Environmental Enrichment Attenuates Aging-Induced BBB Disruption and Cognitive Impairment with Activation of FNDC5/Irisin Signaling

Abstract

Aging disrupts the neurovascular unit (NVU) and blood–brain barrier (BBB), elevates glial inflammatory tone, and compromises hippocampal memory. Environmental enrichment (EE)—a multimodal, lifestyle-based intervention—improves cognition, but its association with BBB/NVU and FNDC5/irisin-related signaling in aging remains incompletely understood. Aged male C57BL/6J mice (21 months old) were housed under EE or standard conditions for 11 weeks. Hippocampal-dependent spatial working memory was assessed using the radial eight-arm maze, and neuronal (NeuN), glial (Iba1, GFAP), and BBB/NVU markers (AQP4 endfoot polarity, occludin, ZO-1, PECAM-1, microvessel length/density) were quantified. FNDC5/irisin-related signaling was evaluated by measuring PGC-1α, FNDC5/irisin, IGF-1, BDNF, pAKT, and serum irisin. EE improved spatial working memory in aged mice, reducing working-memory errors, increasing correct choices before the first error, and enhancing path efficiency. EE attenuated the age-related decline of NeuN(+) neurons in the hippocampal CA1 and CA3 regions and suppressed microglial and astrocytic activation. EE strengthened BBB/NVU integrity by restoring AQP4 endfoot polarity, increasing occludin, ZO-1, and PECAM-1, and increasing cortical microvessel length and density. At the molecular level, EE upregulated the PGC-1α–FNDC5/irisin–IGF-1 axis and was accompanied by increased cortical BDNF and pAKT levels, as well as elevated circulating irisin, changes that occurred in parallel with NVU stabilization and reduced glial activation. EE mitigates age-related cognitive decline in association with coordinated neuronal, glial, vascular, and FNDC5/irisin-related signaling changes, supporting BBB/NVU preservation and cognitive resilience during aging.

1. Introduction

Globally, population aging is accelerating due to declining birth rates and increased life expectancy [1,2]. This demographic shift has heightened interest in not only extending lifespan but also promoting healthy aging by preserving cognitive and physical function throughout later life. Aging triggers systemic physiological decline, with the brain particularly vulnerable to structural and functional deterioration that manifests as progressive cognitive impairment [1,3]. These changes, including reduced cerebral blood flow, synaptic loss, and neurovascular unit (NVU) disruption, elevate the risk of neurodegenerative diseases such as Alzheimer’s disease and Parkinson’s disease, even among cognitively intact older adults [4].

Age-related reductions in cerebral blood flow are closely associated with decreased capillary density and damage to the NVU [5]. Maintaining NVU integrity is essential for healthy brain aging. The NVU comprises neurons, astrocytic endfeet, pericytes, and endothelial cells. Endothelial cells, a key component of the NVU, form the structural barrier of the blood–brain barrier (BBB) through tight junction proteins such as occludin and ZO-1 [6]. The BBB preserves cerebral homeostasis by preventing the entry of harmful substances and pathogens [7]. Pericytes, which encase endothelial cells along the basement membrane, regulate BBB permeability, stabilize vessel structure, and modulate cerebral blood flow [8]. Astrocytic endfeet, which express aquaporin-4 (AQP4) water channels in a polarized manner, cover over 99% of brain capillaries and play crucial roles in mediating neuron–endothelial interactions and maintaining BBB integrity [9]. However, aging disrupts the integrity of the NVU. In aged endothelial cells, occludin and ZO-1 expression is reduced, resulting in increased BBB permeability [10], and perivascular cell coverage is markedly diminished [11]. Aged astrocytes exhibit flattened and abnormal morphologies and show reduced neuronal support [12].

In response to these age-associated pathological changes, environmental enrichment (EE) has emerged as a promising non-pharmacological intervention. EE entails housing rodents in large cages equipped with diverse objects, tunnels, and running wheels, thereby enhancing sensory, social, cognitive, and voluntary physical activity compared to standard housing [13]. In aged rodents, EE promotes hippocampal neurogenesis [14] and improves learning and memory performance [15]. Furthermore, EE enhances cerebrovascular integrity, including increased angiogenesis, tight junction protein expression (occludin, claudin-5), and reduced BBB permeability in both ischemic and neonatal hypoxia-ischemia models [16,17]. However, most studies have focused on disease models such as stroke or traumatic brain injury, and research examining EE’s impact on NVU/BBB structure and function in aging remains limited.

FNDC5, a transmembrane protein whose secreted ectodomain is termed irisin, is upregulated in skeletal muscle via PGC-1α during physical activity [18,19]. Irisin crosses the BBB and enhances cognitive function by promoting hippocampal BDNF and other neuroprotective genes [20,21]. It also exerts anti-inflammatory and antioxidant effects in the brain. Additionally, irisin has been shown to protect BBB integrity by maintaining tight junction proteins and modulating endothelial function [22,23]. Considering these mechanisms, EE—characterized by increased physical activity compared with standard housing—may stimulate FNDC5 expression. Notably, circulating irisin levels decline with age [24], making it essential to evaluate whether EE can restore irisin expression in aging models.

The present study aimed to determine whether EE enhances FNDC5/irisin signaling in aged mice and whether such molecular changes are associated with stabilization of the NVU and BBB and with reduced glial activation. Additionally, we evaluated whether EE ameliorates age-related deficits in hippocampal-dependent spatial working memory. By elucidating links between EE-induced FNDC5/irisin signaling and neurovascular and cognitive protection, this study explores the potential of EE as a feasible non-pharmacological intervention for mitigating cognitive aging.

2. Results

2.1. Environmental Enrichment (EE) Attenuates Aging-Related Cognitive Impairment in the Radial Eight-Arm Maze Test

To determine whether EE mitigates age-related working memory impairment, we assessed performance in the radial eight-arm maze test. Compared with young controls, aged mice showed impaired performance, evidenced by a greater number of working memory errors and fewer correct choices before the first error (F = 9.839, p < 0.001) (Figure 1A). Notably, aged mice exposed to EE made significantly fewer working memory errors and more correct choices before the first error than age-matched controls (F = 12.941, p < 0.001) (Figure 1B), indicating that EE ameliorates age-related working memory impairment. The total number of entries required to obtain all water rewards increased in aged mice but decreased in EE-exposed aged mice (F = 8.158, p < 0.001) (Figure 1C). Therefore, EE appeared to mitigate aging-related cognitive impairment.

Figure 1.

EE attenuates age-related cognitive impairment in the radial eight-arm maze test. Aged mice showed impaired performance relative to young controls, whereas EE exposure improved cognitive impairment in aged mice. (A) Correct choices before the first error. (B) Number of errors (re-entries into previously visited arms) required to obtain all water rewards. (C) Total number of entries required to obtain all the water rewards. Data are presented as mean ± SEM. Group sizes: young control (Y, 8 weeks old; n = 8), aging control (A, 24 months old; n = 8), and aging + EE (A + EE, 24 months old; n = 8). * indicates a statistically significant difference from the young group; # indicates a statistically significant difference from the aging group (one-way ANOVA followed by Tukey’s post hoc test; exact F and p values are reported in the Section 2).

2.2. EE Attenuates Age-Related Loss of NeuN(+) Neurons in the Hippocampal CA1 and CA3 Regions

In the hippocampus, aging was associated with a marked reduction in NeuN immunoreactivity and fewer NeuN(+) neurons in the hippocampal CA1 (F = 13.236, p < 0.001) and CA3 (F = 39.231, p < 0.001) regions (Figure 2A). Consistent with these observations, quantification revealed significantly fewer NeuN(+) cells in aged mice than in young controls in both regions. EE mitigated this loss: EE-exposed aged mice showed higher NeuN(+) cell counts than standard-housed aged mice in CA1 (p < 0.001) and CA3 (p < 0.001) (Figure 2B). These findings indicate that EE ameliorates age-related neuronal damage in the hippocampal CA1 and CA3 regions.

Figure 2.

EE attenuates age-related loss of NeuN(+) neurons in the hippocampal CA1 and CA3 regions. (A) Representative images of NeuN immunostaining in the CA1 and CA3 regions. (B) Quantification of NeuN(+) cell counts in the hippocampal CA1 and CA3 regions. Aged mice show fewer NeuN(+) neurons than young controls, whereas EE increases NeuN(+) cell counts relative to aged controls in both regions. Data are presented as mean ± SEM. Group sizes: young control (Y, 8 weeks old; n = 8), aging control (A, 24 months old; n = 8), and aging + EE (A + EE, 24 months old; n = 8). * indicates a statistically significant difference from the young group; # indicates a statistically significant difference from the aging group (one-way ANOVA followed by Tukey’s post hoc test; exact F and p values are reported in the Section 2). Scale bar: 50 μm.

2.3. EE Reduces Age-Related Activation of Microglia and Astrocytes in the Hippocampal CA1 and CA3 Regions

Iba1 immunostaining revealed increased microglial labeling in aged mice in the hippocampal CA1 and CA3 regions, reflected by a larger Iba1-positive area compared with young controls (Figure 3A). Quantification of the Iba1-positive area showed that EE-exposed aged mice had reduced microglial labeling compared with standard-housed aged mice in the hippocampal CA1 (F = 30.592, p < 0.001) and CA3 (F = 9.742, p < 0.001) regions (Figure 3B). Representative GFAP immunostaining in the hippocampal CA1 region shows stronger astrocytic labeling in aged mice than in young controls, whereas EE reduces GFAP labeling relative to aged controls (p < 0.001) (Figure 3C). Quantification of the GFAP-positive area in the hippocampal CA1 (F = 32.271, p < 0.001) region shows a significant group effect with aging increasing GFAP and EE reducing it compared with aged controls (p < 0.001) (Figure 3D). Together, these data indicate that EE mitigates age-related microglial and astrocytic activation in the hippocampus.

Figure 3.

EE reduces age-related activation of microglia and astrocytes in the hippocampus. (A) Representative Iba1 immunostaining in the hippocampal CA1 and CA3 regions. (B) Quantification of Iba1-positive area (% area) in hippocampal CA1 and CA3 regions. Aged mice show increased Iba1 labeling relative to young controls, whereas EE reduces the Iba1-positive area compared with aged mice. (C) Representative GFAP immunostaining in the hippocampal CA1 region. (D) Quantification of GFAP-positive area (% area) in the hippocampal CA1 region. Aging increases GFAP labeling, which is attenuated by EE. Data are presented as mean ± SEM. Group sizes: young control (Y, 8 weeks old; n = 8), aging control (A, 24 months old; n = 8), and aging + EE (A + EE, 24 months old; n = 8). * indicates a statistically significant difference from the young group; # indicates a statistically significant difference from the aging group (one-way ANOVA followed by Tukey’s post hoc test; exact F and p values are reported in the Section 2). Scale bar: 50 μm.

2.4. EE Attenuates Aging-Induced Microvascular Damage in the Cortex

Aging was associated with reduced endothelial marker signals in the cortex and a loss of polarity at astrocytic endfeet. PECAM-1 immunoreactivity was lower in aged mice than in young mice, whereas EE restored signal intensity and vascular continuity (Figure 4A). Aquaporin 4 (AQP4) immunostaining in the cortex shows prominent perivascular enrichment in young mice, which becomes diffuse with aging, indicating a loss of astrocytic endfoot polarity; EE visibly restores perivascular AQP4 enrichment. Microvascular skeletonization analysis showed that aging reduced both total microvessel length (F = 27.125, p < 0.001), and vessel density (F = 36.476, p < 0.001), whereas EE significantly increased both metrics compared with age-matched controls (p < 0.001) (Figure 4B). Quantification of perivascular AQP4 labeling shows a significant group effect, with aging reducing perivascular AQP4 (F = 17.094, p < 0.001) and EE increasing it compared with aged controls (p < 0.001). Dual-label immunofluorescence for endothelial (PECAM-1) and astrocytic endfoot (AQP4) markers revealed that aging decreased AQP4 coverage along PECAM-1-positive vessels and reduced their colocalization index. EE significantly increased perivascular AQP4 coverage and its colocalization with endothelial profiles, suggesting restoration of astrocytic endfoot polarity and endothelial–glial coupling (Figure 4C). Western blot bands for occludin, ZO-1, and PECAM-1 show reduced expression with aging and restoration by EE relative to aged controls (Figure 4D). Densitometric quantification, normalized to the loading control and expressed as % of young, showed that aged mice had lower levels of occludin (F = 12.642, p < 0.001), ZO-1 (F = 21.235, p < 0.001), and PECAM-1 (F = 9.601, p < 0.001) than young mice, whereas EE increased each protein relative to aged controls (p < 0.001) (Figure 4E).

Figure 4.

EE attenuates aging-induced microvascular deficits in the cortex. (A) Representative PECAM-1 immunostaining of cortical microvessels illustrating reduced total microvessel length and vessel density with aging, both improved by EE. Microvascular skeletonization quantification of total microvessel length and vessel density; aging reduces both metrics, whereas EE increases them compared with aged controls. (B) Representative AQP4 immunostaining illustrating loss of perivascular enrichment with aging and restoration by EE. Quantification of perivascular AQP4 labeling showing a significant group effect, with aging decreasing and EE increasing perivascular AQP4. (C) Dual-label immunofluorescence for PECAM-1 (endothelium) and AQP4 (astrocytic endfeet) and corresponding quantification showing reduced AQP4 coverage along PECAM-1 vessels and a lower colocalization index with aging, both improved by EE. (D) Representative Western blots for occludin, ZO-1, and PECAM-1. (E) Densitometric quantification normalized to the loading control and expressed as % of young; aging decreases all three proteins, whereas EE increases them relative to aged controls. Data are presented as mean ± SEM. Group sizes: young control (Y, 8 weeks old; n = 8), aging control (A, 24 months old; n = 8), and aging + EE (A + EE, 24 months old; n = 8). * indicates a statistically significant difference from the young group; # indicates a statistically significant difference from the aging group (one-way ANOVA followed by Tukey’s post hoc test; exact F and p values are reported in the Section 2). Scale bars: IHC, 100 µm; IF, 10 µm.

2.5. EE Modulates FNDC5/Irisin-Related Signaling in the Cortex

Aging was associated with a broad reduction in FNDC5/irisin pathway components and neurotrophic. As shown in Figure 5A, aged mice exhibited lower levels of FNDC5/irisin and its upstream regulator PGC-1α, along with reduced IGF-1 and phosphorylated AKT (pAKT). Neurotrophic support and redox defense were likewise diminished, evidenced by decreased BDNF (F = 5.029, p < 0.05). In contrast, EE significantly increased these proteins relative to aged controls, restoring them toward young levels across targets (p < 0.05). Total AKT was comparable among groups, whereas pAKT was selectively reduced by aging (F = 6.685, p < 0.01) and elevated by EE (p < 0.01) (Figure 5B). Consistent with the tissue data, circulating irisin also differed by group (Figure 5C). Serum irisin was lower in aged mice than in young mice (F = 17.184, p < 0.001) and was increased by EE relative to aged mice, approaching young levels (p < 0.001). Together, these findings indicate that aging suppresses the FNDC5/irisin–PGC-1α–IGF-1-BDNF-pAKT axis whereas EE mitigates these age-related declines.

Figure 5.

(A) Representative Western blots for PGC-1α, FNDC5/irisin, IGF-1, pAKT, total AKT, BDNF, and β-actin in the cortex. (B) Densitometric analysis of the blots in (A). PGC-1α, FNDC5/irisin, IGF-1, BDNF, and total AKT were normalized to β-actin. Phosphorylated AKT (pAKT) was quantified as the pAKT/total AKT ratio. All values are expressed as percent of the young group (Y = 100%). Aging reduced PGC-1α, FNDC5/irisin, IGF-1, BDNF, and pAKT, whereas EE increased these proteins relative to aged controls, restoring them toward young levels. Total AKT did not differ among groups, whereas pAKT was selectively reduced by aging and increased by EE. (C) Serum irisin levels measured by ELISA and expressed relative to the Y group. Aging lowered circulating irisin, whereas EE increased it relative to aged controls. Data are presented as mean ± SEM. Group sizes: young control (Y, 8 weeks old; n = 8), aging control (A, 21 months old; n = 8), and aging + EE (A + EE, 21 months old; n = 8). * indicates a statistically significant difference from the young group; # indicates a statistically significant difference from the aging group (one-way ANOVA followed by Tukey’s post hoc test; exact F and p values are reported in the Section 2).

3. Discussion

In the present study, we investigated the potential of EE to mitigate age-related cognitive impairment and associated brain pathology in aged mice, while exploring underlying molecular and vascular mechanisms. We found that EE effectively improved working-memory performance and conferred broad system-level benefits across the aging brain. Structurally, EE preserved NeuN(+) neurons in the hippocampal CA1 and CA3 regions and attenuated age-related neuroinflammation, evidenced by reduced microglial and astrocytic activation. Furthermore, EE restored neurovascular unit integrity by recovering microvascular density, tightening the BBB via upregulated tight junction proteins, and restoring astrocyte endfoot polarity (AQP4). At the molecular level, these systemic effects were accompanied by upregulation of PGC-1α–FNDC5/irisin–IGF-1 axis, along with elevated levels of the neurotrophic factor BDNF in brain tissue and increased circulating irisin. Collectively, these findings support the interpretation that EE is associated with coordinated FNDC5/irisin-related signaling, neurotrophic support, and redox regulation in parallel with preserved neurovascular integrity and cognitive function during aging.

The radial eight-arm maze test is a well-established behavioral assay heavily dependent on the hippocampus for spatial working memory [25]. In our study, EE significantly improved the radial eight-arm maze test performance in aged mice, evidenced by fewer working-memory errors, more correct choices before the first error, and increased efficiency (reduced total entries). The changes observed in latency were proportional to the accuracy metrics, which strongly support a genuine memory-specific enhancement rather than mere alterations in general exploration, locomotion, or motivation. These robust behavioral findings provide crucial and functional evidence of preserved hippocampal integrity by EE. Importantly, our results align with prior reports demonstrating that EE confers cognitive benefits across different pathological models, such as chronic cerebral hypoperfusion [26], suggesting that EE’s neuroprotective effects generalize across various aging- and vascular-related cognitive pathologies.

In the hippocampus, our quantitative analysis revealed that EE significantly attenuated the age-related reduction in NeuN(+) cell counts within the pyramidal layers of CA1 and CA3. This morphological preservation is highly consistent with the observed behavioral rescue and strongly suggests that EE mitigates age-related neuronal loss or, at minimum, preserves neuronal integrity. While extensive neuron loss is characteristic of severe neurodegenerative disorders such as Alzheimer’s disease, normal aging is often defined by a more restricted neuronal vulnerability—chiefly targeting specific cell subtypes—with cognitive decline primarily driven by functional impairments such as synaptic, biochemical, and circuit-level alterations [27]. In support of this concept, previous studies have consistently demonstrated heightened vulnerability of hippocampal neurons during aging [28], and our previous study similarly showed that exercise, a critical component of EE, can mitigate age-related neuronal damage [29]. Collectively, these findings are consistent with EE helping to preserve pre-existing NeuN(+) neuronal populations in CA1 and CA3, rather than reflecting gross neuronal loss and replacement. However, because we did not directly assess adult neurogenesis (e.g., DCX or BrdU labeling), we cannot exclude a contribution of newly generated neurons to the observed cognitive benefits, which will need to be addressed in future studies.

Age-related activation of microglia and astrocytes, as evidenced by increased Iba1 and GFAP expression in the hippocampus, has been reported in other aging mouse models, including naturally aged C57BL/6 mice showing hippocampal microgliosis and astrocytosis [30]. Consistent with reports that EE attenuates neuroinflammation, EE exposure in our study was associated with reduced glial hyperactivity. Under physiological conditions, these glial cells are essential: microglia support synapse maturation and injury surveillance, whereas astrocytes maintain circuit stability through glutamate uptake, ionic/water homeostasis, and vascular regulation. However, in aging, this supportive role gradually shifts toward a more detrimental phenotype. For instance, susceptibility to naturally occurring age-related cognitive impairment in C57BL/6J mice is associated with increased hippocampal MCP-1 despite preserved motor function, implicating a chronic central inflammatory tone [31]. Furthermore, with advancing age, microglia secrete higher levels of pro-inflammatory cytokines (IL-1β, TNF-α, and IL-6) and reactive oxygen species, while astrocytes exhibit reactive changes, including increased GFAP, loss of AQP4 endfoot polarity, and reduced glutamate transporters [32,33]. This persistent glial hyperactivity can depress long-term potentiation, reduce synaptic integrity, destabilize the neurovascular unit, and ultimately impair hippocampal-dependent memory. In our study, the age-related elevation of Iba1 and GFAP in the hippocampus was significantly attenuated, rather than fully normalized, by EE, suggesting that prolonged EE partially mitigates glial hypertrophy and pro-inflammatory tone within the context of chronic aging. We therefore interpret these changes as a phenotypic modulation of microglia and astrocytes toward a less reactive state, rather than a complete reversal of glial aging to a young-like profile. This interpretation is supported by our previous work showing that EE lowers central inflammatory markers and improves auditory function in aged rats [34], and reduces glial reactivity in white-matter inflammatory lesions in chronic cerebral hypoperfusion [35]. Thus, attenuating age-related glial hyperactivity and shifting glia toward a less reactive phenotype may contribute to EE-associated stabilization of the neurovascular unit and preservation of synaptic function, ultimately protecting cognition in aging.

Aging disrupts NVU integrity, including loss of astrocyte endfeet polarity, reduced tight junctions, and fragmentation of microvascular networks [36]. EE restored perivascular AQP4 enrichment and increased occludin, ZO-1, and PECAM-1, accompanied by longer and denser microvessels. These convergent astroglial and vascular changes imply improved endothelial–astrocyte coupling and a more stable barrier, providing a structural basis for reduced neuroinflammation and better neuronal outcomes. Our previous study likewise demonstrated that BBB disruption and NVU breakdown contributed to cognitive impairment in aged mice and that pharmacological stabilization of these elements improved cognition [37]. Consistent with these observations, exercise in chronic cerebral hypoperfusion enhanced cognition in parallel with BBB restoration and reduced glial reactivity [38]. Together, these data support the view that lifestyle-based interventions—whether EE or exercise—confer cognitive benefits via a shared neurovascular mechanism that couples attenuation of microglial/astrocytic activation with stabilization of the BBB and microvascular networks.

EE is a multimodal intervention comprising sensory, cognitive, social, and voluntary physical activity components. Previous studies have reported that long-term EE can enhance daily voluntary physical activity, motor performance, and muscle mass, suggesting that increased physical activity may emerge as a downstream consequence of EE rather than as a fully independent experimental variable [39,40]. In this context, the present EE paradigm should be viewed as a composite intervention in which voluntary exercise, enriched sensory stimulation, cognitive novelty, and social interaction act together, rather than as an exercise-free manipulation. Accordingly, the molecular and cognitive benefits observed in our study are best interpreted as EE-associated adaptations that include contributions from increased voluntary activity, and cannot be attributed specifically to non-exercise components of EE.

In this context, voluntary activity is known to activate the PGC-1α–FNDC5/irisin axis, as exercise paradigms robustly increase PGC-1α and irisin and are frequently accompanied by changes in IGF-1 and BDNF signaling [41,42]. In our study, EE exposure was associated with elevated PGC-1α, FNDC5/irisin, IGF-1, BDNF, and pAKT in the cortex, paralleled by increased circulating irisin despite age-related declines—consistent with exercise-induced FNDC5 cleavage [21]. These findings suggest partial re-engagement of residual signaling plasticity in aged mice under long-term and multimodal EE conditions. Because BDNF expression can be independently regulated through multiple signaling pathways, the observed increase in BDNF in the present study may partly reflect irisin-associated modulation within a broader adaptive response to EE, rather than a strictly linear signaling relationship. Although the PGC-1α–FNDC5/irisin–IGF-1 axis is known to be functionally blunted in aging, our findings indicate that this pathway remains inducible in 21-month-old mice. EE-induced increases in PGC-1α, FNDC5/irisin, IGF-1, BDNF and pAKT restored these molecules toward, but not beyond, young levels, consistent with partial re-engagement of a pathway that retains residual plasticity. Given that our EE protocol combined long-term voluntary activity with sensory, cognitive and social stimulation, we speculate that prolonged multimodal stimulation may provide sufficient chronic drive to recruit this attenuated axis, without implying full rejuvenation of youthful signaling capacity.

These molecular changes coincided with reduced glial activation, restoration of NVU features (AQP4 endfoot polarity, tight junctions, microvessel architecture), and improved radial eight-arm maze performance. Together, these coordinated adaptations support an association between EE exposure and irisin-linked trophic and antioxidant signaling in the NVU of aging brain. FNDC5 encodes a membrane precursor cleaved to release irisin, with αvβ5 integrins validated as functional receptors [43]. We therefore propose a correlation for FNDC5/irisin signaling in EE-associated neurovascular stabilization. Within this framework, FNDC5/irisin signaling may contribute to adaptive compensation in the aging brain, linking sustained EE exposure to partial preservation of BBB integrity, modulation of neuroinflammation and cognitive resilience. Importantly, the present study did not include an exercise-only control group, and therefore the relative contribution of voluntary physical activity versus non-exercise components of EE cannot be directly distinguished. Physical activity is a well-established modulator of neurotrophic signaling, including BDNF, and thus the observed molecular and cognitive effects should be interpreted as arising from a multimodal intervention of EE context rather than from exercise-independent mechanisms.

Although EE concurrently enhanced PGC-1α–FNDC5/irisin–IGF-1 signaling and stabilized BBB/NVU features, our data are correlative. To help establish causality, future work should combine targeted manipulation of FNDC5/irisin with receptor-level interrogation and direct BBB functional assays, ideally within factorial EE paradigms that disentangle voluntary activity from non-exercise components. Additional priorities include cell-type-specific analyses within the NVU (astrocytes, endothelial cells, pericytes, microglia), longitudinal time-course and dose–response studies, consideration of sex and strain differences, and biomarker validation linking circulating irisin to BBB/NVU endpoints and cognition. These approaches will refine mechanistic attribution and strengthen the translational pathway for enrichment-based interventions in brain aging.

Several limitations of the present study should be acknowledged.

First, our EE paradigm was designed as a multimodal intervention combining voluntary physical activity, sensory and cognitive stimulation, and social enrichment. We did not quantify individual wheel-running activity, nor did we include exercise-only or “EE without wheels” control groups. As a result, the relative contributions of voluntary exercise versus non-exercise EE components to the observed molecular and cognitive effects cannot be disentangled, and our findings should be interpreted as associations with the overall EE context rather than as exercise-independent mechanisms.

Second, in the EE cages we relocated the food hopper and water bottle once per week to introduce mild environmental novelty and encourage exploration. Although this manipulation may provide a low level of spatial stimulation, it was not structured as a task-specific learning paradigm and lacked explicit reward contingencies or trial-based working-memory demands; therefore, it is unlikely to have directly trained radial-arm maze performance, although we cannot fully exclude a minor contribution to overall cognitive engagement.

Third, while EE enhanced PGC-1α–FNDC5/irisin–IGF-1–BDNF–pAKT signaling in parallel with improved NVU/BBB and behavioral measures, these data are correlative. We did not perform targeted loss- or gain-of-function manipulations of FNDC5/irisin or its receptors, nor did we directly assay BBB permeability. Future studies using FNDC5/irisin knockout or overexpression models, receptor-level interrogation, and factorial EE designs will be required to establish causal pathways and to determine the specific contribution of irisin signaling to EE-induced neurovascular and cognitive protection.

4. Materials and Methods

4.1. Animals

Eight-week-old male C57BL/6J mice were purchased from Koatech (Pyeongtaek, Gyeonggi-do, Republic of Korea). The aged cohort was maintained until 21 months of age. Prior to environmental enrichment (EE) exposure, mice were randomly assigned to the following groups: young control (Y, 8 weeks old; n = 8), aging control (A, 21 months old; n = 8), and aging + EE (A + EE, 21 months old; n = 8). Animals were housed in a controlled environment (22 ± 2 °C; 60% relative humidity) under a 12 h light–dark cycle, with food and water provided ad libitum. All procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of Kyung Hee University (approval no. KHSASP-25-151; approved on 3 April 2025).

4.2. Environmental Enrichment (EE)

Two housing conditions were used. Before the intervention, all mice were maintained under standard housing in clear polycarbonate cages (26.5 × 32 × 18 cm; L × W × H) with four mice per cage until they reached 21 months of age. The intervention began when mice were 21 months old and continued for 11 weeks; thus, animals were approximately 24 months old at the time of analysis. Beginning at 21 months, the aging + EE group was housed for 11 weeks in multi-level EE cages (83 × 55 × 100 cm) with eight mice per cage to enhance social interaction. Each EE cage contained nesting material and a variety of objects with different colors and shapes, including two running wheels, tubes, a seesaw platform, plastic cubes, wooden blocks, and ladders. To maintain mild novelty and promote exploration, objects were replaced weekly and the positions of the food hopper and water bottle were changed weekly. Cages were cleaned weekly with tap water. Age-matched control mice (aging group) were maintained in standard cages for the corresponding duration.

4.3. Radial Eight-Arm Maze Test

Working memory was assessed in aged mice after completion of the EE intervention. Testing was conducted in week 11 of the housing protocol (i.e., after 11 weeks of standard or EE housing at 21 months of age). The radial eight-arm maze test apparatus for mice consisted of eight arms (60 × 12 × 12 cm) radiating at equal angles from a central octagonal platform. Opaque walls prevented direct visibility of the distal bait wells from the center. Illumination was maintained at 30–40 lux. Before testing, mice were allowed to acclimate to the radial arm maze for two days, and water intake was restricted for 48 h prior to the first experimental day. During the experiment, a drop of water was placed in a recessed hole at the end of each arm of all eight arms. At the beginning of each trial, a mouse was placed on the central platform and allowed to explore freely for up to 8 min or until all eight arms had been visited. An arm entry was defined as advancing beyond the arm’s midpoint with all four paws. Re-entry into a previously visited arm was scored as a working-memory error. Performance measures included the number of correct choices before the first error, the total number of working-memory errors, the total number of arm entries, and the latency to retrieve all baits. The maze was wiped with 70% ethanol between subjects to minimize olfactory cues, and outcome assessments were performed under blinded conditions.

4.4. Tissue Preparation

At the experimental endpoint, mice were deeply anesthetized with isoflurane delivered via a precision vaporizer, and the depth of anesthesia was confirmed by the absence of reflexes. For histology, animals underwent transcardial perfusion with ice-cold phosphate-buffered saline (PBS; 0.1 M, pH 7.4) followed by 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer. Brains were removed, post-fixed in 4% PFA for 24 h at 4 °C, and cryoprotected in 30% sucrose (w/v) in PBS at 4 °C until they sank (typically 24–72 h). Coronal sections were cut at 30 µm on a cryostat (−20 to −23 °C) and stored in antifreeze solution at −25 °C until staining. For protein analyses (Western blot), mice were perfused briefly with ice-cold PBS to minimize blood contamination, and brains were rapidly extracted on ice. Target regions were dissected, snap-frozen in liquid nitrogen, and stored at −80 °C until use.

4.5. Immunohistochemistry and Immunofluorescence

Free-floating coronal brain sections (30 µm thickness) were processed for chromogenic immunohistochemistry (IHC) and immunofluorescence (IF). All solutions were prepared in 0.1 M phosphate-buffered saline (PBS; pH 7.4) unless otherwise specified, and all incubations were performed with gentle agitation. For chromogenic IHC, endogenous peroxidase activity was quenched by incubating sections in 3% hydrogen peroxide (in PBS) for 20 min at room temperature (RT). Sections were rinsed in PBS (3 × 5 min), permeabilized in PBS containing 0.3% Triton X-100 for 10 min, and then blocked for 1 h at RT in blocking buffer composed of 1% bovine serum albumin (BSA) and 5–10% normal donkey serum in PBS to reduce non-specific binding. After blocking, sections were incubated overnight at 4 °C in primary antibody solutions prepared in PBS containing 0.1–0.5% BSA. The following primary antibodies were used at the indicated dilutions: NeuN (1:1000; 24307S, Cell signaling, Danvers, MA, USA), Iba1 (1:500; ab5079, Abcam, Cambridge, UK), glial fibrillary acidic protein (GFAP; 1:1000; ab53554, Abcam, Cambridge, UK), PECAM-1 (1:1000; NS-C348736, LSBio, LifeSpan BioSciences, Seattle, WA, USA), and aquaporin-4 (AQP4; 1:500; ab128903, Abcam, Cambridge, UK). The next day, sections were washed in PBS (3 × 5 min) and incubated for 1 h at RT with species-appropriate biotinylated secondary antibodies (donkey anti-rabbit IgG and/or donkey anti-mouse IgG, 1:200–1:500; Vector Laboratories, Burlingame, CA, USA). Following PBS washes, sections were incubated with the VECTASTAIN Elite ABC reagent (Vector Laboratories) for 1 h at RT according to the manufacturer’s instructions. Signal was developed with 3,3′-diaminobenzidine (DAB; Vector Laboratories, Newark, CA, USA) for 2–5 min under visual control to avoid overdevelopment, and the reaction was stopped by immersion in distilled water. Sections were mounted onto gelatin-coated slides, air-dried, dehydrated through graded ethanol (70%, 95%, 100%; 5 min each), cleared in xylene (2 × 5 min), and coverslipped with a resinous mounting medium. The obtained images were analyzed using the ImageJ 1.53 software program (NIH, Bethesda, MD, USA). For NeuN quantification, images were acquired at 40× magnification from one coronal section per animal, sampling both hemispheres. In the CA1 region, two non-overlapping fields were analyzed per hemisphere, whereas in the CA3 region, two non-overlapping fields were analyzed per hemisphere. Counts from all fields within each region were averaged to obtain a representative NeuN-positive cell count for CA1 and CA3 for each animal. The activation levels of GFAP and Iba-1, as well as the expression of AQP4, were analyzed based on the percentage area (% area). For vascular analysis, the “Vessel Analysis” plugin provided by ImageJ was utilized. Vessel Analysis is a plugin that automatically calculates vascular density metrics. Vascular density is defined as the vessel area divided by the total area (vessel area/total area × 100%) and vascular length density refers to the skeletonized vessel area divided by the total area, multiplied by 100% (skeletonized vessel area/total area × 100%).

For IF, separate sets of sections were rinsed in PBS and permeabilized/blocked for 1 h at RT in a buffer containing 2% BSA, 5–10% normal donkey serum, and 0.3% Triton X-100 in PBS. Sections were then incubated overnight at 4 °C with primary antibodies diluted in PBS containing 0.1% BSA. The following markers were probed: PECAM-1 (1:1000; NS-C348736, LSBio, LifeSpan BioSciences, Seattle, WA, USA), and aquaporin-4 (AQP4; 1:500; ab128903, Abcam, Cambridge, UK). After three PBS washes (5 min each), sections were incubated for 1 h at RT with fluorophore-conjugated secondary antibodies matched to the host species of the primaries (e.g., goat anti-rabbit Alexa Fluor 488 and goat anti-mouse Alexa Fluor 594; 1:500–1:1000; Thermo Fisher/Molecular Probes, Eugene, OR, USA). Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI; 0.5–1 µg/mL in PBS) for 5 min. Sections were rinsed in PBS, mounted onto gelatin-coated slides, and coverslipped with an antifade mounting medium. Fluorescence images were acquired using a confocal microscope (LSM 700, Zeiss, Oberkochen, Germany) under identical acquisition settings across groups. For AQP4 and PECAM-1 colocalization, confocal z-stacks were acquired across the thickness of the 30 µm sections, and maximum-intensity projection images were generated. Quantification of perivascular AQP4 coverage and colocalization with PECAM-1 was performed on these projection images using ImageJ (NIH, Bethesda, MD, USA).

4.6. Western Blotting

Brain tissue was homogenized on ice in RIPA lysis buffer (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with protease and phosphatase inhibitor cocktails (Thermo Fisher Scientific). Lysates were cleared by centrifugation (13,000× g, 30 min, 4 °C), and supernatants were collected. Protein concentrations were determined using a colorimetric assay (Bio-Rad Protein Assay Dye Reagent; Bio-Rad, Hercules, CA, USA) according to the manufacturer’s instructions. Equal amounts of protein (20 µg) were mixed and heated at 95 °C for 5 min, and resolved by SDS–PAGE on 10–12% polyacrylamide gels. Proteins were transferred to polyvinylidene fluoride (PVDF) membranes using a wet transfer system (100 V, 90 min, 4 °C). Membranes were briefly stained with Ponceau S to verify uniform transfer and loading, then rinsed in Tris-buffered saline with 0.1% Tween-20 (TBST; 10 mM Tris-HCl, 150 mM NaCl, 0.1% Tween-20, pH 7.6). Membranes were blocked for 1 h at RT in 5% non-fat dry milk in TBST. For phospho-protein detection, 3% BSA in TBST was used for blocking and antibody dilutions. Membranes were incubated overnight at 4 °C with the following primary antibodies diluted in TBST containing 3% BSA or milk, as appropriate: FNDC5 (1:1000; ab174833, Abcam, Cambridge, UK), PGC-1α (1:1000; NBP1-04676, Novus, Littleton, CO, USA), IGF-1 (1:1000; ab182408, Abcam, Cambridge, UK), BDNF (1:2000; sc-546, Santa Cruz Biotechnology, Dallas, TX, USA), AKT (1:1000; 9272, Cell Signaling, Danvers, MA, USA), phospho-AKT (Ser473; 1:1000; 9271S, Cell Signaling, Danvers, MA, USA), β-actin (1:10,000; sc-47778, Santa Cruz Biotechnology, Dallas, TX, USA). After three washes in TBST (20 min each), membranes were incubated with species-appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies (e.g., donkey anti-rabbit HRP or donkey anti-mouse HRP; 1:5000; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA or Cell Signaling Technology, Danvers, MA, USA) for 1 h at RT. Following additional TBST washes, immunoreactive bands were visualized using enhanced chemiluminescence (ECL; Clarity Max, Bio-Rad) and imaged on a ChemiDoc system (Bio-Rad Laboratories, Hercules, CA, USA) with exposure settings kept within the linear range. Densitometric analysis was performed using ImageJ (NIH, Bethesda, MD, USA).

4.7. ELISA

To quantify serum irisin concentrations, a mouse-specific irisin ELISA kit (ELK Biotechnology, Sugar Land, TX, USA) was used. The assay is based on a sandwich enzyme-linked immunosorbent assay (ELISA) principle. Each microplate was pre-coated with an antibody specific to mouse irisin. Following equilibration of all reagents and samples to room temperature, 100 μL of standards or serum samples were added to each well and incubated at 37 °C for 80 min. After incubation, the contents of each well were discarded, and the plate was washed three times with 200 μL of 1× wash buffer. The plate was then blotted dry using clean absorbent paper. Next, 100 μL of biotinylated antibody working solution was added to each well, followed by incubation at 37 °C for 50 min. The wells were again washed three times, and the plate was blotted dry as before. Subsequently, 100 μL of Streptavidin-HRP working solution was added to each well and incubated at 37 °C for another 50 min. After incubation, the wells were washed five times, and the plate was again dried using absorbent paper (Whatman, Cytiva, Maidstone, UK). Then, 90 μL of TMB substrate solution was added to each well, and the plate was incubated at 37 °C for 20 min in the dark. The enzymatic reaction was stopped by adding stop solution to each well, followed by gentle shaking of the plate for 1 min to ensure mixing. Finally, the optical density (OD) was measured at 450 nm using a microplate reader (Epoch; Bio-Tek Instruments, Winooski, VT, USA), and irisin concentrations were calculated based on the standard curve.

4.8. Statistical Analysis

Statistical analyses were performed using GraphPad Prism version 8.0.1 (GraphPad Software, San Diego, CA, USA). Data were assessed for homogeneity of variances using Bartlett’s and Brown–Forsythe tests. When variance homogeneity was satisfied, one-way ANOVA followed by Tukey’s post hoc test was applied. When the Brown–Forsythe test indicated unequal variances, Welch’s ANOVA with Games–Howell post hoc test was used. p-values less than 0.05 were considered statistically significant. Data are expressed as mean ± standard error of the mean (SEM).

5. Conclusions

Our findings show that EE improved hippocampal-dependent spatial working memory in aged mice and was accompanied by coordinated, multi-level changes with a prominent impact on the BBB and NVU. EE preserved NeuN(+) neurons in the hippocampal CA1 and CA3 regions, reduced microglial and astrocytic activation (Iba1, GFAP), and strengthened BBB/NVU integrity by restoring astrocyte endfoot polarity (AQP4), increasing tight junction proteins (occludin, ZO-1), elevating PECAM-1, and enhancing microvessel architecture and density—features consistent with improved barrier stability and neurovascular coupling. At the molecular level, EE increased PGC-1α, FNDC5/irisin, IGF-1, BDNF, pAKT, and raised circulating irisin. Together, these EE-associated adaptations align with BBB/NVU stabilization, preservation of hippocampal neurons, and improved memory performance in aging. While causality was not tested here, our results provide an integrated dataset linking behavioral benefit with coordinated neuronal, glial, vascular, and irisin-related signaling changes under EE, supporting the potential of lifestyle interventions for healthy brain aging.

Author Contributions

Conceptualization, Y.-J.K.; methodology, Y.J.C.; validation, J.M.L.; formal analysis, D.-E.S.; investigation, J.M.L. and Y.J.C.; resources, S.G.Y. and Y.-J.K.; data curation, D.-E.S.; writing—original draft preparation, J.M.L. and Y.J.C.; writing—review and editing, Y.-J.K.; visualization, J.M.L. and Y.J.C.; supervision, Y.-J.K.; project administration, S.G.Y. and Y.-J.K.; funding acquisition, Y.-J.K. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

The study was conducted according to The Kyung Hee University Guidelines for Institutional Animal Care and Use Committee (protocol code KHSASP-25-151; approved on 3 April 2025).

Informed Consent Statement

Not applicable.

Data Availability Statement

The data supporting the findings of this study are contained within the article. Further information is available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

This study was supported by the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (RS-2023-00208905).