| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Obstetrics, Gynecology and Reproductive Sciences (N.J.M., T.H., C.L.) and Neurobiology (C.L.), Yale University School of Medicine, New Haven, Connecticut 06520; Laboratory of Molecular Neurobiology (T.H.), Biological Research Center, Hungarian Academy of Sciences, 6726 Szeged, Hungary; Neuroscience Program (J.A.J., C.L.J.), Michigan State University, East Lansing, Michigan 48824; and Department of Biomedical Sciences (N.J.M.), University of Guelph, Guelph, Ontario, Canada N1G 2W1
Address all correspondence and requests for reprints to: Neil J. MacLusky, Ph.D., Department of Biomedical Sciences, Ontario Veterinary College, University of Guelph, Guelph, Ontario, Canada N1G 2W1. E-mail: nmaclusk{at}uoguelph.ca.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The mechanisms responsible for the androgen-induced increase in CA1 PSSD remain poorly understood. Because estrogen regulates CA1 PSSD in female rats (9, 10), it was initially believed that local conversion of androgen to estrogen (11, 12) might mediate effects of circulating androgens on hippocampal structure. However, intermediate estrogen formation does not appear to be essential for androgen effects on CA1 synapse density. Thus, in male rats, estrogen treatment has no effect on CA1 PSSD (3), whereas responses to androgen are unaffected by inhibition of aromatase, the enzyme responsible for estrogen biosynthesis (1). Moreover, androgens that cannot be converted to estrogens, such as 5
-dihydrotestosterone (DHT), are effective inducers of CA1 spine synapse formation, in both sexes (2, 3).
An alternate potential mechanism is that the effects of androgen may be mediated via the androgen receptors that are abundantly expressed in hippocampal neurons, including the CA1 pyramidal cells (13, 14). To test for potential androgen receptor involvement, we examined the effects of the androgen antagonist, flutamide, which blocks the growth-promoting effects of androgens in nonneural androgen target tissues. Surprisingly, we found that flutamide itself induced hippocampal spine synapse formation, whereas in combination with DHT, the effects of flutamide on CA1 synaptogenesis were additive rather than antagonistic (1). These experiments were not definitive, however, because under some circumstances flutamide has been reported to be a partial androgen agonist (15, 16). Therefore, the lack of antagonistic effect of flutamide could indicate a different receptor-based mechanism, rather than a lack of androgen receptor involvement.
Another way to test the role of androgen receptors is to make use of genetic models in which expression of this receptor is selectively impaired. Spontaneously occurring mutations in the androgen receptor gene give rise to the syndrome of X-linked testicular feminization (Tfm). In Tfm male rats, a single amino acid substitution in the androgen receptor results in impairment of androgen responses in the brain and reproductive tract, interfering with normal masculinization of the external genitalia (17, 18, 19, 20). In the present study, we used Tfm male rats to test the potential role of androgen receptors in the hormonal regulation of CA1 PSSD. Using rigorous stereological techniques, we have determined the effects of androgen on the number of CA1 spine synapses. Our results indicate that, like the wild-type male, the Tfm male exhibits no measurable CA1 synaptic response to estrogen. Surprisingly, however, their CA1 PSSD responses to DHT treatment are also indistinguishable from those of wild-type males, despite the reduction in transcriptional efficiency of the androgen receptor resulting from the Tfm mutation.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In both experiments, rats were overdosed with an ip injection of sodium pentobarbital 2 d after the second hormone or vehicle injection and perfused transcardially with heparinized saline, followed by a fixative containing 4% paraformaldehyde and 0.1% glutaraldehyde in 0.1 M phosphate buffer (pH 7.35). The brains were removed and postfixed overnight in the same fixative without glutaraldehyde. Then, by random alternation within each treatment group, hippocampi from one side were dissected out, and 100-µm-thick horizontal vibratome sections were cut throughout the entire hippocampus. The n = approximately 100 vibratome sections per hippocampus were divided into 10 sets using systematic, uniformly random sampling.
One randomly sampled portion of sections was used for volume estimation of the CA1 stratum radiatum (CA1sr) in each hippocampus (Fig. 1
) using the Cavalieri principle (21). The CA1sr was point-counted in each sampled section (n = 10 per hippocampus) and the volume of the entire CA1sr was estimated as:
 |
 |
|
, arrows). The height of CA1sr for each of the electron microscopic sampling places was measured (Fig. 1; H1–6) and added up being equal to H(CA1sr) for each hippocampus. Our aim was to study 90 electron microscopic disectors in each hippocampus, which were taken with a height distance of: H(CA1sr)/90. Thus, our sampling method provided 180 electron micrographs for each hippocampus to evaluate.
Each sample was then trimmed for thin sectioning. Approximately four consecutive ultrathin (75 nm) sections were cut from each location, and the height sampling loci were identified. Using a magnification of approximately 11,000x, digitized electron micrographs were taken for the physical disector from adjacent ultrathin sections in a Tecnai-12 transmission electron microscope furnished with a Hamamatsu HR/HR-B CCD camera system (Hamamatsu Photonics, Hamamatsu City, Japan). Identical regions in adjacent ultrathin sections were identified using landmarks, such as myelinated fibers, that were easily recognizable in neighboring sections because of their size. The digitized electron micrographs were then printed using a laser printer. Before data analysis, the printed pictures were coded and the code was not broken until the analysis was completed. Spine synapses were counted according to the rules of the disector technique (23) within a two-dimensional unbiased counting frame with an area of 79 µm2, superimposed onto each electron micrograph. The density of spine synapses corrected for shrinkage in each CA1sr was calculated as:
 |
Qsyn is the total number of synapses for each CA1sr sampled by the disector; ks is the shrinkage correction factor; 2·90 = 180 is the number of evaluated electron micrographs per hippocampi; the section thickness t was measured by the method of Small’s smallest fold (24); and 79 is the area of the counting frame in square micrometers. To estimate the total number of synapses in each animal, the shrinkage-corrected spine synapse density Dsyn was multiplied with the total volume of CA1sr:
 |
Statistical analysis of spine synapse measurements
The approach to the determination of PSSD and total spine synapse number used here differs significantly from that employed in our previous studies (2, 3). The current methodology not only uses a measure of reference volume to allow better discrimination between effects on synapse density and total synapse number, as described above, but also increases both the region of neuropil sampled and the total number of sections counted (Fig. 1). This increased sampling density reduces the variance of measurement for each animal. This both minimizes animal use and allows for better experimental design: with fewer animals, multiple treatment groups can be processed simultaneously, thereby also reducing interanimal variance. Typically, with these methods, SDs for counting CA1 synapses are less than 5% of mean. With a SD of 5% and sample sizes of three per group, a 15% change in mean PSSD can be detected with = 0.05 and 80% power. Hence, for purposes of the present study, the minimum treatment group size was set at three animals.
Results were analyzed using Bartlett’s test for homogeneity of variance, two-way ANOVA, followed by the conservative Tukey-Kramer multiple range test for comparison of individual group means. A criterion for statistical confidence of P < 0.05 (two-tailed) was adopted.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Thus, changes observed in the CA1 PSSD can be interpreted in terms of overall changes in synapse number, rather than effects of the hormones on hippocampal volume or on fixation and embedding-associated artifacts (25).
|
. In agreement with our previous observations in Sprague Dawley male rats (3), wild-type males treated with EB exhibited no significant change in either CA1 PSSD (Fig. 2A) or the total number of spine synapses in the CA1sr (Fig. 2B), compared with vehicle-treated controls. As expected, in normal males there was an almost 100% increase in CA1 PSSD after DHT treatment, consistent with our previous findings (3). The results of estrogen treatment in Tfm males were indistinguishable from those in the wild-type male controls: EB had no significant effect on the density or number of CA1 spine synapses. Surprisingly, however, the effects of DHT were also found to be essentially identical in wild-type male and Tfm animals, with DHT inducing an approximately 100% increase in the density and number of CA1 spine synapses (Fig. 2, A and B).
|
). We have previously shown in normal Sprague Dawley rats that flutamide does not block the effects of DHT on hippocampal synaptogenesis, but instead exerts partial agonist effects (1). The effects of hydroxyflutamide were examined here to determine whether similar partial agonist responses might be observed in Tfm males. The responses of both Tfm and wild-type males to DHT were indistinguishable from the results obtained in experiment 1: in both genotypes CA1 PSSD increased approximately 2-fold after DHT treatment (Fig. 3A). Both genotypes also responded to hydroxyflutamide, which elicited a smaller (
50%) but still statistically significant increase in PSSD. As expected from the lack of the effect of the hormone treatments on reference volume of the CA1 stratum radiatum (Table 1), estimates of the total number of CA1 spine synapses revealed treatment effects that were indistinguishable from those observed in PSSD (Fig. 3B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Androgen-induced increases in CA1 PSSD represent an increase in synapse numbers
Although previous studies have extensively documented the effects of androgen on CA1 PSSD (2, 3), important questions remained to be answered. Theoretically, an increase in PSSD could result from either increased synapse formation, a decrease in the total reference volume of the tissue, or both (25). Gonadal steroids also affect glial structure (4) and hence, conceivably, hormone treatment could result in changes in the total volume of the neuropil that might result in a measured change in PSSD, without any real effect on total synapse number. In the present study, we have modified and improved the stereological procedures used previously for counting hippocampal synapses (2, 3), to provide a more complete analysis of synapse density across the CA1 stratum radiatum, as well as to determine whether these effects include an overall change in reference volume. Our results demonstrate that short-term androgen or estrogen treatment does not significantly affect total hippocampal volume (Table 1). These observations extend our previous studies in females, which showed a similar lack of effect of estrogen on hippocampal volume, despite the dramatic effect of this hormone on spine synapse density in ovariectomized rats (27). We conclude that, at least with respect to short-term androgen and estrogen treatment effects, increases in PSSD reflect overall increases in synapse number, rather than simply a change in reference tissue volume.
Tfm males do not respond to estradiol with increases in CA1 PSSD
In females, estrogen induces a marked increase in CA1 PSSD (9, 10, 26). This response is not observed in males (3). The observations presented here suggest that although the Tfm male rat fails to undergo normal phenotypic masculinization, it resembles normal males as far as the CA1 synaptic response to estrogen is concerned. This conclusion is consistent with a number of previous studies on Tfm male rodents: despite the deficiency that these animals have in androgen receptor function, they nevertheless undergo developmental androgen-induced differentiation of the brain (18, 19). The underlying mechanisms may well involve intracranial aromatization of testosterone to estradiol, a pathway that is relatively unaffected by defective androgen receptor signaling (28). Consistent with this hypothesis, using the Golgi impregnation technique, Lewis et al. (29) reported that neonatal administration of the aromatase inhibitor, letrozole, partially prevents defeminization of the CA1 dendritic response to estradiol in normal males. These observations do not conclusively establish that the lack of effect of estrogen on CA1 spine synapse density in the adult Tfm male is a result of estrogen action during development because, as we show here, the hippocampus of the Tfm male also retains the capacity to respond to DHT. If these androgen response mechanisms also exist in the developing Tfm brain, androgen-mediated effects could conceivably contribute to the loss of estrogen sensitivity observed in these animals. Additional studies will be required to clarify this issue, to determine whether selectively blocking aromatase during development may result in Tfm males that retain the capacity to exhibit increases in CA1 PSSD in response to estrogen administration.
Normal and Tfm males exhibit similar hippocampal synaptic responses to androgen
In the Tfm rat, a single base mutation results in transcription of a defective androgen receptor, with a glutamine for arginine substitution at position 734 in the ligand binding domain of the receptor (30). This reduces the binding capacity of the receptor to approximately 10–15% of normal and reduces, but does not abolish, transcriptional responses to androgen administration. The phenotypic expression of this defect includes essentially complete loss of normal developmental responses to androgen in the reproductive tract. Thus, although these animals express Müllerian-inhibiting hormone and therefore lack a female internal reproductive tract, androgen-dependent differentiation of both the external genitalia and Wolffian duct-derived structures of the male internal reproductive system, such as the seminal vesicles and prostate, fail to occur (17). Many androgen receptor-mediated responses in the central nervous system are also either completely abolished or severely impaired in Tfm males (18, 19). Nevertheless, in the present study, the effects of standard doses of DHT and hydroxyflutamide on CA1 PSSD were found to be essentially identical in Tfm and normal male animals.
Hydroxyflutamide is believed to be the active metabolite of flutamide. The results with hydroxyflutamide, in both normal and Tfm male rats are essentially identical with those we reported previously after flutamide administration to normal male and female Sprague Dawley rats (1). The circulating concentrations of hydroxyflutamide resulting from the two treatment protocols are likely to have been similar because flutamide undergoes extensive first pass metabolism in the liver, resulting in formation of the more metabolically stable hydroxylated metabolite. Hence, hydroxyflutamide is the primary circulating antiandrogen (31). In our previous study (1), rapid conversion of flutamide to its hydroxylated metabolite probably occurred, resulting in similar effects on CA1 spine synapse density in the two sets of experiments.
We have previously suggested two possible explanations for the apparently anomalous responses of CA1 spine synapse density to androgen and flutamide administration (1). Extranuclear androgen receptor variants or androgen receptor-associated coactivator proteins could contribute to the synaptic effects, altering the ligand specificity of the receptor protein. Alternatively, the effects of androgens might not be mediated via androgen receptors at all. The present data do not allow us to reject either of these hypothetical mechanisms. Persistence of androgen effects in androgen receptor-deficient animal models has been noted in a number of studies, frequently being ascribed to androgen action via conversion to estrogenic metabolites (19). Estrogen receptor-mediated responses probably do not contribute to the effects noted here, however, because estrogen does not affect CA1 PSSD in either normal or Tfm males (Fig. 2). It is possible that the defective Tfm androgen receptor could retain sufficient biological activity to elicit synaptogenic effects, even though androgen-mediated morphological changes elsewhere in the central nervous system are clearly impaired (32, 33). We have previously speculated (1) that the effects of flutamide on CA1 PSSD might reflect actions of this drug mediated via membrane-associated androgen receptors, or modulation of responses by androgen receptor coactivator proteins. In prostate cancer cells, hydroxyflutamide exerts androgen-like effects on MAPK activity (15), suggesting a mechanism other than transcriptional effects mediated via nuclear androgen receptor binding. Induction of MAPK has also been implicated in estrogen-mediated regulation of CA1 spine synapse density (34), as well as the rapid testosterone mediated activation of Sertoli cells (35). Because extranuclear androgen receptors have been observed by immunocytochemistry in axons and dendrites in regions of the rat forebrain (36), including the hippocampus (37), some of the effects of androgens on the brain could be mediated via extranuclear signaling pathways. Alterations in androgen receptor coactivator expression can also increase androgen receptor transcriptional activity, enhancing the androgen agonist properties of flutamide (16), providing another potential route via which responses to DHT and flutamide might be retained in Tfm males.
The idea that androgens might exert effects on the brain independent of the nuclear androgen receptor system is not new: several investigators over the last two decades have suggested that there may be alternate pathways of androgen action (38, 39, 40). In addition to the mechanisms described above, the possibility must also be considered that androgens could exert effects via regulation of neurotransmitter function, independent of either nuclear- or cytoplasmic androgen receptors. For example, DHT and other natural androgens can be converted in the brain to 5 androstan-3-17ß-diol, which potentiates the effects of 
-aminobutyric acid (GABA) on the GABA-benzodiazepine-chloride channel complex (40, 41, 42). Flutamide at high doses affects the same target response mechanisms: it has recently been reported that flutamide has weak benzodiazepine-like activity in pentylenetetrazole-treated mice (43). Previous studies have implicated a change in GABAergic innervation of the pyramidal cells in estrogen-induced CA1 spine synapse formation (44, 45). GABA-containing afferents originating from the septum innervate a subpopulation of the GABAergic interneurons in the hippocampus (46), consistent with the hypothesis that hormonal activation of subcortical GABAergic systems could lead to disinhibition of the pyramidal neurons. Estrogen (26) and androgen (47) effects on CA1 spine synapse density are both dependent on afferent subcortical input. Thus, a possible hypothesis to explain the effects of estrogen and androgen on hippocampal PSSD might be that these two hormones, via different mechanisms, both activate subcortical GABAergic afferents to the hippocampus, thereby reducing local inhibitory control over pyramidal cell excitability.
In summary, we have demonstrated that the effects of androgen or estrogen administration on CA1 spine synapse density and numbers in Tfm androgen receptor-deficient male rats are indistinguishable from those in wild-type male animals. These observations provide additional circumstantial support for our hypothesis (1, 48) that the mechanisms mediating gonadal steroid regulation of hippocampal spine synapse density differ significantly from those mediating the growth-promoting effects of these steroids outside the central nervous system. Hence, despite the profound reduction in androgen sensitivity observed in the peripheral androgen target organs of the Tfm male, hippocampal synaptic responses to androgen in these animals remain similar to those of normal males.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
All authors have nothing to declare.
First Published Online January 26, 2006
Abbreviations: DHT, 5 Dihydrotestosterone; EB, estradiol benzoate; GABA, -aminobutyric acid; HFL, hydroxyflutamide; ORCH, orchidectomized; PSSD, density of pyramidal cell dendritic spine synapses; Tfm, testicular feminization mutation.
Received June 6, 2005.
Accepted for publication January 19, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-reduced metabolites in the hippocampus. Behav Neurosci 118:1352–1364[CrossRef][Medline]-androstanediol. I: Modulation of sexual receptivity and promotion of GABA-stimulated chloride flux. Behav Brain Res 79:109–118[CrossRef][Medline]-androstanediol. Neuroreport 15:515–518[CrossRef][Medline]-androstanediol may mediate the effects of testosterone on conditioned place preference. Psychoneuroendocrinology 26:731–750[CrossRef][Medline] and 17ß isomers of estradiol both induce rapid spine synapse formation in the CA1 hippocampal subfield of ovariectomized female rats. Endocrinology 146:287–293This article has been cited by other articles:
 |
|
 |
|
  C. Leranth, K. Szigeti-Buck, N. J. MacLusky, and T. Hajszan Bisphenol A Prevents the Synaptogenic Response to Testosterone in the Brain of Adult Male Rats Endocrinology, March 1, 2008; 149(3): 988 - 994. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T.-V. V. Nguyen, M. Yao, and C. J. Pike Flutamide and Cyproterone Acetate Exert Agonist Effects: Induction of Androgen Receptor-Dependent Neuroprotection Endocrinology, June 1, 2007; 148(6): 2936 - 2943. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Hajszan, N. J. MacLusky, J. A. Johansen, C. L. Jordan, and C. Leranth Effects of Androgens and Estradiol on Spine Synapse Formation in the Prefrontal Cortex of Normal and Testicular Feminization Mutant Male Rats Endocrinology, May 1, 2007; 148(5): 1963 - 1967. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
