Please Scroll Down to See Forums Below 
needtogetaas
New member
Anabolic androgenic steroids (AAS) are synthetic derivatives of testosterone originally designed to promote maximal enhancement of protein synthesis and muscle growth while having minimal androgenic or masculinizing effects (for review, Shahidi, 2001). Although originally devised for therapeutic purposes, AAS are used broadly and illicitly as ergogenic agents. To date, well over 60 different AAS compounds that vary in metabolic fate and physiological effects have been designed and synthesized (for review, [Kammerer 2000] and [Shahidi 2001]). All AAS are thought to have some androgenic activity, and the androgen receptor (AR) binding properties of several of these compounds have been characterized in brain tissue (Roselli, 1998). Three main classes of AAS have been described (Fig. 1). AAS in the first commonly self-administered group are used primarily as injectable compounds and are derived from esterification of the 17β-hydroxyl group of testosterone, for example, testosterone propionate (Testex). Testosterone esters can be hydrolyzed into free testosterone, reduced to 5α-dihydrotestosterone, an androgen with higher biological activity at brain AR than testosterone (Winters 1990 S. Winters, Androgens: Endocrine physiology and pharmacology, NIDA Res Monogr 102 (1990), pp. 113–130. View Record in Scopus | Cited By in Scopus (8)[Winters 1990] and [Kochakian 1993]), or aromatized to estrogens. Molecules that have been 5α-reduced cannot be metabolized into estrogens but may be metabolized into other androgens, such as 3α-androstanediol (3α-diol) ([Winters 1990] and [Kochakian 1993]). Thus, these AAS can be metabolized to compounds that may be agonists or partial agonists at brain AR and estrogen receptors (ER), and also may be metabolized to neurosteriods that have allosteric effects at GABAA and glutamate receptors ([Clark and Henderson 2003] and [Henderson and Jorge 2004]).
Fig. 1. Chemical structures of commonly abused AAS representing the three major classes: Testosterone cypionate (class I), nandrolone decanoate (classes I and II), and 17α-MeT (methyltestosterone), stanozolol, oxymetholone, and methandrostenolone (class III).
View Within Article
19-Nortesosterone derivatives that have a substitution of a hydrogen for the methyl group at C19 comprise the second class of AAS ([Basaria et al 2001] and [Shahidi 2001]) (Fig. 1). Like the testosterone esters in class I, AAS in class II can be aromatized to 17β-estradiol, although not as well as testosterone ([Ryan 1959] and [Winters 1990]). Thus, aromatizable AAS in both class I and class II may have additional and significant CNS effects not only at the AR, but also vis-à-vis the actions of their estrogenic metabolites (Winters, 1990) at brain ER. The most commonly self-administered 19-nortestosterone derivatives include nandrolone decanoate (Deca-Durabolin) and nandrolone phenylpropionate (Durabolin).
The third class of AAS includes those compounds that are alkylated at C17. Commonly self-administered AAS in class III are 17α-methyltestosterone (17α-MeT; Metadren), oxymetholone (Anadrol-50), methandrostenolone (Dianobol), and stanozolol (Winstrol) (Fig. 1). Because alkylation retards metabolism by the liver, AAS in this group are orally active (Basaria et al., 2001). None of the 17α-alkylated steroids is converted into dihydrotestosterone or aromatized to 17β-estradiol, however 17α-alkylated AAS may be converted to other androgenic and estrogenic metabolites (for review, [Kammerer 2000] and [Clark and Henderson 2003]).
For compounds belonging to class I (testosterone esters) or esters of class II compounds (nortestosterone derivatives), the type of ester side chain alters the half-life of these compounds, but not their ultimate metabolic fate. Thus, metabolism of testosterone esters results in the formation of the same compounds that arise from endogenous testosterone. Similarly, all esters of nortestosterone form only nortestosterone derivatives ([Kammerer 2000] and [Basaria et al 2001]). For compounds in both class II and class III, however, the metabolites generated from individual AAS are complex, variable and beyond predictable categorization (Kammerer, 2000). Moreover, the ability of few of these metabolites to interact with classical AR and ER has been described, and, to our knowledge, the interactions of these metabolites with putative membrane steroid receptors or ion channels have not been examined.
Although AAS have clinically beneficial applications, most AAS are taken illicitly and in patterns that constitute abuse. The medical benefits of AAS administered by clinicians stand in sharp contrast to the potential health risks associated with the excessive doses self-administered by elite athletes and a growing number of recreational users (for review, Lukas, 1996). Importantly, in contrast to the low nanomolar levels of endogenous steroids, levels of AAS in serum have been estimated to reach micromolar concentrations in human subjects who chronically abuse these steroids and in the cerebrospinal fluid of male volunteers administered short-term doses of AAS in a clinical setting (for review, Clark and Henderson, 2003; for discussion Yang et al., 2005). Thus, while the AAS are steroids that have the capacity to interact with known neuroendocrine systems, they are unlikely to simply mimic natural androgens and estrogens due to both their distinct chemical structures and to the fact that, in steroid abusers, they are present at supraphysiological levels.
GABAA receptors in the mammalian forebrain
In both human abusers and animal models, the most prominent behavioral effects associated with AAS use are changes in aggression, anxiety and reproductive behaviors; all behaviors for which neural transmission mediated by GABA type A (GABAA) receptors in the basal forebrain play a pivotal role. In particular, GABAergic transmission in the medial preoptic area (mPOA), ventromedial nucleus (VMN) and the medial amygdala (MeA) are implicated in the expression of reproductive behaviors, aggression and anxiety (for review, [Clark and Henderson 2003] and [Henderson and Jorge 2004]). The native GABAA receptor is a pentameric ionotropic transmembrane protein for which 16 different receptor subunit genes (α1–α6, β1–β3, γ1–γ3, δ, ε, π, and θ) have been identified in mammals (for review, Henderson and Jorge, 2004). With respect to subunit composition, although the most common GABAA receptor isoform expressed in the adult brain consists of α1, β2 or β3, and γ2 subunits, α2-containing receptors predominate at all developmental ages in forebrain regions that are crucial for the expression of fear, anxiety and aggression (for review, [Clark and Henderson 2003], [Clark et al 2004] and [Henderson and Jorge 2004]). The type of α subunit expressed in a given brain region will impart differences to the GABAA receptor with respect to GABA affinity, deactivation, desensitization, and allosteric modulation by benzodiazepines, neurosteriods and the AAS (for discussion, [Yang et al 2002] and [Yang et al 2005]; see below). The basal forebrain is also noteworthy in that select GABAA receptor subunits, including the γ1 and the ε subunits, are highly expressed in it, but have limited to minimal expression elsewhere in the brain (for review, [Clark et al 2004] and [Henderson and Jorge 2004]). Of particular interest with respect to AAS actions on sexual and reproductive behaviors is the high level of expression of the ε subunit in gonadotropin-releasing hormone (GnRH) neurons in the VMN and the mPOA (Moragues et al., 2003) since GABAergic control of GnRH pulsatility is essential for pubertal onset and regular estrous cyclicity (for review, Ojeda and Urbanski, 1994). Benzodiazepines, neurosteriods and the AAS elicit significantly different effects at ε- than at γ- or δ-containing receptors ([Davies et al 1997], [Whiting et al 1997], [Thompson et al 2002], [Jones et al 2003] and [Yang et al 2005]), suggesting that inclusion of this subunit in native receptors may confer a unique profile of allosteric modulation of the GABAergic tonus to GnRH neurons.
GABAA receptor subunit gene expression during adolescence
While the popular press has highlighted AAS use by adult body builders and professional athletes, the more insidious use of AAS is among adolescents. Current assessments of children between 12 and 18 years of age indicate a usage prevalence of 2–4%; which is comparable to that for crack cocaine (Johnston et al., 2004). Adolescents are also reported to exhibit a heightened sensitivity to the effects of AAS (for review, Clark and Henderson, 2003). In spite of these facts, nearly all studies to date on AAS effects have focused on adult men or on adult male non-human subjects. As a result, vast gaps exist in our understanding of the physiological and psychological repercussions of AAS use during adolescence, especially in female subjects.
We have shown that GABAA receptor subunit mRNA expression in the basal forebrain changes with development through adolescence (McIntyre et al., 2002). Semi-quantitative reverse transcription coupled with the polymerase chain reaction (RT-PCR) analyses were made for mRNAs corresponding to the α1, α2, α5, γ1, γ2, and ε subunits in the mPOA, VMN and MeA in both male and female C57Bl/6J mice at a time corresponding to approximate onset of adolescence (postnatal day 25: PN25), a mid-point in adolescence (PN36) and in adult animals (>8 weeks). Significant changes in all subunit mRNAs were found to occur between PN25 and adulthood, although changes were not conserved among the different brain regions or between the sexes. The greatest number and greatest extent of changes in subunit mRNA levels were observed between PN25 to PN36, with fewer changes noted between PN36 and adulthood. Region-specific, as well as developmental, differences were also apparent. The mPOA showed the greatest number of significant changes in subunit mRNAs in males both between PN25 and PN36 and between PN36 and adulthood. We have not yet assessed the role of gonadal steroids in promoting these developmental changes (that is, by determining if age-dependent changes are blunted in gonadectomized subjects), but given the documented sensitivity of GABAA receptor subunit expression to steroid hormones (for review, Henderson and Jorge, 2004) and the fact that puberty is characterized by marked changes in the sensitivity of the brain to the actions of gonadal steroids (Sisk and Foster, 2004), it is likely that endogenous steroids contribute to these developmental events.
Regulation of GABAA receptor expression and function by gonadal steroids
The possibility that the AAS may elicit many of their behavioral effects via modulation of forebrain GABAergic transmission is strongly supported by extensive literature demonstrating that natural, gonadal steroids and their derivatives regulate GABAA receptor expression and function (for review, Henderson and Jorge, 2004). In particular, work from our laboratory has shown that GABAA receptor expression and function in the rodent forebrain are dependent upon hormonal state in adult animals. Using in situ hybridization and semi-quantitative RT-PCR approaches, we have shown that levels of α, γ and ε mRNAs in the mPOA and the VMN vary across the estrous cycle in Long-Evans rats and in C57Bl/6J mice ([Clark et al 1998a] and [Jorge et al 2002]). We have further shown that the mean amplitude of GABAA receptor-mediated spontaneous inhibitory synaptic currents (sIPSCs) in the mPOA and the ability of the endogenous androgenic neurosteroid, 3α-diol, to potentiate these events varied significantly across the estrous cycle and between gonadally-intact male and female C57Bl/6J mice (Jorge et al., 2002). Moreover, properties of GABAergic sIPSCs in the AR-null testicular feminized (Tfm) mouse were comparable to those observed in estrous females, suggesting that the sex- and cycle-dependent differences in GABAergic transmission arise, at least in part, from signaling mediated by the AR. Interestingly, levels of ε subunit mRNA also varied with hormonal state (e.g. highest in gonadally-intact males; lowest in diestrous females) with an inverse correlation to the ability of 3α-diol to potentiate sIPSCs (e.g. lowest in gonadally-intact males; highest in diestrous females). Furthermore, the reciprocal relationship between ε subunit mRNA expression and modulation by 3α-diol was also evident between C57Bl/6J wildtype and Tfm male mice, lending support to the hypothesis that endogenous steroids may regulate allosteric modulation of GABAA receptors in the mPOA by altering the expression of the ε subunit (Jorge et al., 2002).
Chronic AAS exposure alters GABAA receptor subunit expression and function
To determine if chronic exposure to AAS altered GABAA receptor expression and if the effects of AAS depended upon the dose of the steroid or on the age or sex of the animal, we performed semiquantitative RT-PCR analysis for α1, α2, α5, γ1, γ2, and ε subunit mRNAs in the MeA, VMN and mPOA for male and female C57Bl/6J mice given doses of the AAS, 17α-MeT (17α-methyl-4-androsten-17β-ol-3-one), that correspond to either a moderate (0.75mg/kg/day) or high (7.5mg/kg/day) human use regimen (Clark et al., 1998b). Injections were made for 4 weeks beginning either in adolescence (PN24) or in adulthood (>8 weeks) (McIntyre et al., 2002). In these, and all subsequent experiments described below, all animal care procedures were approved by the Institutional Animal Care and Use Committee at Dartmouth and are in agreement with the guidelines and recommendations of the National Institutes of Health and the American Veterinary Medical Association. All experiments were performed to minimize the numbers of animals used and their suffering. In all but one case, when we observed significant differences between oil-injected control and 17α-MeT-treated animals, the effect of the AAS was to decrease the levels of the specific subunit mRNAs. In both adolescent and adult animals, significant effects of this AAS were more commonly observed in female than in male mice. In particular, for animals treated with the moderate dose of 17α-MeT, significantly lower levels of mRNAs encoding individual α subunits were observed in treated versus control females, however no significant differences at this dose were observed for any subunit transcript in any brain region in adolescent males. In adult animals, no significant changes were detected in either sex with the moderate dose, however treatment with a concentration of 17α-MeT that reflects high doses abused by human subjects elicited decreases in specific subunit mRNA levels in both sexes. As with adolescent animals, the most prevalent changes in adults were noted for the α subunit isoforms (McIntyre et al., 2002). Results from this study demonstrated, for the first time, not only that chronic exposure to AAS led to significant alterations in the levels of selective GABAA receptor subunit mRNAs, but that the ability of the AAS to induce significant changes depended upon the dose of the AAS and on the age and the sex of the animals. In particular, for animals treated with AAS during adolescence, for those subunits whose levels were significantly changed by chronic administration of the AAS, 17α-MeT, not, vert, similar80% of these changes in mRNA levels were observed in females. Specifically, significant decreases in α2, γ1 and γ2 mRNAs were observed in the VMN and α2 mRNA in the MeA of females given a high dose and significant decreases were observed for α1 in the VMN and α2 and α5 mRNAs in the MeA of females given a moderate dose. In contrast, no significant changes in any subunit mRNA were observed in males given a moderate dose, while γ1 mRNA was significantly reduced and α5 mRNA significantly increased in males given a high dose of this AAS. These data may be of particular relevance with respect to AAS abuse in human populations since notable increases in AAS use in recent years have been reported for adolescent girls (Johnston et al., 2004).
We have recently extended the work reported in McIntyre et al. (2002) to demonstrate that AAS treatment beginning in adolescence (PN21) accentuates sex-specific differences in GABAA receptor expression and function in the medial preoptic nucleus (MPN) of C57Bl/6J mice (Penatti et al., 2005). Specifically, in animals receiving oil injections (control subjects), real time RT-PCR analyses revealed that the expression of α2 and α5 subunit mRNAs was lower and patch-clamp recording revealed that GABAA receptor-mediated sIPSCs were smaller and slower in females than in males. Three to four weeks of treatment with 17α-MeT resulted in a significant decrease in sIPSC frequency in female mice (44% of control), but no significant effect in males. Real time RT-PCR analysis indicated that treatment with 17α-MeT beginning at PN21 and extending for 6 weeks also resulted in decreased α subunit mRNA expression in female mice to 70%, 63% and 81% of control for α1, α2 and α5, respectively; differences that attained significance for the α2 subunit mRNA and the α2 subunit protein, as assessed by immunocytochemistry. In contrast, this treatment regimen had no significant effect on α subunit mRNA levels in the MPN of male mice. Six weeks of treatment also had no effect on the numbers of interneurons, as assessed by calbindin, calretinin and parvalbumin immunoreactivity. These data suggest that the diminished sIPSC frequency observed in female mice treated with 17α-MeT reflects either a change in the numbers of presynaptic terminals impinging on MPN neurons or in the efficacy of presynaptic release from those terminals, but not to an AAS-induced change in the numbers of interneurons within the MPN. Taken together, our results suggest that chronic exposure during adolescence to the AAS, 17α-MeT, promotes diminished GABAergic function in the MPN of female, but not male, mice. In doing so, this AAS enhances the sexually dimorphic nature of GABAergic transmission in the MPN, a forebrain region crucial for the expression of aggression and sexual behaviors.
In sum, our studies indicate that with respect to changes in forebrain GABAergic systems, female mice, and in particular, adolescent female mice, are more sensitive than are male mice to the effects of 17α-MeT, which has a relatively high binding affinity for the AR (for review, Clark and Henderson 2003 A.S. Clark and L.P. Henderson, Behavioral and physiological responses to anabolic-androgenic steroids, Neurosci Biobehav Rev 27 (2003), pp. 413–436. Abstract | Article | PDF (340 K) | View Record in Scopus | Cited By in Scopus (50)Clark and Henderson, 2003). It will be of interest to determine if other AAS with greater propensity to have biological actions that may be ER-mediated, such as stanozolol (Whitney and Clark, 2001), have more marked effects in male subjects. These data will have important implications for delineating the sex-specific actions of the different AAS on CNS function.
AAS-dependent changes in GABAA receptor expression: implications for receptor function
A wide range of psychoactive agents, including the benzodiazepines, ethanol, endogenous neurosteroids, and the AAS, shares, as an important mechanism of action, allosteric modulation of the GABAA receptor. The ability of these compounds to elicit acute changes in receptor function depends upon subunit composition (for review, [Clark et al 2004] and [Henderson and Jorge 2004]). Three commonly abused AAS, 17α-MeT, stanozolol and nandrolone, were found to allosterically modulate sIPSCs in a region-specific manner in the rodent brain. Specifically, we found that these AAS reversibly enhanced the amplitude and duration of sIPSCs in the VMN, but diminished amplitudes of currents in the mPOA in neonatal female Sprague–Dawley rats (Jorge-Rivera et al., 2000). In contrast, AAS had no effect on sIPSCs recorded from cerebellar Purkinje neurons of neonatal Balb/c mice (Yang et al., 2002). To determine if differences in subunit composition contribute to the region-specific differences in the ability of AAS to modulate synaptic responses, we analyzed the effects of 17α-MeT on responses elicited from recombinant GABAA receptors of known, but varied, subunit composition. We found that for responses elicited from recombinant α1β3γ2 receptors by applications of GABA that mimic synaptic conditions, 17α-MeT had no effect on peak current, time constants of deactivation, desensitization or recovery from desensitization (Yang et al., 2002). In contrast, for recombinant α2β3γ2 receptors, 17α-MeT enhanced peak current, slowed deactivation and diminished desensitization (Yang et al., 2005). These results indicate that α subunit composition regulates the ability of the AAS to modulate GABAA receptor-mediated responses, as it does for the neurosteroids and the benzodiazepines. While α subunit composition is thus important for the actions of a wide variety of allosteric modulators, our studies also have indicated that the mechanisms by which the AAS alter GABAergic currents are distinct from these other psychoactive compounds ([Yang et al 2002] and [Yang et al 2005]). We have also shown that substitution of the δ for the γ2 subunit in α2-containing receptors abrogates modulation by 17α-MeT (Yang et al., 2005), while substitution of the ε subunit reverses the modulation to inhibition (Jones et al., 2003). Thus, changes in GABAA receptor subunit composition that arise as a consequence of chronic AAS use may alter the sensitivity of the CNS to the actions of the AAS themselves, as well as other endogenous and exogenous agents that have a significant impact on anxiety, aggression and reproductive behaviors.
AAS-dependent changes in GABAA receptor expression: implications for behavior
In adult rats, the AAS, dihydrotestosterone propionate, 17α-MeT, stanozolol, and nandrolone decanoate all inhibit estrogen-induced female sexual behavior at doses that reflect high abuse concentrations in human subjects (Blasberg et al., 1998). The effects of these AAS on estrogen-induced sexual behavior require activity at the AR, since their actions can be blocked by the AR antagonist, flutamide. The neural basis for these effects is not known, but GABAergic transmission in the basal forebrain is critical for display of female sexual behaviors (for review, Clark and Henderson, 2003), and AAS-dependent, AR-mediated changes in GABAA receptor expression and subsequent function may contribute to these inhibitory effects of AAS. The results of these studies from adult animals also raise the question of how AAS might alter the neuroendocrine axis during adolescence. Changes in brain structure and function in response to endogenous steroids are known to occur well beyond the classical perinatal critical period. It is particularly noteworthy that significant developmental changes in the architecture and function of the forebrain ([Davis et al 1995] and [Spear 2000]), including changes in GABAA receptor expression (McIntyre et al., 2002) and function (Sim et al., 2000), occur during adolescence concomitant with dramatic changes in steroid hormones. Thus adolescence may represent a period of time in which the nervous system is primed for steroid-dependent changes, and exogenous use of AAS may impart effects in adolescence that would not be elicited in adulthood.
To determine if AAS have deleterious effects on neuroendocrine function during puberty, female Long-Evans rats were injected daily beginning on PN21 and continuing for 30 days with 17α-MeT (0.5–5mg/kg), stanozolol (0.05–5mg/kg) or methandrostenolone (0.5–5mg/kg). Three distinct temporal measures of the maturation of the hypothalamic–pituitary–gonadal (HPG) axis were taken: vaginal opening (as an indictor of pubertal onset), first estrus (as an indicator of maturation of the HPG axis) and estrous cyclicity (as an indicator that neuroendocrine adult patterning has been established). All three AAS advanced vaginal opening; 17α-MeT and stanozolol delayed the day of first estrus; and all three AAS suppressed estrous cyclicity (Clark et al., 2003). Interestingly, concomitant studies with the AR antagonist, flutamide, demonstrated that effects of the three AAS on pubertal landmarks were variably dependent upon AR signaling, with effects on first estrus shown to be dependent on AAS action at the AR, while effects on vaginal opening and suppression of estrous cyclicity were not ([Clark et al 2003] and [Whitney and Clark 2001]). The AR-independence of suppression of vaginal cyclicity in pubertal animals contrasts with the requirement for AR signaling for AAS suppression of behavioral receptivity in gonadectomized, estrogen-primed adults (Blasberg et al., 1998) and highlights the fact that the actions of the AAS on reproductive endpoints not only reflect complex interactions between central and peripheral targets, but also that the effects of the AAS may be quite disparate in adolescent versus adult subjects.
Conclusions
Despite two decades of well-documented studies of the influence of AAS on behavior, the effects of these synthetic steroids on brain function are only beginning to be explored. The GABAergic system is an attractive candidate for mediating many of the behavioral effects of the AAS, and data from our laboratories indicate that the AAS elicit both acute and chronic changes in this signaling system. It will be of interest to determine if some of the immediate effects noted by AAS users, such as decreases in anxiety and an enhanced sense of well-being, arise, at least in part, from allosteric enhancement of forebrain GABAergic circuits (Fig. 2). In contrast, enhanced aggression, anxiety and inhibition of reproductive behaviors may require chronic exposure and down-regulation of GABAA receptor expression via genomic mechanisms where cell-specific effects of AAS may be conferred by relative levels of expression of AR and the two ER isoforms, ERα and ERβ (Fig. 2). Additionally, the ability of the AAS to signal through membrane ER and AR has not been explored. These G protein-coupled receptors are postulated to interact with a myriad of different signaling pathways (for review, Segars and Driggers 2002 J.H. Segars and P.H. Driggers, Estrogen action and cytoplasmic signaling cascades. Part I: membrane-associated signaling complexes, Trends Endocrinol Metab 13 (2002), pp. 349–354. Abstract | Article | PDF (372 K) | View Record in Scopus | Cited By in Scopus (87)Segars and Driggers, 2002) that may have immediate effects on cell physiology, may alter the interactions of nuclear AR and ER with steroid-sensitive genes, or may lead to post-translational modification of ion channels, including the GABAA receptors (Fig. 2). Beyond delineating the different cellular mechanisms by which the AAS can alter neural function, it will also be important to determine how age, sex and the chemical characteristics of the different AAS influence the expression of non-reproductive behaviors, such as aggression, in order to fully understand the complexity of actions of these steroid compounds. Finally, while attention has mainly been drawn to the deleterious actions of AAS associated with abuse, we also need to explore the beneficial effects these steroids may have with respect ameliorating cognitive symptoms in diseases, such as fibromyalgia and chronic wasting disorders, and the decline in brain function that accompanies ageing.
Fig. 2. Illustration of a neuronal soma delineating the complexity and multiplicity of known and potential AAS actions. (I) AAS can be metabolized to estrogens, and thus have the ability to interact with not only AR, but also ERα and ERβ, to regulate gene transcription; interactions that can result in changes in GABAA receptor subunit gene expression. (II) AAS act as allosteric modulators of GABAA receptors; for example, as shown by inset of representative whole-cell current responses, potentiating currents at α2-containing receptors. Changes in neuronal activity that result from AAS-induced changes in GABAergic signaling can themselves alter gene expression, including expression of GABAA receptor subunit genes. (III) AAS are also likely to interact with heterotrimeric G protein-coupled membrane AR and ER (serpentine receptor, symbolized by the snake), although such interactions have not yet been tested, and the untested nature of these particular signaling pathways is represented by a dotted line and a question mark. Membrane ER and AR are known to interact with a wide range of signaling pathways (for review, Segars and Driggers, 2002), which may alter gene expression and may also have reciprocal interactions with signaling through GABAA receptors, including posttranslational modifications of receptors (phosphorylation, symbolized by the P), as has been shown to occur for the endogenous neurosteroids (Tasker 2000 J. Tasker, Coregulation of ion channels by neurosteriods and phosphorylation, Sci STKE 2000 (59) (2000), p. PE1. View Record in Scopus | Cited By in Scopus (4)Tasker, 2000).
Fig. 1. Chemical structures of commonly abused AAS representing the three major classes: Testosterone cypionate (class I), nandrolone decanoate (classes I and II), and 17α-MeT (methyltestosterone), stanozolol, oxymetholone, and methandrostenolone (class III).
View Within Article
19-Nortesosterone derivatives that have a substitution of a hydrogen for the methyl group at C19 comprise the second class of AAS ([Basaria et al 2001] and [Shahidi 2001]) (Fig. 1). Like the testosterone esters in class I, AAS in class II can be aromatized to 17β-estradiol, although not as well as testosterone ([Ryan 1959] and [Winters 1990]). Thus, aromatizable AAS in both class I and class II may have additional and significant CNS effects not only at the AR, but also vis-à-vis the actions of their estrogenic metabolites (Winters, 1990) at brain ER. The most commonly self-administered 19-nortestosterone derivatives include nandrolone decanoate (Deca-Durabolin) and nandrolone phenylpropionate (Durabolin).
The third class of AAS includes those compounds that are alkylated at C17. Commonly self-administered AAS in class III are 17α-methyltestosterone (17α-MeT; Metadren), oxymetholone (Anadrol-50), methandrostenolone (Dianobol), and stanozolol (Winstrol) (Fig. 1). Because alkylation retards metabolism by the liver, AAS in this group are orally active (Basaria et al., 2001). None of the 17α-alkylated steroids is converted into dihydrotestosterone or aromatized to 17β-estradiol, however 17α-alkylated AAS may be converted to other androgenic and estrogenic metabolites (for review, [Kammerer 2000] and [Clark and Henderson 2003]).
For compounds belonging to class I (testosterone esters) or esters of class II compounds (nortestosterone derivatives), the type of ester side chain alters the half-life of these compounds, but not their ultimate metabolic fate. Thus, metabolism of testosterone esters results in the formation of the same compounds that arise from endogenous testosterone. Similarly, all esters of nortestosterone form only nortestosterone derivatives ([Kammerer 2000] and [Basaria et al 2001]). For compounds in both class II and class III, however, the metabolites generated from individual AAS are complex, variable and beyond predictable categorization (Kammerer, 2000). Moreover, the ability of few of these metabolites to interact with classical AR and ER has been described, and, to our knowledge, the interactions of these metabolites with putative membrane steroid receptors or ion channels have not been examined.
Although AAS have clinically beneficial applications, most AAS are taken illicitly and in patterns that constitute abuse. The medical benefits of AAS administered by clinicians stand in sharp contrast to the potential health risks associated with the excessive doses self-administered by elite athletes and a growing number of recreational users (for review, Lukas, 1996). Importantly, in contrast to the low nanomolar levels of endogenous steroids, levels of AAS in serum have been estimated to reach micromolar concentrations in human subjects who chronically abuse these steroids and in the cerebrospinal fluid of male volunteers administered short-term doses of AAS in a clinical setting (for review, Clark and Henderson, 2003; for discussion Yang et al., 2005). Thus, while the AAS are steroids that have the capacity to interact with known neuroendocrine systems, they are unlikely to simply mimic natural androgens and estrogens due to both their distinct chemical structures and to the fact that, in steroid abusers, they are present at supraphysiological levels.
GABAA receptors in the mammalian forebrain
In both human abusers and animal models, the most prominent behavioral effects associated with AAS use are changes in aggression, anxiety and reproductive behaviors; all behaviors for which neural transmission mediated by GABA type A (GABAA) receptors in the basal forebrain play a pivotal role. In particular, GABAergic transmission in the medial preoptic area (mPOA), ventromedial nucleus (VMN) and the medial amygdala (MeA) are implicated in the expression of reproductive behaviors, aggression and anxiety (for review, [Clark and Henderson 2003] and [Henderson and Jorge 2004]). The native GABAA receptor is a pentameric ionotropic transmembrane protein for which 16 different receptor subunit genes (α1–α6, β1–β3, γ1–γ3, δ, ε, π, and θ) have been identified in mammals (for review, Henderson and Jorge, 2004). With respect to subunit composition, although the most common GABAA receptor isoform expressed in the adult brain consists of α1, β2 or β3, and γ2 subunits, α2-containing receptors predominate at all developmental ages in forebrain regions that are crucial for the expression of fear, anxiety and aggression (for review, [Clark and Henderson 2003], [Clark et al 2004] and [Henderson and Jorge 2004]). The type of α subunit expressed in a given brain region will impart differences to the GABAA receptor with respect to GABA affinity, deactivation, desensitization, and allosteric modulation by benzodiazepines, neurosteriods and the AAS (for discussion, [Yang et al 2002] and [Yang et al 2005]; see below). The basal forebrain is also noteworthy in that select GABAA receptor subunits, including the γ1 and the ε subunits, are highly expressed in it, but have limited to minimal expression elsewhere in the brain (for review, [Clark et al 2004] and [Henderson and Jorge 2004]). Of particular interest with respect to AAS actions on sexual and reproductive behaviors is the high level of expression of the ε subunit in gonadotropin-releasing hormone (GnRH) neurons in the VMN and the mPOA (Moragues et al., 2003) since GABAergic control of GnRH pulsatility is essential for pubertal onset and regular estrous cyclicity (for review, Ojeda and Urbanski, 1994). Benzodiazepines, neurosteriods and the AAS elicit significantly different effects at ε- than at γ- or δ-containing receptors ([Davies et al 1997], [Whiting et al 1997], [Thompson et al 2002], [Jones et al 2003] and [Yang et al 2005]), suggesting that inclusion of this subunit in native receptors may confer a unique profile of allosteric modulation of the GABAergic tonus to GnRH neurons.
GABAA receptor subunit gene expression during adolescence
While the popular press has highlighted AAS use by adult body builders and professional athletes, the more insidious use of AAS is among adolescents. Current assessments of children between 12 and 18 years of age indicate a usage prevalence of 2–4%; which is comparable to that for crack cocaine (Johnston et al., 2004). Adolescents are also reported to exhibit a heightened sensitivity to the effects of AAS (for review, Clark and Henderson, 2003). In spite of these facts, nearly all studies to date on AAS effects have focused on adult men or on adult male non-human subjects. As a result, vast gaps exist in our understanding of the physiological and psychological repercussions of AAS use during adolescence, especially in female subjects.
We have shown that GABAA receptor subunit mRNA expression in the basal forebrain changes with development through adolescence (McIntyre et al., 2002). Semi-quantitative reverse transcription coupled with the polymerase chain reaction (RT-PCR) analyses were made for mRNAs corresponding to the α1, α2, α5, γ1, γ2, and ε subunits in the mPOA, VMN and MeA in both male and female C57Bl/6J mice at a time corresponding to approximate onset of adolescence (postnatal day 25: PN25), a mid-point in adolescence (PN36) and in adult animals (>8 weeks). Significant changes in all subunit mRNAs were found to occur between PN25 and adulthood, although changes were not conserved among the different brain regions or between the sexes. The greatest number and greatest extent of changes in subunit mRNA levels were observed between PN25 to PN36, with fewer changes noted between PN36 and adulthood. Region-specific, as well as developmental, differences were also apparent. The mPOA showed the greatest number of significant changes in subunit mRNAs in males both between PN25 and PN36 and between PN36 and adulthood. We have not yet assessed the role of gonadal steroids in promoting these developmental changes (that is, by determining if age-dependent changes are blunted in gonadectomized subjects), but given the documented sensitivity of GABAA receptor subunit expression to steroid hormones (for review, Henderson and Jorge, 2004) and the fact that puberty is characterized by marked changes in the sensitivity of the brain to the actions of gonadal steroids (Sisk and Foster, 2004), it is likely that endogenous steroids contribute to these developmental events.
Regulation of GABAA receptor expression and function by gonadal steroids
The possibility that the AAS may elicit many of their behavioral effects via modulation of forebrain GABAergic transmission is strongly supported by extensive literature demonstrating that natural, gonadal steroids and their derivatives regulate GABAA receptor expression and function (for review, Henderson and Jorge, 2004). In particular, work from our laboratory has shown that GABAA receptor expression and function in the rodent forebrain are dependent upon hormonal state in adult animals. Using in situ hybridization and semi-quantitative RT-PCR approaches, we have shown that levels of α, γ and ε mRNAs in the mPOA and the VMN vary across the estrous cycle in Long-Evans rats and in C57Bl/6J mice ([Clark et al 1998a] and [Jorge et al 2002]). We have further shown that the mean amplitude of GABAA receptor-mediated spontaneous inhibitory synaptic currents (sIPSCs) in the mPOA and the ability of the endogenous androgenic neurosteroid, 3α-diol, to potentiate these events varied significantly across the estrous cycle and between gonadally-intact male and female C57Bl/6J mice (Jorge et al., 2002). Moreover, properties of GABAergic sIPSCs in the AR-null testicular feminized (Tfm) mouse were comparable to those observed in estrous females, suggesting that the sex- and cycle-dependent differences in GABAergic transmission arise, at least in part, from signaling mediated by the AR. Interestingly, levels of ε subunit mRNA also varied with hormonal state (e.g. highest in gonadally-intact males; lowest in diestrous females) with an inverse correlation to the ability of 3α-diol to potentiate sIPSCs (e.g. lowest in gonadally-intact males; highest in diestrous females). Furthermore, the reciprocal relationship between ε subunit mRNA expression and modulation by 3α-diol was also evident between C57Bl/6J wildtype and Tfm male mice, lending support to the hypothesis that endogenous steroids may regulate allosteric modulation of GABAA receptors in the mPOA by altering the expression of the ε subunit (Jorge et al., 2002).
Chronic AAS exposure alters GABAA receptor subunit expression and function
To determine if chronic exposure to AAS altered GABAA receptor expression and if the effects of AAS depended upon the dose of the steroid or on the age or sex of the animal, we performed semiquantitative RT-PCR analysis for α1, α2, α5, γ1, γ2, and ε subunit mRNAs in the MeA, VMN and mPOA for male and female C57Bl/6J mice given doses of the AAS, 17α-MeT (17α-methyl-4-androsten-17β-ol-3-one), that correspond to either a moderate (0.75mg/kg/day) or high (7.5mg/kg/day) human use regimen (Clark et al., 1998b). Injections were made for 4 weeks beginning either in adolescence (PN24) or in adulthood (>8 weeks) (McIntyre et al., 2002). In these, and all subsequent experiments described below, all animal care procedures were approved by the Institutional Animal Care and Use Committee at Dartmouth and are in agreement with the guidelines and recommendations of the National Institutes of Health and the American Veterinary Medical Association. All experiments were performed to minimize the numbers of animals used and their suffering. In all but one case, when we observed significant differences between oil-injected control and 17α-MeT-treated animals, the effect of the AAS was to decrease the levels of the specific subunit mRNAs. In both adolescent and adult animals, significant effects of this AAS were more commonly observed in female than in male mice. In particular, for animals treated with the moderate dose of 17α-MeT, significantly lower levels of mRNAs encoding individual α subunits were observed in treated versus control females, however no significant differences at this dose were observed for any subunit transcript in any brain region in adolescent males. In adult animals, no significant changes were detected in either sex with the moderate dose, however treatment with a concentration of 17α-MeT that reflects high doses abused by human subjects elicited decreases in specific subunit mRNA levels in both sexes. As with adolescent animals, the most prevalent changes in adults were noted for the α subunit isoforms (McIntyre et al., 2002). Results from this study demonstrated, for the first time, not only that chronic exposure to AAS led to significant alterations in the levels of selective GABAA receptor subunit mRNAs, but that the ability of the AAS to induce significant changes depended upon the dose of the AAS and on the age and the sex of the animals. In particular, for animals treated with AAS during adolescence, for those subunits whose levels were significantly changed by chronic administration of the AAS, 17α-MeT, not, vert, similar80% of these changes in mRNA levels were observed in females. Specifically, significant decreases in α2, γ1 and γ2 mRNAs were observed in the VMN and α2 mRNA in the MeA of females given a high dose and significant decreases were observed for α1 in the VMN and α2 and α5 mRNAs in the MeA of females given a moderate dose. In contrast, no significant changes in any subunit mRNA were observed in males given a moderate dose, while γ1 mRNA was significantly reduced and α5 mRNA significantly increased in males given a high dose of this AAS. These data may be of particular relevance with respect to AAS abuse in human populations since notable increases in AAS use in recent years have been reported for adolescent girls (Johnston et al., 2004).
We have recently extended the work reported in McIntyre et al. (2002) to demonstrate that AAS treatment beginning in adolescence (PN21) accentuates sex-specific differences in GABAA receptor expression and function in the medial preoptic nucleus (MPN) of C57Bl/6J mice (Penatti et al., 2005). Specifically, in animals receiving oil injections (control subjects), real time RT-PCR analyses revealed that the expression of α2 and α5 subunit mRNAs was lower and patch-clamp recording revealed that GABAA receptor-mediated sIPSCs were smaller and slower in females than in males. Three to four weeks of treatment with 17α-MeT resulted in a significant decrease in sIPSC frequency in female mice (44% of control), but no significant effect in males. Real time RT-PCR analysis indicated that treatment with 17α-MeT beginning at PN21 and extending for 6 weeks also resulted in decreased α subunit mRNA expression in female mice to 70%, 63% and 81% of control for α1, α2 and α5, respectively; differences that attained significance for the α2 subunit mRNA and the α2 subunit protein, as assessed by immunocytochemistry. In contrast, this treatment regimen had no significant effect on α subunit mRNA levels in the MPN of male mice. Six weeks of treatment also had no effect on the numbers of interneurons, as assessed by calbindin, calretinin and parvalbumin immunoreactivity. These data suggest that the diminished sIPSC frequency observed in female mice treated with 17α-MeT reflects either a change in the numbers of presynaptic terminals impinging on MPN neurons or in the efficacy of presynaptic release from those terminals, but not to an AAS-induced change in the numbers of interneurons within the MPN. Taken together, our results suggest that chronic exposure during adolescence to the AAS, 17α-MeT, promotes diminished GABAergic function in the MPN of female, but not male, mice. In doing so, this AAS enhances the sexually dimorphic nature of GABAergic transmission in the MPN, a forebrain region crucial for the expression of aggression and sexual behaviors.
In sum, our studies indicate that with respect to changes in forebrain GABAergic systems, female mice, and in particular, adolescent female mice, are more sensitive than are male mice to the effects of 17α-MeT, which has a relatively high binding affinity for the AR (for review, Clark and Henderson 2003 A.S. Clark and L.P. Henderson, Behavioral and physiological responses to anabolic-androgenic steroids, Neurosci Biobehav Rev 27 (2003), pp. 413–436. Abstract | Article | PDF (340 K) | View Record in Scopus | Cited By in Scopus (50)Clark and Henderson, 2003). It will be of interest to determine if other AAS with greater propensity to have biological actions that may be ER-mediated, such as stanozolol (Whitney and Clark, 2001), have more marked effects in male subjects. These data will have important implications for delineating the sex-specific actions of the different AAS on CNS function.
AAS-dependent changes in GABAA receptor expression: implications for receptor function
A wide range of psychoactive agents, including the benzodiazepines, ethanol, endogenous neurosteroids, and the AAS, shares, as an important mechanism of action, allosteric modulation of the GABAA receptor. The ability of these compounds to elicit acute changes in receptor function depends upon subunit composition (for review, [Clark et al 2004] and [Henderson and Jorge 2004]). Three commonly abused AAS, 17α-MeT, stanozolol and nandrolone, were found to allosterically modulate sIPSCs in a region-specific manner in the rodent brain. Specifically, we found that these AAS reversibly enhanced the amplitude and duration of sIPSCs in the VMN, but diminished amplitudes of currents in the mPOA in neonatal female Sprague–Dawley rats (Jorge-Rivera et al., 2000). In contrast, AAS had no effect on sIPSCs recorded from cerebellar Purkinje neurons of neonatal Balb/c mice (Yang et al., 2002). To determine if differences in subunit composition contribute to the region-specific differences in the ability of AAS to modulate synaptic responses, we analyzed the effects of 17α-MeT on responses elicited from recombinant GABAA receptors of known, but varied, subunit composition. We found that for responses elicited from recombinant α1β3γ2 receptors by applications of GABA that mimic synaptic conditions, 17α-MeT had no effect on peak current, time constants of deactivation, desensitization or recovery from desensitization (Yang et al., 2002). In contrast, for recombinant α2β3γ2 receptors, 17α-MeT enhanced peak current, slowed deactivation and diminished desensitization (Yang et al., 2005). These results indicate that α subunit composition regulates the ability of the AAS to modulate GABAA receptor-mediated responses, as it does for the neurosteroids and the benzodiazepines. While α subunit composition is thus important for the actions of a wide variety of allosteric modulators, our studies also have indicated that the mechanisms by which the AAS alter GABAergic currents are distinct from these other psychoactive compounds ([Yang et al 2002] and [Yang et al 2005]). We have also shown that substitution of the δ for the γ2 subunit in α2-containing receptors abrogates modulation by 17α-MeT (Yang et al., 2005), while substitution of the ε subunit reverses the modulation to inhibition (Jones et al., 2003). Thus, changes in GABAA receptor subunit composition that arise as a consequence of chronic AAS use may alter the sensitivity of the CNS to the actions of the AAS themselves, as well as other endogenous and exogenous agents that have a significant impact on anxiety, aggression and reproductive behaviors.
AAS-dependent changes in GABAA receptor expression: implications for behavior
In adult rats, the AAS, dihydrotestosterone propionate, 17α-MeT, stanozolol, and nandrolone decanoate all inhibit estrogen-induced female sexual behavior at doses that reflect high abuse concentrations in human subjects (Blasberg et al., 1998). The effects of these AAS on estrogen-induced sexual behavior require activity at the AR, since their actions can be blocked by the AR antagonist, flutamide. The neural basis for these effects is not known, but GABAergic transmission in the basal forebrain is critical for display of female sexual behaviors (for review, Clark and Henderson, 2003), and AAS-dependent, AR-mediated changes in GABAA receptor expression and subsequent function may contribute to these inhibitory effects of AAS. The results of these studies from adult animals also raise the question of how AAS might alter the neuroendocrine axis during adolescence. Changes in brain structure and function in response to endogenous steroids are known to occur well beyond the classical perinatal critical period. It is particularly noteworthy that significant developmental changes in the architecture and function of the forebrain ([Davis et al 1995] and [Spear 2000]), including changes in GABAA receptor expression (McIntyre et al., 2002) and function (Sim et al., 2000), occur during adolescence concomitant with dramatic changes in steroid hormones. Thus adolescence may represent a period of time in which the nervous system is primed for steroid-dependent changes, and exogenous use of AAS may impart effects in adolescence that would not be elicited in adulthood.
To determine if AAS have deleterious effects on neuroendocrine function during puberty, female Long-Evans rats were injected daily beginning on PN21 and continuing for 30 days with 17α-MeT (0.5–5mg/kg), stanozolol (0.05–5mg/kg) or methandrostenolone (0.5–5mg/kg). Three distinct temporal measures of the maturation of the hypothalamic–pituitary–gonadal (HPG) axis were taken: vaginal opening (as an indictor of pubertal onset), first estrus (as an indicator of maturation of the HPG axis) and estrous cyclicity (as an indicator that neuroendocrine adult patterning has been established). All three AAS advanced vaginal opening; 17α-MeT and stanozolol delayed the day of first estrus; and all three AAS suppressed estrous cyclicity (Clark et al., 2003). Interestingly, concomitant studies with the AR antagonist, flutamide, demonstrated that effects of the three AAS on pubertal landmarks were variably dependent upon AR signaling, with effects on first estrus shown to be dependent on AAS action at the AR, while effects on vaginal opening and suppression of estrous cyclicity were not ([Clark et al 2003] and [Whitney and Clark 2001]). The AR-independence of suppression of vaginal cyclicity in pubertal animals contrasts with the requirement for AR signaling for AAS suppression of behavioral receptivity in gonadectomized, estrogen-primed adults (Blasberg et al., 1998) and highlights the fact that the actions of the AAS on reproductive endpoints not only reflect complex interactions between central and peripheral targets, but also that the effects of the AAS may be quite disparate in adolescent versus adult subjects.
Conclusions
Despite two decades of well-documented studies of the influence of AAS on behavior, the effects of these synthetic steroids on brain function are only beginning to be explored. The GABAergic system is an attractive candidate for mediating many of the behavioral effects of the AAS, and data from our laboratories indicate that the AAS elicit both acute and chronic changes in this signaling system. It will be of interest to determine if some of the immediate effects noted by AAS users, such as decreases in anxiety and an enhanced sense of well-being, arise, at least in part, from allosteric enhancement of forebrain GABAergic circuits (Fig. 2). In contrast, enhanced aggression, anxiety and inhibition of reproductive behaviors may require chronic exposure and down-regulation of GABAA receptor expression via genomic mechanisms where cell-specific effects of AAS may be conferred by relative levels of expression of AR and the two ER isoforms, ERα and ERβ (Fig. 2). Additionally, the ability of the AAS to signal through membrane ER and AR has not been explored. These G protein-coupled receptors are postulated to interact with a myriad of different signaling pathways (for review, Segars and Driggers 2002 J.H. Segars and P.H. Driggers, Estrogen action and cytoplasmic signaling cascades. Part I: membrane-associated signaling complexes, Trends Endocrinol Metab 13 (2002), pp. 349–354. Abstract | Article | PDF (372 K) | View Record in Scopus | Cited By in Scopus (87)Segars and Driggers, 2002) that may have immediate effects on cell physiology, may alter the interactions of nuclear AR and ER with steroid-sensitive genes, or may lead to post-translational modification of ion channels, including the GABAA receptors (Fig. 2). Beyond delineating the different cellular mechanisms by which the AAS can alter neural function, it will also be important to determine how age, sex and the chemical characteristics of the different AAS influence the expression of non-reproductive behaviors, such as aggression, in order to fully understand the complexity of actions of these steroid compounds. Finally, while attention has mainly been drawn to the deleterious actions of AAS associated with abuse, we also need to explore the beneficial effects these steroids may have with respect ameliorating cognitive symptoms in diseases, such as fibromyalgia and chronic wasting disorders, and the decline in brain function that accompanies ageing.
Fig. 2. Illustration of a neuronal soma delineating the complexity and multiplicity of known and potential AAS actions. (I) AAS can be metabolized to estrogens, and thus have the ability to interact with not only AR, but also ERα and ERβ, to regulate gene transcription; interactions that can result in changes in GABAA receptor subunit gene expression. (II) AAS act as allosteric modulators of GABAA receptors; for example, as shown by inset of representative whole-cell current responses, potentiating currents at α2-containing receptors. Changes in neuronal activity that result from AAS-induced changes in GABAergic signaling can themselves alter gene expression, including expression of GABAA receptor subunit genes. (III) AAS are also likely to interact with heterotrimeric G protein-coupled membrane AR and ER (serpentine receptor, symbolized by the snake), although such interactions have not yet been tested, and the untested nature of these particular signaling pathways is represented by a dotted line and a question mark. Membrane ER and AR are known to interact with a wide range of signaling pathways (for review, Segars and Driggers, 2002), which may alter gene expression and may also have reciprocal interactions with signaling through GABAA receptors, including posttranslational modifications of receptors (phosphorylation, symbolized by the P), as has been shown to occur for the endogenous neurosteroids (Tasker 2000 J. Tasker, Coregulation of ion channels by neurosteriods and phosphorylation, Sci STKE 2000 (59) (2000), p. PE1. View Record in Scopus | Cited By in Scopus (4)Tasker, 2000).
