Figure 2.

Path Integration Uses Motor-Related Movement Signals

(A and E) Schematic of track designs used to test a decrease (A) or an increase (E) in the gain between motor and visual reference frames. For standard trials, for every 60 cm mice run on the treadmill, the visual track moves 60 virtual units (VU). On reduced gain trials, for every 60 cm mice run, the visual track moves 30 VU. For increased gain trials the visual track moves 120 VU for every 60 cm mice run.

(B and F) Example plots of stop locations from single mice for trials in which the gain between treadmill movement and visual update of the track is reduced by 0.5 (B) or increased by ×2 (F). The trial number refers to all trials, but for clarity only data from gain change trials are shown.

(C and G) Average of Z scored stop locations across all mice for control probe trials (×1) and trials on which the gain is reduced (C) or increased (G). Averaged data are plotted as ± SEM (N = 5 mice for ×0.5 gain, N = 4 mice for ×2 gain).

(D and H) To quantify the effects of the gain change we compared, for each trial type, the ratio of stops in the location of the reward zone in the visual reference frame (orange) to the sum of the number of stops in the reward zone in the visual and motor reference frames (green). The ratio is modified by reducing (t(4) = 3.7, p = 0.021, paired t test) (D) or increasing gain (t(3) = 6.5, p = 0.0073) (H). Error bars indicate SEM.

Thus, on trials with reduced gain (B–D), or increased gain (F–H), stops occur in anticipation of the reward zone location in the motor reference frame.