| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
and -ß in the Infundibular Nucleus of the Human Hypothalamus Are Related to the Occurrence of Alzheimer’s Disease Neuropathology
Netherlands Institute for Brain Research (A.H., D.F.S.), Amsterdam 1105 AZ, The Netherlands; and Department of Obstetrics and Gynecology (A.H.), Faculty of Medicine, University of Indonesia, Jakarta 10430, Indonesia
Address all correspondence and requests for reprints to: Dick F. Swaab, M.D., Ph.D., Professor of Neurobiology, Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands. E-mail: d.swaab{at}nih.knaw.nl.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and -ß in the infundibular nucleus of the hypothalamus was studied immunocytochemically in 28 control subjects and 14 patients with Alzheimer’s disease (AD). A shift was found from more nuclear staining of ER in young female controls to more cytoplasmic staining in elderly female controls, whereas no such change was observed in elderly male controls. The shift of ER from nucleus to cytoplasm in elderly female controls was accompanied by a relative absence of AD neuropathology, i.e. hyperphosphorylated tau stained by hyperphosphorylated tau protein (AT8). In contrast, male and female AD patients showed more nuclear ER and a much stronger AD neuropathology. It is proposed that the shift of ER from nucleus to the cytoplasm may reflect activation of neurons and that hyperactivity decreases the risk that neurons in the course of aging develop AD neuropathology. In contrast, the presence of nuclear ER seems to predispose to reduced activity and increases the risk of some neurons to develop AD neuropathology. ERß in basket-like terminals was preferentially observed in elderly male controls and AD patients, a novel phenomenon. This suggests that the presence of basket-like ERß may reflect reduced activity, which is-associated with an increase in hyperphosphorylated tau staining. However, the neurons inside the basket-like ERß showed signs of hyperactivity and did not stain for AT8. All AT8-positive neurons in the infundibular nucleus contained MSH as a marker for proopiomelanocortin neurons. These neurons produce ß-endorphin that inhibits GnRH release. Because they diminish in activity in postmenopausal women, this may contribute to the hyperactivity of GnRH neurons. The regulation of the gonadal axis may thus be affected by AD neuropathology independent of AD neuropathology in cognition-related brain structures. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is tempting to relate the sex difference in neurofibrillary pathology in the infundibular nucleus to the strong activation of this structure in postmenopausal women due to the disappearance of the estrogen feedback. Activation of neurons seems, in general, to diminish the risk of developing AD neuropathology during the course of aging, a phenomenon we paraphrased as use it or lose it (6, 7). The total number of neurons in the infundibular nucleus of pre- and postmenopausal women is not different, whereas the mean neuronal volume increases up to 40% in postmenopausal women due to an increase in neuronal size. Hypertrophy and hyperactivity in the infundibular nucleus of postmenopausal women are found in neurons expressing estrogen receptors (ERs), neurokinin B (NKB), or substance P (SP) gene transcripts (8, 9). Hypertrophy of neurons in this nucleus is thus not a compensation for a loss of neurons but rather the result of activation due to the loss of the estrogen feedback in menopause (8, 10). In addition, some hypertrophy, although to a much lesser degree, also occurs in elderly men (11).
Sex steroid hormones and peptides influence gonadotropin secretion via a-negative feedback on the pituitary and hypothalamus (12, 13, 14, 15, 16). Loss of ovarian function with menopause and reduced testosterone production with age in men results in release of inhibition effects of gonadal steroids on the pituitary and the hypothalamus resulting in an increase in GnRH secretion and gonadotropins serum levels (17, 18, 19, 20, 21). Although these levels remain higher than that recorded in younger population, a declined gonadotropin serum levels is observed with age in elderly subjects. In women, although not in all studies (22), a decline in both FSH and LH levels occurs with age after menopause (19, 23), and decreased gonadotropins secretion with age is also observed in aged men (24). This indicates that gonadotropin secretion is influenced by not only sex steroid and peptide feedback but also by age. Either in men or in women, ER is mediated by the feedback mechanism of gonadal steroids on the hypothalamus (8, 25). GnRH-producing neurons are generally considered not to contain ER (26, 27, 28), and it is presumed that the inhibitory influence of estrogens on gonadotropin secretion is mediated by interneurons containing tachykinins, such as SP or NKB, rather than by a direct effect of estrogens on GnRH-producing neurons (9, 29, 30, 31). On the other hand, because ERß immunoreactivity was reported to be present in GnRH-expressing neurons in rat and estrogen treatment reduced the percentage of GnRH-expressing neurons, ERß may also mediate direct effects of estrogens on GnRH-producing neurons (32).
-MSH is a marker for endorphin neuron that inhibits GnRH and diminishes in activity in postmenopausal women (33) and is thus presumed to increase the vulnerability of these neurons for the development of AD neuropathology. -MSH, neuropeptide Y (NPY), somatostatin (SS), and galanin (GAL) neurons are, moreover, also major neuron population in the infundibular nucleus that is presumed to be affected by AD neurofibrillary pathology.
The aim of the present study was to determine whether alterations in ER and/or ERß expression in the infundibular nucleus of the hypothalamus occur in postmenopausal women during aging in men and in AD and whether such changes are-related to the occurrence of hyperphosphorylated tau as an early neuropathological hallmark of AD neurofibrillary pathology. Moreover, we determined what neuron populations in the infundibular nucleus might be affected by AD neurofibrillary pathology.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Postmortem material was obtained from The Netherlands Brain Bank (coordinator Dr. Rivka Ravid). Permission was obtained for a brain autopsy and the use of brain material and clinical information for research purposes. Hypothalami of 42 subjects were collected: 28 controls and 14 AD patients. The controls consisted of seven young females and seven young males ranging in age between 21 and 38 yr and seven elderly males and seven elderly females ranging in age between 78 and 90 yr (Table 1
). The 28 controls did not suffer from a primary neurological or psychiatric disease. In addition, the hypothalami of 14 sex- and age-matched AD patients were studied (Table 2). These patients, ranging in age between 69 and 89 yr, met the clinical criteria of probable AD according to the National Institute of Neurological and Communicative Disorders and Stroke–Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) (34). All control and AD brains were investigated in the same systematic way by the neuropathologist W. Kamphorst (see Ref. 35 for details). The distribution of the Alzheimer changes over the brain was estimated according to Braak and Braak (1).
|
|
, ERß, and hyperphosphorylated tau protein (AT8) immunocytochemical staining and for dual-label immunofluorescence of AT8 with neuropeptide-expressing neurons such as -MSH neurons, NPY neurons, SS neurons, and GAL neurons. Immunocytochemistry and specificity of the antisera
A polyclonal rabbit anti-ER antiserum (MC-20) that recognizes the carboxyl terminus epitope of the ER (cat. no. sc-542, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and a polyclonal goat anti-ERß antiserum (N-19), directed against an amino acid sequence mapping at the amino terminus of human ERß (cat. no. sc-6820, Santa Cruz Biotechnology), were used in the present study. The staining procedures and-specificity tests for ER and ERß antisera have been previously described extensively by Ishunina et al. (39) and Kruijver et al. (40, 41). No staining was observed after omitting the MC-20 antiserum or after adsorption of MC-20 to its blocking peptide (39). In a spot blot test, MC-20 recognized its blocking peptide on nitrocellulose paper by showing the expected concentration gradient (Santa Cruz Biotechnology; blocking peptide, cat. no. sc-542, lot no.C059). In addition, with two different anti-ER antisera, C-314 (N terminus directed; Santa Cruz Biotechnology; cat. no. sc-786; antibovine ER; lot no. J278) and MC-20 (C terminus directed), a similar distribution pattern was displayed in the human hypothalamus (40). Western blot with the ER antiserum MC-20 on the human hypothalamic tissue showed a-specific band around the expected 68 kDa of ER with no such band around the 54 kDa of ERß (40, 42, 43). Western blot with the ERß antiserum N-19 on human hypothalamic tissue recognized a protein band around expected 54 kDa weight and did not recognize the protein band 68 kDa, i.e. ER (41, 42). In a spot blot, it was also confirmed that the antiserum N-19 recognized the homologous blocking peptides, whereas an adsorption test with the homologous peptide resulted in elimination of the staining (39). Moreover, staining of adjacent sections with the antiserum against the C terminus of the ERß (L-20, Santa Cruz Biotechnology, cat. no. sc-6822) (44) revealed the same staining pattern as the antiserum against the N terminus of ERß used in the present study (39). ERß cytoplasmic staining was observed in granulosa cells and follicles of the human ovary, a localization that is consistent with a study in the rat (45).
In human testis, Leydig and connective tissue cells showed nuclear ERß staining, which is also in agreement with a study in rat (46). The differences in distribution showed by the ER antiserum MC-20 and the ERß antiserum N-19 in the hypothalamus, pituitary, ovary, and testis as described extensively by Ishunina et al. (39) and Kruijver et al. (40, 41) supported the-specificity of the antisera. The ERß antiserum appeared thus to be-specific for ERß and not cross-react with ER and the ER antiserum is-specific for ER and is not cross-reactive with ERß. This series of observations demonstrated that the ER and -ß antisera used in our study were-specific. The intensity of staining was estimated semiquantitatively by means of light microscopy according to the following scale: strong (2++), moderate (2+), weak (+), very weak (±), and absent (–) (39). Nuclear ER staining and cytoplasmic ER staining was distinguished.
For immunocytochemical staining of hyperphosphorylated tau, a primary monoclonal antiserum AT8 directed against the phosphorylated tau epitopes serine 202 and threonine 205 (47, 48) was used. This antiserum was used to recognize hyperphosphorylated tau as an early marker for the neurofibrillary AD pathology. The staining procedure was performed according to Schultz et al. (2), and semiquantitative estimation for neuropathological changes in AT8-staining was done according to the following scale: severe (2 + 2+), marked (2++), moderate (2+), mild (+), and no discernible changes (0) (2).
Immunofluorescence
The procedure of double labeling using confocal laser-scanning microscope for AT8 and MSH, NPY, SS, or GAL localization was based on Wolswijk (49). Polyclonal rabbit antihuman MSH (4372, 23.04.75, Netherlands Institute for Brain Research) (50), polyclonal rabbit antiporcine NPY (Niepke, 26.11.88) that recognizes NPY (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36) (38), polyclonal rabbit antisomatostatin-28 (S309) (51, 52), and polyclonal rabbit anti-GAL (GAALTJE, 9.06.93, Netherlands Institute for Brain Research) that recognizes GAL (33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61) were used. AT8 was stained green by fluorescein isothiocyanate, and MSH, NPY, SS, or GAL was stained red by Cy3.
Statistical analysis
All differences were tested using the nonparametric Mann-Whitney U test, corrected for ties. A P < 0.05 was considered to be significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
immunoreactivity (ER-IR)
Control subjects.
Throughout the infundibular nucleus of all controls, many neurons contained ER. In addition, there was also a clear expression of ER-IR in some of ependymal cells lining the third ventricle and in endothelial and smooth muscle cells of the blood vessel wall. In the infundibular nucleus in young controls, ER-IR was localized preferentially in the cell nucleus (Fig. 1, A and B), but cytoplasmic staining was also observed (Table 1
and Fig. 1B). No significant sex difference was present for nuclear ER-IR or cytoplasmic ER-IR in young controls.
|
-IR distribution mainly in the nucleus, whereas only few neurons showed staining in the cytoplasm (Fig. 1D), comparable with the staining in young male and female controls (Table 1). In elderly female controls, a clear shift from nuclear to cytoplasmic ER staining was observed, compared with young female controls, consisting of a significant decrease in nuclear ER staining (P = 0.035) and an increase in cytoplasmic ER staining (P = 0.021) (Table 1 and Fig. 1C). Both normal-sized and hypertrophied neurons contained ER-IR, mainly in the cytoplasm. The infundibular neurons in the elderly female controls were clearly hypertrophied, whereas neuronal hypertrophy was much less pronounced in elderly males (Fig. 1, C and D).
In elderly female controls, the hypertrophied neurons were accompanied by enlarged nucleolar size and contained in a high proportion of large spheroid bodies inside the nucleus (Fig. 1C). These nuclear spheroids stained often for ER in a similar or even more intense way as the cytoplasm also in neurons that did not have ER nuclear staining (Fig. 1C).
Only few of such phenomena were observed in elderly male controls and none of them in young controls (Fig. 1, A, B, and D).
AD patients.
In AD patients, we observed a significant increase in nuclear ER-IR (P = 0.005) and decrease in cytoplasmic ER-IR (P = 0.011), compared with age- and sex-matched controls (Table 2 and Fig. 1, E and F). No significant difference of nuclear ER/ß-IR, cytoplasmic ER/ß-IR, basket-like ERß-IR, or AT8-IR was observed either in relation to a different degree of AD severity (Braak stage IV-V vs. Braak stage VI) or a different apolipoprotein E (ApoE) genotype, i.e. with or without an ApoE 
4 allele. The neurons in the infundibular nucleus of AD patients were generally not hypertrophied, as observed in nondemented postmenopausal women. However, in some neurons the nuclear spheroids were also present, more in female than in male AD patients.
ERß-immunoreactivity (ERß-IR)
Control subjects.
In young controls, a negative to moderate staining intensity of nuclear ERß-IR was observed in females (Fig. 2A) and a negative to weak intensity of nuclear ERß-IR in male controls. In the cytoplasmic compartment, a negative to weak staining intensity of ERß-IR was observed in female and male controls (Fig. 2A). No significant differences between young male and female controls were observed in the nucleus or cytoplasm. A negative to very weak staining intensity of nuclear and cytoplasmic ERß-IR was found in elderly female and male controls (Fig. 2C). No significant differences between elderly male and female controls were observed for the nuclear or in cytoplasmic staining.
|
and Fig. 2, B and D–F). None of the young controls showed NPTs. The basket-like ERß-staining appeared in a very weak to moderate intensity in male and female young controls (Tables 3 and 4 and Fig. 2B) and did not show a significant sex difference. Although NPT-ERß staining was found in elderly male controls, none of elderly female controls showed such a NPT-ERß staining (Table 3). Only one elderly female control had a weak basket-like ERß-staining. The infundibular nucleus of elderly male controls contained more basket-like patterns of ERß-staining than elderly female controls (P = 0.034) (Tables 3 and 4 and Fig. 2D).
|
|
). No significant sex difference was, however, observed for NPT-ERß in AD patients.
|
and Fig. 2, E and F), whereas no significant sex difference was present. The occurrence of basket-like ERß-staining was significantly increased in female AD patients, compared with age-matched female controls (P = 0.003), whereas no significant difference was observed between male AD patients and age-matched male controls (Tables 3 and 4 and Fig. 2D). The neurons that are located inside the basket-like ERß staining terminals tended to be larger and have a larger nucleus than the surrounding neurons (Fig. 2, E and F).
Hyperphosphorylated tau (AT8 immunostaining)
Three types of cytoskeletal abnormalities associated with AD were distinguished (Fig. 3), i.e. somatic NFTs, NPTs, and dystrophic neurite on neurons and blood vessels staining with AT8. This AD-related staining was completely absent in young controls. Two elderly male controls did not have any cytoskeletal abnormality, whereas four elderly male controls showed a mild degree of AT8 staining and one was stained moderately. In contrast, only one elderly female control had a mild degree of AT8 staining, whereas the rest showed no AT8 staining at all. This sex difference was statistically significant (P = 0.035). All AD patients showed AT8 staining in the infundibular nucleus, from mild to marked. There was no sex difference in AT8 staining present in the infundibular nucleus of AD patients (Table 2).
|
, C and D). It was remarkable that only a small number of neurons in the infundibular nucleus were positive for AT8. Most of these neurons contained densely stained NFTs, which filled the entire cytoplasm, whereas the nucleus was dislocated to the periphery of the cell. In addition, a small number of affected neurons in the infundibular nucleus of the hypothalamus showed only slightly granular cytoplasmic AT8 staining (Fig. 3, C and D), and a small number of AT8-positive neurons in the infundibular nucleus showed a delicate fibrillary AT8 staining (Fig. 3, C and D). We did not find ghost tangles.
Using two adjacent sections, we observed an obvious overlapping in the distribution of nuclear ER-containing neurons and AT8-positive neurons. In contrast, a distinct difference in distribution was found between the ERß basket-like staining pattern and AT8-positive neurons in the infundibular nucleus. In a study in six AD patients (three males and three females) using dual immunofluorescence labeling, we confirmed the colocalization of AT8 in all AT8-positive neurons with MSH (Fig. 4, A–D), whereas no colocalization existed with AT8 with NPY, SS, or GAL neurons in the infundibular nucleus (Fig. 4, E–G).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
staining in smooth muscle cells and endothelial cells of blood vessels. This observation is in full agreement with Kruijver et al. (40), who also observed the presence of ER in the smooth muscle cells and endothelial cells of blood vessels in the human hypothalamus. Moreover, the observation of only some of ependymal cells of the third ventricle showed ER staining is also in agreement with Donahue et al. (28), who observed the presence of ER in the ependymal cells of the third ventricle of the human hypothalamus based on an immunocytochemical study using another antiserum H222 (a monoclonal antibody raised against ER derived from MCF-7 human breast cancer cells), and Langub and Watson (53), who observed the presence of ER in some of ependymal cells encircling the third ventricle in the preoptic area and the median eminence of the guinea pig using electron microscopy and also H222.
The presence of ER-containing neurons in the infundibular (arcuate) nucleus of the hypothalamus has been reported in several mammalian species such as guinea pig, hamster, rat, and human (28, 40, 54, 55, 56). Estrogens are involved in different key functions of this nucleus, such as reproduction, sexual behavior and sexual differentiation of the brain (57, 58, 59, 60, 61), body weight homeostasis, and metabolism (38, 62, 63). Like the other members of the steroid receptor family, ERs are activated after estrogen binding and translocated to the nucleus as homo/heterodimer, in which they act as transcription factors (64, 65, 66, 67, 68). The receptors are continuously shuttling between the nucleus and cytoplasm (69). Moreover, increasing information on additional nonclassic mechanisms of estrogen action is becoming available. On estrogen binding, ERs in the cytoplasm, either in the cell body or processes, may couple to second messenger systems to regulate a variety of cellular events, such as the MAPK signaling pathways or phosphatidylinositol 3-kinase signaling pathways (see Refs. 70 and 71 for reviews). Moreover, other signaling pathways involving growth factors and/or membrane-initiated responses may also contribute to estrogen actions (reviewed in Refs. 72, 73, 74, 75). Alternatively, a drop in sensitivity of estrogen-regulated ER expression or an alteration of intracellular trafficking (76) might be another possible explanation for the cytoplasmic ER location in the current study.
In postmenopausal women, the depletion of ovarian follicles leads to a marked and abrupt decline in the production of estrogens (77), followed by a strong elevation of gonadotropin levels, as a reaction to the diminished-negative feedback of the gonadal steroids on the infundibular nucleus neurons (78, 79, 80, 81). Evidence supports, although not in all studies (22), a decline in both FSH and LH levels that occurs with age after menopause (19, 23, 82). The highest serum levels of gonadotropins in nondemented postmenopausal women were found at 50–60 yr, whereas during aging a slight age-associated decrease in serum FSH and LH levels was reported (24). A significant decline, particularly FSH, was observed in women of 80–90 yr of age, and a sharp reduction in both FSH and LH levels was observed over the ninth decade (24). Despite this age-associated gonadotropin decline, the serum FSH and LH in postmenopausal women were higher than those observed in premenopausal women, even in the ninth decade of life (24). In addition, an increased GnRH secretion with age was deduced from peripheral endocrine studies in a group of 70- to 80-yr-old nondemented postmenopausal women (21), whereas a high GnRH gene expression was observed in the medial basal hypothalamus of postmenopausal women until the eighth decade (83). In elderly men the gradually decreasing plasma testosterone levels are also followed by an elevation of gonadotropin plasma levels (84), but these changes are much less pronounced and occur more gradually than in women (85, 86). The higher levels of plasma estradiol and testosterone in healthy elderly men, compared with age-matched women, may protect the aspect of explicit memory in normal aging (87). The relationship between serum gonadotropin levels and age in nondemented men is different from that recorded in women. Serum gonadotropin levels remain stable from 50 to 70 yr and showed thereafter an age-associated increase up to around 80–94 yr when the highest serum levels of FSH and LH were reached, followed by a not statistically significant gradual age-associated decline (24, 82). During aging, nondemented men always showed a lower mean gonadotropin level, compared with that observed in postmenopausal women (24).
There is evidence that during a critical illness, the response of the hypothalamic-pituitary axis follows a biphasic pattern, with an acute and chronic phase (88). An elevated LH level was observed during the acute stress of surgery or myocardial infarction, whereas FSH and inhibin levels remained normal (89, 90, 91). In prolonged critically ill men in intensive care conditions, the pulsatility fraction of LH release was particularly attenuated (92). Moreover, abnormally low values of gonadotropins were observed in hospitalized postmenopausal women with prolonged severe illness (93). In addition, significantly higher serum gonadotropin levels were observed in patients with AD, compared with these in age- and sex-matched nondemented subjects (94, 95), suggesting that a decreased sensitivity of estradiol feedback on the pituitary level may occur in AD patients.
There are a number of parameters that indicate neuronal hyperactivity in the infundibular nucleus of postmenopausal women, such as an increased cell size (8, 10, 11) and expression of mRNAs of various peptides (9, 83). The hypertrophied neurons in the infundibular nucleus of postmenopausal women appears, in addition, from the presence of enlarged nucleoli and the nuclear spheroids, phenomena that were not observed in young controls (Fig. 1, A–C). Nucleolar augmentation, multiplication, and vacuolization has been described earlier in the nerve cells of the subventricular part of the infundibular nucleus in postmenopausal women and hypergonadotropic hypogonadal women, whereas in younger, still-fertile women, these manifestations were only rarely present in the infundibular nucleus and were absent in the subventricular nucleus (96). These nucleolar changes remain visible up to the age of 111 yr. In elderly men, far fewer nucleolar changes were present than observed in postmenopausal women (96). The nucleolus is a membraneless organelle within the nucleus whose major function is ribosome biogenesis (97). Protein synthesis mediated by ribosome is crucial for cell growth, proliferation, and adaptation to changing environments (98, 99). An increase in nucleolar volume can be taken as a sign of increased protein synthesis, hence of increased neuronal metabolic activity (96, 100, 101, 102).
The presence of nuclear spheroids is a remarkable phenomenon to which hardly any attention has been paid in the literature. The nuclear speroids have been described in a case of panhypopituitarism due to a craniopharyngioma (103) and in postmenopausal women and patients with Sheehan syndrome, conditions that are known to be accompanied by hyperactivity of the infundibular nucleus (104). The nuclear spheroids contain, according to an electron microscopic investigation, cytoplasmic constituents (104). Indeed, we often found these nuclear spheroids in subjects who showed a very intense staining of ER as found in the cytoplasma (Fig. 1C). The exact functional implication of these nuclear spheroids is not known, but it is clear that they appear in case of strong activation of the infundibular neurons, i.e. in postmenopausal women. The postmenopausal shift from nuclear ER in young women to mainly cytoplasmic ER in the hypertrophied hyperactive neurons suggests that the neurons are activated after diminishment of the inhibitory nuclear effects of estrogens in the postmenopausal state, and the nuclear spheroids might indicate an enhancement of neuronal activity, which is related to nonclassical mechanism of estrogen action (71, 72). Most of the hypertrophic neurons in the infundibular nucleus contain nuclear spheroids, and because increased cell size has a positive correlation with enhanced neuronal activity (105, 106), nuclear spheroids may also play a role in the mechanism of enhanced of neuroendocrine activity in these cells.
The infundibular nucleus of female AD patients seems to be less affected by hyperphosphorylated tau than that of male AD patients. This is in contrast to the general observation that women are more affected by AD than men (107, 108), although this observation is not without controversy (109). Interestingly, we also observed more hyperphosphorylated tau in the nucleus basalis of Meynert (NBM) of AD women than AD men (110). Apparently, structures related to cognitive functions such as the NBM are affected in a different sexually dimorphic way from a neuroendocrine structure like the infundibular nucleus.
The present study shows for the first time that ER and -ß distribution changes in the infundibular nucleus in a very different way in aging and AD. Neurons in this nucleus of postmenopausal female controls are strongly activated and hypertrophied and appeared to express ER mainly in the cytoplasm, in contrast to the preferential nuclear localization as we observed in the less active infundibular neurons of young controls, elderly male controls, and AD patients (Fig. 1, C–F).
The staining intensity of nuclear ER in the infundibular nucleus of young males was found to be similar to that in young females (Tables 1 and 4). Conversion of androgens into estrogens by the enzyme aromatase is a key mechanism by which testosterone regulates many physiological and behavioral processes in the brain, including the activation of male sexual behavior, sexual differentiation of the brain, and negative feedback effects of steroid hormones on GnRH production (111, 112, 113). In contrast to the decline in estradiol levels in women in menopause, a gradual and more modest decrease in testosterone levels takes place in elderly men (85, 114). The ongoing aromatization process of testosterone (114) seems to maintain the brain estrogen levels in elderly men at a higher level than in elderly women because in elderly male control patients, a similar amount of nuclear ER-IR was found as in young men (Tables 1 and 4).
Surprisingly, in elderly AD patients of both sexes, more nuclear ER expression was found than in sex- and age-matched controls (Fig. 1, E and F). The phenomenon of increased nuclear ER in AD seems to be a more general phenomenon because up-regulation of this receptor was also observed in the NBM (115), diagonal band of Broca (116), and medial mamillary nucleus (117) of AD patients. This was an unexpected finding because it concerned elderly patients, and, in addition, circulating estradiol levels are lower rather than higher in AD (118, 119).
Different types of neurons in the infundibular nucleus contain ERs. Besides the hypertrophied NKB and SP neurons (8, 9), 
-amino butyric acid (GABA)-ergic neurons, opioid-, GAL-, and other neuropeptide-producing neurons in the infundibular nucleus of the hypothalamus have been shown to express this receptor in rodents (56, 120). In the monkey ER mediates the inhibitory effects of estradiol on NKB and SP neurons (121). Our data suggest that the inhibitory nuclear effect of estrogens on the human infundibular nucleus is mediated by ER rather than ERß and that the inhibitory effect mediated by ER may remain present in elderly male controls. In AD patients, decreased activity of the infundibular neurons may result from the AD disease process or the changes in sex hormone levels. In addition, the decreased amount of nuclear ER expression in the infundibular nucleus neurons, as observed in the nondemented postmenopausal women, seems to mediate hyperactivity and is accompanied by a lower risk of developing AD neuropathology in these neurons.
The observation that only one of the seven elderly female controls showed mild AD neuropathology in the infundibular nucleus, whereas 71% of the elderly male controls showed mild to moderate AD neuropathology is in full agreement with the work of Schultz et al. (2, 3), who reported AD neurofibrillary pathology in a much larger series of patients, only in small number (6–10%) of the females and in up to 90% of the males (2, 3). The lower amount of AD changes in nondemented postmenopausal women, compared with age-matched nondemented men, is accompanied by a number of observations that indicate that in nondemented postmenopausal women, there is a much stronger neuronal activation after the removal of inhibitory effects of estrogen. The observation of Schultz et al. (2) and our data in the infundibular nucleus are thus in full agreement with the concept that a higher neuronal metabolic activity as observed in nondemented postmenopausal women may protect the neurons against the development of AD neuropathology, a phenomenon we paraphrased use it or lose it (6, 7). The data in the infundibular nucleus of nondemented controls show, moreover, that the regulation of the gonadal axis may be affected by AD neuropathology independent of AD neuropathology in cognition-related brain structures.
The basket-like ERß distribution that was observed preferentially in AD patients and male subjects is a novel phenomenon (Fig. 2, B and D–F). The neurons inside the basket tended to be larger and have a larger nucleus and nucleolus than the surrounding neurons, which suggests that the neuron inside a basket is strongly activated. Our ongoing study showed that the basket-like ERß-containing nerve terminals coexpress glutamate decarboxylase. Therefore, we hypothesize that the presences of ERß in GABAergic nerve terminals mediate an inhibition of the secretion of the inhibitory neurotransmitter GABA from the synapses and induce in this way an activation of the neuron that is surrounded by the basket. Moreover, in our ongoing experiments, we observed so far the presence of basket-like ERß nerve terminals surrounding GAL-expressing neurons. In addition, our current research suggests that some neuron populations remain very active in AD and others decline in activity (7). This complex phenomena need to be further studied.
In the activation of the infundibular nucleus during the course of aging and in menopause, a decreased activity of proopiomelanocortin (POMC) neurons seems to play a crucial role in various species. In vitro perifusion of human postmortem hypothalami showed that administration of the opiate receptor antagonist naloxone increased the release of GnRH, whereas opiates inhibited this process (122). This indicates that endogenous opioid peptides, in particular ß-endorphin, inhibit GnRH secretion. Moreover, a decreased activity of POMC neurons was observed in the infundibular nucleus of postmenopausal women by in situ hybridization (33), and an age-associated decrease was found in the number of neurons expressing ß-endorphin or its precursor in the hypothalamus of aged male and female rodents (123, 124, 125). No reversal of the lower numbers of neurons expressing the POMC gene transcript in aged ovariectomized cynomolgus monkey was found after estrogen replacement therapy, which indicates that the age-related decreased activity of the neurons is caused by factors other than just the low circulating estrogen levels (121). In the present study, using MSH as a marker for POMC neurons in the infundibular nucleus, we indeed observed colocalization of MSH in all AT8-stained neurons (Fig. 4, A–D). This shows that the POMC neurons, which are less active in aged individuals (33), are more vulnerable to the development of AD changes, which is again in full accordance with our use-it-or-lose-it concept (6, 7). In contrast, NPY-, SS-, and GAL-containing neurons were not stained by AT8 (Fig. 4, E and F). It suggests that the presence of AD neuropathology in the infundibular nucleus may affect the functions of some neurons in the infundibular nucleus in relation to the regulation of the gonadal axis (33, 95, 126).
Although ApoE-4 gene is a risk factor for AD, our relatively small sample did not show any clear relationship between the presence of this genotype and AD neuropathology in the infundibular nucleus. According to Schultz et al. (2), the formation of AD neuropathology in the infundibular nucleus of elderly controls develops independently of the development of AD changes in the cortex because the AD neuropathology was observed in the infundibular nucleus of controls that did not have any other AD neuropathology (i.e. Braak stage 0 subjects). This observation was confirmed when we observed a mild to moderate AD neuropathology in five (71%) elderly male controls, whereas a mild AD neuropathology was only present in one of the elderly female controls. On the other hand, in female AD patients, we found an increase of AD neuropathology in the infundibular nucleus, compared with elderly controls.
In conclusion, sex differences in hormonal alteration in aging are associated with sex differences in the occurrence of AD-like neuropathology, i.e. hyperphosphorylated tau. The cause and effect of this relationship is currently studied in patients with abnormal levels of sex hormones.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: AD, Alzheimer’s disease; ApoE, apolipoprotein E; AT8, hyperphosphorylated tau protein; ER, estrogen receptor; ER-IR, ER immunoreactivity; ERß-IR, ERß-immunoreactivity; GABA, -amino butyric acid; GAL, galanin; NBM, nucleus basalis of Meynert; NFT, neurofibrillary tangle; NKB, neurokinin B; NPT, neuropil thread; NPY, neuropeptide Y; POMC, proopiomelanocortin; SP, substance P; SS, somatostatin.
Received May 19, 2003.
Accepted January 6, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in the human basal forebrain. Brain Res 856:142–151[CrossRef][Medline]
and ß immunoreactivity in the human supraoptic nucleus in relation to sex and aging. J Clin Endocrinol Metab 85:3283–3291 distribution in the human hypothalamus in relation to sex and endocrine status. J Comp Neurol 454:115–139[CrossRef][Medline]
distribution. J Comp Neurol 466:251–277[CrossRef][Medline]
protein in neuronal processes of cultured hippocampal neurons. Neuroendocrinology 71:34–42[CrossRef][Medline]
-MSH in fetal development and the presence of an -MSH-like compound in nervous tissue. Front Horm Res 4:170–178[Medline]
-MSH, in the control of sexual behavior in the female rat. Peptides 19:1309–1316[CrossRef][Medline]
