C2: Universal tissue access

The blood-brain barrier (BBB) is not a single structure but three overlapping mechanisms that together maintain a carefully isolated chemical environment for the central nervous system.

Specialised endothelial cells with tight junctions. The endothelial cells lining brain capillaries are sealed together by tight junction protein complexes (claudins, occludins, junctional adhesion molecules) that prevent molecules from passing between cells (the paracellular route that leaky peripheral capillaries permit). To cross, a molecule must go through the cells.

Active efflux transporters. Even molecules that enter the endothelial cells are actively pumped back out by efflux transporters — P-glycoprotein (P-gp), breast cancer resistance protein (BCRP), and the multidrug resistance proteins. These pumps have broad substrate specificity and eject a large fraction of small-molecule drugs that would otherwise diffuse in. This is why many drugs that work elsewhere in the body simply never reach therapeutic concentration in the brain.

Restricted transcytosis. The endothelial cells have very low rates of the vesicular transport (pinocytosis) that lets peripheral capillaries shuttle material across.

Crossing the BBB always involves a tradeoff against its protective function. Brute-force methods (focused ultrasound with microbubbles, hyperosmotic agents like mannitol) temporarily disrupt the tight junctions and allow drug entry — but during the open window, they also admit whatever else is circulating, including potential pathogens and toxins. Trojan-horse methods (antibody shuttles targeting the transferrin or insulin receptor, AAV capsids selected for transcytosis) ride existing transport mechanisms without disrupting the barrier, but are payload-size limited and become immunogenic on repeated dosing. Direct injection (intrathecal, intraventricular, intraparenchymal) bypasses the barrier entirely but is invasive and distributes unevenly.

No approach is yet routine for any-payload, any-time CNS delivery — which is precisely what AIHS-grade universal tissue access would require.

Adeno-associated virus (AAV) is a small, non-pathogenic virus repurposed as a gene-therapy delivery vehicle. Certain serotypes — most notably AAV9 and engineered derivatives like AAV-PHP.B and AAV-PHP.eB — can cross the blood-brain barrier with modest efficiency after systemic administration.

The mechanism: the viral capsid binds receptors on brain capillary endothelial cells, undergoes receptor-mediated transcytosis through the cells, and is released into the brain parenchyma where it transduces neurons and glia, delivering its DNA payload. Zolgensma (onasemnogene abeparvovec, approved 2019 for spinal muscular atrophy) is the most-deployed AAV-CNS therapy and demonstrates that the route works clinically.

The limits are significant and shape what AAV can contribute to AIHS:

Immunogenicity. AAV provokes neutralising antibody responses, so most patients can receive a given serotype only once before pre-existing antibodies block re-dosing. Many patients also have pre-existing immunity from natural AAV exposure, excluding them from treatment. For AIHS — which would need repeated, sustained delivery over a treatment course — single-dose-only is a fundamental constraint.

Payload size. AAV's genome capacity is limited to ~4.7 kilobases, too small for many genes of interest and for sophisticated multi-component circuits.

Uneven distribution. Transduction within the brain is non-uniform — some regions and cell types are reached far better than others.

Capsid engineering programmes (directed evolution of new serotypes, rational capsid design) are producing variants with improved tropism and reduced immunogenicity, but the fundamental constraints of viral delivery persist. AAV-CNS will likely remain a workhorse modality through Phase II while these limits are chipped away rather than eliminated.

The BBB receives the bulk of attention, but it is one of several privileged compartments, and they differ markedly in both difficulty and clinical motivation.

The eye is, in practice, the most tractable. The blood-retina barrier is similar in principle to the BBB, but the eye's surgical accessibility changes everything: intravitreal injection delivers therapy directly to the retina, bypassing the barrier. Several gene therapies are already FDA-approved for ocular use (Luxturna for inherited retinal dystrophy). The eye is also immunologically privileged, which reduces some delivery complications.

Joint cartilage is avascular — it has no blood supply at all — so systemic delivery via the bloodstream simply cannot reach it. Therapies rely on intra-articular injection and local diffusion through the cartilage matrix, which is slow and limited by the dense extracellular matrix. The challenge here is diffusion physics rather than an active barrier.

The testis is protected by the blood-testis barrier, but it is rarely a therapeutic target — most interventions aim to preserve rather than alter fertility — so the motivation to solve delivery to it is limited, and it receives correspondingly little research attention.

Each site has its own constraint pattern: active efflux and tight junctions for the brain, surgical bypass available for the eye, diffusion limits for cartilage, low motivation for the testis. The BBB is unique in being both clinically critical (an enormous unmet need for CNS therapies across neurodegeneration, psychiatric disease, brain cancer, and stroke) and physically the most aggressive barrier. This combination is why C2 attention concentrates there, and why AIHS deployment is more constrained by BBB-crossing capability than by access to any other privileged site. Significant NIH BRAIN Initiative investment reflects this priority.