Non-invasive modulation of meningeal lymphatics ameliorates ageing and Alzheimer's disease-associated pathology and cognition in mice

Historically, the brain was considered immune privileged due to the presumed absence of a lymphatic drainage system. In 2015, meningeal lymphatic vessels (mLVs) were identified in the dura, forming a network that clears macromolecular waste and inflammatory mediators, directs immune cell transport, and coordinates immune responses in the CNS. Emerging work links mLV function to ageing and multiple neurological diseases including Alzheimer's disease (AD), Parkinson's disease, traumatic brain injury, subarachnoid hemorrhage, and CNS viral infection, with transport capacity influencing disease progression. AD features β-amyloid (Aβ) aggregation and neurofibrillary tangles causing neuronal dysfunction and cognitive decline. mLV function declines with ageing and AD, potentially exacerbating cognitive dysfunction. While intracisterna magna delivery of VEGF-C can enhance mLV function and improve learning and memory, invasive approaches are ill-suited for chronic neurodegeneration. Because mLVs lie superficially in the dura, transcranial neuromodulation may non-invasively modulate mLV drainage. This study tests the hypothesis that near-infrared (NIR) light can enhance mLEC function and mLV drainage, thereby improving cognition and alleviating AD-related pathology. The authors propose that enhancing mitochondrial metabolic homeostasis in mLECs by NIR light promotes cell adhesion and junction integrity, boosting meningeal lymphatic transport and ameliorating CNS pathology.

Prior work established structural and functional meningeal lymphatic vessels in the dura and implicated their roles in macromolecule clearance, immune cell trafficking, and CNS immune coordination. mLV dysfunction has been associated with ageing, AD, PD, TBI, SAH, and neurotropic viral infections, where altered mLV transport impacts disease outcomes. In AD models, mLV impairment correlates with cognitive decline. Viral-mediated VEGF-C delivery via intracisterna magna increases mLV function, enhances drainage of CNS toxins, and improves learning and memory, supporting mLVs as a therapeutic target. Photobiomodulation at 1267 nm has been reported to enhance clearance of injected Aβ and red blood cells via brain lymphatics, although mechanisms were unclear. Cytochrome c oxidase (CCO) is considered a primary photoreceptor for red/NIR light (630–900 nm), where photons enhance respiratory chain efficiency and ATP synthesis, potentially restoring cell function.

Animal models: C57BL/6J mice were used. Aged cohort: 15–17 months, male and female. Young controls: 1.5 months. AD models: 5xFAD (6 months, male) and APPswe/PS1ΔE9 (APP/PS1, 11 months, male) with WT littermates as controls. Housing under 12 h light/dark, 20–23 °C, 50–60% humidity. Light treatment: Non-invasive transcranial NIR at 808 nm (continuous wave) delivered to intact skull with hairless scalp over 0.8 cm². Power densities: 10, 20, or 50 mW/cm² for 10 min per session (doses 6–30 J/cm²), 3 sessions/week for 4 weeks. Surface scalp temperature monitored; no significant heating effect detected. Control mice received identical handling and anesthesia with indoor lighting for 10 min. Light penetration: Ex vivo measurement through skull with hairless scalp indicated ~25% transmittance at 808 nm, independent of output power. Behavioral testing: Open field (mobility/anxiety), novel object location (NOL), novel object recognition (NOR), Y-maze (novel arm exploration), and Morris water maze (MWM) for spatial learning and memory. Lymphatic drainage assays: CSF tracers OVA-A647 or OVA-ICG administered via intracisterna magna (10 µL; 0.5 mg/mL for OVA-A647; 12.5 µg/mL for OVA-ICG) at 2.5 µL/min. In vivo NIR-II fluorescence imaging tracked tracer kinetics in superficial cervical lymph nodes (sCLNs) and deep CLNs (dCLNs). Ex vivo fluorescence quantification in CLNs and brain at 2 h post-injection. Vessel morphology and blood flow: Immunohistochemistry of whole-mount meninges with LYVE-1 to assess mLV coverage and diameter; CD31 for blood vessels. Laser speckle contrast imaging (LSCI) measured cerebral blood flow. mLV ablation: Visudyne (0.5 mg/mL, 10 µL, i.c.m.) followed after 15 min by 689 nm photoconversion at five calvarial sites (600 mW/cm², 83 s/spot; 50 J/cm²) to ablate mLVs. These mice then underwent the same NIR treatment protocol. Neuropathology: Brain sections stained for Aβ1–42, Iba1 (microglia), NeuN (neurons), synaptophysin (Syn), and MAP2 (dendrites). Quantified Aβ area fraction, microglial morphology (branch length, branch number, endpoints, cell number), neuronal counts, Syn and MAP2 area fractions. Mitochondrial assays: Transmission electron microscopy (TEM) of mLECs in mLVs to assess cellular junctions/arrangement and mitochondrial morphology/length. Mitochondrial superoxide (MitoSOX) and membrane potential (TMRE) in mLECs of aged and AD mice. Transcriptomics: RNA-seq on hippocampus (n=4/group), meninges (n=3/group), and sorted mLECs (FACS; DAPI− CD45− CD31+ PDPN+; n=3/group). Differential expression with DESeq2; GO/KEGG enrichment via clusterProfiler; Hallmark GSEA; visualization with ggplot2/pheatmap. Statistics: two-tailed t-test, one-/two-way ANOVA with Sidak’s multiple comparisons; significance at P<0.05.

General: 808 nm transcranial light enhanced mLV drainage and improved cognition in aged and AD mice without affecting locomotion or anxiety; cerebral blood flow remained unchanged. The 20 mW/cm² setting showed stable efficacy. Aged mice (15–17 months): - mLV drainage to dCLNs (OVA-A647 area in dCLNs) increased vs untreated aged: 10 mW/cm² 7.103 ± 1.443% vs 2.653 ± 0.760% (P=0.00997); 20 mW/cm² 10.338 ± 1.057% vs 2.653 ± 0.760% (P=0.00006); 50 mW/cm² 8.297 ± 0.748% vs 2.653 ± 0.760% (P=0.00163). - mLV structure (LYVE-1 area fraction) restored toward young levels: +10 mW/cm² 1.605 ± 0.166% vs aged 1.004 ± 0.089% (P=0.00112); +20 mW/cm² 1.727 ± 0.111% (P=0.00006); +50 mW/cm² 1.691 ± 0.063% (P=0.00013). - Cognition improved: NOR preference (aged + light) 58.92 ± 2.813% vs aged 41.38 ± 3.456% (P=0.00033); improved NOL and Y-maze novel arm time; OF unchanged. AD models: - 5xFAD (6 months, male): NOR index improved 63.466 ± 4.501% vs 42.987 ± 3.839% (P=0.00047). MWM acquisition latency decreased vs untreated (P=0.0000001); probe performance improved; swim speed unchanged. - APP/PS1 (11 months, male): NOR improved 65.312 ± 3.974% vs 44.732 ± 6.231% (P=0.01716). MWM latency reduced (P<0.0000001). - WT controls: no significant change with light (e.g., NOR WT + light 59.981 ± 3.492% vs WT 55.494 ± 2.693%, P=0.63265; MWM P=0.99929). Pathology in 5xFAD/APP-PS1: - Aβ deposition in hippocampus (HPC) and prefrontal cortex (PFC) reduced in light-treated AD mice, approaching WT levels. - Microglial activation decreased with increased branch length, branch number, and endpoints; Iba1+ cell counts showed no significant change. - Neuronal and synaptic integrity improved: NeuN+ neuron counts and Syn area fraction increased to near WT; MAP2 area fraction increased; dendrite organization more orderly. Similar neuroprotection observed in APP/PS1 mice. Hippocampal RNA-seq (AD): - GO enrichment: ion transport, cell migration, neurogenesis/development altered by light. - KEGG: PI3K-AKT, apelin, TGF-β, oxytocin signaling, α-linolenic acid metabolism altered. - Hallmark GSEA: down-regulation of apoptosis, coagulation, hypoxia, p53 pathway, and inflammatory responses (complement, IL2-STAT5, TNF-α via NF-κB). mLV drainage and structure in AD: - 5xFAD: Increased OVA-ICG outflow to CLNs by NIR-II FLI (8.225 ± 0.925% vs 2.249 ± 0.302%, P=0.00092). LYVE-1+ coverage not reduced in 5xFAD, but light increased mLV diameter (29.105 ± 0.894 µm vs 26.024 ± 0.727 µm, P=0.04623). - APP/PS1: Light increased CLN tracer signals and LYVE-1+ area fraction (lymphangiogenesis); meningeal blood vessel distribution unchanged. Necessity of mLVs for benefit: - Visudyne/photoconversion ablated mLVs eliminated cognitive benefits of light in aged and 5xFAD mice (e.g., NOR: aged Vis./photo. + light 39.780 ± 6.939% vs Vis./photo. 42.339 ± 5.873%, P=0.95789; 5xFAD Vis./photo. + light 46.170 ± 3.061% vs Vis./photo. 40.767 ± 4.361%, P=0.47442). - sCLN tracer kinetics enhanced by light in intact mice but flat and low after mLV ablation regardless of light. - dCLN fluorescence at 2 h increased with light in 5xFAD (13548.915 ± 552.890 vs 9431.965 ± 568.433, P=0.00006) but not after ablation (Vis./photo. + light 9141.019 ± 505.416 vs Vis./photo. 9415.882 ± 741.405, P=0.94270). Glymphatic enhancement by light was also abrogated by mLV ablation. mLEC ultrastructure and mitochondria: - TEM showed disrupted mLEC junctions/arrangement and rounded mitochondria with fractured/fuzzy cristae in AD; light restored compact arrangement, elongated oval/rod mitochondria with clear cristae; mitochondrial length increased (significant). - MitoSOX decreased and TMRE membrane potential trended upward in mLECs of aged/AD mice after light, indicating improved mitochondrial function. Transcriptomics of meninges and mLECs: - Meninges: Up-regulation of genes for cell proliferation, adhesion, differentiation, development, transport; no distinct Vegfc change but increased VEGF pathway-related genes (Nckap1l, Nck1, Calm2, Cdc42). Tight junction/adhesion components up-regulated. KEGG: increased cell cycle, mismatch repair, vascular smooth muscle contraction; decreased AD and neurodegeneration pathways. Hallmark GSEA: upregulated oxidative phosphorylation, cell cycle control, DNA repair, immune regulation; OXPHOS genes (Cox7c, Cox17, Atp5pb, Mfn2, Immt, Iscu) enriched. - mLECs (FACS): Increased Lyve1, Prox1, and Tgfa; Pecam1 unchanged; mitochondrial metabolism gene programs up. These data support enhanced oxidative phosphorylation and lymphangiogenesis in mLECs after light.

The data support that non-invasive 808 nm transcranial photobiomodulation enhances meningeal lymphatic drainage and improves cognition in aged and AD mice through functional restoration of meningeal lymphatic endothelial cells. Mechanistically, light likely activates cytochrome c oxidase in mitochondria, boosting oxidative phosphorylation and ATP production, thereby improving mLEC metabolic homeostasis, cell proliferation/adhesion, and tight junction integrity. This restores compact mLEC arrangement and expands/normalizes mLV structure and function, enhancing CSF/macromolecule clearance, reducing Aβ burden and microglial activation, and preserving neurons and synapses, culminating in improved learning and memory. The benefits are dependent on intact mLVs, as ablation abolishes drainage enhancement and cognitive gains, and also negates glymphatic improvements, suggesting coordinated CSF circulation between meningeal lymphatic and glymphatic systems. Light did not enhance cognition in WT mice, consistent with a ceiling effect when baseline drainage is intact. Transcriptomic alterations in hippocampus, meninges, and mLECs align with reduced inflammatory/stress pathways and elevated oxidative phosphorylation, cell cycle and repair programs, supporting a global shift toward neuroprotection and tissue homeostasis.

This study demonstrates that non-invasive, transcranial near-infrared light can target superficially distributed meningeal lymphatics to enhance drainage, ameliorate AD-associated pathology, and improve cognition in aged and AD mouse models. Photobiomodulation restores mitochondrial metabolism and cellular junction organization in mLECs, expanding/normalizing mLV function and promoting clearance of Aβ and inflammatory mediators. The cognitive benefits are mLV-dependent and occur without altering cerebral blood flow. These findings identify mLV-targeted phototherapy as a promising strategy for neurodegenerative disease intervention. Future work should assess preventive or periodic treatment at early disease stages, delineate contributions of meningeal immune modulation and potential direct cortical effects, and evaluate translational applicability.

The mechanism is inferred from mitochondrial, ultrastructural, and transcriptomic correlates rather than direct causal manipulation of cytochrome c oxidase activity. Benefits were demonstrated in mouse models; translational relevance to humans remains to be established. Light did not improve cognition in WT mice, suggesting efficacy may depend on pre-existing drainage deficits. Some enrichment analyses reported unadjusted P values, and experiments primarily assessed short-term outcomes following a 4-week treatment. Other mechanisms (e.g., meningeal immune modulation or direct cortical effects) may contribute and were not fully dissected.