A cholesterol-coupled N-acetyl-aspartyl-glutamate metabolic network facilitates the neuroprotective impact of estradiol in neurons

doi:10.21203/rs.3.rs-4489289/v1

Download PDF

Article

A cholesterol-coupled N-acetyl-aspartyl-glutamate metabolic network facilitates the neuroprotective impact of estradiol in neurons

https://doi.org/10.21203/rs.3.rs-4489289/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Sex differences have been demonstrated in Alzheimer's disease (AD), yet the intrinsic cellular changes underlying the enhanced disease vulnerability observed in postmenopausal women remain elusive. We demonstrate that sustained loss of peripheral estradiol is correlated with accelerated cognitive and memory decline. The resulting brain transcriptomic and metabolomic changes suggest the impairment of ERRα. Estradiol supports ERRα activity via its actions on neuronal cholesterol homeostasis. Consequently, this prevents truncation of the TCA cycle at succinate dehydrogenase, which would otherwise cause a net catabolic shift of N-acetyl-aspartyl-glutamate (NAAG), driven by an adaptive aspartate-dependent response that attempts to reconstruct a “mini-cycle”. The free glutamate released alongside the net catabolism of NAAG is stochastically released presynaptically, thereby increasing spontaneous neuronal activities. Coupled with the bioenergetic incompetency that occurs during estradiol-loss, this slowly depletes cellular ATP and increases susceptibility to energy crises triggered by additional excitatory insults, ultimately contributing to the female-biased vulnerability to AD.

Biological sciences/Neuroscience/Cellular neuroscience

Biological sciences/Cell biology/Mechanisms of disease

Health sciences/Diseases/Neurological disorders/Dementia/Alzheimer's disease

Understanding the pivotal role of sex in health and disease is a cornerstone of personalized medicine^1,2. Females are proposed to be at the epicentre of the late-onset Alzheimer's disease (LOAD) epidemic. While some studies of North and South American cohorts have not found a consistent sex difference in the incidence of late-onset Alzheimer's disease (LOAD), other studies, particularly those conducted in Europe and Asia, indicate a higher disease incidence among women, especially after the age of 80 years^3,4. Both human autopsy studies and mouse models of AD have consistently demonstrated that females exhibit higher levels of amyloid beta (Aβ) plaques and neurofibrillary tangles, indicating that the underlying neuropathology is fundamentally different between the two sexes^5–7. For a given level of neuropathology demonstrated at autopsy, females have consistently exhibited more rapid declines in cognition and accelerated hippocampal atrophy compared to males^8,9. These observations collectively underscore the significant challenge that women living with dementia present globally.

Age is widely recognized as the primary risk factor for LOAD¹⁰. Therefore the higher prevalence of the disease in females has traditionally been attributed to their longer life expectancies¹¹. However, the age factor alone fails to fully account for the lower prevalence of the disease in young women; in fact, the opposite trend holds true among postmenopausal women¹². Alongside the loss of ovarian functions, the endocrine transition from estrogen cycling during the reproductive phase to the rapidly diminished levels of the hormone in the post-menopausal phase is linked to mild cognitive decline (MCI), a prodromal phase of LOAD¹³. Adult brain neurons, particularly the pyramidal neurons located within the disease-relevant brain regions such as the hippocampus and cortex, are reported to be estrogen-sensitive¹⁴. Previous studies have indicated that significant reductions in brain and cerebrospinal fluid levels of estrogen are specific to female LOAD patients, as compared to age-matched non-demented controls^15–17. These findings suggest that the loss of neuronal estrogen receptor (ER)-mediated signalling may transform a sex-biased neuroprotective benefit into disease susceptibility in postmenopausal women.

17β-estradiol (E2) is the major estrogen in the body predominantly produced by the ovary¹⁸. It is also a highly lipophilic compound which allows it to directly cross the blood-brain barrier via lipid-facilitated diffusion¹⁹ and execute its role in regulating reproduction via the hypothalamic-pituitary-gonadal axis. In addition to its classic role, estrogen receptor (ER) signalling is also playing significant metabolic roles in preserving the integrity of energy homeostasis¹⁸. At the periphery, research in both preclinical and clinical settings has clearly revealed that states of estrogen deficiency, either brought on by menopause or ovariectomy, are associated with decreased energy expenditure, metabolic issues, and obesity^20,21. At the central nervous system, however, the majority of the work has focused on estrogen's regulatory role in the hypothalamus, which in turn regulates peripheral metabolic health¹⁸. The direct effect of ER signalling on local brain metabolic status, and whether this is relevant to the sex-dimorphic changes observed in late-onset Alzheimer's disease (LOAD), remains unclear.

In this study, we found that female disease-affected subjects in the Religious Orders Study and Memory and Aging Project (ROSMAP) cohort exhibited more pronounced brain transcriptomic and metabolomic changes. The impact of peripheral estrogen loss on the brain was subsequently simulated in an accelerated ovarian failure (AOF) mouse model, revealing an unexpected regulatory connection between neuronal cholesterol metabolism and its bioenergetic balance. Dyshomeostasis of such results in a rewired N-acetyl-aspartyl-glutamate (NAAG) metabolic axis, which leads to higher spontaneous postsynaptic activities to chronically exhaust cellular ATP. These changes have negative impact on neuronal resilience against additional excitatory stimulus of insults, resulting in altered electrophysiological properties at brain circuitry level and behavioural phenotypes.

Female patients with late-onset Alzheimer's disease (LOAD) exhibited a more pronounced decline in the ERRα-regulated bioenergetic network within neurons

Previous neuroimaging and blood biomarker studies have suggested that brain hypometabolism contributes to the sex-dimorphic effects observed in LOAD^22,23. To examine this in detail, bulk transcriptomic analysis was performed on brain tissue samples from 613 individuals who participated in the ROSMAP study²⁴ - including 133 non-dementia males, 219 non-dementia females, 88 LOAD males, and 173 LOAD females subjects (Fig. 1a-d). The phenotypic status of these samples (LOAD vs. non-dementia) was defined based on multiple clinicopathological features, including neuritic plaque load (CERAD), neurofibrillary tangle pathology (Braak stage), and cognitive status (Cogdx and DCFDX) (Fig. 1a). DEG analysis between sex-specific LOAD versus ND samples revealed a significantly higher number of DEGs in females than males (DEG_{LOAD vs. ND}: female = 6,615 genes; male = 440 genes) (Fig. 1b, Supplementary Table 1). While the small number of differentially expressed genes (DEGs) identified in the male cohort failed to cluster into any meaningful pathways (Supplementary Table 1), the upregulated DEGs in the female LOAD group were enriched for pathways related to neuroinflammation, including TNFα signalling via NF-κB, IL-2/STAT5 signalling, TGFβ signalling, and IL-6/JAK/STAT3 signalling. In contrast, the pathways enriched by the set of downregulated DEGs in females were primarily metabolic, such as oxidative phosphorylation (OXPHOS) and mTORC1 signalling (Fig. 1c).

Among the significant pathways identified, oxidative phosphorylation (OXPHOS) was the highest ranked (smallest p-value). Further analysis of the most likely common transcriptional regulators of the enriched genes suggested estrogen-related receptor alpha (ERR1/ERRα) as a key regulator (Fig. 1d, Supplementary Table 2). ERRα, also known as NR3B1, is a nuclear receptor that shares substantial DNA sequence homology with estrogen receptor alpha (ERα), despite estradiol (E2) being an unlikely endogenous ligand for ERRα²⁵. To further investigate the relevance of ERRα activity in different brain cell types, its relative gene expression level was evaluated in a single-nucleus transcriptomic dataset²⁶, which indicated that neurons generally express the highest levels of the ESRRA transcript (Fig. 1e). The importance of ERRα activities in neurons was further validated by the similar enrichment of PPARGC1A transcript, which encodes its co-activator PGC1α in these cells (Fig. 1e). As compared to non-dementia (ND) controls, while no obvious differences in ESRRA transcript levels were found in LOAD (Supplementary Fig. 1, Note: although p < 0.05, the Log2FC values were all ≤ |0.07|, too small to be considered significant), immunohistochemistry analyses however revealed obvious reductions in nuclear signals of ERRα, particularly among affected female subjects (Fig. 1f). This suggests that the loss of ERRα nuclear activities at the post-translational level may serve as a link to the female-specific changes observed in the disease, potentially related to the dramatic hormonal changes that occur during the menopausal transition.

Cognitive and memory functions correlate with the residual estradiol levels after the induction of accelerated ovarian failure (AOF) which emulates menopause

Menopause in women is characterized by a sharp decline in estrogen (E2) levels within a relatively short period, whereas aging men do not experience a comparable change in testosterone levels²⁷. To investigate how dramatic endocrine changes associated with menopause may affect brain function, menopause was artificially induced in laboratory mice using 4-vinylcyclohexene diepoxide (VCD). VCD is an occupational chemical that causes selective destruction of small pre-antral ovarian follicles by accelerating the natural, apoptotic process of follicular atresia, while leaving the overall ovarian anatomy intact in rodents^28,29. Young adult mice (3 months old, postnatal day 90) were chosen for this study, such that any unpredictable confounding effects that could arise from chronological aging could be minimized (Fig. 2a). While the VCD treatment had no noticeable effect on body weight change (Fig. 2b), it resulted in elevated circulating levels of follicle-stimulating hormone (FSH) (Fig. 2b) but an opposite effect to estradiol (Fig. 2d) - the primary form of estrogen during reproductive years. These findings observed on experimental day 90 indicates a successful induction of accelerated ovarian failure, mimicking the menopausal transition in these test animals. To test whether the dramatic endocrine changes initiated in the periphery (i.e., loss of estradiol) may alter brain function, a battery of behavioural tests was performed on the animals. While animals subjected to VCD exposure showed no obvious changes in motor function (Supplementary Fig. 2a-b), they did exhibit declines in spatial learning and memory (Fig. 2e), as well as short-term working memory (Fig. 2f). The estradiol-deficient animals also spent significantly less time in the centre zone of the open-field test, indicating a mild anxiety-like behaviour (Fig. 2g). Notably, the degree of changes in these behavioural measures was correlated with the residual circulating estradiol levels after treatment (Fig. 2e-f). The neuroprotective effects of estradiol were further supported by immunohistochemistry analysis, which revealed obvious neurite degeneration in the hippocampal regions of VCD-treated animals (Fig. 2h). These behavioural and anatomical changes were correlated with declines in neuronal function, as evidenced by significant impairments in field excitatory postsynaptic potentials (fEPSPs) in the Schaffer collateral pathway (Fig. 2i) and long-term potentiation (LTP) (Fig. 2j).

Estradiol depletion impairs of the ERRα signalling network in the brain

The behavioural and electrophysiological findings demonstrated a functional connection between circulating estradiol and the brain. To delineate the details at the cellular and molecular levels, a bulk transcriptomic analysis was performed using total cerebral cortex tissues harvested from the animals. With the brain transcriptome profiles of the VCD-treated and vehicle-treated groups being clearly distinct from one another (Fig. 3a), a total of 342 upregulated and 542 downregulated transcripts were identified (Fig. 3b, Supplementary Table 3). While the upregulated differentially expressed genes (DEGs) were not cluster into any meaningful pathways, many of the significantly downregulated genes were implicated in mitochondrial energetics networks, including oxidative phosphorylation (OXPHOS), thermogenesis, and the citrate cycle (TCA cycle). Additionally, these downregulated genes were associated with a list of neurodegenerative disorders, such as Parkinson's disease, prion disease, Huntington's disease, Alzheimer's disease, and amyotrophic lateral sclerosis (Fig. 3c, Supplementary Table 4). Further gene overlapping analysis suggested that the majority of genes clustered in the OXPHOS pathway were indeed common to those clustered in various neurodegenerative disorders (Fig. 3d). Moreover, subsequent transcription factor analyses of all downregulated DEGs revealed again the ERR1/ERRα as the primary upstream regulator of these genes, with at least one occurrence of the highly conserved ERRα binding motif TGACCTY in the regions spanning 4 kb cantered on their transcription starting sites [-2kb, +2kb] (Fig. 3e). In agreement with the findings observed with human brain tissues (Fig. 1f), protein signals of ERRα, as well as that of its essential co-activator PGC1α were predominantly enriched in neuronal nuclei under normal conditions (Fig. 3f). However, systemic VCD administration triggered a loss of such signals (Fig. 3f). While this was unlikely due to changes at the transcription level (Supplementary Fig. 3a), significant reductions in the protein-protein interactions between ERRα and PGC1α were observed. This may also have promoted the default degradation of the intrinsically disordered PGC1α protein within the cells (Fig. 2g, Supplementary Fig. 3b).

Previous studies have revealed that the ligand-binding domain (LBD) and activation function-2 (AF2) domains of ERRα are involved in interacting with the 3rd LXXLL motif of PGC1α³⁰. Additionally, previous research has shown that the cholesterol binding - the natural ERRα ligand - to the LBD may concurrently enhance the binding affinity between ERRα and PGC1α³¹, although the precise details remain unclear. By in silico simulation, we provided further structural evidence that helps explain this phenomenon. First, we simulated a complete, open structure of the ERRα ligand-binding domain (Open ERRα) by merging the previously reported, yet incomplete, open structures of the ERRα LBD extracted from the Protein Data Bank (PDB: 7E2E and 2PJL)^32,33 (Supplementary Fig. 3c) with the simulated missing alpha-helix segment reconstructed by the AlphaFold2³⁴ (Supplementary Fig. 3d). The resulted combined structure then constituted a complete, open LBD structure that allows possible cholesterol binding. We then further docked this cholesterol-bound ERRα conformation against the AlphaFold2-simulated 3rd LXXLL motif (amino acids 208–216) of PGC1α³⁴ using the HADDOCK 2.4 algorithm³⁵. Compared to the docking result generated from the closed, yet complete ERRα LBD structure (PDB: 1XB7)³³, which prohibits cholesterol ligand binding (Supplementary Fig. 3c), cholesterol binding to the complete, open LBD structure resulted in conformational remodelling of substructures near the cholesterol-binding pocket (Supplementary Fig. 3e). This remodelling is predicted to facilitate more hydrophobic interactions and hydrogen bond formation between ERRα and the 3rd LXXLL motif of PGC1α, thereby supporting their protein-protein interaction and downstream ERRα transcriptional activities (Fig. 3h). With brain tissues harvested from the AOF mouse model, we found that systematic VCD treatment resulted in diminished levels of cholesterol being bound to ERRα (Fig. 3i). These results hinted that the post-translational modulation on ERRα signalling is likely related to cholesterol availability in these cells.

ERα signalling activates ERRα signalling by sustaining cholesterol homeostasis

The diminished ERRα-bound cholesterol in the AOF mouse model suggested a broad metabolic reprogramming effect, likely which is likely initiated by the dramatic loss of estrogen receptor signalling. Unbiased metabolomics (Fig. 4a-b, Supplementary Table 5) and targeted differential metabolic gene expression analysis (Fig. 4c, Supplementary Table 6) consistently suggested an impairment of cholesterol biosynthesis, as accompanied by disruption of central carbon metabolism (i.e., glycolysis and TCA cycle) (Fig. 4d). In addition to serving as a major energy source, glucose may also contribute carbons to cholesterol biosynthesis in the brain³⁶ through a series of reactions. These include glycolysis and the subsequent conversion of the glycolytic end product pyruvate to acetyl-CoA inside the mitochondria. The acetyl-CoA can then fuse with oxaloacetate to form citrate, which can be re-exported to the cytoplasm. There, citrate lyase can regenerate acetyl-CoA, which can then commit to the mevalonate pathway for cholesterol biosynthesis (Fig. 4d). In the adult brain, the two primary mechanisms for meeting neuronal cholesterol requirements are: (1) the de novo neuronal biosynthesis of cholesterol, and (2) the lipoprotein-mediated transfer of the metabolite from neighbouring glial cells³⁷. To determine if specific ER (estrogen receptor) signalling is involved, the key dysregulated metabolites and genes belonging to the related metabolic steps (as labelled in red in Fig. 4d) were re-evaluated using a mature primary neuronal culture model, which was done with and without treatment using a long-lasting ester derivative of 17β-estradiol (i.e., 100 nM estradiol cypionate, E2) and specific ER antagonists (i.e., ERα: 100 nM MPP dihydrochloride; ERβ: 100 nM PTHPP; GPER: 100 nM G-15) (Fig. 4e). Targeted metabolite analyses revealed that ERα (estrogen receptor alpha) was primarily responsible for mediating the effect of estradiol on neurons (Fig. 4e). Furthermore, based on initial predictions made using the ENCODE, CHEA (Table S7), and ChIP-atlas databases (Supplementary Fig. 4a), chromatin immunoprecipitation analysis showed that 5 of the dysregulated metabolic genes identified from the AOF model were indeed ERα targets (Fig. 4e). These target genes were also having high expression in neurons relative to other brain cell types (Supplementary Fig. 5). These ERα target genes included Pdha1, which encodes a key component of the pyruvate dehydrogenase complex responsible for converting pyruvate to acetyl-CoA; so as the Dhcr24 and Cyp51 genes that respectively encode the 24-dehydrocholeserol reductase and lanosterol 14-alpha demethylase of the cholesterol biosynthesis pathway (Fig. 4e). In addition to the de novo cholesterol biosynthesis process, neurons can also directly uptake cholesterol from the circulating apolipoprotein E (ApoE) chaperones released by the neighbouring astrocytes³⁸. Indeed, gene expression key neuronal ApoE receptors such as Lrp1 and Vldlr were also sensitive to ERα signalling (Fig. 4e, Supplementary Table 7). Further supporting this, ChIP-PCR revealed that estradiol treatment induced robust ERα binding to the predicted binding sites in the promoter regions of these receptor genes, and that such binding could be suppressed by MPP dihydrochloride, a small molecule antagonist that selectively inhibits estradiol binding to ERα (Fig. 4e, Supplementary Fig. 4)³⁹. The selective decline in expression of these genes (except for LRP1) in female patients was validated using the ROSMAP brain transcriptomics data. Furthermore, a greater degree of this gene expression decline was associated with more severe cognitive impairment (Fig. 4g). These data not only revealed the existence of a glucose-cholesterol biosynthetic axis in neurons and its importance for maintaining cognitive function, but also highlighted the upstream regulatory role of ERα signalling in this process. To further validate these findings, a stable isotope tracing analysis was performed to trace down the immediate fates of glucose-¹³C₆ in neurons under the chronic influence of estradiol (Fig. 4h). At 4 hours after the isotope exposure, isotopologue profiling revealed that estradiol pre-programmed the cells to have a robust flux of glucose-derived pyruvate towards mevalonate biosynthesis - the key precursor of cholesterol (Fig. 4h). The necessity of ERα signalling for this effect was further validated, as co-pretreatment of the cells with MPP dihydrochloride effectively prevented the increased flux of glucose-derived metabolites towards mevalonate biosynthesis (Fig. 4h). At the downstream levels of these events, the relative abundances of ERRα-bound cholesterol (Fig. 4i), the robustness of the ERRα interaction with PGC1α (Fig. 4j, Supplementary Fig. 6), and ERRα nuclear transcription activities (Fig. 4k) were all found to be regulated in a similar manner.

Loss of ERRα dysregulates the neuronal NAAG metabolic axis

ERRα is a master regulator of cellular energy metabolism⁴⁰. However, the detailed metabolic reprogramming events and their impact on cellular resilience against stress, particularly in neurons, remained unclear following the loss of ERRα. To investigate so, adeno-associated virus (AAV) carrying a microRNA-30 (miR30)-based short hairpin (shRNA) to specifically knocked down ERRα specifically in neurons⁴¹ was stereotaxically injected into both the medial prefrontal cortex and hippocampal CA1-2 regions of the C57BL/6 mice (Fig. 5a-b, Supplementary Fig. 7a). A battery of behavioural assessments and brain tissue investigation were then conducted at approximately 3 months after the initial injection of the AAV which the lag time allow recovery and the shRNA effect to take place. The successful knockdown of Esrra was first validated by both transcript and protein levels analyses in the injected tissue region, which were dissected and harvested under a fluorescent microscope (Supplementary Fig. 7a-b). As expected, Esrra knockdown in these target brain regions resulted in poorer cognitive and memory performances (Supplementary Fig. 7c). Subsequent brain histological analysis, using the ectopically expressed GFP signals as visual guides, revealed significant neurite losses upon Esrra knockdown (Fig. 5b). This finding at least in part explained the behavioural functional changes observed in these animals. With the GFP + brain tissues enriched, transcriptomic (Fig. 5c-d, Supplementary Table 8) and metabolomic profiling (Fig. 5e, Supplementary Table 9) assays were conducted. The integrated transcriptomic and metabolomic analysis revealed that the TCA cycle was the top affected pathway upon Esrra knockdown in neurons (Fig. 5f). Further profiling of the TCA cycle's series of biochemical reactions revealed diminished abundances of metabolites (i.e., succinate and fumarase) beyond the step catalysed by succinyl-CoA ligase (SUCLG1) (Fig. 5g), hinting that the downstream reactions of the TCA cycle could be affected. As a possible homeostatic response to such challenges, metabolic reprogramming that utilizes aspartate aminotransferase to connect the segment between oxaloacetate (4C) and α-ketoglutarate (5C) was likely enhanced, as suggested by the concomitant reduction of aspartate levels in the system (Fig. 5g). To validate the increased reliance of aspartate to the TCA cycle, isotopologue tracing with labelled ¹³C₄-aspartate was conducted in a primary neuronal culture model, such that the (1) aspartate aminotransferase facilitated transfer of amino group from aspartate (4C) to α-ketoglutarate (5C), (2) the conversion of α-ketoglutarate to glutamate as well as (3) its re-entrance to the TCA cycle via the action of glutamate dehydrogenase could all be detected [Fig. 4h(1)]. The results validated that this "mini version" of the TCA cycle was activated in neurons depleted of functional ERRα, either due to the silencing effect of Esrra shRNA or the treatment with the ERRα inverse agonist, thiadiazoleacrylamide XCT-790⁴². As these cells became more dependent on the 4-carbon backbone of aspartate to "complete the cycle", elevated levels of M + 4 citrate were observed as compared to neurons in the control treatment groups [Fig. 4h(2)]. Furthermore, when the assay time was prolonged, M + 4 citrate and M + 3 glutamate became the dominant forms of their respective species [Fig. 4h(2–3)]. This was because the labelled aspartate continued to feed the mini-TCA cycle, resulting in the de novo generation of these isotopologues [Fig. 4h(4)]. Intriguingly, the resulting loss in the levels of exogenously labelled M + 4 aspartate was not observed for its endogenous M + 0 isotopologue [Fig. 4h(4)]. Moreover, even though this rewired mini-TCA cycle predicts a 1:1 ratio of glutamate utilized versus glutamate regenerated in the reactions (i.e., no net loss), the resulting unlabelled M + 0 isotopologue [Fig. 4h(3)] as well as its total quantity [Fig. 4h(5)] were however elevated instead. These observations hinted that an unknown replenishment mechanism was turned on for these endogenous amino acids when Esrra was lost.

With reference to the unbiased metabolomics data (Table S9), it was speculated that an enhanced breakdown and/or reduced biosynthesis of N-acetyl-aspartyl-glutamate (NAAG) contributed to the observed homeostatic changes. NAAG is a neuron-specific dipeptide synthesized from an initial conjugation of an acetyl-CoA-derived acetyl group to aspartate (i.e., N-acetyl-aspartate, NAA), followed by a second conjugation reaction with glutamate⁴³. Indeed, our ¹³C₄-aspartate isotopologue tracing experiment revealed that the contributions of both aspartate and glutamate carbons to the formations of NAA and NAAG were significantly reduced in neurons deficient of functional ERRα, regardless of their sources of origin (i.e., endogenous [M + 0 glutamate and M + 0 aspartate] or exogenous [M + 3 glutamate and M + 4 aspartate]) (Fig. 5i). The diminished pre-existing unlabelled forms of NAA and NAAG indeed supported the notion of pro-catabolic shift, as evidenced by their reduced total content (Table S7). Additionally, the relatively lower levels of the newly synthesized versions of heavily labelled neuropeptides also indicated a diminished anabolic biosynthesis as well (Fig. 5i). Consequently, the net NAAG-NAA catabolic shift could have unleashed the unlabelled M + 0 forms of aspartate and glutamate, as detected in the isotopologue study [Fig. 5h(4)]. Concordantly, diminished ERRα signalling axis observed previously in the AOF mouse model (Fig. 3e-i) was also accompanied by reductions in the total levels of succinate, fumarate, NAA, and even NAAG, as evident from the metabolomics data (Supplementary Fig. 8, Supplementary Table 5).

In clinical studies, higher levels of NAAG have been associated with better brain functional outcomes in patients of various neurological conditions, such as psychosis⁴⁴; schizophrenia⁴⁵ and multiple sclerosis⁴⁶. In this study, the relevance of NAAG to better brain function was further validated in human LOAD samples. According to the dorsolateral prefrontal cortex metabolomics profile data extracted from the ROSMAP cohort [N = 339 for disease-affected (AD + MCI), N = 153 for non-dementia]⁴⁷, sex-specific analyses revealed more dramatic changes in metabolites among female affected subjects compared to males (Fig. 5j, Supplementary Table 10). Subsequent metabolite set enrichment analysis revealed that the majority of the differentially changed metabolites were likely part of the amino acid metabolism, including the metabolism of aspartic acid (Fig. 5k). Further analysis revealed that while no significant correlations between deregulated metabolites and cognitive impairment were found in males, a list of amino acid- and lipid-related metabolites were identified in the female group (Fig. 5l). Notably, the reduction in NAAG was the one most correlated with cognitive decline in female subjects (Fig. 5l-m), supporting the idea that the metabolic homeostasis of this neuron-specific metabolite is an important factor in determining the sex-biased functional outcome in LOAD. In contrast, the total levels of brain glutamate and aspartate, which could be contributed by various brain cell types, showed less consistent trends compared to the changes observed specifically in neurons (Fig. 5n-o). Nevertheless, brain glutamate levels were still induced and this was correlated with more severe cognitive decline in both sexes (Fig. 5o). This elevation in glutamate was more pronounced in females, so as the trend of changes across different cognitive scores was also more consistent in females (Fig. 5o).

Elevated spontaneous postsynaptic activity, coupled with bioenergetic incompetency due to loss of ERRα, confer heightened neuronal vulnerability to excitotoxic insults

Previous studies have shown that NAAG exhibits neuroprotective effects against NMDA-receptor-mediated excitotoxicity, likely through its partial antagonist action at the NMDA receptor⁴⁸. The current data suggest that the acquired aspartate-dependence results in a net hydrolytic effect on NAAG, which would concurrently increase the free glutamate pools within neurons - the major excitatory neurotransmitter in the brain. Spontaneous synaptic vesicle fusion is a fundamental feature of all synapses, and these random release events typically activate receptors within a single postsynaptic site, giving rise to miniature postsynaptic currents⁴⁹. To investigate whether the metabolic changes occurring during ERRα loss would affect the electrophysiological properties of neurons, whole-cell patch clamp recordings were performed. Compared to scrambled shRNA-treated controls, Esrra-knockdown resulted in a significant increase in the frequency of spontaneous miniature excitatory postsynaptic currents (mEPSCs), even without any obvious changes in their amplitude (Fig. 6a). Likewise, similar patterns of changes were observed in neurons chronically exposed to the ERRα antagonist XCT-790 (Fig. 6a). In a separate group of neurons, the spontaneous miniature inhibitory postsynaptic currents (mIPSCs) were evaluated. Both Esrra-knockdown and XCT-790 treatment did not change frequency nor the amplitude of mIPSCs, as they were recorded using patch pipettes with CsCl replacing K-gluconate and with excitatory postsynaptic currents blocked by DNQX (Fig. 6b). The solo increase in mEPSC frequency suggested presynaptic changes. Supporting this notion, the relative abundance of the presynaptic vesicular glutamate transporter VGLUT1 was elevated in the ERRα-deficient neurons. In contrast, the abundance of postsynaptic glutamatergic receptors was not significantly altered (Fig. 6c). Since VGLUT1 mediates the transport of intracellular glutamate into secretory vesicles⁵⁰, its upregulation would likely increase the probability of stochastic, spontaneous glutamate release, thereby leading to the observed changes in miniature excitatory postsynaptic currents.

Postsynaptic currents are a major energy-consuming process in the brain⁵¹. The pronounced increase in mEPSC frequency suggested that a net increase in the number of spontaneous postsynaptic firing events per second could have occurred chronically in the ERRα-deficient neurons. This would impose a significant metabolic demand on these cells. While this minor increment in the spontaneous neuronal activity would unlikely affect the energy levels of a healthy neuron, as they have reserved respiratory capacity to compensate for such additional metabolic demands⁵²; here even at the resting condition in the absence of any external stimulation, the total ATP levels in the ERRα-deficient neurons were already lower than that observed in the control neurons (Fig. 6d). As hinted by the TCA cycle analysis (Fig. 5g-h), the diminished commitment to reactions beyond succinyl-CoA suggests reduced activities of the succinate dehydrogenase (SDH) enzyme complex. According to the bulk transcriptomics data (Fig. 5c-d), expression levels of Sdha and Sdhd—two of the four major subunits of the enzyme—were significantly reduced (Table S8). The same patterns of changes were observed as well in primary neuronal cells subjected to Esrra knockdown and XCT-790 treatment (Supplementary Fig.S9b). According to the ENCODE (Supplementary Table 11) and ChIP-atlas (Supplementary Fig. 9a) databases, Sdha and Sdhd appear to be the downstream targets of ERRα, and was further validated by ChIP-PCR experiments (Supplementary Fig. 9c) and enzyme functional assays (Fig. 6f) in both Esrra knockdown or XCT-790 treatment models. More importantly, the ROSMAP study also indicated that the reduction in SDHA expression was unique and correlated with cognitive decline in females (Fig. 6e), highlighting that the fidelity of SDH function is one of the important factors contributing to female vulnerability in LOAD.

In addition to its central role in the TCA cycle, SDH also serves as Complex II of the electron transport chain (ETC). As the sole enzymatic complex that regulates the flux of both the TCA cycle and the ETC, the functional capacity of SDH is considered a source of the reserved respiratory capacity, and thus a key determinant of the ATP production rate in response to a sudden increase in energy demand⁵³. To assess whether the recovery of ATP in ERRα-deficient neurons was suboptimal, dynamic changes in the mitochondrial oxygen consumption rate (OCR) was first evaluated (Fig. 6g). Both the basal respiration rate (Fig.S9d) and the reserved respiratory capacity (Fig. 6g) were significantly reduced in the ERRα-deficient groups. These respiratory inefficiencies, when coupled with the incresed spontaneous post-synaptic firing, likely explained the relatively lower levels of total ATP even when the neurons were at rest (Fig. 6d). These neurons were then synaptically challenged for 30 seconds with glutamate, which induces large fluxes of sodium and potassium and thus stimulates the accelerated hydrolysis of ATP by the Na+/K+-pump⁵⁴. Real-time monitoring of mitochondrial ATP dynamics using the ATP-Red live cell dye revealed that while glutamate treatment resulted in similar magnitudes of ATP depletion relative to the corresponding baseline levels across all tested groups (Supplementary Fig. 9e), the subsequent 30-minute recovery of mitochondrial ATP was significantly less efficient in the ERRα-deficient groups (Fig. 6h). At 4 hours after the glutamate challenge, when the neurons were expected to have better recovered, nearly 16% of these neurons instead became apoptotic via the activation of caspase 3/7 (Fig. 6i). Together, these data revealed that neurons lacking active ERRα are more vulnerable to external excitatory stimulus or insults that lead to elevated energy demand and bioenergetic stresses.

The exact mechanism of sex biases in brain diseases by far remains unclear, but likely involves an interplay between the peripheral endocrine system and the central nervous system. Our study highlighted how the long-term loss in estradiol production, initiated at the peripheral, affects the female brain physiology and function. In neurons, ERα signalling unexpectedly regulates cholesterol homeostasis. This, in turn, alters the ERRα network. The latter not only governs cellular bioenergetics, but also regulates the status of N-acetyl-aspartyl-glutamate (NAAG). Ultimately, these changes affect the intrinsic electrophysiological properties and energy homeostasis of the cell. These conclusions, drawn from both in vivo and in vitro studies, align with observations from human brain data, supporting that an altered ERα-ERRα-NAAG metabolic axis imposes neuronal vulnerabilities to excitotoxic insults, and contributes to the female-biased predisposition towards LOAD.

In this study, we were surprised to uncover that estrogen has a direct regulatory effect on cholesterol biosynthesis and transport in neurons. This phenomenon was indeed hinted in earlier cancer cell studies, which reported that the cholesterol biosynthesis pathway is commonly upregulated in estrogen-sensitive cancer cell lines which influences cell growth and apoptosis^55–57. Here, we provide direct evidence that estrogen not only upregulates key lipoprotein receptors to facilitate cholesterol uptake, but also the metabolic genes that support cholesterol biosynthesis from glucose carbons. The loss of this regulatory effect, resulting either from an accelerated ovarian atrophy, pharmacological inhibition of ERα signalling, or a natural menopausal transition, could have contributed to a greater extent of reduced brain glucose metabolism in female brains when compared to males⁵⁸. In clinics, cerebral hypometabolism occurs frequently at the prodromal stage of LOAD and is associated with grey matter atrophy and cognitive decline^59,60. Here our data may provide an explanation on how this metabolic decline is more aggravated in female subjects. In the accelerated ovarian failure mouse study, neurodegeneration and brain functional decline were observed, and the degree of decline was correlated to the residual circulating levels of estradiol. This observation once again highlighted the intimate linkage between the status of the brain and the peripheral organs, as the latter are also sensors of environmental exposures. As a steroid hormone that passes through the blood-brain barrier, our findings revealed that the endocrine effects of estradiol on the brain are direct and long-lasting. This therefore could potentially make circulating estradiol level an early blood biomarker for cognitive and memory decline in aging female subjects.

Cholesterol is an essential component for neuronal physiology. Depletion of cholesterol in neurons impairs synaptic vesicle exocytosis, neuronal activity, and leads to dendritic spine and synapse degeneration^59,61,62. While it is generally assumed that cholesterol uptake is the primary mechanism in mature neurons, de novo cholesterol biosynthesis is actually upregulated in these cells during stressful conditions, such as surviving in an APOE4 microenvironmentt⁶³ or at times when myelin degeneration occurs³⁷. Our study revealed an underappreciated role of cholesterol as a direct regulator of mitochondrial bioenergetics, acting as an endogenous ligand of the ERRα – the master regulator of mitochondrial biogenic gene netowrk⁶⁴. Consistent with the previous findings^31,65, our data revealed that cholesterol binding to the transcription factor ERRα facilitates its interaction with the cofactor PGC1α, and enhances the nuclear activity of ERRα. Our study model also explained that this effect is dependent on ERα signalling-mediated cholesterol biosynthesis at the upstream, consistent with a previous finding which suggested that female subjects are more susceptible to the loss of ERRα function⁶⁶. The cholesterol-ERRα linkage revealed how a neuroprotective hormone like estrogen may crosstalk with the mitochondrial biology. Specifically, estrogen enhances the utilization of glucose carbons in females through its interaction with ERRα. Cholesterol dysregulation in the brain is tightly linked to the risk and pathogenesis of LOAD. Genes involved in cholesterol transport in different brain cells, such as APOE, TREM2, ABCA7, INPP5D, CLU, SPI1, and SORL1, have all been reported as genetic risk factors of the disease⁶⁷. Given the finding that the APOE4 genotype, the greatest genetic risk factor for LOAD, may have greater penetrance in females⁶⁸, it is speculative that the loss of ERα-mediated support on neuronal cholesterol biosynthesis and mitochondrial metabolism after the menopausal transition may contribute to this sex-biased vulnerability.

The importance of ERRα in regulating mitochondrial metabolism has been well established. Our findings extend this current knowledge by illustrating how the mature neurons may adapt to the effects imposed by the loss of ERRα function. Integrated transcriptomic and metabolomic profiling analyses revealed that the TCA cycle is the most affected metabolic pathway upon Esrra knockdown, which is due to the loss in expression of specific functional units of the succinate dehydrogenase (SDH) complex involved in the cycle. The resulting 'truncated TCA cycle' adopted to utilizes aspartate as a source of 4-carbon backbone. Through the action of aspartate aminotransferase, aspartate is deaminated to form oxaloacetate, whereas the amino group is transferred to α-ketoglutarate to complete the cycle. Indeed, a similar form of truncated TCA cycle and increased aspartate-dependence were reported in cancer cells with SDH deficiency⁶⁹ and even physiologically in the normal prostate⁷⁰. Our isotopologue tracing analyses also revealed that the adaptive dependence on aspartate may subsequently alter its participation in the biosynthesis of NAAG, a neuropeptide that is uniquely synthesized in neurons. Traditionally, NAAG has been considered a suppressor of glutamatergic synaptic activity, serving as a retrograde neurotransmitter that activates the inhibitory Gi/G0-coupled metabotropic glutamate receptor 3 located at the presynaptic regions⁷¹. This indeed provides the postsynaptic neuron with a feedback mechanism to inhibit excessive glutamate signalling⁷¹. Furthermore, our data revealed that the loss of this NAAG-mediated feedback mechanism, due to the net decline in NAAG biosynthesis, leads to an elevated pool of glutamate within the neurons. This in turn results in marked inductions of the frequency of the spontaneous mEPSCs, which indicates an increase in presynaptic neurotransmitter release. This was likely due in part to the reduced feedback inhibition by NAAG, or an enhanced spontaneous and stochastic release of glutamate, as evidenced by the elevated abundance of presynaptic VGLUT1 signals in these cells. It is well known that the spontaneous firing of action potentials accounts for 60–80% of the metabolic energy consumed by the brain^72,73. Our data revealed that the persistent increase in the frequency of mEPSCs would impose additional energy demands on these neurons, even in the absence of external stimuli. This renders them exquisitely vulnerable to any insults that leads to metabolic inefficiency. Unfortunately, SDH is also the Complex II of the oxidative phosphorylation (OXPHOS) system. The insufficient activity of this enzymatic complex would also compromise mitochondrial respiratory functions, resulting in inefficient ATP biosynthesis, and elevated susceptibility to cell death, particularly at times when energy expenditure and demand are surged by external excitatory stimulus^54,74. Pathological damage to glutamatergic neurons, particularly the cell bodies and neurites in layers III and IV of the neocortex, so as those glutamatergically-innervated cortical and hippocampal neurons, are particularly evident in LOAD brains⁷⁵. Our findings therefore offer an explanation for the increased vulnerability of female neurons to the same level of excitotoxic insults compared to male cells. This sex-based difference in susceptibility may contribute to the observed sex-biased disadvantage in disease progression and functional decline. The clinical relevance of the NAAG finding was further confirmed in ROSMAP cohort, as the reduction of brain NAAG content was the most correlated with cognitive decline in female subjects. While elevated levels of glutamate were found in both male and female subjects and were also correlated with cognitive decline, this elevation was much more profound and consistent among the female subjects.

It is important to consider that the menopausal transition and the associated decline in estrogen signalling are natural and inevitable changes that occur in all aging women⁷⁶. Given that the average human lifespan in developed countries is projected to exceed 80 years by 2050⁷⁷, this implies that more than half of a woman's lifetime will be spent in an estrogen-deficient state. Our study therefore offered important insights at both the cellular and molecular levels on how the loss of ER signalling may result in increased neuronal vulnerability to excitotoxic insults in key brain regions affected in LOAD.

Acknowledgements

The authors would like to gratefully acknowledge the following parties. The results published here are in whole or in part based on data obtained from the AD Knowledge Portal, related to the ROSMAP metadata sets. The human brain tissue samples utilized for immunohistology analyses was obtained from the University of Maryland Brain and Tissue Bank which is a Brain and Tissue Repository of the NIH NueroBioBank, as requested by Dr. Karl Herrup. We would also like to acknowledge the Addgene and depositors of the plasmid materials for their generosities.

The work was supported, in part, by grants from the following: The Hong Kong Research Grants Council (RGC)-General Research Fund (GRF) (PI: GRF16100219 and ECS24107121) (all to K.H-M.C.) and the RGC-Collaborative Research Equipment Grant (CREG) (Co-I: C4002-21E) (K.H-M.C.); the National Natural Science Foundation-Excellent Young Scientists Fund 2020 (Ref: 32022087) and Young Scientist Fund 2023 (32300643) (K.H-M.C.). All the funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Contributions

J.K-L.S. performed in vivo studies. D.W. performed bioinformatics analyses. A.Z.P. performed in vitro experiments. G.C-N.W. performed in silico simulations and docking studies. R.P.H. contributed plasmid reagents. K.H. contributed human brain samples. K.H-M.C. contributed new reagents and analytical tools. K.H-M.C. conceptualized and designed the research, conducted the experiments, analysed all the data and wrote the paper. R.P.H., K.H. and K.H-M.C. discussed the data and offered suggestions to the study.

Ethics declarations

The authors declare no competing interests.

Rocca, W. A., Mielke, M. M., Vemuri, P. & Miller, V. M. Sex and gender differences in the causes of dementia: a narrative review. Maturitas 79, 196–201, doi:10.1016/j.maturitas.2014.05.008 (2014).
Miller, V. M. Why are sex and gender important to basic physiology and translational and individualized medicine? Am J Physiol Heart Circ Physiol 306, H781-788, doi:10.1152/ajpheart.00994.2013 (2014).
Mielke, M. M. et al. Consideration of sex and gender in Alzheimer's disease and related disorders from a global perspective. Alzheimers Dement 18, 2707–2724, doi:10.1002/alz.12662 (2022).
O'Neal, M. A. Women and the risk of Alzheimer's disease. Front Glob Womens Health 4, 1324522, doi:10.3389/fgwh.2023.1324522 (2023).
Carroll, J. C. et al. Sex differences in beta-amyloid accumulation in 3xTg-AD mice: role of neonatal sex steroid hormone exposure. Brain Res 1366, 233–245, doi:10.1016/j.brainres.2010.10.009 (2010).
Oveisgharan, S. et al. Sex differences in Alzheimer's disease and common neuropathologies of aging. Acta Neuropathol 136, 887–900, doi:10.1007/s00401-018-1920-1 (2018).
Yue, M., Hanna, A., Wilson, J., Roder, H. & Janus, C. Sex difference in pathology and memory decline in rTg4510 mouse model of tauopathy. Neurobiol Aging 32, 590–603, doi:10.1016/j.neurobiolaging.2009.04.006 (2011).
Barnes, L. L. et al. Sex differences in the clinical manifestations of Alzheimer disease pathology. Arch Gen Psychiatry 62, 685–691, doi:10.1001/archpsyc.62.6.685 (2005).
Koran, M. E. I., Wagener, M., Hohman, T. J. & Alzheimer's Neuroimaging, I. Sex differences in the association between AD biomarkers and cognitive decline. Brain Imaging Behav 11, 205–213, doi:10.1007/s11682-016-9523-8 (2017).
Guerreiro, R. & Bras, J. The age factor in Alzheimer's disease. Genome Med 7, 106, doi:10.1186/s13073-015-0232-5 (2015).
Seshadri, S. et al. Lifetime risk of dementia and Alzheimer's disease. The impact of mortality on risk estimates in the Framingham Study. Neurology 49, 1498–1504, doi:10.1212/wnl.49.6.1498 (1997).
Beam, C. R. et al. Differences Between Women and Men in Incidence Rates of Dementia and Alzheimer's Disease. J Alzheimers Dis 64, 1077–1083, doi:10.3233/JAD-180141 (2018).
Caldwell, C. C., Yao, J. & Brinton, R. D. Targeting the prodromal stage of Alzheimer's disease: bioenergetic and mitochondrial opportunities. Neurotherapeutics 12, 66–80, doi:10.1007/s13311-014-0324-8 (2015).
Shughrue, P. J. & Merchenthaler, I. Estrogen is more than just a "sex hormone": novel sites for estrogen action in the hippocampus and cerebral cortex. Front Neuroendocrinol 21, 95–101, doi:10.1006/frne.1999.0190 (2000).
Callahan, M. J. et al. Augmented senile plaque load in aged female beta-amyloid precursor protein-transgenic mice. Am J Pathol 158, 1173–1177, doi:10.1016/s0002-9440(10)64064-3 (2001).
Barron, A. M. & Pike, C. J. Sex hormones, aging, and Alzheimer's disease. Front Biosci (Elite Ed) 4, 976–997, doi:10.2741/E434 (2012).
Chow, H. M. et al. CDK5 activator protein p25 preferentially binds and activates GSK3beta. Proc Natl Acad Sci U S A 111, E4887-4895, doi:10.1073/pnas.1402627111 (2014).
Mahboobifard, F. et al. Estrogen as a key regulator of energy homeostasis and metabolic health. Biomed Pharmacother 156, 113808, doi:10.1016/j.biopha.2022.113808 (2022).
Banks, W. A. Brain meets body: the blood-brain barrier as an endocrine interface. Endocrinology 153, 4111–4119, doi:10.1210/en.2012-1435 (2012).
Mauvais-Jarvis, F., Clegg, D. J. & Hevener, A. L. The role of estrogens in control of energy balance and glucose homeostasis. Endocr Rev 34, 309–338, doi:10.1210/er.2012-1055 (2013).
Ko, S. H. & Jung, Y. Energy Metabolism Changes and Dysregulated Lipid Metabolism in Postmenopausal Women. Nutrients 13, doi:10.3390/nu13124556 (2021).
Park, J. C. et al. Sex differences in the progression of glucose metabolism dysfunction in Alzheimer's disease. Exp Mol Med 55, 1023–1032, doi:10.1038/s12276-023-00993-3 (2023).
Sarica, A. et al. Sex Differences in Conversion Risk from Mild Cognitive Impairment to Alzheimer's Disease: An Explainable Machine Learning Study with Random Survival Forests and SHAP. Brain Sci 14, doi:10.3390/brainsci14030201 (2024).
Bennett, D. A. et al. Religious Orders Study and Rush Memory and Aging Project. J Alzheimers Dis 64, S161-S189, doi:10.3233/JAD-179939 (2018).
Deblois, G. & Giguere, V. Functional and physiological genomics of estrogen-related receptors (ERRs) in health and disease. Biochim Biophys Acta 1812, 1032–1040, doi:10.1016/j.bbadis.2010.12.009 (2011).
Mathys, H. et al. Single-cell transcriptomic analysis of Alzheimer's disease. Nature 570, 332–337, doi:10.1038/s41586-019-1195-2 (2019).
Hagg, S. & Jylhava, J. Sex differences in biological aging with a focus on human studies. Elife 10, doi:10.7554/eLife.63425 (2021).
Brooks, H. L., Pollow, D. P. & Hoyer, P. B. The VCD Mouse Model of Menopause and Perimenopause for the Study of Sex Differences in Cardiovascular Disease and the Metabolic Syndrome. Physiology (Bethesda) 31, 250–257, doi:10.1152/physiol.00057.2014 (2016).
Kappeler, C. J. & Hoyer, P. B. 4-vinylcyclohexene diepoxide: a model chemical for ovotoxicity. Syst Biol Reprod Med 58, 57–62, doi:10.3109/19396368.2011.648820 (2012).
Takacs, M. et al. The asymmetric binding of PGC-1alpha to the ERRalpha and ERRgamma nuclear receptor homodimers involves a similar recognition mechanism. PLoS One 8, e67810, doi:10.1371/journal.pone.0067810 (2013).
Wei, W. et al. Ligand Activation of ERRalpha by Cholesterol Mediates Statin and Bisphosphonate Effects. Cell Metab 23, 479–491, doi:10.1016/j.cmet.2015.12.010 (2016).
Shinozuka, T., Ito, S., Kimura, T., Izumi, M. & Wakabayashi, K. Discovery of a Novel Class of ERRalpha Agonists. ACS Med Chem Lett 12, 817–821, doi:10.1021/acsmedchemlett.1c00100 (2021).
Kallen, J. et al. Evidence for ligand-independent transcriptional activation of the human estrogen-related receptor alpha (ERRalpha): crystal structure of ERRalpha ligand binding domain in complex with peroxisome proliferator-activated receptor coactivator-1alpha. J Biol Chem 279, 49330–49337, doi:10.1074/jbc.M407999200 (2004).
Mirdita, M. et al. ColabFold: making protein folding accessible to all. Nat Methods 19, 679–682, doi:10.1038/s41592-022-01488-1 (2022).
van Zundert, G. C. P. et al. The HADDOCK2.2 Web Server: User-Friendly Integrative Modeling of Biomolecular Complexes. J Mol Biol 428, 720–725, doi:10.1016/j.jmb.2015.09.014 (2016).
Barkai, A. I. Flow of glucose carbon into cholesterol and phospholipids in various regions of the adult rat brain: enhanced incorporation into hypothalamic phospholipids. J Neurosci Res 6, 585–595, doi:10.1002/jnr.490060505 (1981).
Berghoff, S. A. et al. Neuronal cholesterol synthesis is essential for repair of chronically demyelinated lesions in mice. Cell Rep 37, 109889, doi:10.1016/j.celrep.2021.109889 (2021).
Getz, G. S. & Reardon, C. A. Apoprotein E as a lipid transport and signaling protein in the blood, liver, and artery wall. J Lipid Res 50 Suppl, S156–161, doi:10.1194/jlr.R800058-JLR200 (2009).
Yuseran, H., Hartoyo, E., Nurseta, T. & Kalim, H. Molecular docking of genistein on estrogen receptors, promoter region of BCLX, caspase-3, Ki-67, cyclin D1, and telomere activity. J Taibah Univ Med Sci 14, 79–87, doi:10.1016/j.jtumed.2018.10.003 (2019).
Audet-Walsh, E. & Giguere, V. The multiple universes of estrogen-related receptor alpha and gamma in metabolic control and related diseases. Acta Pharmacol Sin 36, 51–61, doi:10.1038/aps.2014.121 (2015).
Liu, B., Xu, H., Paton, J. F. & Kasparov, S. Cell- and region-specific miR30-based gene knock-down with temporal control in the rat brain. BMC Mol Biol 11, 93, doi:10.1186/1471-2199-11-93 (2010).
Eskiocak, B., Ali, A. & White, M. A. The estrogen-related receptor alpha inverse agonist XCT 790 is a nanomolar mitochondrial uncoupler. Biochemistry 53, 4839–4846, doi:10.1021/bi500737n (2014).
Moffett, J. R., Ross, B., Arun, P., Madhavarao, C. N. & Namboodiri, A. M. N-Acetylaspartate in the CNS: from neurodiagnostics to neurobiology. Prog Neurobiol 81, 89–131, doi:10.1016/j.pneurobio.2006.12.003 (2007).
Zink, C. F. et al. Association of Missense Mutation in FOLH1 With Decreased NAAG Levels and Impaired Working Memory Circuitry and Cognition. Am J Psychiatry 177, 1129–1139, doi:10.1176/appi.ajp.2020.19111152 (2020).
Jessen, F. et al. N-acetylaspartylglutamate (NAAG) and N-acetylaspartate (NAA) in patients with schizophrenia. Schizophr Bull 39, 197–205, doi:10.1093/schbul/sbr127 (2013).
Rahn, K. A. et al. Inhibition of glutamate carboxypeptidase II (GCPII) activity as a treatment for cognitive impairment in multiple sclerosis. Proc Natl Acad Sci U S A 109, 20101–20106, doi:10.1073/pnas.1209934109 (2012).
Batra, R. et al. The landscape of metabolic brain alterations in Alzheimer's disease. Alzheimers Dement, doi:10.1002/alz.12714 (2022).
Thomas, A. G., Vornov, J. J., Olkowski, J. L., Merion, A. T. & Slusher, B. S. N-Acetylated alpha-linked acidic dipeptidase converts N-acetylaspartylglutamate from a neuroprotectant to a neurotoxin. J Pharmacol Exp Ther 295, 16–22 (2000).
Atasoy, D. et al. Spontaneous and evoked glutamate release activates two populations of NMDA receptors with limited overlap. J Neurosci 28, 10151–10166, doi:10.1523/JNEUROSCI.2432-08.2008 (2008).
Takamori, S., Rhee, J. S., Rosenmund, C. & Jahn, R. Identification of a vesicular glutamate transporter that defines a glutamatergic phenotype in neurons. Nature 407, 189–194, doi:10.1038/35025070 (2000).
Faria-Pereira, A. & Morais, V. A. Synapses: The Brain's Energy-Demanding Sites. Int J Mol Sci 23, doi:10.3390/ijms23073627 (2022).
Desler, C. et al. Is There a Link between Mitochondrial Reserve Respiratory Capacity and Aging? J Aging Res 2012, 192503, doi:10.1155/2012/192503 (2012).
Dhingra, R. & Kirshenbaum, L. A. Succinate dehydrogenase/complex II activity obligatorily links mitochondrial reserve respiratory capacity to cell survival in cardiac myocytes. Cell Death Dis 6, e1956, doi:10.1038/cddis.2015.310 (2015).
Chow, H. M. et al. ATM is activated by ATP depletion and modulates mitochondrial function through NRF1. J Cell Biol 218, 909–928, doi:10.1083/jcb.201806197 (2019).
Messa, C. et al. Estrogenic regulation of cholesterol biosynthesis and cell growth in DLD-1 human colon cancer cells. Scand J Gastroenterol 40, 1454–1461, doi:10.1080/00365520510024007 (2005).
Simigdala, N. et al. Cholesterol biosynthesis pathway as a novel mechanism of resistance to estrogen deprivation in estrogen receptor-positive breast cancer. Breast Cancer Res 18, 58, doi:10.1186/s13058-016-0713-5 (2016).
Coradini, D. De novo cholesterol biosynthesis: an additional therapeutic target for the treatment of postmenopausal breast cancer with excessive adipose tissue. Explor Target Antitumor Ther 3, 841–852, doi:10.37349/etat.2022.00116 (2022).
Zhao, L., Mao, Z., Woody, S. K. & Brinton, R. D. Sex differences in metabolic aging of the brain: insights into female susceptibility to Alzheimer's disease. Neurobiol Aging 42, 69–79, doi:10.1016/j.neurobiolaging.2016.02.011 (2016).
Costantini, L. C., Barr, L. J., Vogel, J. L. & Henderson, S. T. Hypometabolism as a therapeutic target in Alzheimer's disease. BMC Neurosci 9 Suppl 2, S16, doi:10.1186/1471-2202-9-S2-S16 (2008).
Chen, Y. et al. Evaluating the association between brain atrophy, hypometabolism, and cognitive decline in Alzheimer's disease: a PET/MRI study. Aging (Albany NY) 13, 7228–7246, doi:10.18632/aging.202580 (2021).
Liu, Q. et al. Neuronal LRP1 knockout in adult mice leads to impaired brain lipid metabolism and progressive, age-dependent synapse loss and neurodegeneration. J Neurosci 30, 17068–17078, doi:10.1523/JNEUROSCI.4067-10.2010 (2010).
Liu, Q. et al. Amyloid precursor protein regulates brain apolipoprotein E and cholesterol metabolism through lipoprotein receptor LRP1. Neuron 56, 66–78, doi:10.1016/j.neuron.2007.08.008 (2007).
Li, X. et al. Astrocytic ApoE reprograms neuronal cholesterol metabolism and histone-acetylation-mediated memory. Neuron 109, 957–970 e958, doi:10.1016/j.neuron.2021.01.005 (2021).
Fan, W. & Evans, R. PPARs and ERRs: molecular mediators of mitochondrial metabolism. Curr Opin Cell Biol 33, 49–54, doi:10.1016/j.ceb.2014.11.002 (2015).
Ghanbari, F., Mader, S. & Philip, A. Cholesterol as an Endogenous Ligand of ERRalpha Promotes ERRalpha-Mediated Cellular Proliferation and Metabolic Target Gene Expression in Breast Cancer Cells. Cells 9, doi:10.3390/cells9081765 (2020).
Cui, H. et al. Behavioral disturbances in estrogen-related receptor alpha-null mice. Cell Rep 11, 344–350, doi:10.1016/j.celrep.2015.03.032 (2015).
Ahmed, H. et al. Brain cholesterol and Alzheimer's disease: challenges and opportunities in probe and drug development. Brain 147, 1622–1635, doi:10.1093/brain/awae028 (2024).
Valencia-Olvera, A. C., Maldonado Weng, J., Christensen, A., LaDu, M. J. & Pike, C. J. Role of estrogen in women's Alzheimer's disease risk as modified by APOE. J Neuroendocrinol 35, e13209, doi:10.1111/jne.13209 (2023).
Lussey-Lepoutre, C. et al. Loss of succinate dehydrogenase activity results in dependency on pyruvate carboxylation for cellular anabolism. Nat Commun 6, 8784, doi:10.1038/ncomms9784 (2015).
Fregeau-Proulx, L. et al. Multiple metabolic pathways fuel the truncated tricarboxylic acid cycle of the prostate to sustain constant citrate production and secretion. Mol Metab 62, 101516, doi:10.1016/j.molmet.2022.101516 (2022).
Morland, C. & Nordengen, K. N-Acetyl-Aspartyl-Glutamate in Brain Health and Disease. Int J Mol Sci 23, doi:10.3390/ijms23031268 (2022).
Mazzoni, A. et al. On the dynamics of the spontaneous activity in neuronal networks. PLoS One 2, e439, doi:10.1371/journal.pone.0000439 (2007).
Raichle, M. E. Neuroscience. The brain's dark energy. Science 314, 1249–1250 (2006).
Bandara, A. B., Drake, J. C. & Brown, D. A. Complex II subunit SDHD is critical for cell growth and metabolism, which can be partially restored with a synthetic ubiquinone analog. BMC Mol Cell Biol 22, 35, doi:10.1186/s12860-021-00370-w (2021).
Albin, R. L. & Greenamyre, J. T. Alternative excitotoxic hypotheses. Neurology 42, 733–738, doi:10.1212/wnl.42.4.733 (1992).
Ilankoon, I., Samarasinghe, K. & Elgan, C. Menopause is a natural stage of aging: a qualitative study. BMC Womens Health 21, 47, doi:10.1186/s12905-020-01164-6 (2021).
Wise, J. Global life expectancy to increase by almost five years by 2050, study predicts. BMJ 385, q1126, doi:10.1136/bmj.q1126 (2024).

There is NO Competing Interest.

Download PDF

Version 1

posted

You are reading this latest preprint version

A cholesterol-coupled N-acetyl-aspartyl-glutamate metabolic network facilitates the neuroprotective impact of estradiol in neurons

Status:

Version 1

Abstract

Figures

Introduction

Results

Estradiol depletion impairs of the ERRα signalling network in the brain

ERα signalling activates ERRα signalling by sustaining cholesterol homeostasis

Loss of ERRα dysregulates the neuronal NAAG metabolic axis

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1