Long-term perfusion of the cerebroventricular system of dogs without leakage to the peripheral circulation

Methods developed previously for studying the effect of cerebroventricular injection or ventriculocisternal perfusion of test substances are unsatisfactory because the test substance is not confined to the central compartment. Most likely the test substance enters the peripheral circulation via the arachnoid villi. The purpose of this paper is to describe a method for perfusing the cerebroventricular system of conscious dogs without passage of test substances to the peripheral circulation. With the method described, the mean (±SE) cerebroventricular pressure in conscious dogs was 7.4 ± 0.8 cmH₂O (n = 16), and the mean (±SE) production of cerebrospinal fluid (CSF) was 25 ± 0.3 μl/min (n = 16). Endogenously occurring migrating myoelectric complexes (MMCs) of the small intestine were recorded in dogs before catheters were implanted in the left and right lateral ventricles and the fourth ventricle and after catheter implantation during cerebroventricular perfusion with artificial CSF alone or with CSF containing sulfated (S-CCK- OP) or nonsulfated cholecystokinin octapeptide (NS-CCK-OP). Only cerebroventricular perfusion with S-CCK-OP (1.2 pmol · kg^-1 · min^-1; n = 20) replaced spontaneously occurring MMCs with a fed-like pattern of myoelectric activity. The results suggest that replacement of the fasting pattern of myoelectric activity with a fed-like pattern in the fasted dog was mediated by CCK-A receptors located in one or more brain nuclei surrounding the third ventricle.