ELEVATED TESTOSTERONE INDUCES APOPTOSIS IN NEURONAL CELLS

contrast, knock down of InsP 3 R type 3 neither modified the apoptotic response nor the Ca 2+ signals. These results support our hypothesis that elevated testosterone alters InsP 3 R type 1-mediated intracellular Ca 2+ signaling and that the prolonged Ca 2+ signals lead to apoptotic cell death. These effects of testosterone on neurons will have long-term effects on brain function.


INTRODUCTION
Neurosteroids have been implicated as components essential for the normal function of the central nervous system (CNS) (1)(2)(3)(4). The gonadal steroid hormones are required for reproductive function, but androgens also affect areas of the brain that are not primarily involved in reproduction such as the hippocampus (5), preoptic area, amygdala and medial hypothalamic area (6). At physiological levels, androgens are involved in neuronal differentiation, neuroprotection, neuronal survival and development (7)(8)(9). These responses occur slowly (over hours) and are mediated through the intracellular androgen receptor (AR). In the developing brain, androgens are capable of changing the ultra structural characteristics of the neuronal plasma membrane with a relatively fast pace (10,11). Recently, we have shown that nanomolar levels of testosterone induce rapid intracellular Ca 2+ increases in neuroblastoma cells (within seconds), which begin as Ca 2+ transients in the cytosol, propagate as waves of Ca 2+ in the cytoplasm and nucleus, and develop into an oscillatory pattern (12). These Ca 2+ signals depend on an interplay between Ca 2+ efflux from the endoplasmic reticulum (ER) through inositol 1,4,5 trisphosphate-sensitive Ca 2+ release channels (InsP 3 R) and Ca 2+ reuptake into the ER by Ca 2+ pumps. This new testosterone-induced pathway in neuronal cells leads to neurite outgrowth (12), an essential event in neuronal differentiation (13). These results suggest an important physiological mechanism for the action of testosterone in neurons at physiological concentrations. However, it is unknown how these cells respond to high plasma levels of this neurosteroid, generally administrated exogenously to achieve an increase in muscle mass (14,15) or for replacement therapy (16). In vivo administration of large doses of androgens has been Testosterone-induced neuronal apoptosis correlated with neurobehavioral changes like hyper-excitability, supra-aggressive nature and suicidal tendencies (17,18). These behavioral changes could be the outward manifestation of neuronal damage resulting from exposure to high concentrations of testosterone.
In this investigation, we evaluated the hypothesis that high concentrations of testosterone can induce deleterious effects in neurons. We show that high levels of testosterone initiate an apoptotic program in neuroblastoma cells. Apoptosis is the normal and controlled process of cell death (19). Precise control of apoptosis is critical in many processes of life, including development and disease. The onset of apoptosis and its progression can be altered by both endogenous and exogenous insults (20). Apoptosis is characterized by many physical, cellular and physiological parameters which include membrane blebbing, caspase activation, change in membrane potential and DNA damage-associated nuclear deformation (19,21,22). Prolonged elevated cytosolic calcium (Ca 2+ ) concentrations can initiate the apoptotic program in many cell types (23,24) including neurons (25). The testosterone-induced apoptosis described here occurs through overactivation of intracellular Ca 2+ signaling pathways. The progression into cell death could be attenuated by addition of inhibitors of the InsP 3 R signaling pathway. Moreover, we found that InsP 3 R type 1, the predominant isoform of the InsP 3 R found in neurons (26), is the key integrator of the testosterone response in both normal and hyperstimulated Ca 2+ signaling in neuroblastoma cells. It is this pathway that has been implicated in other pathophysiological conditions. For example, over-stimulation of the apoptotic program in neurons has been associated with several neurological illnesses such as Alzheimer disease (27,28) and Huntington's disease (29)(30)(31).
Our results suggest that the responses to elevated testosterone can be compared to these pathophysiological conditions.

Cell cultures:
The human neuroblastoma cell line (SH-SY5Y; ATCC) was cultured in 1:1 DMEM:Ham's F12 medium supplemented with 10% (v/v) heat inactivated fetal bovine serum, 5% non-essential aminoacids, 100 U penicillin and 50 µg/mL streptomycin, in a 95% air-5% CO 2 humidified incubator at 37°C. Cells were grown on 22 mm X 22 mm gelatin-coated glass coverslips for the Ca 2+ measurements or in 100-mm petri-dishes for biochemical and spectrochemical assays. Cultured cells were washed once with PBS before stimulation with testosterone (from concentrated stocks made in ethanol). The final ethanol concentration (<0.01%) had no effect on intracellular Ca 2+ concentration or biochemical determinations (32) Cell viability: Cells were exposed to various concentrations of steroid hormones for different times. The number of viable cells was determined using a standard trypan blue exclusion assay. Cells were washed twice with PBS and then resuspended with EDTA. Cells were stained with 0.5% trypan blue, and the number of viable and non-viable cells was determined in a haemacytometer. Results are expressed as percent viability (100 x number of viable cells/number of total cells per well) for each concentration of hormone tested. Cell viability was also assessed using the TOX-2 kit (Sigma) according to the manufacturer's protocol. This method determines the ability of metabolically active cells to reduce the yellow salt 3-(4,5-dimethylthiazol-2-yl)-2,5-difenyl tetrazolium bromide (XTT) to an orange formazan dye. Therefore, the conversion only occurs in living cells and the amount of orange formazan formed directly correlates to the number of living cells. Neuroblastoma cells, grown in 60 mm petri dishes in culture medium without phenol red and exposed to different steroid hormone concentrations, were washed with PBS and incubated with the XTT solution (final concentration 0.3 mg/ml), according to the kit specifications. After this incubation period, quantification of formazan dye formed was determined using a spectrophotometer at a wavelength of 450 nm minus the absorbance at 690 nm. Results are expressed as percentages with respect to controls (untreated cells). Ca 2+ Imaging: Cells were loaded with 5 µM Fluo4-AM at 37°C for 30 min in "imaging solution" (in mM): 135 NaCl, 5 KCl, 2 MgCl 2 , 2 CaCl 2 , 10 HEPES, 5.6 mM Glucose, pH 7.4. After loading, cells were washed twice with the imaging solution and placed on an inverted microscope (Axiovert 100, Zeiss) connected to a laser scanning imaging system (LSM 510 META, Zeiss). Cells were stimulated with the hormone diluted in the imaging solution. A 488 nm excitation wavelength was focused through a 40X Apochromat water immersion objective lens (NA 1.2, Zeiss). In order to identify the cells expressing the siRNA, cells were cotransfected with DsRed-pCMV-cyto construct, which produced red fluorescence protein (DSR), and were examined using fluorescence excitation at 568 nm. Regions of interest (ROI) with the same pixel dimensions were identified and analyzed using Image J software (NIH, Bethesda, MA). The inhibitors were added during the dye incubation; times and concentrations are indicated in Results. The fluorescence intensity ratio (F/Fo) was plotted as a function of time.
Annexin-V labeling: Cells were incubated in different concentrations of testosterone for 6 h. After treatment, cells were washed twice with PBS and once with fixation buffer (10 mM HEPES/NaOH, pH 7.4, 140 mM NaCl, 5 mM CaCl 2 ) and were incubated with FITCconjugated Annexin-V (0.5 µl per 100 µl of fixation buffer; BD Biosciences) for 15 min at room temperature. The fluorescence of FITC was visualized by laser scanning confocal microscopy (LSM 510 META, Zeiss). Annexin-V fluorescence positive cells were analyzed by the Image J program using the particle analysis macro (NIH, Bethesda) and normalized to total cell count.
DNA fragmentation assay: Neuroblastoma cells (1 x 10 6 cells) growing on 100 mm plates were washed once in PBS; harvested in 0.5 ml of extraction buffer (50 mM Tris-HCl pH 8.0, 10 mM EDTA, 0.5% SDS, 0.5 mg/ml proteinase K) and were incubated at 55°C for 30 min in presence of 20 mg/ml RNAse A. The lysate was centrifuged at 15,000 g for 10 min and aqueous fractions were extracted with phenol/chloroform two to three times. Nucleic acids were precipitated with 0.1 vol 3 M sodium acetate and 2.5 vol 100% ethanol. Pelleted DNA was washed with 70% ethanol and was resuspended in TE buffer (10 mM Tris-HCl, 1 mM EDTA pH 8.0) kept at 37°C. DNA was run on a 2% agarose gel containing ethidium bromide, visualized with a UV transilluminator, recorded with a UVP gel-doc system, and analyzed by scanning densitometry.
Assay of proteolytic activity of caspases: Six hours after incubation with 100 nM to 10 µM testosterone, 1 x 10 6 neuroblastoma cells were washed with PBS and collected in cell lysis buffer (50 mM HEPES, pH 7.5, containing 10% sucrose and 1% Triton X-100). Following removal of cell debris and mitochondria from the lysate by centrifugation, the supernatant was diluted 2-fold with PBS, and incubated at 37°C for 30 min in the presence of 10 mM DTT. Ac-DEVD-AFC (at the final concentration of 15 µM), dissolved in DMSO, was added to the mixture and the sample was incubated for another 30 min. The mixture was then transferred to a quartz 1 cm square cuvette and AFC fluorescence (excitation: 400 nm, and emission: 505 nm) was measured (Quantamaster spectrofluorometer, Photon Technology International, NJ). For colorimetric measurement of caspase activity, Ac-DEVD-pNA (Calbiochem) was added at a final concentration of 200 µM to lysates generated from identical number of cells under different experimental conditions. Caspase activity was monitored as optical absorbance at 405 nm (Spectronic Genesys2 spectrophotometer) in a time dependent manner. In the caspase inhibition experiments, 200 nM of Ac-DEVD-CHO (BD Biosciences-Pharmingen) was added to the mixture prior to Ac-DEVD-AFC or Ac-DEVD-pNA for 30 min.
Western blot: Cells lysates containing 40 µg of protein were separated by SDS-PAGE in a 4-20% linear gradient followed by electrophoretic transfer onto PVDF membranes for 3 h at 400 mA. The following primary antibodies and their dilution were used: anti-InsP 3 R type 1 (1:2,000; custom produced by Research Genetics), anti-InsP 3 R type 3 (1:1,000; BD Biosciences), ERK2 (1:1,000, Santa Cruz). Membranes were incubated with primary antibodies overnight at 4°C. After incubation with horseradish peroxidase-conjugated secondary antibodies (1:10,000; BioRad) for 2 h at room temperature, the bands were visualized by enhanced chemiluminescence system (Pierce). Membranes were stripped and re-probed with ERK2 antibody which was used as a sample loading control. Blots were quantified by scanning densitometry.
siRNA:. siRNA for InsP 3 R type 1 and type 3 were kindly provided by Dr. M. Nathanson (Dept of Internal Medicine, Yale University). The siRNA templates were obtained from Ambion and the sequence was AAAGCACCAGCAGCTACAACTCCTGTCT C of the human InsP 3 R type 1 and AAACAAGTTTGCCAGCACCAT for InsP 3 R type 3. Transfection of cells with siRNA was performed using RNAiFect (Qiagen) and downregulation of either type 1 or type 3 InsP 3 R was confirmed by immunofluorescence and western blot analysis.
Statistics: Data are expressed as the mean ± SEM. Colorimetrical measurements of caspase activity were fit to a linear equation to estimate the rate of activation. In order to compare the difference between basal and post-stimulated points, we carried out ANOVA and the statistic differences were determined by Bonferroni posttest. P < 0.05 was considered statistically significant.

High concentration of testosterone induces neurotoxicity
Previously we found that physiological concentrations of testosterone (nanomolar range) induced neuronal differentiation in neuroblastoma cells (12), but it was not clear that supra-physiological levels of this hormone (micromolar range) would have the same effect. Therefore, we treated neuroblastoma cells with different concentrations of testosterone and compared the effect on several parameters, including cell survival. When treated with low concentrations of testosterone (100 nM), cell viability did not change to any appreciable extent over the time range studied (Fig. 1A). Exposure of the cells to 1 µM testosterone induced a significant decrease in cell viability over 24 h. At 10 µM testosterone was even more lethal (Fig. 1A). To confirm the results determined by the trypan blue exclusion assay, we monitored the effect of testosterone using the XTT cell viability assay. The response after 12 h of treatment was selected for comparison because the higher concentrations of testosterone (1 µM and 10µM) induced a significant deleterious effect at this time point using the previous assay. As predicted, 100 nM testosterone did not affect cell viability whereas 1 µM and 10 µM testosterone significantly decreased cell viability (Fig. 1B). Using a similar range of concentrations of estrogen (17β-estradiol), we saw no effect on cell viability, indicating that the decrease in cell viability was specifically triggered by high concentrations of testosterone (Fig 1 B).

High concentrations of testosterone induce apoptosis
Cell viability as determined using either trypan blue exclusion or XTT methods does not discriminate between cell death by apoptosis or necrosis. To identify the mechanism of cell death induced by testosterone, we examined annexin-V labeling, caspase activation, and DNA laddering, all markers of apoptosis. Untreated cells did not show fluorescence due to annexin-V labeling ( Fig. 2A) whereas staurosporine (1 µM), a well-established inducer of apoptosis, produced high levels of annexin-V positive cells, depicted in green (Fig. 2B). In the bright-field image it is possible to observe the morphological changes induced by staurosporine ( Fig. 2B; superimposed images are shown in the right panel). When neuroblastoma cells were exposed to 100 nM testosterone, very few annexin-V positive cells were detected after 6 h of incubation ( Fig. 2C; 4 ± 1 %, n = 780 cells). Elevated concentrations of testosterone significantly increased the number of annexin-V positive cells: 21 ± 3 % for 1 µM (n = 834 cells; p<0.05) and 34 ± 8 for 10 µM (n = 600 cells; p<0.01) testosterone were annexin-V positive ( Fig. 2 D, E) where p values are calculated for treatment as compared with the control. These percents for apoptotic cell death are consistent with the number of dead cells identified by the cell viability assay. Moreover, bright-field images also show morphological changes associated with apoptosis; cells have plasma membrane blebs and were more round in shape, with loss of their processes (Fig. 2 D, E). To further confirm the specificity of the response to testosterone, cells were exposed to high concentrations of estrogen (1 µM -10 µM). There was no significant labeling of annexin-V detected (data not shown), indicating that the apoptotic death pathway was specifically activated by testosterone.
Apoptotic cell death can be defined by morphological and biochemical characteristics such as DNA fragmentation (33) that can be visualized on an agarose gel. We observed that 1 µM and 10 µM testosterone induced considerable DNA damage in neuroblastoma cells within 6 hours (Fig. 3A, lanes 4, 5) whereas treatment with 100 nM testosterone for the same duration did not induce DNA damage (Fig. 3A, lane 3). To be certain that the death of neuroblastoma cells is indeed due to a canonical apoptotic pathway, it is necessary to show that testosterone activates aspartate-specific proteases (caspases) (34). Pretreatment of neuroblastoma cells with the cell permeable caspase inhibitor (Ac-DEVD-CHO with 16 Nterminal amino acid residues of the signal peptide of Kaposi fibroblast growth factor that confers cell permeability to the peptide) abolishes the testosterone effect on cell viability at both concentrations tested earlier (Fig. 3B). To directly follow caspase activity, the hydrolysis of a specific fluorogenic substrate, Ac-DEVD-AFC, in cell-free extracts prepared from neuroblastoma cells was recorded after cells were treated with testosterone. The steadystate fluorescence emission spectrum of the mitochondria-free extract (Fig. 3C), recorded after the addition of Ac-DEVD-AFC, indicates hydrolysis of the substrate and release of the AFC fluorophore. Prior addition of the caspase inhibitor, Ac-DEVD-CHO, competitively inhibits AFC release (Fig. 3C) suggesting the specific activation of aspartate protease(s). The cell-free extracts prepared from untreated neuroblastoma cells produced no caspase activity. In addition, the time course of caspase activation as a function of increasing concentrations of testosterone was monitored by using a colorimetric caspase substrate Ac-DEVD-pNA. There was an increase in caspase activity in cells treated with either 1 µM or 10 µM testosterone whereas treatment with 100 nM testosterone elicited caspase activity comparable with untreated cells (Fig. 3D). All three characteristic parameters, annexin-V labeling, DNA fragmentation and caspase activation, were induced by treatment with testosterone supporting the conclusion that cells treated with high levels of testosterone (micromolar concentrations) die by initiation of the apoptotic cascade.

High concentrations of testosterone increases intracellular Ca 2+
An incremental rise in cytosolic Ca 2+ has been suggested as a critical step in activating neuronal apoptosis (22). We examined the testosteroneinduced Ca 2+ responses in neuroblastoma cells by measuring Ca 2+ changes at the single cell level in cells loaded with the Ca 2+ -sensitive fluorescent dye, Fluo4-AM. Stimulation of neuroblastoma cells with different testosterone concentrations (100 nM, 1 µM, and 10 µM) induced Ca 2+ responses with different temporal patterns (Fig. 4). At 100 nM testosterone, increases in Ca 2+ , which were absent at the time of hormone application, were observed and the level of Ca 2+ returned to the basal level between repetitive oscillations ( Fig. 4A; n=70 cells from 5 independent cultures). We previously determined that Ca 2+ signals evoked by testosterone were dependent on the generation of InsP 3 as well as on the activity of the InsP 3 R (12). At higher concentrations of testosterone, the response was more irregular, often with a persistent Ca 2+ rise which was maintained as long as hormone was present in the medium. At 1 µM, an initial rise was observed that was followed by a slow return to baseline (Fig. 4B, n=83 cells, from 4 independent cultures), whereas at 10 µM, long-lasting Ca 2+ increase were seen where the return to basal level occurred only after several minutes (Fig. 4C, n=91 cells, from 4 independent cultures). The duration of the Ca 2+ spike increased as the hormone concentration was raised (Fig. 4D), showing a change from repetitive Ca 2+ spikes to a prolonged Ca 2+ transient. The peak amplitudes of the Ca 2+ transients were conserved (Fig. 4). These results demonstrate a loss of the normal oscillatory Ca 2+ pattern when testosterone levels are elevated.
The ability of testosterone to induce either Ca 2+ oscillations or sustained Ca 2+ responses was attenuated by InsP 3 R inhibitors. When the cells where pre-incubated with either 2-APB (20 µM) or Xestospongin-C (5 µM), the testosteroneinduced Ca 2+ oscillations at 100 nM were abolished ( Fig. 4A; n=46 and n=57, respectively). The addition of InsP 3 R inhibitors also altered the Ca 2+ response patterns evoked by elevated concentrations of testosterone. At 1 µM or 10 µM testosterone, the Ca 2+ response was significantly attenuated or inhibited ( Figure  4B, C; n=35 and n=48, respectively). Taken together, these results suggest that testosterone regulates InsP 3 R-evoked Ca 2+ increases by controlling the Ca 2+ release from internal stores.

InsP 3 R is involved in the apoptotic pathway induced by testosterone
Our results suggest that the InsP 3 R is directly involved in the apoptotic response induced by testosterone. We tested this hypothesis by monitoring the apoptosis parameters in the presence of InsP 3 R blockers. Cells treated with 1 or 10 µM testosterone in the presence of these blockers had significantly lower staining for annexin-V as compared to cells treated with testosterone alone (compare Fig. 5A, B with Fig.  2D, E). We found that 2-APB and Xestospongin-C could also protect the cells from apoptosis-associated DNA damage (Fig. 5C, lane 3 and 4 and Fig. 5D). We also compared caspase activity in the presence of 2-APB and Xestospongin-C and found that treatment with InsP 3 R blockers suppressed the activation of caspases by testosterone at both concentrations tested (see Fig. 5E for 1 µM). The rates of caspase activation by testosterone were significantly lower when cells were treated with 2-APB and Xestospongin-C (Fig. 5F). Taken together, these results confirm that testosteroneinduced apoptosis is mediated though InsP 3 R and that InsP 3 R blockers can protect cells from progression into this cell death pathway.

InsP 3 R type 1 is involved in both normal and dysregulated Ca 2+ signaling
Neuroblastoma cells express mainly InsP 3 R type 1 (35), but the presence of significant amounts of InsP 3 R type 3 has also been described (36). To establish which isoform of InsP 3 R is important for the testosterone-induced Ca 2+ increases and subsequent activation of the cell death program, neuroblastoma cells were transfected with specific siRNA against InsP 3 R type 1 or type 3. These siRNA probes have not shown any cross-reactivity among the subtypes (37). There was a significant reduction in the immunosignal for InsP 3 R protein in either InsP 3 R type 1 siRNA transfected cells ( Fig  6A, 6B). In order to determine the molecular contribution of InsP 3 R type 1 and type 3 to the testosterone-induced Ca 2+ response, Ca 2+ transients were recorded in InsP 3 R type 1 or type 3 knocked down cells. To identify the successfully transfected cells, we co-transfected the siRNA with a plasmid DNA coding for the red fluorescent protein (DSRed). We found that neuroblastoma cells knocked down for InsP 3 R type 3 did not modify their ability to evoke Ca 2+ oscillations or prolonged transients in response to testosterone (Fig 6C-E,  black lines). These responses to testosterone were similar to those recorded in untransfected cells (compare Fig 6C-E with Fig 4). In contrast, cells knocked down for the InsP 3 R type 1 did not respond to testosterone (Fig. 6C-E, gray lines), but they still responded to extracellular application of ATP (data not shown). The magnitude of the response to thapsigargin, an inhibitor of the Ca 2+ pump on the ER, was similar (data not shown), suggesting that store loading was unchanged by InsP 3 R knock down. These data indicate a critical role of InsP 3 R type 1 but not type 3 in controlling the Ca 2+ signaling in response to testosterone, and regulation of its neuronal function.

DISCUSSION
Testosterone is the main endogenous anabolic/androgenic steroid hormone and it plays fundamental roles in development, differentiation and cellular growth (38,39). In neurons, testosterone acts as a neurosteroid and can induce changes at the cellular level which in turn lead to changes in behavior, mood, and memory (40). Both neuroprotective (41,42) as well as neurodegenerative (43) effects of androgens have been reported. Testosterone levels are substantially increased in human subjects using elevated doses of anabolic steroids to increase muscle mass (14), which can negatively affect their behavior (17) or induce degenerative processes (40). The cellular mechanisms of these physiological and deleterious effects of testosterone have not been elucidated.
In the present study we have demonstrated for the first time that the treatment of neuroblastoma cells with elevated concentrations of testosterone for relatively short time periods (6 h to 12 h) induces a decrease in cell viability by activation of a cell death program. Low concentrations of testosterone had no effects on cell viability, whereas at high concentrations the cell viability decreased with incremental increases in hormone concentration. Using three characteristic parameters, it was possible to show that this action of testosterone was through the apoptotic program. Being lipid soluble molecules, steroids could influence membrane fluidity (and phosphatidylserine accessibility), as shown in lipid bilayer experiments with milimolar concentrations of steroids (44). In addition, aromatase, an enzyme that converts testosterone into the estrogen (17β-estradiol) has been reported in the CNS and neuroblastoma cells (45) . To rule out these possibilities, we simulated the cells with estrogen and found that equally elevated concentrations of estrogen had no effect on cell viability (Fig. 1B), suggesting that the response was specific for elevated testosterone concentration and not due to its metabolism to estrogen (Fig. 1B).
Previously, we demonstrated that testosterone (10 nM -100 nM) induces intracellular Ca 2+ oscillations in the cytosol and nucleus, which are an important mediator of downstream events such as neurite outgrowth (12). In several cell models and using different apoptotic stimuli, it has been shown that an intracellular Ca 2+ increase can trigger apoptosis (23,25). In general, this occurs when the Ca 2+ signal differs from the pattern expected for a specific stimulus. For example modifying the amplitude or duration of the Ca 2+ signals can lead to cell death (28). Here we show that testosterone can induce concentration-dependent patterns of Ca 2+ signals. At low concentrations (100 nM), testosterone produces intracellular Ca 2+ oscillations, which could be an important physiological mechanism for the action of androgen in neurons. When the concentration of testosterone is increased, the oscillatory pattern is lost and transient and longlasting responses appear. These testosteroneinduced, long-lasting rises in intracellular Ca 2+ signals would be sufficient to initiate the apoptotic response.
Recently, a mechanism of apoptosis has been proposed which involves an interaction between cytochrome-c and the InsP 3 R (46,47). It has been suggested that the release of cytochrome-c from mitochondria into the cytosol during the early stages of apoptosis could induce changes in InsP 3 -mediated Ca 2+ signals. The model relies upon a binding between cytrochrome-c and the InsP 3 R causing a lack of regulation and producing an exaggerated Ca 2+ release from the ER (46)(47)(48). The results presented in this study are consistent with this model which shows that InsP 3 R blockers can inhibit the apoptotic response induced by testosterone. There are several reports which demonstrate that androgens can activate PTX-sensitive G proteins (49,50). Also, other G-protein receptor agonists can induce an apoptotic program via PLC-InsP 3 R-Ca 2+ increase. For example, G-protein activation and subsequent interaction of Gβγ with PLC mediates the proapoptotic effects of ethanol in developing neural crest (51). The pro-apoptotic effects of the hormone angiotensin II also has been shown to mediate Ca 2+ release from internal stores via G-protein coupled receptor (52). Moreover, overexpression of wildtype Gαq, which further increased Gq signaling, produced initial hypertrophy in cardiomyocytes, which rapidly progressed to apoptotic death (53).
Moreover, the importance of the InsP 3 R pathway for cell survival has been shown in cell models where deletion of InsP 3 Rs is associated with less cellular damage and more resistance to apoptosis (54,55). Depending on the cell type studied, different subtypes of InsP 3 R have been associated with apoptosis progression. In developing and regenerating neurons, InsP 3 R type 1 is concentrated in the growth cone suggesting its participation in neurite outgrowth (56), whereas InsP 3 R type 3 has been shown to be in neuron terminals in limbic and basal forebrain regions including olfactory tubercle, central nucleus of the amygdala, and bed nucleus of the stria terminalis (26). Elevated expression of InsP 3 R type 3 has been implicated in developmental apoptosis in many cell types like early postnatal cerebellar granule cells, dorsal root ganglia, embryonic hair follicles, and intestinal villi (57). Recently, InsP 3 R type 3 was shown to modulate apoptosis in epithelial cells (58). In contrast, InsP 3 R type 1 was required for activation of the apoptosis pathway in T cells (54,55) and in the neuroblastoma cells studied in this report. There are increasing numbers of reports showing differential distribution and different roles of InsP 3 Rs (59,60) in agreement with the hypothesis that specific isoforms of the InsP 3 R will be critical for maintaining cell health; in the case of testosterone, the deleterious effects appear to be mediated through InsP 3 R type 1.
Both beneficial and pathological effects of androgens are observed clinically. Depletion of androgen (below normal plasma levels) increases kainate-induced neuronal loss (61). Interestingly, high levels of testosterone also produced serious side effect and have been related to neurobehavioral changes like hyper excitability, supra-aggressive nature and suicidal tendencies (17,18). Externally administrated supra physiological levels of testosterone can stimulate muscular bulk in human subjects (14,15). Although this effect could be useful for patients with muscular physio-pathologies, these kind of studies generally use blood testosterone concentrations ranging from 0.1 to 0.5 mM (16). Despite the complexity of the blood brain barrier, the lipid permeable nature of testosterone and the duration of these treatments (more than weeks) suggest that significant amounts of this hormone would be accessible to neuronal tissues. The effects of testosterone we describe here at the single cell level will have long term effects at the system level. Importantly, this study suggests that androgen supplementation for additional performance, rather than clinical needs, requires a careful reexamination. Thus, it appears that the beneficial effects of androgens are carefully regulated over a narrow (nanomolar) range of concentration, which is crucial for normal neuronal functions. A deviation from this concentration range in either direction could be deleterious for neuronal health.
The rapid effects of testosterone on intracellular Ca 2+ signaling in neurons occur in the absence of the androgen receptor (12). There are additional effects of testosterone in neurons which require the androgen receptor (62). For example testosterone activation of the androgen receptor increases the expression of cytoskeleton proteins such as tubulin (63) and neuritin (64) which are important for neurite outgrowth and neuronal differentiation. Mutations of the androgen receptor that cause aggregation of polyglutamine-expanded protein induce cytotoxicity with normal levels of testosterone (65,66), a primary cause of Kennedy syndrome (67). The effect of high testosterone concentration on cell function in these altered cells in not known. In addition, specific regions of the CNS contain 5α-reductase, a key protein for the action of testosterone (68), suggesting that in some cases local concentrations of testosterone could be higher than the plasma concentrations. Moreover, differential expression of androgen receptor in the CNS could also generate a range of testosterone sensitivity among neuronal cells (5).
Together these results suggest that elevated concentrations of testosterone can induce programmed cell death which is characterized by a decrease of cell viability, an increase in the number of annexin-V positive cells, DNA fragmentation and caspase activation. Apoptotic parameters are blocked or diminished by InsP 3 R inhibitors and the specific knocking down of InsP 3 R type 1, indicating that this isoform is a key regulator in this event. Overall, the conclusions from these studies are that normal levels of testosterone are necessary for the normal Ca 2+ response and to maintain homeostasis, but elevated concentrations produce altered Ca 2+ signal leading to deleterious effects in neurons.

Figure 1: High concentrations of testosterone cause cytotoxicity in neuroblastoma cells. A)
Time course of neuroblastoma cell survival as a function of various testosterone concentrations. Cells were incubated with 100 nM ( ), 1 µM ( ) or 10 µM ( ) of testosterone (Testo) for the indicated times. Untreated cells were used as a negative control ( ). Cell viability was assessed for trypan blue exclusion (where 100% is complete exclusion) and data were averaged for 1 X 10 6 cells. B) Cells were treated with various concentrations of either testosterone or 17β-Estradiol (Estro) and were scored for viability using an XTT assay 12 h after treatment with the hormones. The values represent the mean ± SEM of three independent experiments. (* p<0.05; ** p<0.01).  Cells were treated with three concentrations of testosterone (Testo) for 6 hr; and DNA fragmentation was assessed by 2% agarose gel electrophoresis and ethidium bromide staining. B) Neuroblastoma cell viability was determined by an XTT assay. Cells were pre-incubated with the cell permeable caspase inhibitor and exposed to 1 µM or 10 µM of testosterone for 12 hr. C) The fluorescence spectrum, recorded 30 min after addition of Ac-DEVD-AFC to the extract prepared from 1 µM testosterone induced cells. Ac-DEVD-AFC was added to a final concentration of 150 nM. The caspase dependent release of AFC following hydrolysis of the substrate is monitored on excitation at 400 nm. A distinctive emission peak was observed between 475 and 505 nm. The extracts were prepared from 1 x 10 6 cells ~6 hr after treatment with testosterone (black line), or from cells without any treatment (gray line). Addition of the caspase inhibitor Ac-DEVD-CHO competitively blocked the AFC release. Lysates from the testosterone treated cells which were incubated first with Ac-DEVD-CHO and than subjected to same treatment as in (C) did not exhibit the AFC release (dotted line). D) Concentration dependence of testosterone on caspase activation. For these experiments, a colorigenic substrate of caspase, Ac-DEVD-pNA was used. Lysates generated from cells treated with 100 nM ( ), 1µM ( ) and 10 µM ( ) testosterone were incubated with the substrate and the pNA absorbance was plotted as a function of time. Lysates from untreated cells were used as negative control ( ). Data points were fit to straight-line (gray lines) to calculate the rate of caspase activation (1.1 ± 0.05 and 1.4 ± 0.02 per second for 1 µM and 10 µM testosterone treatments respectively). The same amount of protein was used in each assay. The values represent the mean ± SEM of three independent experiments (* p<0.05; ** p<0.01).  and Xestospongin-C (Xesto-C, 5 µM), were used to determine the role of the InsP 3 R in testosteroneinduced apoptosis. A, B) Annexin-V labeling in the presence of InsP 3 R inhibitors. Cells were treated with high (1 µM (A) and 10 µM (B)) concentrations of testosterone in presence of 2APB or Xesto-C and scored for annexin-V positive cells. On addition of these blockers, few cells were positive to annexin-V and the cell morphology did not change with respect to the control cells; compare these images with the responses shown in Fig. 2. One representative set of images is shown for 1 µM and 10 µM testosterone. C) DNA fragmentation in the presence of InsP 3 R blockers. Genomic DNA isolated from neuroblastoma cells treated with 1 µM testosterone in the presence of 2APB (lane 3) and Xesto-C (lane 4) show significantly less DNA damage than that of 1 µM testosterone alone (lane 2). Cells without treatment (lane 1) and cells treated with 1 µM staurosporine (lane 5) were used as negative and positive control experiments, respectively. DNA damage was quantified by scanning densitometry and was normalized by total DNA intensity in each lane. D) The percentage DNA degraded by 1 µM testosterone (52 ± 11%), in the presence of 2APB (14 ± 2%), or Xesto-C (8 ± 6%) with respect to the response after treatement with staurosporine (86 ± 6%) and untreated cells (12 ± 3%). E, F) Testosterone-induced caspase activity in the presence of InsP 3 R inhibitors. Using a colorigenic caspase substrate, Ac-DEVD-pNA, the time course of caspase activation was followed in lysates from cells treated with testosterone in the presence of InsP 3 R blockers. (E) Caspase activation induced by 1 µM testosterone ( ) is inhibited by 2APB ( ) or Xesto-C ( ) treatment. Addition of the caspase inhibitor, Ac-DEVD-CHO, also abolishes the caspase activity ( ). These data were fit to the linear functions (gray lines) to yield rates of caspase activation. (F) Rates of caspase activation are severely affected by InsP 3 R blockers at both concentrations of testosterone tested. At 1 µM testosterone, the rate of caspase activation was 1.1 ± 0.1 per second versus 0.2 ± 0.2 per second for 2APB; and 0.2 ± 0.1 per second for Xesto-C. At 10 µM testosterone, the rate of caspase activation was 1.4 ± 0.0 per second versus 0.2 ± 0.1 per second for 2APB; and 0.1 ± 0.1 per second for Xesto-C. ** =p<0.01. In InsP 3 R type 1-siRNA transfected cells the caspase activity was inhibited ( ), whereas knock down of InsP 3 R type 3 had little effect on the caspase activity ( ) in response to testosterone. Quantification of the rates of caspase activity in each of the conditions tested is shown in the inset. C) The response to testosterone in InsP 3 R-siRNA transfected cells depended upon the isoform of the InsP 3 R which was decreased. Knock down of the InsP 3 R type 1 blocked the testosterone-induced Ca 2+ signals at all three concentrations of testosterone tested (gray lanes). Knock down of the InsP 3 R type 3 did not modify the response to testosterone compared with untransfected cells (black lanes; see Fig. 4). Arrows indicate the time of addition of the hormone. ** =p<0.01.