1. Unit activity in the rat lateral hypothalamus (LHA) was recorded during discrimination learning of cue tone (CTS) or cue light (CL) stimulation that predicted reward by glucose or intracranial self stimulation (ICSS), or aversion by weak electric shock or tail pinch. Roles of the catecholaminergic and cholinergic systems in the LHA were investigated by electrophoretic application of dopamine (DA), norepinephrine (NE), acetylcholine (ACh), and their antagonists [spiperone (SPP), phenoxybenzamine (PBZ), phentolamine, propranolol, and atropine (Atr)]. 2. Activity of 264 LHA neurons was recorded. Of these, 234 (89%) responded during CTS learning in one or more phases. Of 121 neurons tested by both rewarding and aversive stimuli, 86 (71%) discriminated reward and aversion and their respective CTSs. 3. Effects of DA on 138, NE on 134, and ACh on 73 neurons were tested. Among these, 67 were tested with all three. DA inhibited 40 and excited 14. NE inhibited 74 and excited 10. ACh excited 35 and inhibited 3. DA-sensitive neurons responded to both NE (P < 0.001) and ACh (P < 0.05) more often than DA-insensitive neurons. In most cases, the effect of DA was similar to the effect of NE, and opposite to the effect of ACh. The inhibitory effect of DA was blocked by SPP, a D2 antagonist, and the excitatory effect of ACh was blocked by Atr. The inhibitory effect of NE was blocked by the beta-antagonist, propranolol, and enhanced by the alpha-antagonist, phentolamine. 4. DA-sensitive neurons responded to both rewarding and aversive stimuli and respective CTS+ and CTS- more often than DA-insensitive neurons (P < 0.01). The effect of DA was usually similar to the effect of rewarding stimuli and their predicting CTS+ and was opposite to the effect of aversive stimuli and their predicting CTS-. 5. The proportion of NE-sensitive neurons that responded to rewarding and aversive stimuli was the same as the proportion of NE-insensitive neurons that responded to the same stimuli. NE-sensitive neurons responded to CTS+ and CTS- more often than NE-insensitive neurons (P < 0.01). The effect of NE was usually similar to the effect of rewarding stimuli and predicting CTS+, and opposite to the effect of aversive stimuli and predicting CTS-. 6. ACh-sensitive neurons responded to aversive stimuli and predicting CTS- more often than ACh-insensitive neurons (P < 0.01), but the response ratio of ACh-sensitive neurons to rewarding stimuli was similar to that of ACh-insensitive neurons. The effect of ACh was usually similar to the effect of aversive stimuli and their predicting CTS-, and opposite to the effect of rewarding stimuli and their predicting CTS+. 7. Neuronal responses to CTS+ predicting rewarding stimuli were blocked by SPP, whereas CTS- predicting aversive stimuli were blocked by Atr. Neuronal responses were similarly acquired for CL predicting rewarding stimuli, and for CTS+. Neuronal responses to CL were also blocked by SPP. 8. Sensitivity of LHA neurons to DA and to single pulse stimulation of an ICSS site was reduced plastically by extinction of CTS+ learning for reward. The sensitivity of LHA neurons to ACh was reduced plastically during extinction of CTS- learning for aversion. 9. The data suggest that multiple afferent inputs from dopaminergic and cholinergic systems to the same LHA neuron are involved in CTS learning of positive and negative reinforcement, respectively. The sensitivity of neurons to DA and ACh was also plastically modulated during the course of CTS learning. NE might be concerned with the learning of CTS by LHA neurons.
