The authors are careful to frame this as a hypothesis rather than a demonstrated mechanism, and that restraint is what makes the review scientifically valuable. Epidemiology provides the clinical backdrop. The Rotterdam Study, a door-to-door survey of 5,559 people aged 55 to 95 conducted in the early 1990s, found active epilepsy prevalence of roughly 0.9 percent, rising with age. Yet cerebrovascular disease, neurodegenerative disorders, tumors, and prior brain injury account for much of the seizure risk in older populations, so the review insists that findings from Alzheimer-associated seizures cannot be imported wholesale as evidence for a mechanism in uncomplicated physiological aging.

The first evidence stream concerns the glycocalyx itself. This hydrated layer of glycoproteins, proteoglycans bearing glycosaminoglycans such as heparan sulfate and chondroitin sulfate, and glycolipids regulates vascular permeability, mechanotransduction, and cell interactions. Recent mouse experiments, highlighted in a 2025 Nature study by Shi and colleagues, established that mucin-type O-glycosylation, the attachment of N-acetylgalactosamine to serine or threonine residues on endothelial proteins, is a genuine determinant of blood-brain barrier integrity during aging. In male C57BL/6 mice, comparisons between 3-month and 21-month animals revealed reduced glycocalyx thickness and mucin-domain labeling alongside increased glycosaminoglycan signals, while measured sialic-acid signals remained unchanged. Critically, knocking down the enzyme C1GALT1, which adds galactose to form core 1 O-glycans, increased leakage of tracer, albumin, and IgG in young mice, and adeno-associated virus-mediated restoration of either C1GALT1 or B3GNT3 in 17-month-old mice improved barrier labeling and reduced tracer leakage eight weeks later.

The second evidence stream involves galectins, a family of beta-galactoside-binding lectins. In a lithium-pilocarpine seizure model in young male Sprague-Dawley rats, the galectin-3 inhibitor TD139, delivered at 8 milligrams per kilogram per day for ten days beginning four hours after seizures, reduced behavioral seizure measures and hippocampal inflammatory markers. In parallel experiments, kainate-exposed BV2 mouse microglia and neonatal rat primary microglia showed reduced NLRP3, ASC, cleaved caspase-1, and gasdermin-D readouts when galectin-3 was blocked pharmacologically or knocked down genetically. These intracellular inflammasome readouts, the review stresses, must not be conflated with extracellular galectin-glycan recognition at an endothelial surface; NLRP3 operates inside cells, and secreted galectin-3 acting through Toll-like receptor 4 on microglia represents a different compartment entirely.

Here lies the central anatomical problem the authors dissect with unusual rigor. Parenchymal microglia sit outside the vascular lumen, whereas the glycocalyx emphasized in the aging experiments faces circulating blood. Endothelial cells, basement membranes, and perivascular structures separate the two compartments, so microglial galectin-3 upregulation during seizures does not establish exposure of luminal endothelial glycans to microglia-derived galectin-3. The review formulates three testable routes: blood-derived galectin-3 could encounter luminal glycans; microglia-derived galectin-3 could act on the abluminal endothelial surface, whose glycan composition may differ from the luminal one; or altered barrier transport could permit intercompartmental exposure. None of these has been demonstrated in the aging epileptic brain.

The chemistry of galectin binding adds further complexity that undermines any simple story. Glycan elongation, ligand density, terminal modifications such as alpha2,3 versus alpha2,6 sialylation, and steric accessibility could each alter binding in different directions. Work by Stowell and colleagues showed that galectin-1, galectin-2, and galectin-3 differentially recognize extended LacNAc-containing and sialylated structures, meaning that desialylation cannot be assumed to increase all galectin binding equally. Because the aging mouse study found bulk sialylation signals unchanged, age-related desialylation is not an established initiating event. A thinner glycocalyx could expose underlying ligands, but it could also remove binding sites outright, or shed soluble glycoproteins that compete for galectin and dampen signaling. The direction of any functional change simply cannot be predicted from thickness measurements alone.

The review also delivers a cautionary lesson about galectin-1, which cannot be treated as a uniformly protective counterpart to galectin-3. Exogenous galectin-1 reduced selected seizure classes and improved inflammatory and neurodegenerative measures in an intrahippocampal pilocarpine model in rats, and recombinant galectin-1 engaged glycosylated CD45 to deactivate inflammatory microglia in other systems. Yet a separate study found that deleting the endogenous Lgals1 gene markedly reduced pilocarpine-associated neuronal death in adult mice, the opposite of what a simple protective model would predict. Cell source, concentration, timing, and subcellular location may reconcile these findings, but no experimentally validated galectin-1 to galectin-3 ratio has been shown to determine epileptogenesis.

What would it take to turn this convergence hypothesis into established biology? The authors propose compartment-resolved experiments with unusual specificity. In vitro, brain endothelial cultures under flow, supported by pericytes and astrocytes, could compare C1GALT1 or B3GNT3 perturbation with rescue conditions while luminal and abluminal galectin-3 exposure is varied independently, using lactose competition and carbohydrate-binding-defective galectin controls to distinguish genuine glycan-dependent binding from nonspecific injury. In vivo, a factorial design in young and aged animals of both sexes would perturb endothelial glycosylation and microglial versus endothelial galectin-3 separately and in combination, with continuous video-electroencephalography tracking not just acute seizures but persistent spontaneous recurrent seizures over long intervals, including after intervention withdrawal. The hypothesis gains support only if endothelial glycan changes precede altered galectin binding, correcting those glycans reverses the binding change, and seizure-related effects depend on galectin recognition in the same compartment.

Therapeutic implications are tantalizing but strictly provisional. Restoring endothelial O-glycosylation is experimentally feasible in aged mice, and galectin-3 inhibition modifies outcomes in a young-rat seizure model, yet neither intervention has been shown to act through the other. Inhaled TD139, also known as GB0139, demonstrated pulmonary target engagement in idiopathic pulmonary fibrosis patients but says nothing about central nervous system exposure. Peripheral glycocalyx repair strategies studied in sepsis, carotid injury, and diabetes offer translational precedent only. Biomarker hopes face similar constraints: circulating galectin-3 and glycosaminoglycan fragments can originate from vascular beds far from the brain, and multimorbidity in older adults compounds the ambiguity. The authors argue that longitudinal panels pairing blood-brain barrier imaging with glycan, inflammatory, and glial measurements, sampled before seizures, postictally, and interictally, would be needed to distinguish antecedent vulnerability from a mere consequence of seizures.

The deeper significance of this review lies in its methodological honesty. At a time when scientific communication often blurs the line between adjacent findings and established pathways, Legradi, Pareo, and Mir explicitly separate demonstrated results, mechanistic precedents from other biological settings, and untested hypotheses. They acknowledge that aging, barrier leakage, galectin changes, and seizures could share an upstream cause rather than form a causal chain, and that a feed-forward circuit, if it exists, requires independent perturbations and mediation experiments to establish its direction and necessity. Until those experiments are done, the glycocalyx-galectin connection remains a rigorously framed convergence hypothesis, but one that now comes with a precise experimental roadmap. For a field in which late-onset epilepsy is rising with demographic aging and often lacks an identifiable cause, defining exactly which experiments could confirm or falsify a sugar-based vascular mechanism may prove as consequential as the hypothesis itself.

Subject of Research: Endothelial glycocalyx remodeling and galectin signaling as a proposed mechanism in aging-associated epilepsy

Article Title: Endothelial glycocalyx remodeling and galectin signaling in aging associated epilepsy

Article References: Legradi, A., Pareo, C., & Mir, M. Y. (2026). Endothelial glycocalyx remodeling and galectin signaling in aging associated epilepsy. Frontiers in Neurology, 17, Article 1961370. https://doi.org/10.3389/fneur.2026.1961370

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1961370

Keywords: blood-brain barrier, endothelial glycocalyx, galectin-3, galectin-1, epileptogenesis, brain aging, neuroinflammation, NLRP3 inflammasome, O-glycosylation, microglia, astrocytes, late-onset epilepsy