The role of the medial caudate nucleus, but not the hippocampus, in a matching-to sample task for a motor response.

A delayed-match-to-sample task was used to assess memory for motor responses in rats with control, hippocampus, or medial caudate nucleus (MCN) lesions. All testing was conducted on a cheeseboard maze in complete darkness using an infrared camera. A start box was positioned in the centre of the maze facing a randomly determined direction on each trial. On the sample phase, a phosphorescent object was randomly positioned to cover a baited food well in one of five equally spaced positions around the circumference of the maze forming a 180-degree arc 60 cm from the box. The rat had to displace the object to receive food and return to the start box. The box was then rotated to face a different direction. An identical baited phosphorescent object was placed in the same position relative to the start box. A second identical object was positioned to cover a different unbaited well. On the choice phase, the rat must remember the motor response made on the sample phase and make the same motor response on the choice phase to receive a reward. Hippocampus lesioned and control rats improved as a function of increased angle separation used to separate the correct object from the foil (45, 90, 135, and 180 degrees) and matched the performance of controls. However, rats with MCN lesions were impaired across all separations. Results suggest that the MCN, but not the hippocampus, supports working memory and/or a process aimed at reducing interference for motor response selection based on vector angle information.