubation with secondary biotinylated antibody (Vector Laboratories) for 1 h at RT. The brain sections were then washed three times and incubated with Absolute Blocking Cocktail solution (ABC solution, Vector Laboratories) for 1 h at RT, washed, and developed with a 3,3'-diaminobenzidine peroxidase substrate kit (Vector Laboratories). The sections were then dried in graded alcohol solutions, cleared with xylene, and mounted with Permount (Sigma-Aldrich) under a coverslip. The immunostained sections were imaged and galectin-3-positive cells were counted in the hippocampus (500 × 500 μm²) in four sections using a microscope slide scanner (Motic).

Sholl analysis for microglial activation

To quantify the microglial activation, we used Fiji ImageJ (FIJI) open-source software and the Sholl analysis plugin [20,21]. After immunohistochemistry for Iba-1 antibody, capture images for Iba-1–positive cells were submitted in 8-bit format. The greatest intensity projections from the Iba-1-positive cells were thresholded to create a binary mask. For Sholl analysis, the maximum radius of the cell soma and the radius beyond the cell’s longest branch were measured, and the number of primary branches was manually counted. For each mouse, at least 30–50 cells were determined.

Double immunofluorescence

Free-floating brain sections (n = 3–4 mice per group) were washed with 0.1 M PBS three times for 5 min and then blocked with Mouse on Mouse (M.O.M.) Blocking Reagent (MKB-2213, Vector Laboratories) or 5% donkey serum in 0.1 M PBS for 1 h at RT. The brain sections were incubated with the primary antibodies; double aquaporin-4 (AQP4, sc-9888, Santa Cruz Biotechnology, 1:100) plus albumin (ab192603, 1:100, Abcam) and LCN2 (AF1857, 1:100, R&D Systems) plus GFAP (1:200, Sigma Aldrich) and [galectin-3 (1:200) plus TREM-2 (1:200)], diluted in 0.1 M PBS with 0.1% bovine serum albumin and 0.3% Triton X-100 overnight at 4°C. The next day, the sections were washed three times in 0.1 M PBS and incubated in the dark for 1 h at RT with donkey secondary antibodies conjugated with Alexa Fluor 488 and 594 (Invitrogen Life Technologies). Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI, Invitrogen). Fluorescence images were acquired using an FV3000 fluorescent microscope (Olympus). The intensity of immunostained AQP4 protein was measured in 12 fields (50 µm × 50 µm) from each section using ImageJ software (Version 1.52a, National Institutes of Health). For the counting of the number of extravascular albumin from hippocampal section, there fields (100 µm × 100 µm) were randomly selected from each section using ImageJ software (Version 1.52a, National Institutes of Health).

Quantitative real-time reverse-transcription PCR (qRT-PCR)

Total RNA was extracted from primary microglial cells using TRIzol reagent (Invitrogen) and reverse-transcribed using the RevertAid First-Strand cDNA Synthesis Kit (Fermentas Inc.) according to the manufacturer’s protocol. qRT-PCR was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems). PCR amplifications were performed using a SYBR Green qPCR Kit (TaKaRa) with specific primers; interleukin (il)-6 (Forward 5’ TAGTCCTTCCTACCCCAATTTCC 3’ and Reverse 5’ TTGGTCCTTAGCCACTCCTTC 3’), il-1β (Forward 5’ GCAACTGTTCCTGAACTCAACT 3’ and Reverse 5’ ATCTTTTGGGGTCCGTCAACT 3’), tumor necrosis factor (tnf)-a (Forward 5’ CCCTCACACTCAGATCATCTTCT 3’ and Reverse 5’ GCTACGACGTGGGCTACAG 3’). Relative quantification was performed using the DDCt formula, and relative mRNA gene expression levels were expressed as fold changes relative to a calibrator sample.

Statistical analysis

PRISM 10.0 was used for statistical analyses (GraphPad Software Inc.). Group differences were determined by one-way ANOVA followed by Tukey’s post-hoc tests. All results are presented as the mean and standard error of the mean (SEM). Significance was defined as a p-value less than 0.05.

Aging increases neurodegeneration in the mouse hippocampus

To investigate neurodegeneration in the hippocampus of adult (6-month) and aged (12-and 24-month) female mice, we determined the protein expression levels of phosphorylated Tau (pTau, Ser202). Western blot analysis revealed a significant increase in pTau protein expression in aged mice compared to adult mice (Fig. 1A). In particular, the ratio of pTau/Tau level was higher in 24-month-old mice compared to 6- and 12-month-old mice. Similar to the ratio of pTau/Tau level, protein aggresome accumulation in the hippocampus gradually increased with age (Fig. 1B). It was highest in the hippocampus of 24-month-old mice. Furthermore, TUNEL analysis showed that only TUNEL-positive cells were observed in 24-month-old mice (Fig. 1B). Cresyl violet staining revealed that 24-month-old mice had less Nissl substance in CA1 pyramidal layers compared to 6-month-old mice (Fig. 1C). Although there was no change in the thickness of DG, the CA1 thickness was lowest in the hippocampus of 24-month-old mice. These findings indicate that hippocampal neurodegeneration increases with aging.

Figure 1. Aging increases neurodegeneration in the mouse hippocampus. (A) Western blot analysis and quantification of pTau (Ser202)/Tau in the mouse hippocampus of 6-, 12-, and 24-month-old mice. β-actin was used as a loading control. (B) Representative images of the ProteoStat staining and TUNEL assay in the hippocampal CA1 region. The highly magnified images in the white boxes within the upper panels are shown in the bottom panels. The intensity of the ProteoStat dye is presented as fold changes. Scale bar = 50 µm (insert, 10 µm). (C) Cresyl violet-stained brain sections. The highly magnified images (CA1 and DG) in the white boxes are shown in the middle and lower panels. Quantification of the thickness of CA1 and DG. Scale bar = 100 µm (lower pannel, 20 µm). Data are shown as mean ± SEM. pTau, phosphorylated Tau; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling; CA1, cornu ammonis1, DG, dentate gyrus. *p < 0.05 versus 6M, ^†p < 0.05 versus 12M.

Aging causes BBB leakage in the mouse hippocampus

Age-related changes in the BBB are characterized by decreased expression of tight junction proteins, increased expression of water channels, and increased BBB permeability to proteins [8,9]. We found that the expression of tight junction protein ZO-1 was significantly decreased in 24-month-old mice compared to 6-and 12-month-old mice (Fig. 2A). AQP4 is highly expressed in astrocyte foot processes surrounding the BBB (Ref). So we performed a double immunofluorescence assay to assess BBB permeability to proteins (Fig. 2B). The intensity of immunostained AQP4 was gradually increased with age (Fig. 2C). It was highest in the hippocampus of 24-month-old mice. In particular, we found that albumin leakage was significantly increased in 24-month-old mice compared to 6-month-old mice (Fig. 2D). These results suggest that BBB leakage may play an important role in the aged mouse hippocampus.

Figure 2. Aging causes BBB-related proteins in the mouse hippocampus. (A) Western blot analysis and quantification of ZO-1 in the mouse hippocampus of 6-, 12-, and 24-month-old-mice. β-actin was used as a loading control. (B) Representative immunofluorescence images of the hippocampal CA1 region showing AQP4 (green) and albumin (red). The white arrowheads indicate extravascular or extracellular albumins. The highly magnified images in the white boxes are shown in right panuls. Nuclei were stained with DAPI. (C) AQP4 fluoresence intensity in Fig. B. (D) Quantification of albumin in the hippocampus. The white arrow head indicates extracellular albumin. Scale bar = 50 μm (insert, 10 μm). Data are shown as mean ± SEM. BBB, blood-brain barrier; ZO-1, zonula occludens-1; AQP4, aquaporin-4; DAPI, 4’,6-diamidino-2-phenylindole. *p < 0.05 versus 6M, ^†p < 0.05 versus 12M.

Aging increases neuroinflammation in the mouse hippocampus

LCN2, which is mainly expressed in the astrocytes that form the BBB, is associated with neuroinflammation and BBB leakage [22]. Compared to 6-month-old mice, 12- and 24-month-old mice had significantly elevated LCN2 expression (Fig. 3A). LCN2 was highest in the hippocampus of 24-month-old mice. Compared to 6-month-old mice, hippocampal GFAP protein was deceased in 12-month-old mice, but it was increased again in 24-month-old mice. Double immunofluorescence revealed that many LCN2-positive astrocytes (GFAP) were observed in the hippocampus of 24-month-old mice (Fig. 3B). Chronically activated microglia in neurodegenerative disorders emit excitotoxicity-like inflammatory cytokines [23]. In addition to hippocampal Iba-1 expression, the proinflammatory IL-1β, a marker of activated microglia, was significantly increased in 24-month-old mice compared to 6-month-old mice (Fig. 3C). To investigate aging-induced morphological changes in microglia, we performed immunohistochemistry to stain Iba-1 and observed cell morphology using Sholl analysis (Fig. 3D). Sholl analysis revealed a decrease in the number of ramification index and intersections in microglial morphology in 24-month-old mice compared to 6- and 12-month-old mice (Fig. 3E, F). These findings suggest that BBB leakage-related neuroinflammation may lead to significant morphologic changes in activated microglia.

Figure 3. Aging increases neuroinflammation and microglial activation in the mouse hippocampus. (A) Western blot analysis and quantification of LCN2 and GFAP in the mouse hippocampus of 6-, 12-, and 24-month-old-mice. β-actin was used as a loading control. (B) Representative immunofluorescence images of the hippocampal CA1 region showing LCN2 (green) and GFAP (red). The highly magnified images in the white boxes are shown. Scale bar = 50 μm (insert, 10 μm). (C) Western blot analysis and quantification of IL-1β and Iba-1 in the hippocampus of 6-, 12-, and 24-month-old-mice. β-actin was used as a loading control. (D) Representative immunohistochemistry images of the hippocampal CA1 region showing Iba-1 and a schematic to graphically illustrate the Sholl analysis of microglial morphology (color dots indicate Sholl intersections and circle lines indicate Sholl sphere radius). The areas in black squares in the left panels are magnified on the middle panels. Sholl analysis images (right panels) of microglia in middle panels. Scale bar = 50 μm (left panel), 10 μm (middle panels), 10 μm (right panels). (E) Ramification index. (F) Average number of intersections at specified distances from the soma in microglia. Data are shown as mean ± SEM. LCN2, lipocalin-2; GFAP, glial fibrillary acidic protein; IL-1β, interleukin-1β; Iba-1, ionized calcium-binding adapter molecule-1. *p < 0.05 versus 6M, ^†p < 0.05 versus 12M.

Aging increases microglial galectin-3 in the mouse hippocampus

Galectin-3 binds to TREM2 and activates microglia, leading to inflammatory responses [15]. Western blot analysis revealed a significant increase in hippocampal galectin-3 and TREM2 expressions in the 24-month-old mice compared to 6-month-old mice (Fig. 4A). Consistent with hippocampal galectin-3 protein expression, serum galectin-3 levels (p-value = 0.0114) were elevated in the 24-month-old mice (1,593.18 ± 32.51) compared to 12-month-old mice (745.77 ± 20.72). Immunofluorescence staining revealed that galectin-3-positive cells colocalized with TREM2-positive microglia in the hippocampus of 24-month-old mice (Fig. 4B). To exclude the galectin-3-positive macrophage in proximity to blood vessels or intravascular macrophage, we additionally performed immunohistochemical study for galectin-3 antibody (Fig. 4C). As shown in Fig. 4C, many galectin-3-positive microglia were significantly increased in 24-month-old mice compared to 6-month-old mice. These results indicate that galectin-3 may play a critical role in microglial activation in the hippocampus of aged mice.

Figure 4. Aging increases microglial galectin-3 in the mouse hippocampus. (A) Western blot analysis and quantification of galectin-3 and TREM2 in the mouse hippocampus of 6-, 12-, and 24-month-old-mice. β-actin was used as a loading control. (B) Representative immunofluorescence images of the hippocampal region showing galectin-3 (green) and TREM-2 (red) staining. Nuclei were stained with DAPI. Scar bar = 50 μm. (C) Representative immunohistochemistry images of the hippocampal region showing galectin-3. The areas in black squares in the upper panel are magnified on the lower panel. Black and red arrows indicate vessel and microglial galectin-3, respectively. Quantification of galectin-3 in the hippocampus. Scar bar = 100 (insert, 30 μm). Data are shown as mean ± SEM. TREM-2, triggering receptor expressed on myeloid cells-2; DAPI, 4’,6-diamidino-2-phenylindole. *p < 0.05 versus 6M, ^†p < 0.05 versus 12M.

LPS treatment increases galectin-3 expression in primary mouse microglial cells

To examine the effect of LPS treatment on galectin-3 expression on primary mouse microglial cells, we performed a Western blot analysis. As expected, like increasing Iba-1 and TREM2 expressions, galectin-3 was also elevated in LPS-treated microglia 12 h after LPS treatment (Fig. 5A). After that time, their protein levels were slightly decreased at 24 h. In addition, proinflammatory il-6, il-1β, and tnf-α mRNA expressions increased 12 h after LPS treatment, whereas il-1β and tnf-α mRNA expression decreased 24 h after LPS treatment (Fig. 5B). These results suggest that increased microglial galectin-3 may be linked to LPS-treated neuroinflammation.

Figure 5. LPS treatment increases galectin-3 expression in primary mouse microglial cells. (A) Western blot analysis of galectin-3, Iba-1, and TREM-2 protein expressions in LPS (500 ng/ml)-treated microglial cells. β-actin was used as a loading control. (B) mRNA expression levels of il-6, il-1β, and tnf-α genes in LPS-treated primary mouse microglial cells using qRT-PCR. Data are presented as the mean ± SEM from five independent experiments. LPS, lipopolysaccharide; Iba-1, ionized calcium-binding adapter molecule-1; TREM-2, triggering receptor expressed on myeloid cells-2; il, interleukin; tnf-α, tumor necrosis factor-α. *p < 0.05 versus 0 h.

Galectin-3 plays a crucial role in microglial activation, which has been associated with neuroinflammation and cognitive impairment. However, the role of galectin-3 in the aged brain is not completely understood. Here, we demonstrated that neurodegeneration, neuroinflammation, and BBB leakage were increased in the hippocampus of 24-month-old female mice compared to 6- and 12-month-old mice. In addition to the upregulation of galectin-3 in LPS-induced primary mouse microglial cells, microglial activation and galectin-3 levels increased with age. Taken together, our results suggest that microglial galectin-3 plays a key role in aging-related neurodegeneration, BBB leakage, and neuroinflammation.

Tau hyper-phosphorylation leads to neurofibrillary tangle formation and subsequent neuronal damage [24]. The age-related increase in pTau we observed suggests that aging may contribute to the risk and severity of neurodegenerative diseases, including AD and PD. Furthermore, Pang et al. [25] showed that Aβ accumulation and increased pTau induced cognitive decline and memory impairment in amyloid precursor protein (APP) knock-in rat models. In this study, we also showed that the expression of pTau increased with age. Together, these results suggest that increased pTau and protein aggresome accumulation in the hippocampus may contribute to aging-related cognitive decline.

Aging-related BBB breakdown exacerbates neuroinflammation, which can damage cells in the brain. ZO-1 is a tight junction protein that helps to anchor the cells of the BBB together, sealing them tightly to prevent harmful substances from entering the brain [26]. We demonstrated an age-dependent decrease in ZO-1 expression, indicating BBB breakdown in the aged hippocampus. AQP4 is a transmembrane water channel protein that facilitates the transport of water between the cerebrospinal fluid and the cytoplasm of astrocytes [27,28]. AQP4 is involved in various physiological processes, including brain water balance, neuroexcitation, astrocyte migration, and neuroinflammation [27]. In the present study, we uncovered increased AQP4 expression in the aged (24-month-old) brain, which may be a compensatory mechanism to repair the BBB leakage that occurs with aging. Albumin and AQP4 immunofluorescence analysis further supported our findings. We found that albumin, a protein that is ordinarily found in the blood, was expressed in and around AQP4-positive astrocytes in the hippocampus of 24-month-old mice, indicating albumin leakage into the brain. Albumin leakage into the hippocampus suggests BBB breakdown, which can potentially lead to neuronal damage and neuroinflammation. Our results provide evidence to support the hypothesis that BBB leakage worsens with age and may be associated with neuroinflammation.

LCN2 is found to be highly upregulated in astrocytes in response to different insults of brain damage and aging-related diseases [29]. In particular, LCN2 is known to be increased in the brains of Aβ oligomer–treated mouse models and postmortem patients with AD [22,30]. Our previous study suggested that astrocytic LCN2 is the systemic inflammatory mediator that plays a causative role in neuroinflammation in the diabetic brain via BBB breakdown [31]. Similarly, we observed an increase in LCN2 with age. LCN2 was shown previously to be localized in astrocytes in Aβ oligomer–treated mice, which are used as a mouse model of AD [22]. In the present study, we also observed an age-related increase in LCN2 expression in astrocytes. Eruysal et al. [32] showed increased LCN2 levels in the serum of pre-clinical patients with AD. Furthermore, a recent study suggested that LCN2 deletion reduces inflammatory factors, such as TNF-α and TREM-2, in the hippocampus [33]. Our results show that hippocampal LCN2 protein levels were increased in 24-month-old mice compared to 6-month-old mice, suggesting a potential association between aging and AD. The pro-inflammatory cytokine IL-1β increases in the hippocampus due to aging, and this increase is more predominant in females than males [34]. In agreement with these findings, we demonstrated that IL-1β levels were increased in 24-month-old mice compared to 6- and 12-month-old mice, suggesting that females may be more vulnerable to AD. Therefore, these data suggest that LCN2 may be crucial for developing therapeutic strategies targeting neuroinflammatory and age-related conditions.

Previous studies reported increased Iba-1-positive clusters and decreased microglia ramification length using senescent accelerated prone 8 (SAMP8) mice, a mouse model of accelerated aging [35]. Microglia are characterized by thin and long ramification in the ramified form, but their ramification shortens and thickens in the activated form [11]. Our study also found increased Iba-1 expression in 24-month-old mice compared to 12-month-old mice. Using Sholl analysis, we showed that the number of activated microglia was significantly higher in 24-month-old mice than in 12-month-old mice. These findings support the concept that microglial activation increases with age, potentially reflecting enhanced immune responses or progressing neuroinflammation in the aging brain.

A previous report revealed patients with AD had higher galectin-3 serum levels than healthy controls, and galectin-3 expression increased in proportion to the severity of memory impairment [36]. In transgenic APP/PS1 mice, endogenous Aβ oligomerization is correlated with an age-dependent rise in hippocampal galectin-3 expression from 3 months to 11 months [18]. Nonetheless, the present study is presented as a preliminary finding to support the role of galectin-3 in the progression of aging. Galectin-3 and TREM-2 interact to exacerbate neuroinflammation in AD [15]. In addition to LPS-induced galectin-3 in primary mouse microglial cells, our findings demonstrated that galectin-3 and TREM-2 expression increased with aging, suggesting they may exacerbate neuroinflammation in the aging brain. Galectin-3 is also known to be associated with phagocytic microglia in LPS-treated activated microglia [37] and has previously been shown to be localized to microglia in a diabetic mouse model treated with Aβ oligomers [38]. Similarly, our results showed that galectin-3 localized to microglia, not astrocytes, which increased with age. These findings suggest that the age-related increase in galectin-3 expression may be associated with phagocytic microglia. In addition, we observed both microglia that merged with galectin-3 and microglia that did not merge in 24-month-old mice, suggesting microglial galectin-3 may be involved in phagocytosis.

In conclusion, these findings suggest that microglial galectin-3 plays a key role in aging-related neuroinflammation and neurodegeneration in aged female mice. However, there are several limitations to this study. First, we found that microglial galectin-3 increases in the hippocampus with age but did not elucidate its precise function in aging. Galectin-3 function with respect to aging can be more clearly elucidated using 24-month-old galectin-3 knockout mice. Second, we need to further investigate the effects of galectin-3 on cognitive decline and memory impairment in aging using behavioral experiments. Finally, additional studies with male mice, in addition to female mice, are necessary to establish the function of galectin-3 in aging, especially concerning sex-specific differences in neuroinflammation and neurodegeneration.

The authors extend their deepest appreciation to all the participants for their invaluable support for this study.

The authors declare no conflicts of interest.

This study was supported by a grant from the Basic Science Research Program through the National Research Foundation of Korea (No. 2021R1A2C2093913, 2022R1A1A01067302, and RS-2023-00219399).