Estrogen, synaptic plasticity and hypothalamic reproductive aging A.J. Hung a , M.G. Stanbury a,1 , M. Shanabrough a , T.L. Horvath a , L.M. Garcia-Segura b , F. Naftolin a, * a Reproductive Neuroscience Unit, Department of Obstetrics and Gynecology, Yale University School of Medicine, P.O. Box 208063, 333 Cedar Street, FMB 335, New Haven, CT 06520 8063, USA b The Cajal Institute of Neuroscience, Madrid 28002, Spain Received 20 August 2002; accepted 21 August 2002 Abstract Unlike primates who undergo ovarian failure and loss of sex steroids at the end of reproduction, aging rodents undergo constant vaginal estrus followed by constant diestrus and finally anestrus, which indicates the absence of responsive ovarian follicles. The latter state is analogous to menopause in women. The timing of the appearance of constant estrus is determined by many factors including estrogen exposure in the brain during development and the number of times that the animal gets pregnant. The chief site of this reproductive aging in rat brains is the arcuate nucleus of the hypothalamus. The transition from normal cycles to constant estrus parallels the females’ gradually decreased ability to respond to administered estradiol with a cycle of inhibition followed by disinhibition of gonadotrophin-releasing hormone. Evidence has accumulated indicating this to be due to a loss of the rat’s ability to respond to markedly elevated estradiol with the usual arcuate nucleus neuro-glial plasticity that supports the estrogen-induced gonadotrophin surge (EIGS). Just as male rats are not capable of an EIGS, aged females loose this ability through repeated EIGS. Experiments indicate that in male rats the hypothalamic synaptology that develops as a result of exposure to testicular androgens in the perinatal period (brain sexual differentiation) is a result of conversion of testosterone from the testes to estrogen in the brain and is therefore due to early estrogen exposure. Aging females appear to reach a synaptology similar to males and constant estrus as a result of repeated exposure to ovarian estrogens during their reproductive careers. The relative role of aging and hormonal factors remains unclear. Morphological evidence is presented that indicates the above effects of estrogen involve changes in hypothalamic arcuate nucleus neurons and glia, including changes in the organization of perikaryal membranes as well as arcuate nucleus synaptology and the load of peroxidase in the astroglia. A possible role for free radicals (reactive oxygen species) in hypothalamic reproductive aging has been proposed. Such a mechanism is supported by evidence that the anti-oxidant vitamin E delays the onset of constant estrus and the accumulation of glial peroxidase in aging female rats. However, since the synaptology and peroxidase load in constant estrus females is independent of the age at which the constant estrus occurs, it appears that the role of (repeated) estradiol exposure is more deterministic of hypothalamic failure than is aging, per se. q 2002 Elsevier Science Inc. All rights reserved. Keywords: Rats; Brain aging; Peroxidase; Heme-oxygenase-1; Astroglia; Estradiol 1. Introduction Fertility depends upon gonadotrophins and their regu- lation by estrogen in both sexes (MacLusky and Naftolin, 1981; Knobil, 1980; Harris and Naftolin, 1970). But, there is a clear dichotomy in gonadotrophin control between the sexes: intact adult males cannot release a surge of luteinizing hormone (LH) of preovulatory proportions in response to estrogen injections, i.e. males are capable only of reciprocal or ‘negative’ gonadotrophin feedback responses to estrogen. Normal females, in addition to reciprocal feedback to estrogen, have an inhibition – disinhibition cycle of gonadotrophin-releasing hormone (GnRH) that is the basis of the estrogen-induced gonado- trophin surge (EIGS) or ‘positive feedback’ in the presence of marked rises in estrogen levels accompanying maturation of ovarian follicles (Naftolin et al., 2001, 1996). In both female rodents and primates rising levels of estrogen at midcycle first inhibit, then disinhibit the release of GnRH. 0531-5565/03/$ - see front matter q 2002 Elsevier Science Inc. All rights reserved. PII: S0531-5565(02)00183-3 Experimental Gerontology 38 (2003) 53–59 www.elsevier.com/locate/expgero 1 Present address: Maria Guadalupe Stanbury, Centro Medico Rafael Guerra Mendez, Avenida Rondon c/c Avenida 5 de Julio, Valencia, Venezuela. * Corresponding author. Tel.: þ 1-203-785-4003; fax: þ1-203-785-4883. E-mail address: frederick.naftolin@yale.edu (F. Naftolin).