This dissertation investigated the diffusion and distribution of antibody-based therapeutics and viral vectors within the brain extracellular space. We hypothesized that for these large biologics diffusion in brain is more hindered relative to free diffusion and that these diffusion measurements could be used to aide CNS drug development.

The diffusion properties of full-length immunoglobulin G (IgG) were determined in free solution and in the cortex of anesthetized rats. The effective diffusion coefficient (D*) of IgG in rat somatosensory cortex was ~10-fold lower than the free diffusion coefficient (D). This hindrance was consistent with steric limitations from the finite ECS width and from binding interactions with endogenous Fc receptors, although factors such as IgG shape or charge may be involved as well. A model for antibody penetration at the brain-cerebrospinal fluid interface based on these diffusion measurements matched the in vivo distribution of IgG in brain after intrathecal infusion.

We further examined methods to improve antibody-based therapeutic distribution in brain utilizing smaller antibody fragments and osmotic manipulation of the brain microenvironment. The diffusion of antibody fragments like bivalent antigen-binding fragments (F(ab')2), monovalent antigen binding fragments (Fab), and single domain antigen binding fragments (sdAb) was hindered ~4- to 5-fold in brain; all fragments were less hindered than full-length IgG in brain. Osmotic manipulation of the brain ECS through mannitol co-injection enhanced the diffusion of full-length IgG and the smaller sdAb fragment in a concentration-dependent manner.

Finally, the diffusion properties of adeno-associated viral (AAV) vectors were explored. The diffusion of AAV8 was significantly hindered in brain compared to free medium (D* was 20-fold lower than D). While D* of AAV2 was not directly measured (due to difficulties obtaining a reliable high titer sample), models based on the vector size and binding properties predicted that AAV2 diffusion in brain would be extremely hindered.

The results presented in this thesis demonstrate that diffusion measurements of biologics in the brain can be used to illustrate important factors involved in local therapeutic distribution, to develop models to predict penetration into the brain, and to guide drug development initiatives such as vector design and novel delivery techniques.