The hypothalamo-pituitary-adrenal (HPA) axis is known to be activated in depressed patients. Although direct evidence is lacking, this activation is hypothesized to be due to hyperactivity of corticotropin-releasing hormone (CRH) neurons of the hypothalamic paraventricular nucleus (PVN). Recent immunocytochemical studies in experimental animals and in humans showed that the number of CRH-expressing neurons correlated with the activity of these neurons. In addition, colocalization of AVP in CRH neurons has been shown to be an index for the secretory activity. Therefore, we estimated the total number of CRH-immunoreactive neurons and their fraction showing colocalization with AVP in the PVN of 10 control subjects and of 6 depressed patients who were diagnosed to be suffering from a major depression or a bipolar disorder. The mean total number of CRH-expressing neurons of the 6 depressed patients was four times higher, and the number of CRH neurons co-expressing AVP was almost three times higher than those in the control group. We also determined the two activity parameters of CRH neurons in the PVN of 2 subjects with a depressive organic mood syndrome or a depressive disorder not otherwise specified. In these two ‘non-major depressed’ subjects, the activity parameters of CRH neurons were comparable to those of control subjects. Our observations strongly support the hypothesis that CRH neurons in the PVN are hyperactivated in major depressed patients. This hyperactivity might be causally related to at least part of the symptomatology of depression.

Immunocytochemical staining, using a monoclonal antibody against corticotropin-releasing hormone, was performed on hypothalami of 13 human subjects between 23 and 91 years of age who had not suffered from a primary neurological or psychiatric disease. Corticotropin-releasing hormone (CRH) immunoreactivity was present in neurons of the paraventricular nucleus (PVN) and in their fibers running to the median eminence. The CRH-positive neurons were scattered throughout the PVN, but in the rostral part relatively few cells were present. There were large individual differences in the number and staining intensity of CRH neurons in the PVN and in the staining intensity of the median eminence. These differences seemed not to be attributed to age, sex, postmortem delay, fixation time or hour of death. In the rat, too, no relationship was found between a postmortem delay of up to 24 h and CRH staining intensity of the median eminence. Since the distribution of CRH-immunoreactive neurons in the human PVN strongly overlap with vasopressin, colocalization of these peptides was investigated in a double label study and indeed found in subjects ranging between 43 and 91 years of age. However, cells staining for only one of the peptides were also observed. The vasopressin cells had a mean cellular profile area which was 2.3 times as large as the CRH cells and 2.2 times as large as the CRH and vasopressin containing neurons. In younger subjects (23-37 years of age) no colocalization of the two peptides was seen. The age-dependent colocalization of CRH with vasopressin is interpreted as a sign of increased activation of the CRH neurons with age.

The human hypothalamus is involved in a wide range of functions in the developing, adult and aging subject and is responsible for a large number of symptoms of neuroendocrine, neurological and psychiatric diseases. In the present review some prominent hypothalamic nuclei are discussed in relation to normal development, sexual differentiation, aging and a number of neuropathological conditions. The suprachiasmatic nucleus, the clock of the brain, shows seasonal and circadian variations in its vasopressin neurons. During normal aging, but even more so in Alzheimer's disease, the number of these neurons decreases. In homosexual men this nucleus is larger than in heterosexual men. The difference between the sexually dimorphic nuclei of men and women arises between the ages of 2-4 to puberty. In adult men this nucleus is twice as large as in adult women. In the process of aging, a sex-dependent decrease in cell number occurs. The vasopressin and oxytocin cells of the supraoptic and paraventricular nucleus are present in adult numbers as early as mid-gestation. Lower oxytocin neuron numbers are found in Prader-Willi syndrome, AIDS and Parkinson's disease. Familial hypothalamic diabetes insipidus is based upon a point mutation in the vasopressin-neurophysin-glycopeptide gene. Parvicellular corticotropin-releasing hormone-containing neurons in the paraventricular nucleus increase in number and are activated during the course of aging. In post-menopausal women, the infundibular or arcuate nucleus contains hypertrophic neurons containing oestrogen receptors. These neurons may be involved in the initiation of menopausal flushes. The nucleus tuberalis lateralis may be involved in feeding behaviour and metabolism. In Huntington's disease the majority of its neurons is lost; in Alzheimer's disease it shows very strong cytoskeletal alterations. Tuberomammillary nucleus neurons contain, e.g., histamine or galanine, and project to the cortex. Strong cytoskeletal changes, as well as plaques and tangles are found in this nucleus in Alzheimer's disease. The various hypothalamic nuclei are probably involved in many functions and symptoms of which only a minority has been revealed.

The hypothalamo-pituitary-adrenal (HPA) axis is activated during aging and even more so in dementia. Increased levels of corticosteroids may be neurotoxic. Therefore we have investigated cortisol levels in cerebrospinal fluid (CSF) of Alzheimer patients and controls. Ventricular postmortem CSF was collected from clinically and neuropathologically well-defined Alzheimer patients (n = 26) and control subjects (n = 21). In the group of Alzheimer patients the mean CSF total cortisol level was 83% higher than that in the controls. In presenile Alzheimer patients (< 65 years of age; n = 13) the CSF-cortisol level was 5 times higher than that of presenile controls (n = 7). In contrast, senile Alzheimer patients (n = 13) and controls of over 65 years of age (n = 14) did not show a significant difference in CSF-cortisol levels. The presence or absence of a difference in the cortisol-CSF levels in, respectively, presenile or senile Alzheimer patients as compared to controls was due to the 3.5-fold rise of CSF-cortisol in control subjects over 65 years of age as compared with controls under 65 years of age. The CSF-cortisol levels in presenile and senile Alzheimer patients were similar. No significant correlation was observed in the Alzheimer patients between age of onset of the dementia and CSF cortisol levels or duration of Alzheimer's disease and CSF cortisol levels. The finding that in senile Alzheimer patients cortisol levels were similar to those of unaffected age-matched controls does not seem to support the cortisol neurotoxicity hypothesis.(ABSTRACT TRUNCATED AT 250 WORDS)

It has been hypothesized that the corticotropin-releasing hormone (CRH) neurons of the hypothalamic paraventricular nucleus (PVN) become hyperactive with age, and even more so in Alzheimer's disease. This hyperactivity could be due to an increased production of CRH per neuron, or an increased number of PVN neurons producing CRH, or both. As a first step in elucidating which of these biological mechanisms might be operative, we have estimated the absolute number of CRH immunoreactive neurons in the PVN of 10 human control subjects between 36 and 91 years of age and 10 Alzheimer patients between 40 and 97 years of age. CRH neurons were immunocytochemically detected in 6 microns paraffin sections with the aid of a highly specific monoclonal antibody to CRH. The antibody signal was amplified by the biotin-streptavidin and alkaline phosphatase methods. The absolute number of CRH neurons in the PVN was obtained by multiplying the number of CRH neurons in a unit volume (NV) by the total volume of the PVN. Two different methods were used to estimate the NV: an unfolding method and a disector method (about three times more time-consuming). Compared to the disector, the unfolding method consistently yielded a lower cell number for all patients by 38% (+/- 2.8%; mean +/- SEM). However, both methods yielded an increase in the absolute number of CRH neurons in control and Alzheimer patients with age. No statistically significant difference in the absolute number of CRH neurons was found between control and Alzheimer patients with both methods. The age-dependent increase in the absolute number of CRH neurons within the PVN of both control and Alzheimer patients is interpreted as a sign of activation of the CRH neurons with age.