Studies in monkeys have demonstrated that the amplitude of a touch-evoked signal generated in the primary somatosensory neocortex predicts behavioral responses in monkeys trained to report the perceived intensity of the touch stimulus (Kulics, et al., 1977; Kulics and Cauller, 1988). This behaviorally relevant signal is the first major surface negative component of the sensory-evoked potential, N1, which occurs approximately 50 ms after the touch stimulus (in humans N1 occurs ~ 100ms poststimulus). N1 follows the primary, P1, component of the sensory-evoked potential (13ms latency in monkeys, 25ms in humans) which is generated by the initial thalamocortical projections to middle layers of the primary sensory areas. The amplitudes of both P1 and N1 are a function of stimulus intensity, but P1 persists during anesthesia and sleep, while N1 is only present during conscious states (Cauller and Kulics, 1988). While the amplitude of P1 is strictly a function of stimulus intensity, the amplitude of N1 predicts behavioral discrimination errors and response latency. For instance, N1 is consistently larger when a monkey incorrectly reports a strong sensation in response to a weak stimulus even when P1 is small indicating the weak input was accurately relayed to cortex. N1 is the shortest latency and most reliable neural correlate of subjective sensation.

In contrast to P1, which is generated by inward currents activated in middle layers by thalamocortical inputs, N1 is generated by inward currents activated by excitatory inputs very near the cortical surface (Cauller and Kulics, 1991). By all accounts, the backward cortico-cortical inputs from higher cortical areas (2, 5, SII and MI) are the most likely source of the superficial inputs responsible for N1 activation. These findings promote a reentrant model of conscious sensory processing with the forward path mediating rapid, high-fidelity distribution of sensory signals constantly interacting with convergent backward inputs reflecting the response of all higher areas engaged by the sensory process (Cauller, 1995).

Tests of the hypothesis that the behaviorally relevant N1 signal is activated by backward cortical inputs require extensive surgical manipulations and anatomical analyses that are not practical in monkeys. Therefore subsequent studies have aimed to identify the homologous N1 or other behaviorally relevant neural correlates in rats. Follow-up studies are underway at UTD to identify the behaviorally relevant components of the sensory-evoked responses in the neocortex of rats. Preliminary studies using chronically implanted transcortical electrodes in somatosensory and auditory areas have found a P1/N1 evoked potential sequence resembling that found in monkeys and humans which appears to be homologous to the behaviorally relevant signal because N1 is abolished by anesthesia while P1 persists (Jackson and Cauller, 1994). We plan to correlate this N1 potential in rats with behavioral performance during stimulus discrimination tasks. Direct stimulation of the secondary cortical areas that project to layer I of the primary areas (see Neuroanatomy Project, Cauller et al., 1998) evokes an excitatory surface negative potential which, like N1, is generated in layer I. To test the hypothesis that the backward cortical projections to layer I are responsible for the behaviorally relevant N1 in rats, we apply blockers in layer I or inhibit the secondary areas during behavioral tasks using chronically implanted microdialysis probes (Jackson and Cauller, 1998). This project has been supported by the Whitehall Foundation.