While the neural mechanisms involved in simple forms of associative learning have been extensively studied, most research has focused upon paradigms in which two stimuli are repetitively paired. In contrast, little is known about the mechanisms underlying conditioned inhibition (CI), where an organism learns that one stimulus signals the absence of a second event. Hermissenda , a sea snail, is an excellent system to study CI, since much information is available as to the cellular and molecular bases of paired learning, and CI can be readily produced using the same natural stimuli. Previous research has shown that CI in Hermissenda, established by repeated trials in which presentations of light and rotation are separated by a fixed, lengthy temporal gap, results in decreased responses to light by ocular Type B photoreceptors. My investigations have determined the ionic bases of the changes in Type B cells and have also identified a second site of CI-related neural plasticity.

Recordings from Type A photoreceptors revealed that CI increased their light response, both for A cells in the intact nervous system, as well as for those that had been isolated from other neurons. This indicates the increase in light response is intrinsic to the Type A photoreceptors. Voltage clamp of the Type B photoreceptors, along with pharmacological blockade of various ionic currents, showed that two potassium currents were increased following CI. Blockade of these potassium currents separately resulted in the reduction or elimination of the learning-induced difference in two measures of light response in Type B photoreceptors. The data on currents from these and past voltage clamp studies were incorporated into a computer model, which simulates the isolated Type B photoreceptor. This model closely mimicked the physiological light response and was able to account for both paired learning and CI when the potassium current amplitudes were altered to match the physiological changes seen following the two types of learning. Light responses from simulations in which one of these potassium currents was blocked were also similar to the physiological responses for naïve, paired training, and CI conditions.