THE EFFECTS OF INTRAHIPPOCAMPAL RAPHE AND/OR SEPTAL GRAFTS IN RATS WITH FIMBRIA-FORNIX LESIONS DEPEND ON THE ORIGIN OF THE GRAFTED TISSUE AND THE BEHAVIORAL TASK USED

Long-Evans female rats sustained electrolytic lesions of the fimbria and the dorsal fornix and, two weeks later, received intrahippocampal suspension grafts of fetal tissue. The grafts were prepared from regions including either the medial septum and the diagonal band of Broca (septal grafts), or the mesencephalic raphe (raphe grafts), or from both these regions together (co-grafts). All rats were submitted to a series of behavioural tests (home cage and open-field locomotion, spontaneous alternation, radial-arm maze and Morris water maze performance) run over two periods after grafting (one to nine weeks and 20-35 weeks). Two weeks after completion of behavioural testing, histological (acetylcholinesterase and Cresyl Violet staining) and/or neurochemical (choline acetyltransferase activity, high-affinity synaptosomal uptake of choline and serotonin, noradrenaline, serotonin and 5-hydroxyindolacetic acid concentrations) verifications were performed on the hippocampus. Compared to sham-operated rats, lesion-only rats exhibited hyperactivity which was transient in a familiar environment (home cage) and lasting in an unfamiliar one (open field), decreased rates of spontaneous T-maze alternation, and impaired memory performance in both the radial-arm maze and the Morris water maze. These rats also showed decreased cholinergic and serotonergic markers with a maximal depletion in the septal two-thirds of the hippocampus. Noradrenaline concentration tended to be increased in the dorsal third of the hippocampus, but was not modified in the other two-thirds. While septal grafts specifically increased the cholinergic markers and raphe grafts the serotonergic ones, neither of these grafts produced a lasting effect on any behavioural variable. Conversely, the co-grafts, which increased both the cholinergic and serotonergic markers in the septal two-thirds of the hippocampus, completely normalized the Morris water maze probe trial performance, but failed to affect any of the other behavioural variables. Our present results confirm that grafts of fetal neurons injected into the denervated hippocampus may induce a neurochemical recovery that depends on the anatomical origin of the grafted cells, and that co-grafting two fetal brain regions allows the combination of their individual neurochemical properties. Furthermore, our results show that these neurochemical effects of the co-grafts may be involved in the recovery of behavioural function observed in the water maze. However, somewhat paradoxically, those effects appear inefficient for inducing any recovery in other behavioural tasks, even in the radial-arm maze, which is assumed to measure similar spatial functions. Finally, it is suggested that the co-grafts might be more efficient in attenuating the spatial reference memory (Morris water maze) rather than the spatial working memory (radial maze) deficits subsequent to fimbria-fornix lesions.