Feeding rats diets containing oils that have a low alpha-linolenic acid [18:3(n-3)] content, such as sunflower oil, results in reduced amounts of docosahexaenoic acid [22:6(n-3)] in all brain cells and organelles compared to rats fed a diet containing soybean oil or rapeseed oil. During the period of cerebral development there is a linear relationship between the n-3 fatty acid content of the brain and that of food until alpha-linolenic acid represents approximately 200 mg/100 g food [0.4% of the total dietary energy for 18:3(n-3)]. Beyond that point brain levels reach a plateau. Similar values are also found for other organs. The level of 22:6(n-3) in membranes is little affected by the dietary quantity of linoleic acid [18:2(n-6)] if 18:3(n-3) represents approximately 0.4% of energy. In membranes from rats fed diets containing sunflower oil, Na+, K(+)-ATPase activity in nerve terminals was 60%, 5'-nucleotidase in whole brain homogenate was 80%, and 2',3'-cyclic nucleotide 3'-phosphodiesterase was 88% of that in membranes from rats fed diets containing soybean oil. A diet low in alpha-linolenic acid leads to anomalies in the electroretinogram, which partially disappear with age. It has little effect on motor activity, but it seriously affects learning tasks as measured with the shuttle box test. Rats fed a diet low in alpha-linolenic acid showed an earlier mortality in response to an intraperitoneal injection of a neurotoxin, triethyltin, than did rats fed a normal soybean oil diet.

Rat brain microvessel endothelial cells were immortalized by transfection with a plasmid containing the E1A adenovirus gene. One clone, called RBE4, was further characterized. These cells display a nontransformed phenotype and express typical endothelial markers, Factor VIII-related antigen and Bandeiraea simplicifolia binding sites. When RBE4 cells were grown in the presence of bFGF and on collagen-coated dishes, confluent cultures developed sprouts that extend above the monolayer and organized into three-dimensional structures. The activity of the blood-brain barrier-associated enzyme, gamma-glutamyl transpeptidase (gamma GTP), was expressed in these structures, not in the surrounding monolayer. Similar results were obtained with the microvessel-related enzyme alkaline phosphatase (ALP). Addition of agents that elevate intracellular cAMP reduced the formation of three-dimensional structures, but every cell inside the aggregates still expressed gamma GTP and ALP activities. Such structures, associated with high levels of gamma GTP and ALP activities, were also induced by astroglial factors, including (1) plasma membranes from newborn rat primary astrocytes or rat glioma C6 cells, (2) C6 conditioned media, or (3) diffusible factors produced by primary astrocytes grown in the presence of, but not in contact with RBE4 cells. RBE4 cells thus remain sensitive to angiogenic and astroglial factors for the expression of the blood-brain barrier-related gamma GTP activity, as well as for ALP activity, and could constitute the basis of a valuable in vitro model of the blood-brain barrier.

No abstract