Neurovascular NF-κB Signaling: Blood–Brain Barrier (BBB) Dysfunction, Hypoperfusion, and Impaired Aβ Efflux

The neurovascular evidence presents a more consistent picture than the neuronal or glial literature. In AD-related models, NF-κB activation in pericytes and cerebral endothelial cells is predominantly associated with blood–brain barrier (BBB) disruption, increased vascular permeability and leukocyte trafficking, reduced cerebral blood flow, and impaired Aβ efflux. This does not imply that physiological vascular NF-κB signaling is inherently harmful, because its normal roles in host defence, vascular repair, and homeostasis were not directly addressed in these studies. Rather, the available evidence specifically implicates disease-conditioned NF-κB signaling in neurovascular dysfunction.

The strongest evidence involves the interaction between APOE4 and brain pericytes. Bell et al. showed that mice expressing human APOE4, as well as Apoe-deficient mice, developed BBB disruption through activation of a pericytic cyclophilin A (CypA)–NF-κB–matrix metalloproteinase 9 (MMP9) pathway, whereas APOE2 and APOE3 maintained greater pathway restraint (Bell et al., 2012). Astrocyte-derived APOE3 suppressed this signaling through lipoprotein receptors, but APOE4 failed to do so. Increased CypA–NF-κB activity enhanced MMP9-mediated degradation of basement-membrane and tight-junction components, increasing BBB permeability and reducing cerebral blood flow. Importantly, these vascular abnormalities preceded detectable neuronal and synaptic dysfunction, and genetic or pharmacological inhibition of CypA attenuated both vascular and neural abnormalities. The findings therefore support a sequence in which APOE4-dependent loss of restraint on pericytic CypA–NF-κB–MMP9 signaling produces BBB and microcirculatory dysfunction that can precede neuronal injury.

Aβ can also activate NF-κB directly in cerebral endothelial cells. Gonzalez-Velasquez et al. found that soluble Aβ1–40 aggregates produced stronger NF-κB nuclear translocation than monomeric or fibrillar Aβ and caused increased endothelial permeability, monocyte adhesion, and transendothelial migration (Gonzalez-Velasquez et al., 2011). NF-κB inhibition prevented these abnormalities, supporting a direct link between soluble Aβ-induced NF-κB activation and loss of endothelial barrier function. The greater activity of soluble aggregates also emphasizes that vascular effects depend on Aβ assembly state rather than simply total peptide concentration, although the endothelial monoculture used in this study does not reproduce the full complexity of the neurovascular unit.

A complementary endothelial pathway may impair removal of Aβ from the brain. Park et al. showed that Aβ1–42 reduced expression of P-glycoprotein (P-gp), an endothelial transporter involved in brain-to-blood Aβ efflux, in brain endothelial cells and near plaques in 5xFAD mice (Park et al., 2014). Mechanistically, Aβ activated the receptor for advanced glycation end products (RAGE) and NF-κB; RAGE blockade or NF-κB inhibition attenuated P-gp loss, whereas RAGE overexpression enhanced it. These findings support an Aβ → RAGE → NF-κB → P-gp reduction pathway. Such signaling could generate a feed-forward cycle in which increasing Aβ activates endothelial NF-κB, reduced P-gp weakens Aβ efflux, and prolonged cerebral Aβ exposure further sustains vascular dysfunction.

Together, these studies suggest several potentially reinforcing neurovascular mechanisms. APOE4 can weaken the BBB through pericytic CypA–NF-κB–MMP9 signaling, soluble Aβ can increase endothelial permeability and leukocyte trafficking through NF-κB, and Aβ–RAGE–NF-κB signaling can reduce P-gp-mediated Aβ efflux. Barrier disruption may consequently expose neural tissue to blood-derived inflammatory factors, while impaired Aβ efflux prolongs exposure of the neurovascular unit to the initiating pathological stimulus. Although this integrated feed-forward model is biologically plausible, no single study has demonstrated the complete sequence in human AD.

Presenilin 1 (PS1) provides an additional molecular connection between AD-related proteins and endothelial NF-κB signaling. Tanaka et al. showed that PS1 can function independently of its γ-secretase activity as a scaffold for a BCR–CK2α–p65 complex in endothelial cells (Tanaka et al., 2018). This complex promoted phosphorylation of RelA/p65 at Ser529 and increased NF-κB-dependent transcription, whereas Psen1 deficiency reduced p65 phosphorylation and recruitment to target promoters. Unlike the APOE4–CypA–MMP9 and Aβ–RAGE–P-gp pathways, however, the study did not establish consequences for BBB permeability, cerebral perfusion, Aβ transport, or neuronal injury. It therefore identifies an endothelial NF-κB regulatory mechanism whose relevance to AD-associated vascular dysfunction remains unresolved.

Overall, disease-associated neurovascular NF-κB signaling is more consistently linked to detrimental outcomes than the corresponding signaling observed in neurons, microglia, or astrocytes. The strongest evidence connects APOE4-dependent pericytic CypA–NF-κB–MMP9 activity to BBB breakdown and hypoperfusion, while endothelial Aβ–NF-κB and Aβ–RAGE–NF-κB pathways increase permeability, leukocyte trafficking, and potentially impair Aβ efflux (Bell et al., 2012, Park et al., 2014, Gonzalez-Velasquez et al., 2011). These findings do not justify global inhibition of vascular NF-κB, because its physiological functions remain important and incompletely examined in AD models. Instead, they support selective targeting of disease-conditioned vascular pathways, particularly APOE4/CypA–NF-κB–MMP9 and Aβ/RAGE–NF-κB signaling, while preserving normal endothelial and perivascular functions.