The sex steroid hormones have pronounced effects in areas of the brain that are not involved in sexual behavior or reproduction. The actions of the sex hormones have been a topic of recent research. One area of interest is their effects on hippocampal neuron spine formation. The paper by Hajszan et al. (1) presented in this issue now reports that dehydroepiandrosterone (DHEA), the adrenal hormone, also alters spine formation, and the effect may be due to its metabolism to estrogen in brain.

The adrenal cortex is the source of a variety of hormones, and the best known of these are the C₂₁ steroids: mineralocorticoids, such as aldosterone, and glucocorticoids, such as cortisol. The C₁₉ steroids of the adrenal gland are androgens, the most potent of this class being testosterone. However, testosterone is synthesized in only small amounts by the adrenal gland (2). Androstenedione, DHEA, and the sulfated derivative of DHEA, DHEA-sulfate (DHEAS), are the most highly circulating adrenal androgens, but these bind androgen receptors with low affinity (3) and thus are considered weak androgens. The role of adrenal androgens, particularly DHEA and DHEAS, in normal physiology has become, in recent years, the subject of intense investigation.

Maturation of the adrenal gland (adrenarche) occurs in humans well before the onset of puberty. This is characterized by the increased secretion of adrenal androgens, without corresponding increases in glucocorticoids or mineralocorticoids (4). Although androstenedione and DHEA are measurable in blood in preadrenarchal children, the amounts are low. Increases in adrenal androgen secretion typically occur in children between 6 and 8 yr of age (4) and, in adulthood, circulating levels of DHEAS are higher than those of any other steroid (5, 6). In addition, DHEA and DHEAS decrease dramatically in aging primates (5, 6). Unlike cortisol, whose secretion is regulated by the hypothalamo-pituitary axis and ACTH, there are few extraadrenal factors that are known to specifically regulate adrenal DHEA synthesis and secretion.

What is the role of DHEA in primate physiology? And can we test these hypotheses using rodent models? A number of recent studies have suggested that DHEA can have beneficial effects and suggest the possibility that DHEA is involved in a number of functions including cognitive function, metabolism, and vascular and immune function (for a review see Ref. 7).

Interestingly, rodent species do not have measurable levels of circulating DHEA when examined before sexual maturation and have very low levels subsequently (8). Thus, the rodent is a potential model in which to explore the effects of high levels of DHEA that are seen in humans without interference from changes in endogenous levels. It can be argued, however, that the lack of circulating DHEAS in rodents also makes it a poor model because treatment with DHEA is not a replacement therapy but is supraphysiological (9).

Nonetheless, given the reported effects of DHEA on primate and rodent physiology, how might these effects of DHEA be mediated at the cellular level? Binding studies demonstrated that DHEA binds the androgen receptor and the estrogen receptor, but with an affinity that is at least an order of magnitude less than that of the endogenous ligand (10, 11). Thus, the effects of DHEA do not appear to be mediated by direct binding of hormone to a currently known receptor type. A recent study by Liu and Dillon (12) has demonstrated a high-affinity G protein-coupled DHEA receptor found in endothelial cells. Such a finding will help in the search for a putative receptor mediating DHEA action.

An alternative explanation for DHEA action on cellular function is through its metabolism to other steroids that do have a high affinity for estrogen and androgen receptors. As shown in Fig. 1⇓, DHEA can be converted to estrogen or testosterone in tissues that have the appropriate steroidogenic enzymes. This theory posits that circulating DHEA exists as a precursor pool for steroid hormone synthesis in appropriate tissues and under appropriate conditions. These steroidogenic enzymes do exist in brain areas such as the hippocampus, thus raising the possibility that DHEA can be produced de novo in brain, or can be converted to another active steroid by local cellular metabolism (13).

In the paper by Hajszan et al. (1), a role for DHEA in the growth of dendritic spines in the hippocampus is shown. Dendritic spines are small bulbous protrusions on short stalks on dendrites. Most excitatory input to the hippocampus occurs on dendritic spines. Recent studies have shown that dendritic spines are highly mobile structures that also appear and disappear over the course of days depending on synaptic inputs (14). Time-lapse images of dendritic spine movements have been made by a number of groups. An example of these can be visualized at the following web site: http://www.fmi.ch/members/andrew.matus/video.actin.dynamics.htm.

Changes in the number of dendritic spines, the size of the spine head, the length of the stalk, and synaptic appositions occur after treatment with estrogen in ovariectomized female rats (15, 16, 17), through activation of N-methyl-d-aspartate receptors (18, 19) and a decrease in γ-aminobutyric acid-mediated inhibition (20). Recent work has also implicated androgen receptors in the modulation of spine density in male rats (21). In the paper by Hajszan et al. (1), the data presented suggest that DHEA treatment increased CA1 spine synapse density, an effect that was blocked by a nonsteroidal aromatase inhibitor. The authors propose that the ability of DHEA to increase spine density is therefore mediated by the aromatization of DHEA in the brain, and could be through activation of either estrogen or androgen receptors.

Although estrogen can increase spine density in female rats, this does not seem to take place in male rats (22). In males, it appears that androgen receptor activation is the prime mechanism for spine growth (21). This raises the question of whether DHEA, a precursor for both estrogens and androgens, can work equally well in both males and females. And, if so, are the intracellular processes mediating this action similar in both sexes?

What are the implications of these results? The hippocampus is a brain region involved in learning, memory, and cognitive function, and it also shows pronounced changes during aging and in pathological disease states, such as Alzheimer’s disease, related to aging and cognition. Estrogen and DHEA have been shown to enhance memory and learning functions (23, 24, 25) and prevent damage due to anoxia (26, 27) and glutamate-induced toxicity (28, 29). DHEA levels and estrogen levels decrease during aging and have also been implicated in etiology and treatment of the damage induced by Alzheimer’s disease (30, 31, 32). The results regarding the changes in spine density in hippocampal dendrites reviewed above are intriguing, but there is still a large gap between these data and how they may be involved, or not, in the changes seen in the behaviors, anoxia, toxicity, and Alzheimer’s disease. These studies and the one reported in this issue by Hajszan et al. (1) are certainly exciting and are motivating to push forward in determining how the spine-altering effects of estrogen, androgens, and the endogenous substance DHEA may be involved in these complex processes.