IntroductionBrain aging is a complex biological process characterized by progressive structural and functional changes in the central nervous system. It is associated with cognitive decline and an increased risk of developing neurodegenerative diseases.1 Neuronal aging, in particular, plays a critical role in mediating the balance between healthy aging and disease progression. Structural and functional impairments in neurons directly affect synaptic plasticity and the stability of neural networks;2,3 therefore, understanding neuronal aging is considered a potential strategy to slow cognitive decline and prevent neurodegeneration. Notably, the compromised integrity of the blood-brain barrier (BBB) during brain aging is especially pronounced in the hippocampus. BBB dysfunction not only allows the infiltration of harmful substances but also damages neurons, contributing to cognitive impairment.4 Additionally, abnormal accumulation of Fe²⁺ in the brain increases with age. In a healthy brain, iron participates in numerous critical physiological processes, including myelination, neurotransmitter synthesis, and the function of antioxidant enzymes. Iron metabolism is tightly regulated to maintain intracellular iron concentrations within a physiologically safe range. However, under conditions such as aging, chronic inflammation, and oxidative stress, iron homeostasis becomes vulnerable to disruption, potentially resulting in pathological iron accumulation. After entering endothelial cells of the BBB, the transferrin receptor (TFRC), which serves as a key gatekeeper for brain iron uptake, mediates the transport of peripheral iron across the BBB into the central nervous system. Its expression displays age-associated dysregulation, implicating disrupted iron homeostasis as a critical factor in brain aging. Fe²⁺ is transported to the brain interstitial fluid via the iron exporter ferroportin (Fpn). However, significant gaps remain in our understanding of how this accumulated Fe²⁺ is delivered to specific brain regions, such as the hippocampus.5,6 In particular, the mechanisms by which Fe²⁺ induces hippocampal neuron dysfunction and age-related degeneration are still poorly understood.Under physiological conditions, iron is irreplaceable in enzymatic reactions as a reducing agent.7 However, elevated intracellular free iron facilitates the fenton reaction, which results in the generation of highly reactive hydroxyl radicals (·OH) and superoxide anions (O₂•−) via electron transfer. This in turn triggers a lipid peroxidation cascade, a process defined as ferroptosis,8 which is pathologically characterized by the abnormal peroxidation of membrane phospholipid polyunsaturated fatty acids (PUFA). Mitochondria function as the primary cellular energy-producing organelles. Reactive oxygen species (ROS), generated as byproducts of the mitochondrial electron transport chain, serve as essential substrates for the Fenton reaction. The consequent accumulation of lipid peroxides constitutes a hallmark biochemical event in ferroptosis and may contribute to neuronal damage.9 Smaller hippocampal volume and higher regional iron concentrations have been reported to be associated with lower memory scores.10 It is reasonable to suggest that ferroptosis in neurons may exacerbate neuronal aging and is worthy of further exploration.The key steps in the process of membrane phospholipid oxidation promoting ferroptosis involve converting fatty acids into lipoyl coenzyme A and PUFA- phospholipids, which are mainly carried out by members of the long-chain family of PUFA acyl-CoA synthetase (ACSL4) and lysophosphatidylcholine acyltransferase 3 (LPCAT3). Phosphatidylethanolamine (PE), a phospholipid with 20:4 and 22:4 acyl tails, exhibits the strongest association with susceptibility to ferroptosis.11 Under normal physiological conditions, PS is maintained in an asymmetric distribution on the inner layer of the cell membrane through phospholipid flip-flop enzymes. However, under pathological conditions, this distribution undergoes irreversible remodeling, which may be an active regulatory step of ferroptosis and directly determines the sensitivity of cells to ferroptosis. ATP11B is the catalytic component of the P4-ATPase flippase complex, catalyzing the hydrolysis of ATP and coupling it with the transport of aminophospholipids,12 including PS and PE.13,14 Current research indicates that ATP11B is involved in regulating synaptic plasticity in neurons,15 and its deficiency results in endothelial cell dysfunction13 and disruption of lipid metabolism in neurons,16 being closely related to brain aging.17 In this study, we identified ATP11B as a key regulatory factor in brain aging through multifaceted mechanisms. Utilizing spatial transcriptomics and single-cell RNA sequencing, we delineated the cellular and spatial molecular landscape influenced by ATP11B. We found that ATP11B deficiency promotes iron transport to hippocampal neurons and blocks neuronal mitochondrial respiration, causing an imbalance in mitochondrial dynamics and activating the hippocampal signaling pathway, thereby driving neuronal ferroptosis and accelerating neuronal aging. Additionally, epigenetic analyses revealed that ATP11B is essential for maintaining chromatin accessibility at genes involved in mitochondrial dynamics. Importantly, the therapeutic strategy of overexpressing ATP11B not only rescued memory ability but also reversed the ultrastructural degeneration of neurons. This study elucidated the mechanisms by which iron ions influence neuronal ferroptosis during aging, emphasizing the crucial role of mitochondrial function in the exacerbation of neuronal aging by ferroptosis. Targeting ATP11B therapeutically could disrupt the vicious cycle of iron homeostasis imbalance and mitochondrial dysfunction during brain aging, which is essential for the prevention and treatment of neurodegenerative disease and provides a potential target for the prevention and rescue of brain aging.ResultsATP11B deficiency enhances aberrantly accumulated Fe2+ in the brain during brain aging, impacting cognitive and memory functions Aging is accompanied by structural and functional impairment of the BBB.18 To investigate the pivotal genes that regulate the uptake and transport of Fe²⁺ into neurons during aging, we initially demonstrated an increase in sodium fluorescein uptake in mice, which serves as an indicator of increased BBB permeability in the brain tissue of aged mice (Fig. 1a). Compared with 3-month-old mice, 18-month-old mice displayed a notable decrease in the mRNA and protein expression levels of Claudin-5, Occludin, and ZO-1, which are key markers associated with BBB integrity, in the hippocampus and cerebral cortex (Fig. 1b-d, supplementary Fig. 1a-b). Subsequently, we investigated the relationship between the expression levels of certain genes and the extent of plasma uptake. The results show that a number of genes exhibited enhanced or inhibitory genetic correlations, including the mitochondrial-related gene mt-Rnr2, the transferrin receptor gene Tfrc, the gene related to central nervous system lipid transport Apoe, which is a core component of plasma lipoproteins, and the transport-related gene Atp11b (Fig. 1e). To identify key regulatory genes, an integrated analysis was conducted across three independent datasets.: correlation analysis of blood-brain substance transport capacity and gene expression, differentially expressed genes (DEGs) between brain endothelial cells (BECs) from young and old mice, and genes encoding transport-related proteins. Candidate genes were prioritized based on correlation analyses. Among the top-ranked genes were ABCG1, ATP11B, CD9, BOK, and ICAM1, with ATP11B emerging as a particularly compelling candidate (Fig. 1f). Quick S et al. showed that deficiency in ATP11B led to cerebral small vessel disease.13 Zhang et al. revealed the relationship between ATP11B and cellular lipid metabolism,16 demonstrating that deficiency in ATP11B accelerated brain aging through gut microbiota.19 Furthermore, functional analysis of aging-related DEGs indicate that ATP11B is closely linked to brain aging (supplementary Fig. 1c-g). The single-cell RNA-seq results show that ATP11B is relatively highly expressed in the choroid plexus, hippocampus, and cerebral cortex compared with other brain regions and tissues (based on Human Protein Atlas and scRNA-seq data) (Fig. 1g, h, supplementary Fig. 1h-l). The permeability of sodium fluorescein in the brain of Atp11b-/- mice was enhanced; and accordingly, the expression levels of tight junction proteins were significantly reduced, indicating that the BBB may be damaged in this group (Fig. 1i-l). Considering that Tfrc, one of the top genes in Fig. 1e, is an iron transporter, we further evaluated the content of Fe2+ in the brain (Fig. 1m, n). Interestingly, Fe2+ content in the brain tissue of aged mice and Atp11b-/- mice was significantly increased in a similar pattern. Notably, Fe²⁺ levels in the brain were comparably elevated in both aged mice and Atp11b-/- mice. which included the hippocampus and cortex brain regions. In addition, the knockout of Atp11b shortened the survival time of mice and accelerated age-related phenotypic decline (reduced vitality and loss of fur) (Fig. 1o, supplementary Fig. 1k). Additionally, results from the three-chamber social test reveal that Atp11b-/- mice exhibit abnormal social behavioral patterns compared with wild-type (WT) mice, with significant differences in the cumulative time spent in different chambers (NS, C, S). These findings suggest impairments in social preference and interaction (Fig. 1p). Furthermore, in comparison with control mice, we found a tendency of decreased latency to fall in Atp11b-/- mice (Fig. 1q). The results of behavioral experiments assessing memory and spatial memory ability indicate that Atp11b-/- mice displayed significant memory impairment as compared with WT mice of the same age. During the Morris water maze trial, Atp11b-/- mice were seen to spend a significantly shorter time in the target quadrant; however, there were no notable differences observed between the two groups of young mice (Fig. 1r, supplementary Fig. 1m). Spatial cognitive performance was assessed using the Y-maze. In both the young and aged groups, Atp11b-/- mice displayed significantly less duration and frequency in the novel arm compared with WT mice (Fig. 1s, t). Findings from the novel object recognition test demonstrate that Atp11b-/- mice, especially in the aged group, displayed a shortened exploration frequency and time for novel objects and were more inclined to repeatedly explore familiar objects, which may be related to hippocampus-dependent memory deficits20 (Fig. 1u, v). These findings indicate that deficiency in ATP11B during the aging process results in brain-localized accumulation of Fe2+, accelerated aging, cognitive decline, and memory impairment.Fig. 1The alternative text for this image may have been generated using AI.Full size imageATP11B deficiency enhances aberrantly accumulated Fe2+ in the brain during brain aging, impacting cognitive and memory functions. a Sodium fluorescein leakage levels in the brain parenchyma of young and aged mice (n = 4). b, c Expression levels of Occludin, ZO-1, and Claudin-5 in the hippocampus and cortex of 3-month- (3 m) and 18-month-old (18 m) mice (n = 4). d Immunoblotting showing the protein expression levels of Claudin-5, Occludin, and ZO-1 in the hippocampus and cortex of 3 m and 18 m mice. Tubulin was used as the internal control. e Spearman’s correlation of 19,899 BEC-expressed genes with blood-brain substance transport capacity. The top and bottom 1% of values are designated as ‘correlates’ and ‘anticorrelates’ respectively, with representative examples shown; the median of the distribution is 0.033. f Venn diagrams were generated using the differential genes of transporters, BEC DEGs, and gene correlation with blood-brain substance transport capacity GCBST. The data were obtained from the NCBI GEO:GSE134058 and GSE142500. g A detailed overview of ATP11B expression in any of the human brain areas is summarized as bar plots for each brain structure. Data source was the Human Protein Atlas. h Expression levels of Atp11b in young and old mice BECs. i Sodium fluorescein leakage levels in the brain parenchyma of WT and Atp11b-/- mice (n = 4). j, k Relative mRNA(j-k) and the protein(l) expression levels of Occludin, ZO-1, and Claudin-5 in the hippocampus and cortex of WT and Atp11b-/- mice (n = 4). m The brain Fe2+ level was evaluated in 3 m and 18 m mice (n = 3). n The brain Fe2+ level was evaluated in WT and Atp11b-/- mice (n = 3). o Kaplan-Meier survival curves depicting the probability of survival over time for WT and Atp11b-/- mice (n = 16). p Comparison of cumulative time between WT and Atp11b-/- mice in different scenarios of the three-chamber social test. The abscissa “NS”, “C”, and “S” represent scenarios with no social stimulus (empty cage), center chamber, and social stimulus (cage with age- and sex-matched mouse, respectively) (n = 10 per group). q Summary graph showing latency to fall during the rotarod test (n = 10 per group). r Long-term memory was assessed using the MWM, with the time spent in the platform zone being quantified (n = 10 per group). s, t Spatial memory was further evaluated using the Y-maze test, recording both the duration and frequency of entries into each arm (n = 10 per group). u, v The object recognition test was employed to measure spatial memory, duration of exploration and frequency directed toward novel objects documented (n = 10 per group). Data are shown as the mean ± SEM. Statistical importance was measured by using a two-tailed unpaired Student’s t-test (a–c, h–k, m–v), Log-rank (Mantel-Cox) test o, and two-way ANOVA followed by Tukey’s post-hoc test for multiple evaluations q. *P