Isolation of the differential effects of chronic and acute stress in a manner that is not confounded by stress severity

Abstract

Firm conclusions regarding the differential effects of the maladaptive consequences of acute versus chronic stress on the etiology and symptomatology of stress disorders await a model that isolates chronicity as a variable for studying the differential effects of acute versus chronic stress. This is because most previous studies have confounded chronicity with the total amount of stress. Here, we have modified the stress-enhanced fear learning (SEFL) protocol, which models some aspects of posttraumatic stress disorder (PTSD) following an acute stressor, to create a chronic variant that does not have this confound. Comparing results from this new protocol to the acute protocol, we found that chronic stress further potentiates enhanced fear-learning beyond the nonassociative enhancement induced by acute stress. This additional component is not observed when the unconditional stimulus (US) used during subsequent fear learning is distinct from the US used as the stressor, and is enhanced when glucose is administered following stressor exposure, suggesting that it is associative in nature. Furthermore, extinction of stressor-context fear blocks this additional associative component of SEFL as well as reinstatement of generalized fear, suggesting reinstatement of generalized fear may underlie this additional SEFL component.

Keywords: Acute stress; Chronic stress; Generalization; Posttraumatic stress disorder; Reinstatement; Stress-enhanced fear learning.

Fig. 1

The acute versus chronic SEFL protocol, in contrast to other common acute versus chronic stress (e.g., immobilization stress) protocols, isolates chronicity as a variable. Depicted: Chronic immobilization stress (IS) differs from acute IS not only in terms of chronicity of the stressor, but also in terms of total stress volume. Conversely, the chronic SEFL resembles acute SEFL in terms of the total amount of time exposed to the stressor context, the pattern of variable inter-stressor context exposure times, as well as total stress volume during stressor presentations, and only differs from acute SEFL in the chronicity of the stressor presentations.

Fig. 2

Impacts of stress chronicity and sex on subsequent fear learning. Depicted: Schematic illustration of the timeline of the experiment and the figure panels in which data from various components of the experiment are displayed (panel A). Percent freezing during stress pretreatment (panels B & C, days 1–15), generalization testing (panel D, day 16), context preexposure (panel E, days 17–19), single-shock baseline (panel F, day 20), and context test (panel G, day 21). Male (M) and female (F) rats received either acute stress exposure (15 footshocks) on day 1 or 15 (AS1 or AS15), chronic stress (CS) exposure (15 footshocks), or identical context exposure with no shock (No Stress; AN & CN). Following stress pretreatment, all rats were exposed to a novel context that shared some similar dimensions to the stress pretreatment context. All groups were then preexposed to a completely novel environment for 8 min/day for 3 consecutive days. Twenty-four hours after the termination of preexposure, all groups received a single footshock in the preexposed context. Then, all groups were tested for contextual fear learning 24 h later. All groups that received footshock during stress pretreatment readily reached asymptotic contextual fear conditioning (panels B & C). Groups that received chronic stress or that received acute stress on day 1 of the chronic stress procedure showed greater generalized fear compared to all other groups in the similar (panel D) or totally novel contexts (panel E). Increased baseline levels of generalized fear prior to the SEFL single-shock exposure were non-significant (panel F). AS and CS groups showed greater freezing behavior during the context test compared to unshocked controls, and CS showed higher freezing as compared to AS (panel G). Error bars denote mean ± SEM. *, ***, **** denotes significance (p ≤ .05, p ≤ .001, and p ≤ .0001, respectively) compared between indicated groups (horizontal square brackets), compared between each Stress group and its respective No Stress control (horizontal line), or compared between a specific Stress group and its respective No Stress control (free-standing).

Fig. 3

Effects of stressor context fear extinction on subsequent contextual and auditory fear learning. Depicted: Schematic illustration of the experiment timeline and the figure panels in which data from the experiment are displayed (panel A). Percent freezing during stress pretreatment (panels B & C, days 1–15), stressor context fear extinction (panel D, days 16–27), context preexposures (panels E & H, days 28–30 and days 33–39, respectively), single-shock baseline (panel F, day 31), context test (panel G, day 32), and tone test (panel I, day 41). Rats received either acute (AS) or chronic (CS) exposure to 15 footshocks, or identical context exposure with no shock (No Stress; AN/CN). Following stress pretreatment, rats received (E) or did not receive (U) extinction training to the stressor context. Following extinction training, all rats were preexposed to a novel context (30 min/day) for 3 consecutive days. Twenty-four hours after preexposure, all groups received a single footshock in the preexposed context. All groups were tested for contextual fear learning 24 h later, then preexposed to a third context (30 min/day) for 7 days, and, finally, tested for fear conditioning to the tone in this context. Rats that received chronic stress exhibited a slower rate of extinction (panel D). CSU rats exhibited higher levels of fear expression to the novel context (panel E). All groups displayed comparably low baseline levels of fear prior to the single-shock exposure (panel F). Stressed rats exhibited enhanced fear learning to the single-shock session. However, extinction training attenuated learning only in the chronic-stress group (panel G). Furthermore, only AS groups exhibited enhanced fear learning to the tone (panel I). Error bars denote mean ± SEM. *, **, and **** denotes significance (p ≤ .05, p ≤ .01, and p ≤ .0001, respectively) compared between indicated groups (horizontal square brackets), compared between each Stress group and its respective No Stress control (horizontal line), or compared between a specific Stress group and its respective No Stress control (free-standing). # denotes significance (p ≤ .05) compared between CSE and ASE (horizontal line).

Fig. 4

Impact of stress chronicity on subsequent fear learning reinforced by a novel aversive stimulus. Depicted: Schematic illustration of the timeline of the experiment and the figure panels in which data from various components of the experiment are displayed (panel A). Percent freezing during stress pretreatment (panels B & C, days 1–15), context preexposure (panel D, days 16–21), context test (panel E, day 23), and open field test (panel F, days 24–25). Rats received either acute (AS) or chronic (CS) exposure to 15 footshocks, or identical context exposure with no shock (No Stress; AN/CN). All groups were preexposed (30 min/day) to a novel environment for 6 consecutive days. Twenty-four hours after the termination of preexposure, all groups received a single startle stimulus in the preexposed context. All groups were tested for contextual fear learning 24 h later. Stressed rats exhibited enhanced contextual fear learning to the startle noise, but there was no difference between AS and CS (panel E). Additionally, stressed rats exhibited decreased exploratory behavior in the open field test as indicated by a decrease in distance travelled, but, again, there were no differences between AS and CS (panel F). Error bars denote mean ± SEM. *, ** denotes significance (p ≤ .05, p ≤ .01) compared between indicated groups (horizontal square brackets), compared between each Stress group and its respective No Stress control (horizontal line), or compared between a specific Stress group and its respective No Stress control (free-standing).

Fig. 5

Post-stress glucose selectively enhances subsequent fear learning in chronically-stressed rats. Depicted: Schematic illustration of the experiment timeline and the figure panels in which data from the experiment are displayed (panel A). Percent freezing during stress pretreatment (panels B & C, days 1–15), context preexposure (panel D, days 16–21), and context test (panel E, day 23). Weights prior to, during, and following stress pretreatment are also reported (panel F). Rats received either acute (AS) or chronic (CS) exposure to 15 footshocks, or identical context exposure with no shock (No Stress; CN). Rats received daily access to a 40% glucose solution in drinking water (G) or drinking water alone (W) for 6 h immediately following the termination of each session of stress pretreatment. All groups were preexposed (30 min/day) to a novel environment for 6 consecutive days. Twenty-four hours after the termination of preexposure, all groups received a single footshock in the preexposed context. All groups were tested for contextual fear learning 24 h later. All groups that received footshock during stress pretreatment readily reached asymptotic contextual fear conditioning (panels B & C). Groups that received chronic or acute stress showed greater generalized fear compared to the unstressed group (panel D). However, chronically-stressed rats that received post-stress glucose (CSG) exhibited markedly higher levels of generalized fear when compared to those that received water (CSW; panel D). An identical trend was observed during the SELF context test (panel E). CS groups showed weight loss during stress pretreatment that was not blocked by glucose while AS groups showed weight loss following stress pretreatment that was blocked by glucose (panel F). Error bars denote mean ± SEM. *, **** denotes significance (p ≤ .05, p ≤ .0001) compared between indicated groups (horizontal square brackets), compared between each Stress group and CNW (horizontal line), or compared between a specific Stress group and CNW (free-standing). # denotes significance (p ≤ .05) between CSG and CSW (horizontal line).