Which specific injury mechanisms are most likely after BBB crossing?

The short version: once spike crosses the BBB, the most likely injury mechanisms are mitochondrial failure in brain endothelial cells, tight junction breakdown, microglial-driven synaptic pruning, and excitotoxicity from failed glutamate clearance.

The evidence here does not track 2P spike specifically, but it maps the injury pathways that any inflammatory or viral protein in the brain follows. The mechanisms converge on four routes.

Mitochondrial dysfunction in the BBB endothelium comes first. A 2020 study showed that inflammatory stress (LPS) drives Drp1 phosphorylation 4-fold higher in brain microvascular endothelial cells, triggering mitochondrial ROS, membrane potential collapse, and cytochrome c release [2]. Blocking Drp1-Fis1 interaction with peptide P110 reduced BBB leakage and mortality in septic mice [2]. Spike protein is known to induce the same mitochondrial stress pathway in endothelial cells. Once the BBB is breached, the endothelial cells that should repair it are themselves damaged.

Tight junction proteins are directly dismantled by ROS and Rho kinase. Reactive oxygen species activate RhoA, which then phosphorylates occludin at T382 and claudin-5 at T207, causing them to disappear from the cell border within 30 to 60 minutes [10, 5]. This is not a slow leak. It is an active, enzyme-driven dismantling of the barrier. The same study showed that ROS increased monocyte migration across the BBB by 38% [10]. Once spike is inside, it keeps the gate open.

Microglia shift from surveillance to synaptic pruning. A 2026 study in a chronic migraine model found that neuronal stress downregulates dopamine receptor DRD2, which triggers CCL2 release from neurons, which activates microglia to engulf and destroy synapses [14]. Blocking CCL2 signaling with INCB3344 restored synaptic density and cognitive performance [14]. A 2023 study showed that microglial lactate transport via MCT4 is required for lysosomal acidification and proper pruning; when MCT4 is lost, pruning fails and synapses accumulate abnormally [16]. Spike protein in the brain would push microglia into the same hyper-phagocytic state.

Glutamate excitotoxicity from failed astrocyte clearance. A 2011 study found that a single mutation in coronavirus spike protein (Y241H in HCoV-OC43) caused paralytic disease in mice by downregulating the glutamate transporter GLT-1 on astrocytes [38]. Without GLT-1, glutamate accumulates in the synapse, overexcites neurons, and kills them. An AMPA receptor antagonist (GYKI-52466) reduced paralysis and partially restored GLT-1 without changing viral replication [38]. Spike protein from any source can produce the same astrocyte dysfunction.

These four mechanisms are not speculative. Each has been demonstrated in animal models with inflammatory or viral triggers that share properties with spike protein. The missing piece is a human study that tracks these endpoints after vaccination. That study has not been run.

My call: the mechanistic evidence is strong and consistent, but the population-level risk after vaccination remains unquantified because the relevant long-term study was never done. Confidence in the mechanisms: high. Confidence in the risk magnitude: not clear.