What can we know from pituitary-adrenal hormones about the nature and consequences of exposure to emotional stressors?

Abstract

Exposure to stress induces profound physiological and behavioral changes in the organisms and some of these changes may be important regarding stress-induced pathologies and animal models of psychiatric diseases. Consequences of stress are dependent on the duration of exposure to stressors (acute, chronic), but also of certain characteristics such as intensity, controllability, and predictability. If some biological variables were able to reflect these characteristics, they could be used to predict negative consequences of stress. Among the myriad of physiological changes caused by stress, only a restricted number of variables appears to reflect the intensity of the situation, mainly plasma levels of ACTH and adrenaline. Peripheral hypothalamic-pituitary-adrenal (HPA) hormones (ACTH and corticosterone) are also able to reflect fear conditioning. In contrast, the activation of the HPA axis is not consistently related to anxiety as evaluated by classical tests such as the elevated plus-maze. Similarly, there is no consistent evidence about the sensitivity of the HPA axis to psychological variables such as controllability and predictability, despite the fact that: (a) lack of control over aversive stimuli can induce behavioral alterations not seen in animals which exert control, and (b) animals showed clear preference for predictable versus unpredictable stressful situations. New studies are needed to re-evaluate the relationship between the HPA axis and psychological stress characteristics using ACTH instead of corticosterone and taking advantages of our current knowledge about the regulation of this important stress system.

Fig. 1

Schematic representation of two patterns of hormonal response to stressful situation differing in intensity. Animals were left undisturbed (controls, C) or exposed to different mild stressful situation with progressive increases in the level of environmental perturbations (from 1 to 4). Whereas plasma levels of corticosterone increases in function of the intensity of the stressful situation, plasma levels of luteinizing hormone (LH) responded to stress, but independently of the intensity of the situation. Original data can be seen in (Armario et al. 1986)

Fig. 2

Schematic representation of the response of several physiological variables to stress differing in intensity. Levels of intensity are progressively increased from the left to the right. Note that the curve representing each variable has a specific shape, indicating different levels of sensitivity to the intensity of stressors and a saturation of the response at different levels of intensity. Particularly noteworthy is the saturation of the corticosterone response with relatively low levels of stressor intensity

Fig. 3

ACTH response is not sensitive to context fear conditioning using the typical 5 min exposure to the context. Animals were exposed to a particular chamber without receiving electric footshocks (controls) or after 30 footshocks (1.5 mA, 3 s, 1 per min). When exposed 8 days later to the same chamber for 5 min, controls rats showed low levels of freezing, whereas previously shocked rats showed high levels of freezing (left panel). Despite this marked differences in behavior, ACTH levels were similar in both control and shocked rats (right panel). This lack of sensitivity of ACTH to fear conditioning is due to the short period of exposure to the context as ACTH reflects context fear conditioning when exposure to the context last for 15 min (Daviu et al. 2010). Means and SEM (n = 16) are represented. ***P < 0.001 vs. controls

Fig. 4

ACTH response did not reflect levels of anxiety in the elevated plus-maze. Time spent in the open arms of the elevated plus-maze (EPM) in a normal population of adult male rats classified in function of the median in low anxiety (LA) and high anxiety (HA). In the right panel can be seen that ACTH response to the EPM did not differ between LA and HA rats. Means and SEM (n = 9) are represented (unpublished data from C. Márquez, R. Nadal and A. Armario). **P < 0.01 vs. LA, NS non significant