The stress system in depression and neurodegeneration: Focus on the human hypothalamus

Abstract

The stress response is mediated by the hypothalamo–pituitary–adrenal (HPA) system. Activity of the corticotropin-releasing hormone (CRH) neurons in the hypothalamic paraventricular nucleus (PVN) forms the basis of the activity of the HPA-axis. The CRH neurons induce adrenocorticotropin (ACTH) release from the pituitary, which subsequently causes cortisol release from the adrenal cortex. The CRH neurons co-express vasopressin (AVP) which potentiates the CRH effects. CRH neurons project not only to the median eminence but also into brain areas where they, e.g., regulate the adrenal innervation of the autonomic system and affect mood. The hypothalamo-neurohypophysial system is also involved in stress response. It releases AVP from the PVN and the supraoptic nucleus (SON) and oxytocin (OXT) from the PVN via the neurohypophysis into the bloodstream. The suprachiasmatic nucleus (SCN), the hypothalamic clock, is responsible for the rhythmic changes of the stress system. Both centrally released CRH and increased levels of cortisol contribute to the signs and symptoms of depression. Symptoms of depression can be induced in experimental animals by intracerebroventricular injection of CRH. Depression is also a frequent side effect of glucocorticoid treatment and of the symptoms of Cushing's syndrome. The AVP neurons in the hypothalamic PVN and SON are also activated in depression, which contributes to the increased release of ACTH from the pituitary. Increased levels of circulating AVP are also associated with the risk for suicide. The prevalence, incidence and morbidity risk for depression are higher in females than in males and fluctuations in sex hormone levels are considered to be involved in the etiology. About 40% of the activated CRH neurons in mood disorders co-express nuclear estrogen receptor (ER)-α in the PVN, while estrogen-responsive elements have been found in the CRH gene promoter region, and estrogens stimulate CRH production. An androgen-responsive element in the CRH gene promoter region initiates a suppressing effect on CRH expression. The decreased activity of the SCN is the basis for the disturbances of circadian and circannual fluctuations in mood, sleep and hormonal rhythms found in depression. Neuronal loss was also reported in the hippocampus of stressed or corticosteroid-treated rodents and primates. Because of the inhibitory control of the hippocampus on the HPA-axis, damage to this structure was expected to disinhibit the HPA-axis, and to cause a positive feedforward cascade of increasing glucocorticoid levels over time. This ‘glucocorticoid cascade hypothesis’ of stress and hippocampal damage was proposed to be causally involved in age-related accumulation of hippocampal damage in disorders like Alzheimer's disease and depression. However, in postmortem studies we could not find the presumed hippocampal damage of steroid overexposure in either depressed patients or in patients treated with synthetic steroids.

Introduction

As pointed out for the first time by Hans Selye in Nature in 1936 (Selye, 1998), stress or ‘noxious agents’ initiate a reaction in the body, which he called the ‘general adaptation syndrome’ (GAS). Selye distinguished three stages that the body passes when responding to stress in the GAS: 1) the first stage is an ‘alarm reaction’, in which the body prepares itself for ‘fight or flight’; 2) the second stage of adaptation (provided the organism survives the first stage), is one in which a resistance to the stress is built; and 3) finally, if the duration of the stress is sufficiently long, the body enters a stage of exhaustion, a sort of aging, due to ‘wear and tear’. Although the stress response of the body is meant to maintain stability or homeostasis, long-term activation of the stress system can have a hazardous or even lethal effect on the body, increasing the risk of obesity, heart disease, depression, and a variety of other illnesses.

The hypothalamo–pituitary–adrenal (HPA) system is the final common pathway in the mediation of the stress response. Briefly, the hypothalamus releases corticotropin-releasing hormone (CRH) in response to a stressor, CRH acts on the pituitary gland, triggering the release of adrenocorticotropin (ACTH) into the bloodstream, which subsequently causes the hormonal end-product of the HPA-axis, corticosteroid release from the adrenal cortex (mainly cortisol in humans). Cortisol normally exerts a negative feedback effect to shut down the stress response after the threat has passed, acting upon the levels of the pituitary and hypothalamus. Cortisol is a major stress hormone that acts on many organs and brain areas through two types of receptors, i.e. the mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR), which have a specific and selective distribution in the brain (Reul and de Kloet, 1985). GRs have been found in multiple brain regions such as the hippocampus, amygdala and prefrontal cortex, which are relevant to cognition. These regions are also actively involved in feedback regulation of the HPA-axis. Moreover, stimulation by corticosteroids can be exerted at the level of the amygdala, the prefrontal cortex and the brain stem (locus coeruleus), interfering with HPA activity and stress effects on memory (Quirarte et al., 1997, Roozendaal, 2002, Fuchs et al., 2004). CRH-expressing neurons in the hypothalamic PVN project not only to the median eminence but also to other brain areas, where they regulate the adrenal innervation and thus the sensitivity of the adrenal for ACTH via the autonomic system (Buijs and Kalsbeek, 2001). Besides the role of a crucial neuropeptide which regulates the HPA-axis, CRH also shows central effects, including cardiovascular regulation, respiration, appetite control, stress-related behavior and mood, cerebral blood flow regulation and stress-induced analgesia (for review see (Swaab, 2003)). Part of the CRH neurons in the hypothalamic PVN co-express arginine vasopressin (AVP). When released together into the portal capillaries, AVP strongly potentiates the ACTH-releasing activity (Gillies et al., 1982, Rivier and Vale, 1983). In addition, circulating AVP from the supraoptic nucleus (SON) may induce ACTH release from the pituitary (Gispen-de Wied et al., 1992).

CRH and AVP mediate ACTH release via different second messenger systems (Won and Orth, 1990, Won et al., 1993, Won et al., 1990). CRH activates G protein-linked adenylate cyclase, leading to cAMP formation and protein-kinase-A activation. AVP works through a specific AVP receptor subtype termed V1b or V3, which is almost exclusively expressed by pituitary corticotrophs (Rene et al., 2000), and activates only phospholipase C. The V3 receptor is required for a normal pituitary and adrenal response to some acute stressful stimuli, and is necessary for a normal ACTH response during chronic stress (Lolait et al., 2006). The number of CRH and AVP-colocalizing neurons is increased during activation, such as in multiple sclerosis (MS) (Erkut et al., 1995), depression and during the course of aging—at least in males (Raadsheer et al., 1993, Raadsheer et al., 1994b, Bao and Swaab, 2007). V3 receptor mRNA levels and coupling of the receptor to phospholipase C are stimulated by glucocorticoids. Consequently, AVP up-regulation may be critical for sustaining the corticotrophic responsiveness in the presence of high circulating glucocorticoid levels during chronic stress or depression (Aguilera and Rabadan-Diehl, 2000). Oxytocin (OXT), on the other hand, inhibits ACTH release, a finding that has been confirmed in humans (Legros, 2001) and that provides an example of the various opposite actions of AVP and OXT. Following different types of corticosteroid treatment in different disorders or during the presence of high levels of endogenous corticosteroids produced by a tumor, we found, in postmortem tissue, not only that CRH-expressing neurons are hard to detect, but also that AVP expression in the SON and PVN is strongly decreased. OXT neurons were not affected, which further illustrates that, in the human brain, selective negative-feedback of cortisol is present in the CRH cells and in cells that co-express AVP (Erkut et al., 2004, Erkut et al., 1998). The glucocorticoid-induced suppression of AVP-synthesis has been proposed to occur at the posttranscriptional level (Erkut et al., 1998, Erkut et al., 2002b).

Stress suppresses the reproductive system at various levels: CRH prevents the release of the hypothalamic luteinizing hormone-releasing hormone (LHRH = GnRH) which signals a cascade of hormones that direct reproduction and sexual behavior. Cortisol and related glucocorticoid hormones not only inhibit the release of LHRH, but also inhibit LH-induced ovulation and sperm release. In addition, glucocorticoids inhibit the testes and ovaries directly, hindering production of the male and female sex hormones (Swaab, 2003). CRH fibers from the PVN do not only run to the portal capillaries but also innervate LHRH neurons in the infundibular nucleus, which may be one of the substrates for CRH control of reproductive functions (Dudas and Merchenthaler, 2003). On the other hand, the hypothalamic–pituitary–gonadal (HPG) axis also exerts extensive effects on the HPA-axis. Sex hormones partially control CRH gene expression. Both the nuclear estrogen receptor (ERα) (Bao et al., 2005) and the nuclear androgen receptor (AR) (Bao et al., 2006) are found to be colocalized with CRH neurons in the human hypothalamic PVN. In addition, an up-regulation of CRH and nuclear ERα was observed in mood disorders, both in males and females (Bao et al., 2005). There are five perfect half-palindromic estrogen-responsive elements (EREs) in the human CRH gene region, which may confer direct estrogenic regulation of human CRH gene expression (Vamvakopoulos and Chrousos, 1993). An androgen-responsive-element (ARE) that initiates a repressing effect of AR on CRH gene expression in the human CRH gene promoter region has also been localized recently (Bao et al., 2006). Ovarian steroids have been found to increase HPA-axis activity, enhance the HPA-axis response to psychological stress, and sensitize the HPG-axis to stress-induced inhibition in human and rhesus monkey (Kirschbaum et al., 1996, Roy et al., 1999).

Sex-related differences in the stress response are well-known from the animal experimental literature, but in humans the findings seemed inconsistent, probably, at least partly, due to the different methods used to stimulate the HPA-axis and to the age of the subjects. Gender-related differences in sex hormone levels further confound the specific role of the gender in HPA-axis responsivity. Over the past decade, however, the situation has become much clearer as a result of the development of psychological tests that generate adequate HPA-axis responses, for example the Trier Social Stress Test (Kirschbaum et al., 1993, Dickerson and Kemeny, 2004, Kudielka and Kirschbaum, 2005). The conclusion of these studies is that sex differences in the basal, unstressed state are subtle but become greatly pronounced following a psychological stressor. Although there are exceptions, in general between puberty and menopause, the HPA-axis and autonomic responses tend to be lower in women compared to men of the same age (Kajantie and Phillips, 2006). Recently, Roca et al. have found that, compared with age-matched women, young to middle-aged (18–45 years) men showed increased stimulated ACTH and cortisol, to either pharmacological (CRH) or physiological (exercise) stressors during pharmacological suppression of the gonadal axis. In addition, the secretion of cortisol after exercise and the initial secretion (0–30 min) of ACTH to either of the stressors were significantly larger in men compared to women. These data demonstrate that sex differences in the HPA-axis exist even in the absence of characteristic sex differences in reproductive steroids (Roca et al., 2005).

It has also been found that elderly men activate the HPA-axis to a greater extent than women in response to psychological stress (Traustadottir et al., 2003, Kudielka et al., 1998, Uhart et al., 2006). Our group has also recently found gender differences in the number of CRH expressing neurons in the human hypothalamic PVN, namely: 1) there is a significant age-related increase of CRH neurons in men, but not in women; and 2) men have significant more CRH neurons than women from the age of 24 years onward (Fig. 1). We also showed that an abnormal hormone status, induced by castration, ovariectomy or sex hormone-producing tumor, was accompanied by changes in CRH neuron number (Fig. 2) (Bao and Swaab, 2007). The increased number of CRH neurons in the human PVN may be interpreted as a sign of activation of the CRH neurons, because in situ hybridization of CRH mRNA analyzed in the same patients gave the same results under such varied chronic circumstances as depression, hypertension and MS (Huitinga et al., 2004, Goncharuk et al., 2002, Raadsheer et al., 1995). Moreover, during aging, cortisol levels in the cerebral spinal fluid (CSF) and in plasma are found to increase progressively between the ages of 20 and 80 years (Laughlin and Barrett-Connor, 2000).

Age-related activation of CRH neurons could be due to a series of factors, such as a decreased function of the hippocampus, which suppresses the activity of the HPA-axis and which is more sensitive to the process of aging than the PVN (Giordano et al., 2005, Miller and O'Callaghan, 2005)(see below). In this respect, it is of interest that a sex difference has also been reported in hippocampal aging, e.g., a significant age-related decline of hippocampal volume was found in men but not in women (Bouix et al., 2005, Pruessner et al., 2001). Increasing insensitivity of the HPA-axis to the feedback of cortisol may be another factor involved in the activation of the HPA-axis during aging (Goncharova and Lapin, 2002). Baseline AVP levels were found to be significantly higher in elderly subjects compared with young subjects (Rubin et al., 2002), which may also stimulate the HPA-axis. In addition, a sex difference was reported in AVP plasma levels. Men have higher AVP levels than women (van Londen et al., 1997, Asplund and Aberg, 1991), which agrees with the observation that the posterior lobe of the pituitary, where AVP and OXT are released into the circulation, is larger in boys than in girls (Takano et al., 1999). These sex differences are explained by the higher metabolic activity we found in AVP neurons in the SON in young men as compared to women, as determined by the size of the Golgi apparatus (Ishunina et al., 1999).

The stress response is strongly influenced by the time of day. The hypothalamic suprachiasmatic nucleus (SCN), the biological clock, is responsible for the rhythmic changes of the stress system. The SCN innervates brain areas in the human hypothalamic region (Dai et al., 1997, Dai et al., 1998a, Dai et al., 1998b) imposing its rhythm also onto the body via three different routes of communication: 1) Via the secretion of hormones; 2) via the parasympathetic autonomous nervous system and 3) via the sympathetic autonomous nervous system. The SCN uses separate connections via either the sympathetic or the parasympathetic system, not only to prepare the body for the impending change in activity cycle but also to sensitize the body and its organs for the hormones that are associated with such a change (Buijs et al., 2006). We have also found CRH fibers in the area of the SCN (Bao et al., 2005, Bao and Swaab, 2007), which suggests the existence of a bi-directional direct anatomical connection between the SCN and the PVN. The function of such an anatomical connection in health and disease deserves further study.