کنترل واکنش پذیری استرس هورمونی توسط سیستم اوپیوئیدی درون زا

Summary Regulations of hormonal stress responses entail the initiation, amplitude and termination of the reaction, as well as its integration with other stress response systems. This study investigates the role of endogenous opioids in the regulation and integration of behavioral, thermal and hormonal stress responses, as these neuromodulators and their receptors are expressed in limbic structures responsible for stress responses. For this purpose, we subjected mice with selective deletion of β-endorphin, enkephalin or dynorphin to the zero-maze test, a mildly stressful situation, and registered behaviors and stress hormone levels. Behavioral stress reactivity was assessed using zero-maze, light–dark and startle-reactivity paradigms. Animals lacking enkephalin displayed increased anxiety-related behavioral responses in each three, dynorphin knockouts in two models, whereas the responses of β-endorphin knockouts indicated lower anxiety level in the zero-maze test. All knockout strains showed marked changes in hormonal stress reactivity. Increase in ACTH level after zero-maze test situation, unlike in wild type animals, failed to reach the level of significance in Penk1−/− and Pdyn−/− mice. Corticosterone plasma levels rapidly increased in all strains, with a lower peak response in knockouts. In wild-type and β-endorphin-deficient mice, corticosterone levels returned to baseline within 60 min after stress exposure. In contrast, mice lacking dynorphin and enkephalin showed longer-lasting elevated corticosterone levels, indicating a delayed termination of the stress reaction. Importantly, the behavioral and hormonal responses correlated in wild-type but not in knockout mice. Hyperthermia elicited by stress was reduced in animals lacking dynorphin and absent in Penk1−/− mice, despite of the heightened behavioral anxiety level of these strains. These results demonstrate an important role on the endogenous opioid system in the integration of behavioral and hormonal stress responses.

Introduction An ancient part of the mammalian brain, the limbic system, organizes and initiates stress responses. It is not an anatomical entity, but rather consists of separate brain areas forming a functional unit. It receives somatosensory information from the thalamus and sensory cortex, efferent signals from the vegetative system, and it has access to the stored information from the hippocampus (Sah et al., 2003). The amygdaloid complex in strong cooperation with other elements of the limbic system integrates and evaluates incoming information, organizes and initiates stress responses (Akmaev et al., 2004). Cortical afferents from the amygdala contribute to behavioral stress responses, while multiple descending pathways influence vegetative responses and reflexes. The limbic system continuously evaluates a broad range of sensory stimuli, and triggers a stress response if the stimulus is perceived as being dangerous (Sah et al., 2003). The hypothalamic paraventricular nucleus (PVN), itself part of the limbic system, regulates the hormonal stress responses under the control of other limbic elements. A dense capillary plexus provides a rapid access of steroid hormones to the PVN and enables the negative-feed back regulation of the corticotropin-releasing hormone (CRH) secreting parvocellular neurons through activation of neuronal nuclear steroid receptors. Local synaptic circuits (Boudaba et al., 1996) and inputs from the zone immediately surrounding the PVN provide the majority of inhibitory inputs (Herman et al., 2002). The PVN receives both excitatory and inhibitory inputs from limbic and extra-limbic structures, either directly or indirectly through interneurons (Herman et al., 2003). The complex control of the activity of the PVN is necessary for the appropriate initiation, amplitude and termination of the hormonal stress response as well as for the integration of the hormonal stress reactivity with other executive stress reactivity pathways. Neuromodulators are thought to play an important role in the regulation of hormonal stress responses. Modulatory peptidergic interneurons expressing β-endorphin, enkephalin and dynorphin are present in the PVN and in limbic areas that modulate PVN activity ( Drolet et al., 2001; Herman et al., 2002). Endogenous opioids can thus modulate PVN activity directly and indirectly. The expression of endogenous opioids in the PVN is increased after stress exposure unrelated to the nature of the stressor ( Palkovits, 2000; Reyes et al., 2003) suggesting that they participate in the regulation of stress reactivity. Pharmacological and genetic studies supported the potential role of endogenous opioids in the regulation of stress reactivity. Treatment with delta receptor antagonist ( Saitoh et al., 2005), or genetic deletion of delta opioid receptors or its ligand enkephalin ( Filliol et al., 2000; Bilkei-Gorzo et al., 2004) led to an increased emotionality, while the disruption of dynorphin/kappa opioid receptor signaling resulted in blunted stress responses ( McLaughlin et al., 2003). Animal studies suggested a modulatory role of β-endorphin-μ-opioid receptor system in endocrine responses to stress ( Vaanholt et al., 2003; Contet et al., 2006). Clinical studies also revealed an association between polymorphism in the μ-opioid receptor and hormonal stress response ( Chong et al., 2006). Expression of the c-Fos proto-oncogene is a marker of neuronal activation in response to various stimuli (Kovacs, 1998) and has been widely used to correlate behavioral phenotypes with neuronal activity (Wersinger et al., 2002; Matys et al., 2004; Bruening et al., 2006). Neuronal circuitries involved in stress reactivity showed a marked increase in c-Fos expression in different stress models. These methods have therefore been used to map out brain regions responding to stress (Duncan et al., 1996). The intensity of stress-induced c-Fos expression can be modulated by anxiolytic and anxiogenic drugs and is thus thought to reflect the animals’ stress reactivity (de Medeiros et al., 2005; Bilkei-Gorzo et al., 2007). In this study, we investigated the contributions of individual opioid peptides to the regulation of hormonal stress response with regard to the initiation, amplitude, termination of the hormonal reactivity and its integration with the behavioral responses. For this purpose, we challenged control animals and mice with a genetic deletion of the enkephalin- (Konig et al., 1996), dynorphin- (Zimmer et al., 2001), and β-endorphin- ( Rubinstein et al., 1996) encoding genes (Penk1, Pdyn and POMC) in the zero-maze test paradigm, followed by measurements of stress-hormone levels and c-Fos expression in the PVN and in the basolateral amygdala.

. Results 3.1. Open-field activity Using one-way ANOVA we found a significant difference in the horizontal activity between the groups (F3, 37=6.693; p=0.001). Post hoc analysis of the data revealed that the distance traveled was slightly, but significantly, reduced in Penk1−/− animals compared with wild-type mice (−23.3%). Genetic deletion of Pdyn or β-endorphin did not influence the animals’ motor activity. The vertical activity (number of rears) was not different between the strains (F3, 40=1.341; p>0.05) ( Figure 1). Locomoter activity. The activity of wild-type and Penk1−/−, Pdyn−/−, ... Figure 1. Locomoter activity. The activity of wild-type and Penk1−/−, Pdyn−/−, β-endorphin−/− mice was evaluated in the open-field arena in dimly lit environment. The horizontal activity of Penk1−/−, but not other knockout mice, was reduced. There was no difference in the vertical activity between the strains. *p<0.05 (one-way ANOVA followed by Dunnett test). Each column represents the mean (±SEM, n=10). Figure options 3.2. Elevated zero maze The behavior of the strains in the zero-maze paradigm differed in each parameter evaluated (latency: F3, 56=3.51; p<0.05; open time: F3, 56=4.62; p<0.01; open distance: F3, 56=13.34; p<0.001). Penk1−/− and Pdyn−/− mice were less active (−22.1% and −23.7%) than wild-type control animals, indicating increased levels of anxiety in these strains ( Figure 2). The latency to step into the closed area was 85% higher compared with wild-type mice in both strains (see also: Konig et al., 1996; Bilkei-Gorzo et al., 2004). Mice from the β-endorphin−/− strain on the other hand spent significantly more time and were more active in the open parts of the zero maze ( Figure 2) without showing any alteration in the latency time. Behavioral reactivity of wild-type and Penk1−/−, Pdyn−/−, β-endorphin−/− mice in ... Figure 2. Behavioral reactivity of wild-type and Penk1−/−, Pdyn−/−, β-endorphin−/− mice in the zero-maze test. Latency of the first entry into the closed area, time spent and activity in the open part was measured for 5 min. Penk1−/− and Pdyn−/− animals showed higher anxiety levels in this model as shown by the increased latency of first area change and reduced activity in the open part. Time spent and activity in the open area in β-endorphin−/− mice was elevated suggesting a reduced anxiety level of this strain. *p<0.05 knockout vs. wild-type (one-way ANOVA followed by Dunnett test). Each column represents the mean (±SEM, n=10). Figure options 3.3. Light–dark test There was a significant genotype effect for open time (F3, 41=6.09; p<0.01; Figure 3A) and open distance (F3, 41=3.74; p<0.05; Figure 3B). Post hoc analysis revealed that only Penk1−/− animals differed significantly from wild-type animals in the open time or open activity, while Pdyn−/− or β-endorphin−/− mice showed no anxiety-related phenotype in this model. Light–dark test time spent (A) and distance traveled (B) in the open area. ... Figure 3. Light–dark test time spent (A) and distance traveled (B) in the open area. Penk1−/− mice showed an increased anxiety in this model, as indicated by a lower time spent and activity in the lit area. (C) Startle-response test. The amplitude of the startle reaction was significantly higher in Penk1−/− and Pdyn−/− animals, suggesting a higher trait anxiety in these strains. *p<0.05; **p<0.01; ***p<0.001, wild-type vs. knockout (one-way ANOVA followed by Dunnett test; n=10–12). Figure options 3.4. Startle-response test The startle reactivity of the strains differed significantly (F3, 53=48.93; p<0.001). The amplitude of the reaction was higher in Pdyn−/− (+107.8%) and in the Penk1−/− animals (+79.5%) compared with wild-type mice. On the other hand, β-endorphin−/− animals did not show any alterations in the startle reactivity ( Figure 3C). 3.5. Stress-induced hyperthermia Stress elicited different changes in the body temperature, as shown by a significant interaction between time after stress and genotype (F9, 108=3.50; p<0.001). The basal temperatures of Penk1−/− and Pdyn−/− mice were higher than that of wild-type animals (37.98±0.07 and 37.97±0.07 vs. 37.36±0.22, respectively (F3, 36=4.37; p<0.05). Evaluating stress-induced changes in the individual strains, we found that the body temperature in wild-type mice was significantly higher (+0.64 °C) at the second time point (10 min after the first measurement) and remained relatively stable during the following observation period (+0.55 °C in the 20th and +0.74 °C in the 30th minute) ( Figure 4). Pdyn−/− mice showed a rather small hypothermic response to stress. Increase in temperature reached the level of significance at the 20th minute (+0.37 °C), but was not significantly different from the control (+0.22 °C) value at the next measurement ( Figure 4). The amplitude and duration of the hyperthermic response to stress in β-endorphin−/− mice was similar as observed in wild-type animals ( Figure 4). Interestingly, we found a total lack of stress-induced hyperthermia in Penk1−/− mice ( Figure 4). Body temperature was measured rectally four times with 10min interval. The ... Figure 4. Body temperature was measured rectally four times with 10 min interval. The procedure itself elicited stress and induced hyperthermia in wild-type animals. Stress-induced hyperthermia was totally absent from Penk1−/− and was diminished in Pdyn−/− mice **p<0.01, ***p<0.001, compared with the control (before stress) value (two-way ANOVA followed by LSD test). Each column represents the mean value (±SEM) of 10 animals. Figure options 3.6. Plasma ACTH and corticosterone levels As shown in Figure 5, there was a significant difference in the ACTH concentrations between the groups (F7, 77=5.87; p<0.001). Post hoc analysis of the data revealed that the deletion of endorphins did not alter basal ACTH concentrations. However, stress induced a significant elevation in the ACTH level in wild-type and β-endorphin animals, while the increase in the ACTH concentration failed to reach the level of significance in Penk1−/− and Pdyn−/− mice. Plasma ACTH levels in wild-type and endorphin knockout animals. Samples were ... Figure 5. Plasma ACTH levels in wild-type and endorphin knockout animals. Samples were taken from animals without stress or 10 min after a 5 min zero-maze stress situation. Stress induced an increase in the ACTH level in wild-type and β-endorphin−/− animals, but not in Penk1−/− or Pdyn−/− mice. Basal ACTH concentrations did not differ between wild-type and endorphin knockout animals. Columns represent the mean (±SEM, n=8–10) value of the group. *p<0.05; **p<0.01, basal vs. stressed in the corresponding genotype (one-way ANOVA followed by Tukey test). Figure options Comparing the basal (stress-free) plasma corticosterone levels in unstressed animals, we found a significant difference between the strains using one-way ANOVA (F3, 36=4.51; p<0.01). Post hoc analysis using Dunnett's test revealed that the genetic deletion of enkephalin resulted in a low resting plasma corticosterone level (5.8±12.01 in Penk1−/− vs. 31.78±15.28 in wild-type mice) ( Figure 6). Basal (not-stressed) plasma corticosterone levels in wild-type and endorphin ... Figure 6. Basal (not-stressed) plasma corticosterone levels in wild-type and endorphin knockout mice. Corticosterone concentration was lower in Penk1−/− but not in other knockout strains. **p<0.01 compared with wild-type, one-way ANOVA followed by Dunnett's test. Columns represent the mean (±SEM, n=10) value of the group. Figure options Exposure of the animals to the zero-maze test situation resulted in an increase in corticosterone levels in all genotypes. However, the amplitude and the duration of the response was quite different between the strains, as shown by a significant interaction between the genotype and time (F12, 147=5.86; p<0.001) ( Figure 7). Wild-type animals displayed the highest corticosterone levels (140±10 ng/ml) of all genotypes after 30 min, which returned to normal levels after 60 min. β-Endorphin-deficient mice showed a much smaller response amplitude (75±10 ng/ml) which also normalized after 60 min ( Figure 7). Penk1−/− mice had a lower peak corticosterone level when compared with wild-type animals, but a longer-lasting response with significantly increased corticosterone levels after 60 min. The amplitude of the hormone response in Pdyn−/− animals was similar as in wild types, but it reached the maximum earlier, soon in the 15th minute after the stress and it normalized later. Penk1−/− and Pdyn−/− mouse strains therefore showed a larger area under the curve (AUC), when compared with wild-type animals (Penk1−/−, 6050; Pdyn−/−, 5832; wild type, 3682), while the AUC value was slightly lower in β-endorphin-deficient animals (3040). Wild-type and endogenous opioid knockout strains showed a different kinetics of ... Figure 7. Wild-type and endogenous opioid knockout strains showed a different kinetics of stress-induced change in plasma corticosterone concentration. *p<0.05; **p<0.01; ***p<0.001, wild-type vs. knockout; +p<0.05; ++p<0.01; +++p<0.001, basal vs. stressed (n=8–10, both two-way ANOVA followed by LSD test). Figure options The behavioral stress reactivity and the amplitude of the hormonal stress responses correlated significantly in wild-type animals as expected: higher activity in the open areas, a sign of lower anxiety, coincided with low level of corticosterone (r=−0,569; p<0.01). This correlation was not present in the knockout strains (r=0.165 in Penk1−/−; r=−0.235 in Pdyn−/− and r=−0.147 in β-endorphin knockout mice: each p>0.05). 3.7. c-Fos-like immunoreactivity c-Fos-like IR was either not present or was minimal (mean value ⩽0.1) in stress-free conditions independent from the genotype. Exposure to the zero-maze test situation induced a significant increase in the intensity of c-Fos-like IR in most of the limbic structures studied except in the lateral and central amygdala. Statistical analysis of the data from the reactive areas revealed that the intensity of c-Fos expression was higher in the wild-type mice than in the knockout strains in the basolateral amygdala (F3, 28=97.8; p<0.001) and PVN of the hypothalamus (F3, 28=32.0; p<0.001) ( Table 2). The difference between the wild-type and β-endorphin knockout mice was significant also in the bed nucleus of stria terminalis (F3, 28=9.35; p<0.001). We found no differences in the expression of c-Fos between the strains in the cingulate cortex, lateral septal nucleus and in the hippocampal areas ( Table 2). 3.8. Morphometry and histology of adrenal glands and hypophysis Deletion of endorphins did not influence the weight of adrenals (H(4)=0.482; p>0.05) or hypophysis (H(4)=3.497; p>0.05) ( Table 1). Histological analysis did not reveal any change in the microscopic structure of the organs ( Table 2). Table 1. Wet weight of hypophysis and adrenal glands. Organ weight in mg Wild type Penk1−/− Pdyn−/− β-End−/− Hypophysis 3.28±0.58 2.78±0.17 2.39±0.40 2.71±0.28 Adrenal gland 3.23±0.57 3.51±0.90 2.51±0.19 3.10±0.51 There was no significant difference between the values according to Kruskal–Wallis ANOVA, n=6–8. Table options Table 2. c-Fos-like IR in limbic structures 90 min after the zero-maze test paradigm. Wild-type Penk1−/− Pdyn−/− β-Endorphin−/− BLA 11.56±1.19 3.36±0.66*** 5.57±1.02*** 4.57±0.72*** BNST 1.78±0.48 2.43±0.46 0.79±0.19 0.34±0.13* CgCx 3.97±0.91 2.61±0.48 4.61±1.78 1.28±0.87 Hip. CA1 0.46±0.16 1.75±0.53 1.70±0.69 1.36±0.38 Hip. CA3 0.21±0.08 1.91±0.49 0.07±0.07 2.32±0.70 Hip. GD 4.61±0.74 4.22±1.36 5.37±0.96 2.82±0.66 LS 2.17±0.72 3.50±1.50 0.36±0.20 0.75±0.37 PVN 35.00±5.58 7.25±1.48*** 16.96±3.12*** 21.84±2.54** The intensity of c-Fos expression was lower in knockout animals as in wild-type mice in the basolateral amygdala (BLA) and paraventricular nucleus (PVN). In the bed nucleus of stria terminalis (BNST) the signal intensity was reduced in β-endorphin knockouts compared to wild types. There was no difference between the genotypes in the cingulate cortex (CgCx), in the hippocampal CA1, CA3 and dentate gyrus areas (Hip. CA1; CA3; GD) and in the lateral septal nucleus (LS). *p<0.05; **p<0.01; ***p<0.001, wild-type vs. knockout (one-way ANOVA followed by Tukey test). Mean number of cells showing c-Fos expression and ±SEM is shown (n=7–8).