Central mediation of the conditioned taste aversion induced in rats by harmaline

The assumption that drugs used as unconditioned stimuli in conditioned taste aversion (CTA) studies act centrally was tested by comparing the effects of systemic and intracerebral injections of harmaline hydrochloride (H) in 340 rats. Intraperitoneal injection of 5-20 mg/kg but not of 2.5 mg/kg H administered 5 min after 15-min saccharin (0.1%) drinking decreased saccharin-water preference in a two-choice retention test, performed 48 h later, from 55% to 20%. Since CTA was not diminished when H (10 mg/kg) was injected into rats anesthetised immediately after saccharin drinking by pentobarbital (40 mg/kg), H (1.7-50 micrograms) was administered intracerebrally to anesthetised rats fixed in the stereotaxic apparatus. Injection of 3-6 micrograms H into the inferior olive elicited CTA comparable to that of systemic injection of 10 mg/kg H. Injections of 6 and 50 micrograms H into cerebellum and bulbar reticular formation elicited weaker CTA while neocortical, hypothalamic and mesencephalic applications were ineffective. CTA could also be elicited when 50 micrograms but not 6 micrograms H was injected into the inferior olive 1 or 2 h after saccharin drinking. This delay-dependent effect and failure of non-contingent H administration to change saccharin preference indicates that the H-induced CTA is not contaminated by a non-specific increase in neophobia. It is concluded that H probably elicits CTA by activation of caudal bulbar structures, including the nucleus of the solitary tract, area postrema and lateral reticular formation.