Transcytosis of protein through the mammalian cerebral epithelium and endothelium. III. Receptor-mediated transcytosis through the blood-brain barrier of blood-borne transferrin and antibody against the transferrin receptor.

Diferric-transferrin (Tf; 80K mol. wt.) and the OX26 antibody (150K mol. wt.) against the transferrin receptor (TfR) were evaluated in the rat at light and ultrastructural levels as potential vehicles for the blood to brain transcellular transfer (transcytosis) of native horseradish peroxidase (40K mol. wt.), which by itself does not cross the blood-brain barrier (BBB). OX26, the Fab fragment of OX26 (50K mol. wt.), and Tf complexed to two ferric ions were conjugated to HRP irreversibly in a 1:1 molar ratio. The indirect immunoperoxidase technique with OX26 as the monoclonal primary antibody applied to the surface of cryostat sections or delivered intravenously to the live rat revealed TfRs on BBB capillaries, arterioles, and venules; TfRs were absent on non-BBB vessels supplying the circumventricular organs (i.e., median eminence, choroid plexus). OX26-HRP and OX26(Fab)-HRP delivered intravenously and diferric-Tf-HRP administered into the carotid artery labeled BBB vessels throughout the CNS without discernible disruption of the BBB or extravasation of the blood-borne probes into the brain parenchyma. No reaction product for the probes was observed in sites deficient in a BBB. Each of the macromolecular conjugates was endocytosed by BBB endothelia and labeled presumptive endocytic vesicles, endosomes, and dense bodies. OX26-HRP and Tf-HRP, but not OX26(Fab)-HRP, appeared to undergo transcytosis through BBB endothelia for subsequent labeling of perivascular cells. Distinct differences in the intracellular and extracellular distributions between OX26-HRP and Tf-HRP were identified: (1) endocytosis and sequestration of blood-borne OX26-HRP within BBB endothelia were more prominent than those for diferric-Tf-HRP; (2) only OX26-HRP labeled the Golgi complex in BBB endothelia; (3) peroxidase labeling of CNS perivascular clefts and perivascular cells in rats receiving diferric-Tf-HRP was conspicuous at less than 1 h postinjection but not so in rats with blood-borne OX26-HRP at 5 min through 6 h postinjection; and (4) peroxidase-labeled CNS neurons and glial cells were identified readily in rats receiving diferric-Tf-HRP. The results suggest that the receptor-mediated, transendothelial transfer of Tf-HRP from blood to brain is more efficient and direct than that of OX26-HRP. Labeling of the Golgi complex in BBB endothelia with blood-borne OX26-HRP implies that the transendothelial transfer of OX26-HRP follows intraendothelial pathways associated with the process of adsorptive transcytosis. A diagram is provided depicting the possible intracellular and transcellular pathways within BBB endothelia available to blood-borne diferric-Tf and OX26 as vectors for delivery into the CNS of non-lipid-soluble macromolecules that otherwise are denied entry by the blood-brain fluid barriers.