The majority of the data presented here were recorded while passively fixating animals experienced a range of different heading

directions that spanned the horizontal and/or vertical plane (Gu et al., 2006 and Takahashi et al., 2007). Specifically, headings relative to straight ahead were 0, ±22.5, ±45, ±90, ±135°, ±180°. Different heading directions and stimulus types (visual or vestibular) were interleaved randomly within a single block of trials. Each distinct stimulus was typically repeated five times (minimum of three repetitions for inclusion). In each trial, a fixation point first appeared at the center of the screen. After fixation was established for 100–200 ms, the motion stimulus began and lasted for 2 s. In the vestibular condition, the motion platform always began its movement from a common central position. Metformin mw The animal was rewarded if they maintained visual fixation throughout Selleckchem Sunitinib the duration of the stimulus. At the end of the trial (or when fixation was broken), the fixation point disappeared and the motion platform moved

back to the original central position during a 2 s intertrial interval. In the visual condition, the random-dot field appeared on the display after fixation was established, and again moved for 2 s. The dots then disappeared and the animal was rewarded for maintaining fixation, followed again by a 2 s intertribal interval. Three animals were trained many only to perform the passive fixation task, whereas five animals had been extensively trained to perform a heading discrimination task (Fetsch et al., 2009, Gu et al., 2007 and Gu et al., 2008a), in which they were asked to report whether their perceived heading was leftward or rightward relative to straight ahead by making a saccade to one of two choice targets. For a subpopulation

of neurons in these trained animals, responses were obtained while the animals performed both the fixation task and the heading discrimination task. We conducted extracellular recordings of action potentials from single neurons in area MSTd. For most recordings, 2 to 4 tungsten electrodes (Frederick Haer, Bowdoinham, ME; tip diameter 3 μm, impedance 1–2 MΩ at 1 kHz) were used to record multiple single neurons simultaneously. In some cases (57 pairs), two to four electrodes were placed inside multiple guide tubes separated by 0.8–25 mm (different hemispheres). In other cases (55 pairs), multiple electrodes were placed inside a single guide tube. The distance between two simultaneously recorded neurons was estimated from both the horizontal and vertical (depth) coordinates (shank diameter = 75 μm). Data from another 67 cell pairs were obtained from previous recordings with a single electrode (Fetsch et al., 2007, Gu et al., 2006, Gu et al., 2007 and Takahashi et al.