The delivery of ciliary neurotrophic factor (CNTF) to the central nervous system has recently been proposed as a potential means of halting or slowing the neural degeneration associated with Huntington's disease (HD). The following set of experiments examined, in detail, the ability of human CNTF (hCNTF) to prevent the onset of behavioral dysfunction in a rodent model of HD. A DHFR-based expression vector containing the hCNTF gene was transfected into a baby hamster kidney fibroblast cell line (BHK). Using a polymeric device, encapsulated BHK-control cells and those secreting hCNTF were transplanted bilaterally into rat lateral ventricles. Eight days later, the same animals received bilateral injections of quinolinic acid (QA, 225 nmol) into the previously implanted striata. A third group received sham surgery (incision only) and served as a normal control group. Bilateral infusions of QA produced a significant loss of body weight and mortality that was prevented by prior implantation with hCNTF-secreting cells. Moreover, QA produced a marked hyperactivity, an inability to use the forelimbs to retrieve food pellets in a staircase test, increased the latency of the rats to remove adhesive stimuli from their paws, and decreased the number of steps taken in a bracing test that assessed motor rigidity. Finally, the QA-infused animals were impaired in tests of cognitive function-the Morris water maze spatial learning task, and the delayed nonmatching-to-position operant test of working memory. Prior implantation with hCNTF-secreting cells prevented the onset of all the above deficits such that implanted animals were nondistinguishable from sham-lesioned controls. At the conclusion of behavioral testing, 19 days following QA, the animals were sacrificed for neurochemical determination of striatal choline acetyltransferase (ChAT) and glutamic acid decarboxylase (GAD) levels. This analysis revealed that QA decreased striatal ChAT levels by 35% and striatal GAD levels by 45%. In contrast, hCNTF-treated animals did not exhibit any decrease in ChAT levels and only a 10% decrease in GAD levels. These results support the concepts that implants of polymer-encapsulated hCNTF-releasing cells can be used to protect striatal neurons from excitotoxic damage, produce extensive behavioral protection as a result of that neuronal sparing, and that this strategy may ultimately prove relevant for the treatment of HD.