Notes on Astrocyte energy and neurotransmitter metabolism in Alzheimer's disease

1. Introduction

  • Alzheimer’s disease (AD) presents with progressive cognitive decline, often beginning as Mild Cognitive Impairment (MCI), with memory loss, poor judgment, and disorientation.
    • MCI is a clinically compensatory phase that may last decades before dementia onset.
    • Histopathological hallmarks: extracellular amyloid-β (Aβ) plaques and intracellular neurofibrillary tangles (NFTs) containing hyperphosphorylated tau.
    • Aβ and NFTs cluster in cortex and hippocampus, leading to aberrant synaptic signaling and neuronal death.
  • Etiology: familial AD linked to APP/PSEN mutations causing Aβ overproduction; most AD cases are sporadic with age, lifestyle, and genetics as risk factors.
  • Early metabolic changes: deviant brain glucose metabolism is an early biomarker and may drive synaptic dysfunction and neurodegeneration.
  • Brain energy metabolism overview:
    • The brain uses about 20\% of total body energy expenditure, despite being ~ 2\% of body weight, with most energy allocated to neurotransmission and restoration of ionic gradients.
    • While glucose is the primary energy source, the brain can use alternative substrates (ketones, amino acids, fatty acids).
  • Glutamate and GABA: the brain’s main excitatory and inhibitory neurotransmitters, respectively. Both are cleared from the synapse primarily by astrocytes, which neutrally recycle them via the glutamate/GABA-glutamine cycle.
  • The cycle links neurotransmission to astrocyte energy metabolism; in AD, astrocyte remodeling perturbs this cycle, diminishing neuronal metabolic support and contributing to synaptic dysfunction and neurodegeneration.
  • Key open questions: how exactly astrocyte metabolic remodeling alters neurotransmitter recycling in AD and whether restoring astrocyte metabolism can slow clinical progression.

2. Astrocyte metabolism is linked to the glutamate/GABA-glutamine cycle

2.1 Neurons and astrocytes are metabolically connected through neurotransmitter recycling

  • Astrocytes are the major site of neurotransmitter clearance; neuron-astrocyte metabolic coupling is essential for rapid, high-fidelity signaling.
  • Glutamate uptake into astrocytes is predominantly via high-affinity transporters EAAT2 (GLT-1, SLC1A2) and EAAT1 (GLAST, SLC1A3); ~1\% of brain protein for GLT-1; ~5–10% of GLT-1 is expressed in presynaptic neurons, most is in astrocyte processes near synapses.
  • GABA uptake occurs via GABA transporters (GATs): neurons express GAT1; astrocytes express GAT1 and GAT3 (predominantly). Other cells (oligodendrocytes, microglia) may express GAT1 but their contribution is uncertain.
  • Glutamate uptake is energetically costly because it cotransports Na+ and H+ and countertransports K+, increasing the need to restore ion gradients via Na+/K+-ATPase.
  • Uptake of glutamate and GABA by astrocytes drains neuronal transmitter pools, which is counteracted by astrocyte synthesis of glutamine (GS-catalyzed): Glu → Gln, which is released to neurons and converted back to Glu by PAG.
  • Glutamine transport between astrocytes and neurons is mediated by SNAT family transporters: SNAT3/SNAT5 in astrocytes, SNAT1/SNAT2 in neurons, SNAT7/SNAT8 for some neuronal uptake; connexin 43 (Cx43) hemichannels may also contribute to glutamine transfer.
  • Astrocytes also release glutamine via other routes and can shuttle amino groups via BCAAs to support glutamate synthesis.
  • In short, neuron-astrocyte exchange of Glu, GABA, and Gln constitutes a tightly coupled nutrient and neurotransmitter cycle intimately tied to astrocyte energy metabolism.

2.2 Glutamate and GABA are substrates of the TCA cycle

  • Glutamate fuels oxidative metabolism in astrocytes; its carbon skeleton links to the TCA cycle via α-ketoglutarate (α-KG).
  • Primary enzymes for glutamate metabolism:
    • Glutamate dehydrogenase (GDH): primarily astrocytic, reversible reaction; oxidative deamination favors glutamate oxidation under physiological conditions but can run in the opposite direction with high ammonium.
    • Aspartate aminotransferase (AAT): active in both neurons and astrocytes; key component of the malate-aspartate shuttle (MAS) for sustaining glycolysis.
    • Alanine aminotransferase (ALAT) and Branched-chain amino acid aminotransferase (BCAT) also participate in brain glutamate homeostasis.
  • Glutamate is converted to glutamine in astrocytes by glutamine synthetase (GS); this fixation of ammonium is ATP-dependent and is a major ammonia sink in the brain (GS activity is central to cerebral ammonium homeostasis).
  • Glutaminase (PAG) in neurons converts glutamine back to glutamate; in GABAergic neurons, this glutamate is a precursor for GABA via glutamate decarboxylase (GAD).
  • GABA metabolism in astrocytes channels through the GABA shunt: GABA → succinic semialdehyde via GABA-T, then to succinate via SSADH, feeding the TCA cycle.
  • Overall, the glutamate/GABA-glutamine cycle is a major metabolic flux whose activity roughly tracks cerebral oxidative glucose metabolism; in cortex, glutamate/glutamine cycling accounts for about 80\% of total neurotransmitter cycling, with the GABA-glutamine cycle contributing the remainder.
  • Both neurons and astrocytes can oxidize glutamate and GABA for energy; astrocytic expression of PC (pyruvate carboxylase) provides an anaplerotic pathway to maintain the TCA cycle when glutamine synthesis is high, preventing depletion of TCA intermediates.

2.3 Glutamine synthesis is linked to astrocyte energy metabolism

  • Glutamine synthesis in astrocytes (GS) is essential to replenish neuronal glutamate and GABA pools; GS inhibition depletes neuronal glutamate and impairs both excitatory and inhibitory transmission.
  • GS is vulnerable to amyloid-β (Aβ) and oxidative stress; GS expression and activity decline with age and more so in AD brains, including near Aβ plaques.
  • Astrocytes rely on GS for ammonium fixation; impaired GS reduces glutamine supply to neurons, thereby constraining neuronal transmitter replenishment.
  • The TCA cycle intermediate α-ketoglutarate is a precursor for both glutamate and glutamine, so extensive glutamine synthesis can deplete astrocytic TCA intermediates; this is counterbalanced by anaplerotic reactions, notably via PC, which converts pyruvate to oxaloacetate and supports de novo glutamine synthesis.
  • PC activity in astrocytes is critical for sustaining glutamine production; flux through PC can account for up to ~20\% of total in vivo glucose oxidation in the awake rat hippocampus, linking energy metabolism to neurotransmitter synthesis.
  • Astrocyte glycogen metabolism is coupled to glutamine synthesis; glycogen stores can buffer energy supply for glutamine production, and glycogen metabolism supports learning and memory. Inhibition of glycogen metabolism impairs memory, whereas glycogen depletion increases glutamine synthesis when GS is active.
  • The glutamate/GABA-glutamine cycle is thus tightly coupled to astrocyte energy metabolism via glycolysis, the TCA cycle, and anaplerotic inputs; neurotransmitter cycling depends on astrocyte metabolic state and capacity for glutamine export.

3. Astrocytes in Alzheimer’s disease

3.1 General pathophysiology of astrocytes

  • Astrocytes show complex disease-specific alterations depending on brain region, stage, and pathology.
  • Reactive astrogliosis: response to blood-brain barrier (BBB) breach and infiltration of factors; astrocytes undergo transcriptional and metabolic remodeling, which can form a scar-like boundary in lesions but is not always the body of the scar.
  • In diseases without BBB breach (e.g., certain AD contexts), astrocytes can still remodel (hypertrophy with intermediate filament upregulation) and may preserve territorial domains (isomorphic astrogliosis).
  • Clasmatodendrosis: fragmentation of astrocytic processes with soma swelling, seen in ischemia, infections, dementia, and other disorders; indicates astrocytic distress.
  • Overall, glial decline and astrocytic dysfunction destabilize brain homeostasis and contribute to neurodegenerative progression.

3.2 Astrocytes display complex pathological profiles during Alzheimer’s disease

  • Human astrocytes are more complex than rodent astrocytes; rodent models may not fully recapitulate human astrocyte pathology (e.g., interlaminar astrocytes are primate-specific and greatly affected in AD).
  • In AD, astrocytes show both atrophic and hypertrophic phenotypes; atrophy often precedes Aβ plaque formation and can contribute to synaptic dysfunction and cognitive deficits by reducing synaptic coverage, neurotransmitter clearance, and metabolic support.
  • Hypertrophic astrocytes surround Aβ plaques and may participate in defense; however, region-specific patterns exist (hippocampus shows more hypertrophy than entorhinal/prefrontal cortices).
  • In vivo imaging (e.g., 11C-DED for astrogliosis, 11C-BU99008) shows dynamic astrocyte reactivity across disease stages, with high reactivity in MCI that declines as AD progresses to dementia.
  • Glial defense decline and dystrophic microglia may contribute to AD decompensation and cognitive decline; aging itself is associated with glial changes that raise risk for neurodegenerative diseases.

3.3 Astrocyte energy metabolism is disrupted in Alzheimer’s disease

  • AD brain shows progressive hypometabolism of glucose by PET with 18F-FDG; however, astrocytic contribution to FDG signal is significant via EAAT2-mediated glutamate uptake, suggesting astrocyte dysfunction can contribute to global hypometabolism.
  • Proteomics in human AD brains implicate glial carbohydrate metabolism as a strong correlate of disease signatures and cognitive impairment, linking energy metabolism with AD pathology.
  • Inflammation and cytokines modulate astrocyte energy metabolism; it remains to be clarified whether metabolic changes drive astrocyte reactivity or vice versa.

3.3.1 Astrocyte glucose uptake and glycolysis

  • GLUT1 (SLC2A1) is the main glucose transporter in astrocytes; its expression is reduced in cortex and hippocampus in AD, and cerebral capillaries show reduced GLUT1, contributing to lower brain glucose availability.
  • Decreased GLUT1 can contribute to hypometabolism and AD pathology by impairing astrocyte glycolysis and subsequent lactate production.
  • AD models show reduced glycolytic capacity in astrocytes: PK (pyruvate kinase) and HK (hexokinase) activities are lowered; 3xTG mice show reduced glycolysis and lactate release; Aβ exposure reduces glycolytic flux and L-serine production, which feeds into neuronal D-serine synthesis and NMDA receptor co-activation.
  • iPSC-derived astrocytes from AD patients show reduced glucose uptake and glycolytic enzyme expression, aligning with human data.
  • Some studies report compensatory hyperglycolysis or reverse metabolic shifts in certain models, highlighting complexity and heterogeneity of astrocyte energy responses in AD.
  • Overall, impaired astrocyte glycolysis and lactate provisioning may compromise neuronal energy metabolism and synaptic function in AD.

3.3.2 Astrocyte TCA cycle and mitochondrial function

  • Mitochondrial dysfunction is a hallmark of AD; in astrocytes, Aβ can initially increase TCA activity, but多数 studies report decreased oxidative metabolism and mitochondrial enzyme activity (e.g., α-ketoglutarate dehydrogenase, succinate dehydrogenase).
  • Mitochondrial oxygen consumption and ATP production are reduced in astrocytes from AD mouse models; some human iPSC-derived AD astrocytes show increased oxygen consumption, suggesting species differences or disease-stage-specific responses.
  • Elevated reactive oxygen species (ROS) in AD astrocytes prompts upregulation of the pentose phosphate pathway (PPP) to generate NADPH for antioxidant defenses and glutathione synthesis, indicating a compensatory redox response.
  • Overall, astrocyte mitochondria show impaired function in AD, contributing to energy deficits and altered neurotransmitter cycling.

3.3.3 Glycogen and pyruvate carboxylation

  • Astrocyte glycogen metabolism is tied to cognitive processes; abnormal glycogen bodies are reported in aging and AD brains.
  • Aβ can acutely increase glycogen content by inhibiting GS (glutamine synthesis), potentially causing glycogen accumulation when glutamine production is hampered.
  • In some AD models, cortical glycogen content decreases, while glycogen degradation pathways (e.g., glycogen phosphorylase activity) increase, signaling altered glycogen homeostasis.
  • GSK-3 overactivity in AD inhibits glycogen synthase, contributing to glycogen imbalances; glycogen metabolism supports astrocyte lactate release and influences glutamate uptake, linking energy state to neurotransmitter cycling.
  • Pyruvate carboxylase (PC) activity, the astrocyte-specific anaplerotic enzyme, is reduced in some AD rodent models, suggesting impaired astrocyte anaplerosis and diminished capacity to sustain de novo glutamine synthesis.

3.3.4 Astrocyte acetate metabolism

  • Astrocytes preferentially metabolize acetate (to acetyl-CoA) and can be studied with 13C-acetate tracers; neuronal acetate metabolism is limited.
  • In early AD models, astrocyte acetate metabolism can be reduced in hippocampus during early Aβ pathology, but later stages may show normalization or hypermetabolic states in some models (3xTG, APP/PS1, P301L tau).
  • In humans, increased in vivo acetate metabolism has been observed in MCI and AD, possibly reflecting compensatory astrocytic metabolism or increased monocarboxylate transporter expression (MCTs) in AD.
  • Acetate metabolism can support astrocyte energy production and feed into glutamine synthesis, potentially preserving neuronal GABA synthesis. Increases in astrocyte ketone production and lactate generation via acetate metabolism can contribute to neuronal energy support during AD progression.

4. Dysfunctional neurotransmitter metabolism and recycling in Alzheimer’s disease

4.1 Glutamine synthesis and ammonium homeostasis are impaired in AD

  • GS is potently inhibited by Aβ and shows reduced expression and activity in cortical tissue and hippocampus of AD patients and AD models.
  • GS loss is prominent around Aβ plaques and endfeet, and GS decrease correlates with Aβ burden.
  • Chronic oxidative stress also impairs GS; GS is more oxidized in MCI, predicting progression.
  • SNAT3 (astrocyte glutamine transporter) and SNAT1/SNAT2 (neuronal transporters) show altered expression in AD, reducing astrocyte-to-neuron glutamine transfer and neuronal uptake; SNAT3 downregulation may contribute to astrocyte glutamine accumulation and impaired ammonium handling.
  • Ammonium homeostasis is perturbed in AD: elevated brain ammonium is observed in AD; ammonium can be fixed by astrocyte GS and by GDH under certain conditions, but elevated NH4+ can shift GDH from oxidative glutamate dehydrogenation toward glutamate synthesis, potentially aggravating ammonium load.
  • Elevated glutamine levels have been reported in AD cortex; this may reflect disrupted glutamine export or neuronal uptake, or increased ammonium fixation; overall, disruption of ammonium homeostasis may contribute to AD pathology.
  • PAG (neuronal phosphate-activated glutaminase) activity is reduced in some AD brains, consistent with decreased neuronal glutaminergic metabolism; however, certain AD models show maintained or increased neuronal glutamine metabolism at later stages, suggesting adaptation.
  • Because glutamine is a key substrate for glutamate and GABA synthesis, impaired astrocyte glutamine supply can broadly disrupt neurotransmitter homeostasis and neuronal energetics.

4.2 Impaired astrocyte glutamate uptake and metabolism may drive excitotoxicity in AD

  • Glutamate uptake and EAAT2/GLT-1 expression are reduced in AD cortex and hippocampus; some reports show unchanged EAAT2, possibly due to splice variants or regional differences.
  • Aβ inhibits glutamate uptake and downregulates EAAT2 expression in cultured astrocytes; neuronal EAAT2 is less affected but its loss can disrupt energy metabolism and cause excitotoxic injury.
  • Excess extracellular glutamate near Aβ plaques correlates with reduced EAAT2 and is associated with impaired synaptic plasticity (LTP) and enhanced LTD; TBOA (a glutamate uptake blocker) mimics these effects, and extracellular glutamate scavenging can reverse LTD in models.
  • Ceftriaxone, which upregulates EAAT2, reduces extracellular glutamate and neurotoxicity, improves cognition in some AD models, and rescues GS and SNAT transporter expression, suggesting restored astrocyte glutamate handling supports neurotransmitter cycling.
  • Neuronal glutamate uptake may still be preserved in certain brain regions, but overall, disrupted astrocyte glutamate clearance contributes to excitotoxic risk and network dysfunction.
  • GDH activity, which promotes oxidative glutamate metabolism under stress, may be compromised by AD pathology, reducing the astrocyte’s ability to modulate glutamate utilization during high demand.
  • Glutamate carbon backbones feed the astrocyte TCA cycle; limited glutamate availability can force astrocytes to use alternatives, further compromising energy metabolism and transporter function.

4.3 Astrocyte GABA homeostasis is altered in AD

  • GABA levels are reduced in AD brains, consistent with loss of GABAergic synapses; GAD activity (glutamate decarboxylase) is also reduced in AD tissue, reducing GABA synthesis.
  • GAT1 and GAT3 expression is decreased in multiple brain regions in AD, contributing to impaired GABA reuptake; BGT1 levels may be increased in AD, but the functional significance requires clarification.
  • Reactive astrocytes in the dentate gyrus can release GABA via MAO-B and Best1 channels; MAO-B activity is increased in AD, and inhibition of MAO-B and Best1 can restore tonic inhibition and improve memory in AD models.
  • Putrescine-derived GABA synthesis in astrocytes via MAO-B can contribute to elevated astrocytic GABA; in some models neuronal GABA metabolism (GABA-T activity) may be preserved or reduced depending on region and stage.
  • Oxidative GABA metabolism is decreased in AD, which can exacerbate astrocyte GABA accumulation and disrupt the inhibitory tone; this contributes to increased neuronal excitability and seizures observed in AD.

5. Alternative energy substrates and astrocyte metabolism as a target in Alzheimer’s disease

  • Astrocyte energy metabolism and glutamate/GABA-glutamine cycling are tightly coupled; metabolic interventions aim to restore neurotransmitter recycling and support synaptic function.
  • Ketone-based strategies: ketone bodies (β-hydroxybutyrate, acetoacetate) are efficient brain fuels that bypass glycolytic blocks (via acetyl-CoA entry into the TCA cycle). A ketogenic approach can support neuronal energy when glycolysis is impaired.
    • Hepatic ketone production can be encouraged via a ketogenic diet (e.g., medium-chain triglycerides C8 and C10) to raise circulating ketones; C8/C10 can be metabolized in astrocytes and neurons, supporting energy and providing substrates for glutamine synthesis.
    • Ketone-based interventions have shown cognitive benefits in aging, MCI, and some AD models and human studies; βHB supplementation with pyruvate can reduce neuronal hyperexcitability and seizures in APP/PS1 mice; pyruvate restores astrocyte glycogen stores.
    • In astrocytes, C8 and C10 metabolism boosts ketone production (C8) and lactate generation (C10) and enhances astrocyte glutamine synthesis, supporting neuronal GABA synthesis.
  • Acetate metabolism as an astrocyte-specific probe:
    • 13C-acetate tracing reveals astrocyte-dominant metabolism; increased acetate metabolism in AD may reflect astrocyte-driven compensation for impaired glycolysis and energy production.
    • Elevated cerebral MCT expression in some AD models and human data may accompany increased acetate catabolism, potentially contributing to the energy balance and glutamine supply.
  • Branched-chain amino acids (BCAAs: Leu, Ile, Val) as nitrogen donors for glutamate synthesis via BCAT; BCAA metabolism can feed into the TCA cycle via BCKDH.
    • Some data show elevated BCAT expression in AD brain; circulating BCAA levels show complex relationships with AD progression (some cohorts show high BCAA linked to slower progression; others show reduced CSF/serum valine in AD; dietary BCAA effects can be model-dependent and sometimes worsen pathology in rodents).
    • BCAA carbon skeletons can be oxidized to feed the TCA cycle; GDH interacts with BCAA metabolism in astrocytes, and GDH dysfunction can affect anaplerosis.
  • Additional strategies to bolster astrocyte metabolism and redox status:
    • Antioxidants to reduce ROS and modulate PPP activity (NADPH production for glutathione regeneration) to alleviate oxidative stress.
    • Lifestyle interventions: caloric restriction, exercise, social and cognitive activity can enhance astrocyte function and synaptic plasticity.
    • Insulin sensitivity improvements and overall metabolic health can support brain energy homeostasis.
    • Circadian rhythm and clock genes: BMAL1 in astrocytes regulates astrocyte activation and GABA uptake; brain circadian disruption is linked to AD. Astrocyte BMAL1 loss increases astrogliosis and Aβ burden in some models, though cell-type specificity matters; circadian control of GS and GAT expression suggests complex links between metabolism, circadian biology, and AD pathology.
  • Practical takeaway: targeting astrocyte energy metabolism offers a multifaceted route to restore neurotransmitter recycling and synaptic function, potentially slowing AD progression.

6. Conclusions

  • AD is associated with extensive remodeling of astrocyte energy metabolism, including glycolysis, mitochondrial function, glycogen handling, and acetate metabolism.
  • The glutamate/GABA-glutamine cycle is tightly coupled to astrocyte energetics; disruptions in GS, EAATs, SNATs, PAG, GDH, AAT, PC, and GABA metabolism collectively impair neurotransmitter homeostasis and synaptic function.
  • Astrocyte dysfunction reduces neuronal glutamine supply, impairs glutamate and GABA replenishment, and can contribute to excitotoxicity and altered inhibitory signaling, promoting network dysfunction and neurodegeneration.
  • Therapeutic strategies that restore astrocyte energy metabolism or enhance glutamine supply (e.g., ketone-based therapies, acetate metabolism modulation, BCAA pathways, EAAT2 function, GS activity, circadian regulation) hold promise for slowing progression of AD by rebalancing neurotransmitter cycling and neuronal energetics.
  • A comprehensive approach that combines metabolic interventions with lifestyle and circadian rhythm management may provide the most robust benefit in AD treatment and prevention.