1. Recent conceptualizations of the neural systems used during navigation have classified two types of sensory information used by animals: landmark cues and internally based (idiothetic; e.g., vestibular, kinesthetic) sensory cues. Previous studies have identified neurons in the postsubiculum and the anterior thalamic nuclei that discharge as a function of the animal's head direction in the horizontal plane. The present study was designed to determine how animals use head direction (HD) cells for spatial orientation and the types of sensory cues involved. 2. HD cell activity was monitored in the postsubiculum and anterior thalamic nucleus of rats in a dual-chamber apparatus in an experiment that consisted of two phases. In the first phase, HD cell activity was monitored as an animal moved from a familiar environment to a novel environment. It was hypothesized that if HD cells were capable of using idiothetic sensory information, then the direction of maximal discharge should remain relatively unchanged as the animal moved into an environment where it was unfamiliar with the landmark cues. In the second phase, HD cells were monitored under conditions in which a conflict situation was introduced between the established landmark cues and the animal's internally generated sensory cues. 3. HD cells were initially recorded in a cylinder containing a single orientation cue (familiar environment). A door was then opened, and the rat entered a U-shaped passageway leading to a rectangular chamber containing a different prominent cue (novel environment). For most HD cells, the preferred direction remained relatively constant between the cylinder and passageway/rectangle, although many cells showed a small (6-30 degrees) shift in their preferred direction in the novel environment. This directional shift was maintained across different episodes in the passageway/rectangle. 4. Before the next session, the orientation cue in the cylinder was rotated 90 degrees, and the animal returned to the cylinder. The cell's preferred direction usually shifted between 45 and 90 degrees in the same direction. 5. The rat was then permitted to walk back through the passageway into the now-familiar rectangle. Immediately upon entering the passageway, the preferred direction returned to its original (prerotation) orientation and remained at this value while the rat was in the rectangle. When the rat was allowed to walk back into the cylinder, one of three outcomes occurred: 1) the cell's preferred direction shifted, such that it remained linked to the cylinder's rotated cue card; 2) the cell's preferred direction remained unchanged from its orientation in the rectangle; or 3) the cell's preferred direction shifted to a new value that lay between the preferred directions for the rotated cylinder condition and rectangle. 6. There was little change in the HD cell's background firing rate, peak firing rate, or directional firing range for both the novel and cue-conflict situations. 7. Simultaneous recordings from multiple cells in different sessions showed that the preferred directions remained "in register" with one another. Thus, when one HD cell shifted its preferred direction a specific amount, the other HD cell also shifted its preferred direction the same amount. 8. Results across different series within the same animal showed that the amount the preferred direction shifted in the first Novel series was about the same amount as the shifts observed in subsequent Novel series. In contrast, as the animal experienced more Conflict series, HD cells tended to use the cylinder's cue card less as an orientation cue when the animal returned to the rotated cylinder condition from the rectangle. 9. These results suggest that HD cells in the postsubiculum and anterior thalamic nuclei receive information from both landmark and idiothetic sensory cues, and when both types of cues are available, HD cells preferentially use the landmark cues as long as they are perceived
Head direction cells recorded in the anterior thalamic nuclei of freely moving rats
