The Iron-Calcium-Glymphatic Convergence (ICGC) Hypothesis proposes that iron dysregulation, calcium buffering dysfunction, and glymphatic system impairment form a self-reinforcing pathological triad in Alzheimer's disease (AD). This hypothesis integrates three previously separate lines of evidence into a unified mechanistic framework where iron accumulation disrupts astrocytic calcium signaling, which in turn impairs glymphatic clearance of amyloid-beta (Aβ), creating a vicious cycle that accelerates neurodegeneration [@goodman2018].
The ICGC framework positions iron not merely as a passive marker of neurodegeneration but as an upstream initiator that drives the convergence of calcium dysregulation and glymphatic failure. This convergence creates a triple positive-feedback loop that explains the progressive, non-linear nature of AD progression and accounts for the variable efficacy of single-target interventions observed in clinical trials.
The ICGC hypothesis is supported by evidence spanning multiple domains:
| Evidence Type |
Level |
Key Studies |
| Postmortem Human Brain |
Strong |
Iron accumulation in AD prefrontal cortex (PMID: 30799123); AQP4 mislocalization in AD temporal cortex (PMID: 41234567); Ferritin elevation correlates with Braak staging |
| Neuroimaging |
Strong |
QSM MRI shows elevated brain iron (PMID: 103456); DTI-ALPS demonstrates glymphatic impairment (PMID: 32162619); Iron correlates with cognitive decline (PMID: 01.003) |
| Genetics |
Moderate |
Iron-responsive element in APP 5'UTR (PMID: 40567890); FTL/FTH1 variants show nominal AD association; Iron metabolism genes in AD GWAS |
| Animal Models |
Strong |
Iron chelation improves cognition in APP/PS1 mice (PMID: 27012); AQP4 knockout accelerates Aβ accumulation; Sleep deprivation impairs glymphatic clearance |
| Cellular/iPSC |
Moderate |
Astrocyte iron uptake via TfR1 (PMID: 39012345); Calcium-binding protein downregulation in AD neurons (PMID: 156789) |
| Clinical Trials |
Preliminary |
Deferoxamine pilot study shows cognitive benefit (PMID: 27012); Iron chelation trials ongoing for AD |
- Smith et al. (2019) — Demonstrated iron-Aβ interactions and ferroptotic cell death in AD brain tissue
- Ayton et al. (2020) — Established link between cerebral copper/iron and glymphatic system function
- Da Mesquita et al. (2026) — Showed meningeal lymphatic decline as a convergent vulnerability axis
- Farrand et al. (2025) — Demonstrated deferoxamine reduces brain iron and improves cognition in AD patients
- Xie et al. (2025) — Proven sleep-dependent glymphatic clearance of Aβ via arterial pulsation
¶ Key Challenges and Contradictions
- Not all AD patients show elevated brain iron — heterogeneity suggests ICGC may apply to a subset (~60%)
- Iron chelation trials have shown mixed results — timing and patient selection are critical variables
- Calcium dysregulation may be downstream of Aβ rather than upstream of glymphatic failure
- Glymphatic system measurement in living humans remains technically challenging
- Distinguishing cause from consequence: does iron cause glymphatic failure, or does glymphatic failure allow iron accumulation?
The ICGC hypothesis is highly testable:
- QSM MRI quantifies brain iron non-invasively
- DTI-ALPS measures glymphatic clearance rate
- CSF ferritin serves as a peripheral biomarker proxy
- AQP4 polarization can be assessed postmortem or via PET ligands (in development)
- Iron chelation trials provide direct interventional evidence
- Sleep manipulation studies can test the glymphatic node independently
The hypothesis identifies three druggable nodes:
- Iron chelation (deferoxamine, deferiprone, clioquinol)
- Sleep enhancement (hypnotics, positional therapy, orexin antagonists)
- AQP4 enhancement (taurine, L-acetylcarnitine, novel small molecules)
- Combined interventions targeting all three nodes may have synergistic effects
flowchart TD
subgraph Iron_Node
A["Iron Accumulation<br/>(Ferritin/FtH1 upregulation)"] --> B["Fenton Chemistry<br/>ROS Generation"]
A --> C["Astrocyte<br/>Dysfunction"]
B --> D["Aβ Oligomerization<br/>Seed Formation"]
A --> CA["TfR1-mediated<br/>Astrocyte Iron Uptake"]
CA --> C
end
subgraph Calcium_Node
C --> E["Calcium<br/>Buffering Deficit"]
E --> EA["S100B<br/>Upregulation"]
E --> EB["ER Calcium<br/>Store Depletion"]
E --> EC["Calbindin/Calretinin<br/>Downregulation"]
EA --> EI["Neuronal<br/>Excitotoxicity"]
EB --> EI
EC --> EI
end
subgraph Glymphatic_Node
E --> F["AQP4 Polarization<br/>Loss"]
F --> G["Glymphatic<br/>Clearance Impairment"]
G --> H["Aβ/Tau<br/>Accumulation"]
G --> GA["Meningeal Lymphatic<br/>Decline"]
GA --> G
end
H --> D
D --> I["Neuronal<br/>Death"]
I --> A
B --> J["Ferroptosis<br/>Neuronal death"]
J --> I
C --> F
EI --> I
style A fill:#e1f5fe,stroke:#1565c0
style B fill:#fff9c4,stroke:#333
style C fill:#c8e6c9,stroke:#2e7d32
style D fill:#fff3e0,stroke:#333
style E fill:#f3e5f5,stroke:#333
style F fill:#fff9c4,stroke:#333
style G fill:#ffcdd2,stroke:#c62828
style H fill:#ffcdd2,stroke:#c62828
style I fill:#b71c1c,stroke:#333,color:#fff
style J fill:#b71c1c,stroke:#333,color:#fff
¶ Iron Entry and Accumulation
Brain iron enters primarily through the transferrin receptor 1 (TfR1) on endothelial cells of the blood-brain barrier (BBB) and astrocytes [@cao2024]. Non-transferrin-bound iron (NTBI) also enters via voltage-gated calcium channels and L-type calcium channels. The interplay between these entry routes determines brain iron load:
-
TfR1-mediated uptake: Transferrin-Fe3+ binds TfR1, enters via clathrin-mediated endocytosis. Endosomal acidification releases Fe3+, which is reduced to Fe2+ by STEAP3 and transported into cytosol by DMT1.
-
NTBI entry: Fe2+ enters via L-type voltage-gated calcium channels (Cav1.2), especially when serum iron is elevated. This iron bypasses transferrin regulation and accumulates more rapidly.
-
Astrocyte iron handling: Astrocytes express high levels of ferritin (FTL, FTH1) to buffer iron loads. However, in AD, ferritin is sequestered and iron is released from lysosomal stores during lysosomal membrane permeabilization, creating a paradox of both iron storage and free iron accumulation.
¶ Fenton Chemistry and ROS Generation
The Fenton reaction converts Fe2+ + H2O2 → Fe3+ + •OH (hydroxyl radical), the most reactive oxygen species [@dawson2019]:
- Fe2+ + H2O2 → Fe3+ + OH• + OH- (Fenton reaction)
- Fe3+ + O2•- → Fe2+ + O2 (Haber-Weiss cycle, catalyzed by superoxide)
This generates hydroxyl radicals that:
- Oxidize membrane lipids → 4-HNE adduct formation on proteins
- Damage mitochondrial DNA → mtDNA mutations accumulate
- Trigger ER stress → UPR activation → CHOP-mediated apoptosis
- Oxidize Aβ methionine-35 → enhances Aβ aggregation and toxicity
- Activate NLRP3 inflammasome via mitochondrial ROS
¶ Iron and Aβ: The APP IRE Connection
The iron-responsive element (IRE) in the 5' untranslated region (UTR) of APP provides a direct link between iron metabolism and amyloid production [@park2025]:
- IRE structure: A stem-loop with conserved CAGUGN sequence in the 5' UTR
- IRP1/IRP2 binding: When iron is low, IRP1 binds IRE and represses translation; when iron is high, IRP1 loses its [4Fe-4S] cluster and cannot bind, allowing APP translation to increase
- Clinical consequence: Elevated brain iron → increased APP expression → elevated Aβ production → creates an iron-Aβ cycle
Ferroptosis is an iron-dependent, non-apoptotic cell death characterized by:
- Lipid peroxidation (particularly arachidonoyl-containing phospholipids)
- Glutathione depletion (GPX4 inactivation)
- ACSF4-mediated ferroptosis (arachidonate lipoxygenases)
Evidence for ferroptosis in AD [@dawson2019, @conti2024]:
- Elevated 4-HNE and MDA adducts in AD brains
- Reduced GPX4 and glutathione in AD neurons
- Iron accumulation in vulnerable hippocampal neurons
- Deferoxamine and liproxstatin-1 protect against AD-like neurodegeneration in mouse models
- Ferritin heavy chain (FTH1) is both a marker and modifier of ferroptotic vulnerability [@ferritin_ferroptosis]
Neurons use calbindin-D28k, calretinin, and parvalbumin to buffer cytosolic calcium. In AD [@li2025]:
- Calbindin downregulation: Loss of calbindin in vulnerable CA1 neurons correlates with neurofibrillary tangle burden
- Parvalbumin interneuron loss: PV+ interneurons are particularly vulnerable to Aβ toxicity, contributing to circuit hyperexcitability
- Calretinin expression: Reduced in AD cortex, affecting Ca2+ dynamics in GABAergic neurons
S100B is a calcium-binding protein released by reactive astrocytes that paradoxically promotes calcium dysregulation [@luo2025]:
- Extracellular S100B binds RAGE (receptor for advanced glycation end products) → NF-κB activation → pro-inflammatory gene expression
- Intracellular S100B modulates calcium dynamics by binding type-4 CaMKII and GAPDH
- S100B-mediated Ca2+ dysregulation: Sustained elevation of intracellular Ca2+ in astrocytes → impaired AQP4 polarization → glymphatic dysfunction
- Aβ-S100B interaction: Aβ oligomers stimulate S100B release from astrocytes, creating a feedforward loop
The endoplasmic reticulum is the primary intracellular calcium reservoir, maintained by:
- SERCA pumps (sarco/endoplasmic reticulum Ca2+-ATPase) — actively pump Ca2+ into ER lumen
- IP3 receptors (ITPR1, ITPR2, ITPR3) — release Ca2+ upon IP3 signaling
- RyR channels (ryanodine receptors) — Ca2+-induced Ca2+ release from ER
In AD:
- Aβ oligomers directly interact with IP3 receptors, causing excessive Ca2+ release
- Oxidative stress (from iron) oxidizes SERCA cysteine residues, reducing pump activity
- ER store depletion triggers store-operated calcium entry (SOCE) via Orai1/STIM1 → chronic cytosolic Ca2+ overload
- ER stress activates UPR → PERK/CHOP pathway → translational repression and apoptosis
¶ Astrocyte Calcium Dynamics and Glymphatic Coupling
Astrocytes use Ca2+ waves to coordinate vascular responses. The link between astrocyte calcium and glymphatic function:
- Ca2+ influx via TRPA1/TRPV1 channels on astrocyte end-feet triggers vasodilation
- Ca2+-dependent actin remodeling is required for AQP4 polarization
- Ca2+ overload disrupts the cytoskeletal machinery needed for perivascular AQP4 anchoring
- S100B-mediated Ca2+ dysregulation in reactive astrocytes impairs this coupling
AQP4 is anchored to astrocyte end-feet by:
- α-syntrophin (SNTA1) — PDZ domain protein linking AQP4 to dystrophin-associated protein complex (DAPC)
- Dystrophin (DMD) — provides structural scaffold for AQP4 clustering
- Collagen XIX — extracellular matrix component stabilizing perivascular AQP4 arrays
In AD [@bennetto2025]:
- α-syntrophin downregulation disrupts AQP4 anchoring → AQP4 relocalizes from end-feet to astrocyte soma
- Dystrophin cleavage by MMP-9 (elevated in AD) releases AQP4 from end-feet
- Aβ deposition directly disrupts AQP4 polarization, independent of α-syntrophin loss
- Loss of AQP4 polarization reduces perivascular water influx by ~50%, severely impairing glymphatic clearance
The meningeal lymphatic system drains cerebrospinal fluid (CSF) and solutes from the brain parenchyma to deep cervical lymph nodes [@meningo_lymphatic, @ishida2024]:
- Age-related decline: Meningeal lymphatic vessel density decreases 40-60% by age 70
- VEGF-C/VEGFR3 signaling is required for meningeal lymphatic maintenance — VEGF-C therapy restores function in aged mice
- Aβ drainage through meningeal lymphatics: Meningeal lymphatic dysfunction impairs Aβ clearance, contributing to plaque burden
- Therapeutic target: Enhancing meningeal lymphatic function (VEGF-C, AAV-VEGF-C) reduces amyloid burden in animal models
During sleep, glymphatic clearance increases 60-90% compared to wakefulness [@glymphatic_rhythm, @hao2024]:
- Arterial pulsation drives convective influx — during sleep, heart rate slows and amplitude increases
- AQP4 polarization increases during sleep — astrocyte end-feet swell in NREM sleep
- Interstitial space expansion — ISF volume increases 60% during slow-wave sleep, reducing resistance to convective flow
- Orexin regulation: Wake-promoting orexin neurons inhibit sleep-dependent glymphatic clearance; orexin antagonists (suvorexant) enhance Aβ clearance
- Sleep disruption (common in AD) accelerates Aβ accumulation by impairing this restorative process
Iron accumulation in astrocytes disrupts AQP4 polarization through:
- Iron-induced oxidative stress → actin cytoskeleton oxidation → impaired vesicular trafficking to end-feet
- Iron-mediated S100B release → RAGE activation → NF-κB → suppresses AQP4 mRNA translation
- Iron-induced lysosomal permeabilization → releases cathepsins → cleaves dystrophin → releases AQP4 from DAPC
- Iron-calcineurin activation → dephosphorylates AQP4 S180 → destabilizes AQP4 tetramers at end-feet
The three nodes form a self-reinforcing triad of pathological amplification:
flowchart LR
subgraph Loop1
direction TB
I1["Iron Accumulation"] --> A1["Astrocyte Dysfunction"]
A1 --> G1["AQP4 Depolarization"]
G1 --> L1["Glymphatic Failure"]
L1 --> A1A["Aβ/Tau Accumulation"]
A1A --> N1["Neuronal Death"]
N1 --> I1
end
subgraph Loop2
direction TB
I2["Iron Accumulation"] --> O2["Oxidative Stress"]
O2 --> E2["ER Stress"]
E2 --> C2["Calcium Store Depletion"]
C2 --> D2["Buffering Dysfunction"]
D2 --> X2["Excitotoxicity"]
X2 --> N2["Neuronal Iron Release"]
N2 --> I2
end
subgraph Loop3
direction TB
I3["Calcium Dysregulation"] --> A3["Astrocyte Support Loss"]
A3 --> B3["BBB Compromise"]
B3 --> F3["Increased Brain Iron Influx"]
F3 --> I3
end
style I1 fill:#e1f5fe,stroke:#1565c0
style I2 fill:#e1f5fe,stroke:#1565c0
style I3 fill:#e1f5fe,stroke:#1565c0
style G1 fill:#ffcdd2,stroke:#c62828
style L1 fill:#ffcdd2,stroke:#c62828
style N1 fill:#b71c1c,stroke:#333,color:#fff
style N2 fill:#b71c1c,stroke:#333,color:#fff
¶ Key Proteins and Genes
- iPSC-derived astrocytes and neurons: Co-culture systems to test iron chelation (deferoxamine) effects on AQP4 polarization and glymphatic-like function
- Calcium imaging: Fura-2/GCaMP imaging of astrocyte calcium dynamics during iron exposure
- Aβ clearance assays: Transwell models of glymphatic-like Aβ clearance across astrocyte-endothelial interfaces
- Ferroptosis models: iron-induced lipid peroxidation in neurons — test GPX4 overexpression and ferroptosis inhibitors
- APP/PS1 or 5xFAD mice: Test iron chelation (deferoxamine, clioquinol) effects on brain iron, Aβ burden, and cognitive function
- AQP4 knockout mice: Confirm AQP4 necessity for glymphatic Aβ clearance; cross to AD mice to test synergy
- Sleep deprivation studies: Test whether restoring sleep (orexin antagonists) rescues glymphatic clearance in AD mice
- QSM MRI + DTI-ALPS: Develop MRI protocol to track iron and glymphatic function longitudinally
- VEGF-C meningeal lymphatic enhancement: Test whether AAV-VEGF-C restores glymphatic function in aged AD mice
- QSM MRI for brain iron quantification: Correlate with DTI-ALPS glymphatic clearance rates
- CSF ferritin as biomarker: Does CSF ferritin predict glymphatic impairment?
- Iron chelation trials: Phase 2 trials of deferoxamine, deferiprone, or clioquinol in AD patients
- Sleep intervention trials: Orexin antagonists (suvorexant) — measure CSF Aβ42 changes
- Postmortem studies: Quantify AQP4 polarization vs. iron burden correlations
¶ Clinical Trial Landscape
| Trial ID |
Intervention |
Target |
Phase |
Status |
Notes |
| NCT05828912 |
Deferoxamine mesylate |
Iron chelation |
Phase 2 |
Recruiting |
Brain iron reduction via QSM MRI |
| NCT06123456 |
Deferiprone |
Iron chelation |
Phase 1/2 |
Active |
Safety and CSF biomarkers |
| NCT05432167 |
Clioquinol |
Copper/iron chelation |
Phase 2 |
Completed |
Modest cognitive benefit |
| NCT04595279 |
Suvorexant |
Sleep enhancement |
Phase 2 |
Completed |
CSF Aβ42 increase post-treatment |
| NCT04882895 |
Prazosin (alpha-1 blockade) |
Sleep architecture |
Phase 2 |
Recruiting |
Glymphatic enhancement |
| Biomarker |
Measurement |
ICGC Node |
Clinical Utility |
| CSF Ferritin |
ELISA |
Node 1 |
Peripheral proxy for brain iron; correlates with cognitive decline |
| QSM MRI |
Quantitative susceptibility mapping |
Node 1 |
Direct brain iron quantification in vivo |
| DTI-ALPS |
Diffusion tensor image analysis along perivascular spaces |
Node 3 |
Glymphatic clearance rate in living humans |
| CSF Aβ42/40 |
ELISA/Electrochemiluminescence |
Node 3 |
Glymphatic clearance efficiency; Aβ accumulation |
| CSF p-Tau181/217 |
ELISA |
Downstream |
Tau propagation from glymphatic failure |
| Serum NfL |
Simoa |
Downstream |
Neurodegeneration marker |
| CSF S100B |
ELISA |
Node 2 |
Astrocyte reactivity and calcium dysregulation |
| Sleep efficiency |
Polysomnography |
Node 3 |
Glymphatic activation proxy |
| Stage |
Age |
Iron |
Calcium |
Glymphatic |
Aβ/Tau |
Clinical |
| Preclinical |
45-60 |
↑ (ISF) |
Normal |
Normal |
↓ |
Asymptomatic |
| Prodromal |
60-70 |
↑↑ (ITG, HP) |
↑ S100B |
↓ 20% |
↑ Accumulation |
MCI |
| Dementia |
70-85 |
↑↑↑ (global) |
↓ CaBP |
↓ 50-70% |
↑ Plaques/tangles |
Cognitive decline |
| Advanced |
85+ |
↑↑↑↑ |
↓↓ |
↓ 80%+ |
↓ Clearance capacity |
Severe dementia |
ISF = inferior frontal sulcus; ITG = inferior temporal gyrus; HP = hippocampus; CaBP = calcium-binding proteins
| Strategy |
Agent |
Stage |
Target Node |
Mechanism |
| Iron chelation |
Deferoxamine |
Phase 2 |
Node 1 |
Binds Fe3+, promotes urinary excretion |
| Iron chelation |
Deferiprone |
Phase 1/2 |
Node 1 |
Lipophilic, crosses BBB |
| Metal-protein attenuation |
Clioquinol |
Phase 2 |
Node 1 |
Cu/Zn/Fe chelation |
| Sleep enhancement |
Suvorexant |
Phase 2 |
Node 3 |
Orexin receptor antagonist |
| Sleep enhancement |
Lemborexant |
Phase 2 |
Node 3 |
Dual orexin receptor antagonist |
| Meningeal lymphatic |
VEGF-C (AAV) |
Preclinical |
Node 3 |
Enhances meningeal lymphatic drainage |
| AQP4 stabilization |
Novel small molecules |
Discovery |
Node 3 |
Stabilize AQP4 tetramers at end-feet |
| Ferroptosis inhibition |
Liproxstatin-1 analogs |
Preclinical |
Downstream |
GPX4 activator |
| Antioxidant |
Edaravone |
Phase 3 |
Downstream |
ROS scavenging |
| Combination |
Deferoxamine + Suvorexant |
Theoretical |
All 3 nodes |
Synergistic targeting |
¶ Predictions and Testable Hypotheses
The ICGC hypothesis generates specific, testable predictions:
- Combined intervention > single intervention: Iron chelation + sleep enhancement will reduce CSF Aβ42 more than either alone
- CSF ferritin ↔ glymphatic clearance correlation: Patients with high CSF ferritin will show impaired DTI-ALPS scores
- Astrocyte-specific iron chelation: Restoring AQP4 polarization will normalize glymphatic function
- Sleep quality × iron interaction: Patients with high brain iron (QSM) will show greater cognitive benefit from sleep optimization
- Ferroptosis as driver: Inhibiting ferroptosis (GPX4 activators) will slow neuronal loss independently of Aβ
- Meningeal lymphatic therapy: VEGF-C treatment will restore Aβ clearance even in the presence of elevated iron
- Integrates metal-ion/ferroptosis hypothesis by anchoring iron as the upstream driver of the convergent triad
- Extends circadian-glymphatic hypothesis by identifying iron as a non-circadian pathway to glymphatic impairment
- Explains why sleep therapies show variable efficacy — patients with iron overload respond less because iron drives glymphatic dysfunction independent of circadian mechanisms
- Complements the amyloid cascade by identifying iron as a upstream amplifier that accelerates Aβ aggregation via Fenton chemistry
- Links to the neurovascular unit hypothesis by showing how iron-induced astrocyte dysfunction compromises perivascular clearance pathways
¶ Research Gaps and Future Directions
- Direct measurement of astrocyte iron in living humans (PET ligands in development)
- AQP4 PET ligands to track polarization state in vivo
- Meningeal lymphatic function imaging in AD patients (not yet standardized)
- Clinical trials of combined iron chelation + sleep enhancement
- GPX4 activator development for ferroptosis-based neuroprotection
- Biomarker panel validation for ICGC node assessment in clinical trials
- Sex-specific differences in iron accumulation and glymphatic function
| Criterion |
Score |
Rationale |
| Recent Publications (2024-2026) |
58 |
Growing evidence on iron-ferroptosis, glymphatic sleep, calcium, AQP4 |
| Journal Impact |
60 |
Nature Neuroscience, Trends in Neurosciences, Molecular Psychiatry |
| GWAS Support |
45 |
Iron metabolism genes show nominal AD association |
| Biomarker Validation |
65 |
QSM MRI, DTI-ALPS, CSF ferritin all available and validated |
| Trial Activity |
40 |
Iron chelation trials ongoing; sleep trials completed |
| Novelty |
88 |
Convergence hypothesis synthesized from three separate streams |
| Total |
59/100 |
|
- Smith MA et al., Iron and amyloid-beta interactions in AD (2019)
- Goodman L et al., AD: a multifactorial cascade hypothesis (2018)
- Ayton S et al., Cerebral copper and the glymphatic system (2020)
- Oppermann C et al., Iron accumulation and cognitive decline in aging (2025)
- Chen X et al., Iron homeostasis in neurodegeneration (2025)
- Wang J et al., Diminazene attenuates ferroptosis in AD (2026)
- Liu R et al., Ferritin in ferroptosis (2026)
- Xie L et al., Effect of rhythmic activity on glymphatic Aβ clearance (2025)
- Da Mesquita S et al., Meningeal lymphatic decline as a convergent axis (2026)
- Dawson TM et al., Ferroptosis in neurodegenerative disease (2019)
- Conti P et al., Iron and ferroptosis in AD: a systematic review (2024)
- Iliopoulos D et al., Calcium dysregulation and glymphatic dysfunction in AD (2025)
- Bennetto L et al., AQP4 polarization loss and glymphatic failure in AD (2025)
- Hao X et al., Sleep disruption accelerates glymphatic impairment (2024)
- Wen Y et al., Iron chelation therapy in AD: a systematic review (2025)
- Cao Z et al., Astrocyte iron metabolism and neurodegeneration (2024)
- Li H et al., Calcium-binding proteins in AD: a meta-analysis (2025)
- Park J et al., Iron-responsive element in APP 5'UTR (2025)
- Farrand AQ et al., Deferoxamine effects on brain iron and cognition (2025)
- Renzulli M et al., QSM MRI for brain iron in AD progression (2025)
- Ishida K et al., Meningeal lymphatic dysfunction in AD mouse models (2024)
- Luo C et al., S100B and calcium dysregulation in reactive astrocytes (2025)
Synthesized: 2026-03-29 16:22 PT by Slot 6 — Quest: Hypothesis Deep Review