Search

Summary Several observations suggest that neuronal shrinkage rather than cell death is the major phenomenon in neurodegenerative diseases. In order to make this distinction, smaller cells should also be included in cell counts. Also, morphometric determination of total cell numbers of brain structures is required. Morphometry was performed on the locus coeruleus using a newly developed method to delineate this nucleus from five patients who had died with Alzheimer's disease, five with Parkinson's disease, five with amyotrophic lateral sclerosis and from five control subjects who had died from causes that would not have affected the locus coeruleus. The length and volume of the locus coeruleus and its total number of large pigmented neurons, small unpigmented neurons and glial cells were determined. Since reliable delineation of the boundaries of the locus coeruleus is a requirement for the determination of total cell numbers, an image analyser-assisted procedure was developed. In Alzheimer's disease we found an 82% decrease in the number of large pigmented neurons and a 39% decrease of small unpigmented neurons. In Parkinson's disease, we found a 39% decrease of large pigmented neurons but also a 44% (though not significant) increase of small unpigmented neurons, which is indicative of a shift from large pigmented neurons to small unpigmented neurons in Parkinson's disease. The large pigmented/small unpigmented neuron number ratio was greatly and significantly reduced in Alzheimer's disease, Parkinson's disease and amyotrophic lateral sclerosis. These findings support the hypothesis that the decrease of large pigmented neurons of the locus coeruleus in some neurodegenerative diseases is not entirely due to cell death, but rather to cell shrinkage and a loss of phenotype. This hypothesis may have consequences for the development of therapeutic strategies since atrophied cells can be activated. On the other hand our data confirm that, at least in Alzheimer's disease, large pigmented neurons do also undergo cell death.

The earlier described age-related decreases in vasopressin innervation of extra-hypothalamic rat brain regions were found to coincide with a decrease in vasopressin expression in the cells where these fibers originated. A significant age-related decrease in the number of vasopressin-immunoreactive cell bodies was found in the medial amygdala and locus coeruleus of senescent Brown-Norway (BN/BiRij) rats (33 months) when compared to middle-aged (19 months) and young (3 months) rats. In addition, total testosterone plasma levels were significantly reduced in middle-aged and old rats as compared to young animals and the number of vasopressin-immunoreactive cells in both the medial amygdala and locus coeruleus correlated significantly with the decreased testosterone levels in a similar way as found earlier for vasopressin terminals.

Vasopressin binding sites were determined in the choroid plexus of five Alzheimer's disease patients and five non-demented controls using the125I-labelled linear V1a-antagonist. The Alzheimer's disease patients showed a twofold increase in the density of vasopressin binding sites, whereas the increase in the affinity constant Kd did not reach significance.

The distribution of vasopressin and oxytocin immunoreactive fibers was examined in the pontine parabrachial nucleus of the human brain using purified polyclonal antibodies. The results revealed a striking predominance of vasopressin in this brain region. No obvious density difference, either in vasopressin or in oxytocin innervation, was found between Alzheimer's disease patients and matched controls. The present study corroborates other reports that suggest that in Alzheimer's disease the vasopressin innervation in the caudal part of the human brain is not affected.

The study of an increasing number of processes occurring in the human brain can be carried out on autopsy material. The availability of this material, whether fresh, frozen or fixed, makes it possible to develop methodologies for studying the neuroanatomical and neurochemical aspects of the human brain. It has also become possible in recent years to correlate functional changes with neurochemical changes and with neuroanatomical abnormalities in disease states. Some compounds and structures are damaged irreversibly within minutes after death and some brain components are known to disintegrate within seconds. This led to the widespread idea that autopsy material would not be suitable for basic research purposes and would not supply the necessary answers on the various fundamental questions regarding processes occurring in normal or diseased brain. However, from data published in recent years in which autopsy material has been routinely used, it becomes more and more evident that this is a misconception. There is an increasing number of reports based on the use of normal and pathological human brain tissue obtained by autopsies in spite of the fact that there is a worrying continuous decline in autopsy rate which causes serious concern among scientists world-wide (Anderson and Hill, 1989). It also became evident that when using the proper fixation procedures, sufficient structural integrity is retained in the tissue to allow morphological and morphometrical studies (Swaab and Uylings, 1988). Electron microscopic examination of synaptosomal preparations from post-mortem human brain showed them to be only slightly less pure than preparations from fresh tissue although there was some degree of damage (Hardy et al., 1982). Agonal state effects the stability of brain compounds and causes brain hypoxia. This again forms a tremendous difficulty for the study of human neurological and psychiatric diseases as one of the frequent causes of death is bronchopneumonia which leads to brain hypoxia and results in pronounced lactic acidosis. The Netherlands Brain Bank has succeeded to partly circumvent some of the serious problems encountered in providing human tissue for research by performing rapid autopsies with an average post-mortem delay of 2-4 h. This has become possible by a close collaboration of numerous nursing homes in Amsterdam and its vicinity and with the neuropathologists of the Free University in Amsterdam. We also measure the pH of the tissue as indicator of agonal state in order to reveal unsuitable specimens. The human hypothalamus contains various nuclei manifesting a wide variety of changes in different conditions.(ABSTRACT TRUNCATED AT 400 WORDS)

Enhanced immunocytochemical staining of vasopressin-containing neurons was observed after incubation of rat brain slices in Ringer medium for 6 h at room temperature, as compared to brain tissue fixed immediately after death. Hypothalamic vasopressin neurons in the supraoptic nucleus, the paraventricular nucleus and the suprachiasmatic nucleus stained more intensely after a postmortem interval of 6 h. Extrahypothalamic vasopressin neurons (VP) in the bed nucleus of the stria terminalis, the medial amygdala and the locus coeruleus proved to be stained as well. Extrahypothalamic VP neurons in the locus coeruleus could, until now, only be visualized after in vivo pretreatment with colchicine. In addition, staining was observed at two new sites, the dorsal raphe nucleus and the lateral septum. Staining of VP was corroborated by application of different antibodies directed against the intact vassopressin molecules as well as by antibodies directed against the other parts of the vasopressin precursor molecule, i.e. neurophysin andglycopeptide. The specificity of the VP-staining was validated by using pre-immune serum and using Brattleboro rat brain tissue, resulting in a negative staining in both cases. Furthermore, homogenated punches of the suprachiasmatic nucleus were submitted to iso-electric focusssing on polyacrylamide gel, followed by press blotting and subsequent immunocytochemical staining for vasopressin. Iso-electric focussing enabled us to characterized and quantify peptides in the suprachiasmatic nucleus. The vasopressin content increased 6 h postmortem, while c-terminal glycopeptide and neurophysin levels remained stable. Similar results were observed in the suprachiasmatic nucleus from decapitated rats whose brains were left intact in the skull for 6 h at room temperature.

Following two weeks of cerebroventricular administration of arginine-vasopressin (AVP) by Accurel implants, two types of binding sites for this peptide were immunocytochemically visualized in blood vessels in the brain of Brattleboro (di/di) rats: (1) endothelial cells of capillaries were stained with the highest density in hippocampus, striatum, and locus coeruleus (LC), whereas only few such stained cells were present in the septum and cerebral cortex. (2) Bound AVP was also present on endothelial cells and pericyte-like cells in larger blood vessels in striatum and the LC. Both types of vasopressin binding site staining on blood vessels were dose-dependent and could be further enhanced by additional in vitro preincubation of the fixed sections with AVP. This staining was not present in rats implanted with either oxytocin or alpha-melanocyte-stimulating hormone.