Rapid Spatial Learning Controls Instinctive Defensive Behavior in Mice

Abstract

Instinctive defensive behaviors are essential for animal survival. Across the animal kingdom, there are sensory stimuli that innately represent threat and trigger stereotyped behaviors such as escape or freezing [1-4]. While innate behaviors are considered to be hard-wired stimulus-responses [5], they act within dynamic environments, and factors such as the properties of the threat [6-9] and its perceived intensity [1, 10, 11], access to food sources [12-14], and expectations from past experience [15, 16] have been shown to influence defensive behaviors, suggesting that their expression can be modulated. However, despite recent work [2, 4, 17-21], little is known about how flexible mouse innate defensive behaviors are and how quickly they can be modified by experience. To address this, we have investigated the dependence of escape behavior on learned knowledge about the spatial environment and how the behavior is updated when the environment changes acutely. Using behavioral assays with innately threatening visual and auditory stimuli, we show that the primary goal of escape in mice is to reach a previously memorized shelter location. Memory of the escape target can be formed in a single shelter visit lasting less than 20 s, and changes in the spatial environment lead to a rapid update of the defensive action, including changing the defensive strategy from escape to freezing. Our results show that although there are innate links between specific sensory features and defensive behavior, instinctive defensive actions are surprisingly flexible and can be rapidly updated by experience to adapt to changing spatial environments.

Keywords: defensive behavior; escape; freezing; innate behavior; mouse; shelter; spatial learning; spatial memory.

Figure 1

Escape Behavior Is a Goal-Directed Action to Reach Safety (A) Video frames from one trial showing escape to a previously explored shelter after stimulation with an expanding spot projected from above, between the mouse and the shelter location (on-path). Yellow lines indicate the mouse’s trajectory during the preceding 2 s. (B) Example trajectories from several mice, recorded between stimulus onset and the end of flight, showing that flight path and target are independent of stimulus position or quality (number of animals = 10 on-path, 16 on-top, and 15 sound). (C) Initial position of mice in all trials plotted in relation to the shelter location. (D) Accuracy of reaching the shelter during escape. Bars show average accuracy and circles are individual accuracy data points as function of distance to the shelter. (E) Total displacement during escape for 100% accurate flights plotted against linear distance to the shelter. (F) Video frames from one trial during initiation of escape from an expanding spot on-top, highlighting the initial head rotation preceding the initiation of running. The yellow line indicates head direction, and the dashed white line is the reference line between the current mouse position and the shelter. (G) Head angles measured between the white and yellow lines illustrated in (F) for 100% accurate flights, showing that the head is pointing toward the position of the shelter before the distance to the shelter is covered. Circles indicate the initial angles for different trials, lines indicate average head rotation profile, and shaded areas indicate the SD (n = 59 trials from 38 animals). (H) Raster plots showing speed profile of trials in several mice stimulated with sound when exploring the arena (left) or when the same mice were inside an over-ground shelter (right). (I) The probability of flight is dramatically reduced when animals are already inside a shelter. For all relevant panels, the blue circle with “S” identifies the shelter location and dark gray, light gray, and red indicate data from stimulation with spot on-path, spot on-top, and sound, respectively. See also Figure S1 and Movie S1.

Figure 2

Memory of Shelter Location Guides Defensive Flight (A) Video frames from one trial showing escape from aversive sound immediately after the outside of the arena had been rotated, together with local cues (panels on the outside, color-coded for clarity). The dashed yellow line marks the diameter of the fixed platform, and the dashed blue circle shows shelter location before rotation. (B) Trajectories from different mice after arena rotation, showing escape toward the previous shelter location (dashed blue circle). (C) Escape behavior is not significantly changed by arena rotation (accuracy = 102% ± 1%, linearity = 96% ± 2% of control). Reaction time is also not affected (93% ± 14%). p > 0.1 for all comparisons, paired t test between pre- and post-rotation; n = 8 animals. (D) Head rotation profile during escape initiation is not affected by arena rotation (p = 0.39, paired t test between pre- and post-rotation for distance at 10°). Post-rotation angles are measured between the mouse position and the shelter position before rotation. The shaded area indicates the SD. (E) Plot showing when the mouse leaves the initial target hole area after the flight. Red indicates flights after rotation, and blue indicates flights in control conditions where the shelter target was missed. The shaded area indicates the SEM. (F) Video frames from one trial showing sound-evoked flight to a shelter in the center of the arena and persistence of escape to the arena center after the shelter has been removed. (G) Escape trajectories for different mice before (left) and after (right) a shelter in the arena center was removed. (H) Speed profile for escape responses when the shelter is in the periphery (blue; from the same dataset shown in Figure 1) and after the shelter has been removed from the arena center. See also Figure S2 and Movie S2.

Figure 3

Shelter Location Memory Is Formed Rapidly and Supports Fast Updates of Defensive Actions (A) Raster plot showing periods of time inside the shelter from the onset of arena exploration and threat stimulus presentation. An example raster from a regular assay for comparison (as shown in Figure 1) with multiple entries in the shelter during the exploration phase is shown at the top. (B) Average (bars) and data points (circles) for accuracy and linearity of escape after shelter single visits. (C) Time to initiate escape is negatively correlated with the total amount of time spent in the shelter before stimulation. Gray circles are data from the minimum time assay, and blue circles are data from the regular assay. The black line is a regression line fit to all data points, and the shaded area is 95% confidence interval for the regression. (D) Escape trajectories after the original shelter has been closed (dashed blue circle) and a new one open in a different position (blue circle with “S”) for the first and last trials (left and right red, respectively) and the median trial (center red). Trajectories in blue (right) are for secondary flights, which immediately follow escapes to the original location. (E) Evolution of escape behavior after shelter location has been moved, as in (D), showing the fraction of flights across all mice that reach the new shelter location, for the first, three quartiles (Q1–Q3), and last trials. (F) Video frames from one mouse in an arena with the shelter closed, showing freezing behavior in response to a slowly expanding spot projected on top. (G) Raster plots showing speed profiles upon threat stimulation before (bottom) and after the shelter hole has been opened (top) for slowly expanding spots. Trials have been aligned by reaction time (dashed line). See also Figure S3 and Movie S3.