Endogenous event-related potentials in monkey: the role of task relevance, stimulus probability, and behavioral response

Abstract

Monkeys were trained in auditory discrimination tasks resembling human paradigms in which long-latency endogenous components, such as P300, are typically recorded. Morphological, topographical, and functional properties of the monkey event-related potentials (ERPs) were analyzed to determine similarities and differences with human ERPs reported in the literature. ERPs were recorded from epidural electrodes in monkeys trained to produce operant responses. In a conditional discrimination (CD) task, tone pips (2 kHz or 6 kHz, 40 msec duration, and 60 dB above nHL) were presented every 4–8 sec. Target tones presented during ‘time-in’ (TI) were rewarded when followed by a response in the correct post-stimulus interval (400–3000 msec). In contrast, tones presented during ‘time-out’ (TO) were not rewarded. Under both conditions, tones elicited an initial frontally dominant triphasic complex (P56-N92-P157). Additionally, TI target tones followed by a response elicited a large negativity (N358) having maximal amplitude over mid-frontal regions and followed by a parietally distributed positivity (P658). The scalp distribution and covariation with task requirements of N358 resemble those reported for the human ‘O’ wave. ERPs were also recorded in an auditory oddball paradigm in which tone pips (2 kHz and 6 kHz, 40 msec duration, and 60 dB above nHL) were presented in random order every second. Monkeys trained in the CD paradigm, along with additional subjects, were trained to make delayed responses following target tones embedded in a background of different-pitch tones. Tone probabilities were varied in different sessions from 90-10, 70-30, to 50-50 to assess the effects of probability. Background and target tones elicited a triphasic complex (P52-N110-P159) similar in latency and distribution to that recorded in the CD task. Additionally, target tones in this paradigm elicited a long-latency positive component (LPC) that exhibited an inverse relationship with stimulus probability. LPC had an onset latency of approximately 150–200 msec, a duration of approximately 300 msec, and multiple peaks (P244 and P376). These data indicate the importance of stimulus context in eliciting long-latency endogenous activity. It further suggests that strong analogies exist between monkey and human potentials recorded under similar paradigms. The effects of task relevance, stimulus probability, and the act of producing behavioral responses are similar to the effects of these variables on analogous human potentials. Therefore, this monkey model would allow for future invasive studies of the neural substrates underlying complex cognitive functions, given the relationship between human ERP components, such as P300, and specific cognitive processes.