Steady potential correlates of positive reinforcement and sleep onset in the cat; ‘reward contingent positive variation’ (RCPV)

Abstract

Summary Nine cats were implanted with epidural non-polarizable Ag-AgCl electrodes over the parieto-occipital and frontal cortex. In addition, similar electrodes were implanted in subjacent white matter of the posterior marginal, medial suprasylvian, and anterior ectosylvian gyri for transcortical recording. Two bone marrow reference electrodes were also used: one in the posterior aspect of the bone overlaying the frontal sinus, and another in the occipital crest. In ‘acute’ anesthetized preparations it was found that the transcortical recording was free from contamination caused by eye-motion potentials. The chronically implanted subjects, kept on 23 h food and water deprivation schedule, were trained to press a lever for 0.5–1 ml of milk reward. Potentials caused by lapping or licking were recorded from an electrode implanted at the bottom of the frontal sinus or in the temporal bone. Eye movements were monitored from an electrode placed in the orbital bone immediately behind the eyeball. After 2–3 weeks of training, all subjects displayed bursts of high-voltage slow-wave activity of 5–9.5 c/sec over the primary and secondary visual projections during consumption of milk reward i.e. , characteristic post-reinforcement synchronization (PRS). The PRS activity was always associated with 200–700 μV positive steady potential (SP) shift over the visual cortex in either transcortical or surface bipolar recording with reference to the anterior ectosylvian gyrus or frontal sinus. This SP shift was contingent upon the presence of light in the test chamber, although the subjects were trained and habituated to perform in the dark; the SP shift was also contingent upon the relative quality and taste of food reward. Therefore, it was termed ‘reward contingent positive variation’ (RCPV). Eye movements, which in several instances could be correlated with relatively small (30–60 μV) positive lambda waves over the visual cortex, did not significantly contribute to the RCPV. In satiated subjects during sleep onset or grooming activity a similar phasic positive SP shift occurred over the same cortical area; it was always associated with or shortly followed by a burst of high-voltage slow-wave activity of 6–8 c/sec; the latter SP shift was termed ‘sleep onset positive variation’ (SOPV). The occurrence of the SOPV was facilitated by the presence of ambient light, but, in distinction to the RCPV responses, it was not absolutely contingent on it. Two relatively small amplitude but rather consistent mirror image reversals of RCPV and SOPV responses were observed: one in the subjacent white matter 3 mm beneath the striate and parastriate cortex, and another at the surface in the immediate periphery of the visual projections, i.e. , over the anterior portions of the ectosylvian, suprasylvian and marginal gyri. The phasic occurrence of SOPV immediately prior to the development of delta slow-wave sleep patterns resulted usually in a cumulative and more stable positive SP shift over the striate cortex. During the peak of a cumulative SOPV a potential difference of approximately 600–700 μV was observed, when measured epidurally across the parieto-occipital cortex between the anterior ectosylvian and posterior marginal gyri. This potential became more diffuse and partially dissipated with the development of slow-wave delta sleep patterns. The PRS, RCPV and SOPV phenomena have been interpreted as processes subserving internal inhibition.