Neurosteroid DHEA

NEUROSTEROIDS BIND WITH HIGH AFFINITY AND ACTIVATE NERVE GROWTH FACTOR (NGF) RECEPTORS, PREVENTING NEURONAL APOPTOSIS.

ABSTRACT

Neurosteroid DHEA, produced by neurons and glia, affects multiple processes in the brain, including neuronal survival and neurogenesis during development and in aging. However no specific receptor has been reported to date for this important neurosteroid. We provide evidence that DHEA binds with high affinity to pro-survival TrkA and pro-death p75NTR membrane receptors of neurotrophin NGF, acting as a neurotrophic factor: 1) the anti-apoptotic effects of DHEA were reversed by siRNA against TrkA, or by a specific TrkA inhibitor, 2) [3H]DHEA displacement experiments showed that DHEA bound with high affinity on membranes isolated from HEK293 cells transfected with the cDNAs of TrkA and p75NTR receptors (IC50: 0,9 and 5.6 nM respectively). Membrane binding of DHEA on HEK293TrkA and HEK293p75NTR cells was also shown with flow cytometry and immunofluorecence microscopy, using the membrane impermeable DHEA-BSA-FITC conjugate, 3) immobilized DHEA pulled down recombinant and naturally expressed TrkA and p75NTR receptors, 4) DHEA induced TrkA phosphorylation, and NGF receptor-mediated signaling; Shc, Akt, and ERK1/2 kinases down-stream to TrkA receptors and TRFA6, RIP2 and RhoGDI effectors of p75NTR receptors, 5) DHEA rescued from apoptosis TrkA receptor positive sensory neurons of dorsal root ganglia in NGF null embryos and compensated NGF in rescuing from apoptosis NGF receptor positive sympathetic neurons of embryonic superior cervical ganglia. Our findings suggest that DHEA and NGF cross-talk via their binding to NGF receptors to afford brain shaping and maintenance during development. Phylogenetic findings on the evolution of neurotrophins, their receptors and CYP17, the enzyme responsible for DHEA biosynthesis, combined with our data support the hypothesis that DHEA served as a phylogenetically ancient neurotrophic factor.

INTRODUCTION

Dehydroepiandrosterone (DHEA) is a steroid, produced in adrenals, in neurons and in glia [1]. The physiological role of brain DHEA appears to be local i.e. paracrine, while that produced from adrenals, which represents the almost exclusive source of circulating DHEA, is systemic. The precipitous decline of both brain and circulating DHEA with advancing age has been associated to aging-related neurodegenerative diseases [1,2]. It is experimentally supported that DHEA protects neurons against noxious conditions [3-6]. DHEA exerts its multiple pro-survival effects either directly modulating at micromolar concentrations g-aminobutiric acid type A (GABAA), N-methyl-D-aspartate (NMDA) or sigma1 receptors, or following its conversion to estrogens and androgens. We have recently shown that nanomolar concentrations of DHEA protect sympathoadrenal PC12 cells from apoptosis [7]. PC12 cells do not express functional GABAA or NMDA receptors and cannot metabolize DHEA to estrogens and androgens [8]. The anti-apoptotic effect of DHEA in PC12 cells is mediated by highly affine (Kd: 1 nM) specific membrane binding sites [9]. Activation of DHEA membrane binding sites results in an acute but transient sequential phosphorylation of the pro-survival MEK/ERK kinases which, in turn, activate transcription factors CREB and NFkB, which afford the transcriptional control of anti-apoptotic Bcl-2 proteins. In parallel, activation of DHEA membrane binding sites induces the phosphorylation of PI3K/Akt kinases, leading to phosphorylation/deactivation of the pro-apoptotic Bad protein, and protection of PC12 cells from apoptosis [10].

In fact, the anti-apoptotic pathways in sympathoadrenal cells initiated by DHEA at the membrane level strikingly resemble those sensitive to neurotrophin nerve growth factor. NGF, promotes survival and rescuing from apoptosis of neural crest deriving sympathetic neurons (including their related sympathoadrenal cells), and sensory neurons involved in noniception. NGF binds with high affinity (Kd: 0.01 nM) to transmembrane tyrosine kinase TrkA receptor and with lower affinity (Kd: 1.0 nM) to p75NTR receptor, a membrane protein belonging to TNF receptor superfamily [11]. In the presence of TrkA receptors, p75NTR participates in the formation of high affinity binding sites and enhances NGF responsiveness leading to cell survival signals. In the absence of TrkA, p75NTR generates cell death signals. Indeed, docking of TrkA by NGF initiates receptor dimerization, and phosphorylation of cytoplasmic 490 and 785 tyrosine residues on the receptor. Phosphotyrosine-490 interacts with Shc and other adaptor proteins resulting in activation of PI3K/Akt and MEK/ERK signaling kinase pathways [11]. These signals lead to the activation of prosurvival transcription factors CREB and NFkB, the subsequent production of anti-apoptotic Bcl-2 proteins and prevention of apoptotic cell death of sympathetic neurons and sympathoadrenal cells, including PC12 cells [12].

Intrigued by the similarities in the prosurvival signaling of DHEA and NGF, both initiated at the membrane level, we set out to examine in the present study whether the anti-apoptotic effects of DHEA are mediated by NGF receptors. To address this issue we employed a multifaceted approach designing an array of specific experiments. We used RNA interference (RNAi) to define the involvement of TrkA and p75NTR receptors in the anti-apoptotic action of DHEA. We assessed membrane binding of DHEA in HEK293 cells transfected with the TrkA and p75NTR plasmid cDNAs, using binding assays, confocal laser microscopy and flow cytometry. To investigate the potential direct physical interaction of DHEA with NGF receptors we tested the ability of immobilised DHEA to pull-down recombinant or naturally expressed TrkA and p75NTR receptors. Finally, we examined the ability of DHEA to rescue from apoptosis NGF receptor sensitive dorsal root ganglia sensory neurons of NGF null mice, and NGF deprived rat superior cervical ganglia sympathetic neurons in culture [13]. We provide evidence that DHEA directly binds to NGF receptors to protect neuronal cells against apoptosis, acting as a neurotrophic factor.

RESULTS

RNA interference against TrkA receptors reverses the anti-apoptotic effect of DHEA.

To test the involvement of NGF receptors in the anti-apoptotic effect of DHEA in serum deprived PC12 cells we have used the RNAi technology. A combination of three different sequences of siRNAs for TrkA and two different shRNAs for p75NTR transcripts [14] were selected. The effectiveness of si/shRNAs was shown by the remarkable decrease of TrkA and p75NTR protein levels in PC12 cells, observed by immunobloting analysis, using GAPDH as reference standard ( 1b). Scrambled siRNAs were ineffective in decreasing TrkA and p75NTR protein levels and did not significantly alter the effect of DHEA (data not shown). FACS analysis of apoptotic cells (stained with Annexin V) has shown that DHEA and membrane impermeable DHEA-BSA conjugate at 100 nM diminished the number of apoptotic cells in serum deprived PC12 cell cultures from 53.5±17.6% increase of apoptosis in serum free condition (control) to 6±1.4% and 13±5.2%, respectively (n:8, P<0.01 versus control) ( 1a). Decreased TrkA expression in serum-deprived PC12 cells with siRNAs resulted in the almost complete reversal of the anti-apoptotic effect of DHEA and DHEA-BSA conjugate ( 1a). Co-transfection of serum deprived PC12 cells with the si/shRNAs for TrkA and p75NTR receptors did not modify the effect of the TrkA deletion alone. Furthermore, transfection of serum deprived PC12 cells with the shRNAs against p75NTR receptor alone did not significantly alter the anti-apoptotic effect of DHEA, suggesting that the anti-apoptotic effect of DHEA is primarily afforded by TrkA receptors.

Transfection of serum-deprived PC12 cells with the siRNAs against the TrkA transcript fully annulled the ability of DHEA to maintain elevated the levels of anti-apoptotic Bcl-2 protein ( 1b). Again, transfection with the shRNA against p75NTR receptor alone did not significantly affect Bcl-2 induction by DHEA, further supporting the hypothesis that TrkA is the main mediator of the anti-apoptotic effect of DHEA in this system.

It appears that the ratio of TrkA and p75NTR receptors determines the effect of DHEA on cell apoptosis and survival. Indeed, both NGF and DHEA induced apoptosis of nnr5 cells, a clone of PC12 cell line known to express only pro-death p75NTR receptors ( 1c), confirming the pro-apoptotic function of this receptor. Blockade of p75NTR expression by shRNA almost completely reversed the pro-apoptotic effect of both agents. The anti-apoptotic effect of NGF and DHEA was remarkably restored after transfection of nnr5 cells with the TrkA cDNA, the efficacy of reversal being proportionally dependent on the amount of transfected TrkA cDNA ( 1c). DHEA was also controlling the response of NGF receptor positive cells, by regulating TrkA and p75NTR receptor levels, mimicking NGF. Serum deprived PC12 cells were exposed to 100 nM of DHEA or 100 ng/ml of NGF for 12, 14 and 48 hours, TrkA and p75NTR protein levels were measured in cell lysates with immunoblotting, using specific antibodies against TrkA and p75NTR proteins and were normalized against GAPDH. Both NGF and DHEA significantly increased pro-survival TrkA receptor levels in the time frame studied, i.e. from 12 to 48 hours (n:5, P<0.01) ( 1d). Furthermore, DHEA and NGF significantly decreased p75NTR receptor levels between 24 to 48 hours of exposure (n:5, P<0,01).

We have also tested the anti-apoptotic effects of DHEA in neural crest deriving superior cervical ganglia (SCG), a classical NGF/TrkA sensitive mammalian brain tissue, containing primarily one class of neurons, principal sympathetic neurons. Indeed, NGF and TrkA receptors are absolutely required for SCG sympathetic neuron survival during late embryogenesis and early postnatal development [13, 15]. TrkC receptors are barely detectable after E15.5, and no significant TrkB receptors are present in the SCG at any developmental stage [16]. Organotypic cultures of rat SCG at P1 were incubated in the presence of 100ng/ml NGF, or in the same medium as above but lacking NGF and containing a polyclonal rabbit anti-NGF-neutralizing antiserum in the absence or the presence of 100nM DHEA. Withdrawal of NGF resulted in a rapid degeneration of ganglia, effect which was completely reversed in the presence of DHEA ( 2a). We repeated these experiments using dispersed sympathetic neurons in culture, isolated from rat SCGs at P1. Deprivation of NGF strongly increased the number of apoptotic sympathetic neurons stained with Annexin V, while DHEA effectively compensated NGF by decreasing the levels of apoptotic neurons, effect which was blocked by a specific TrkA inhibitor thus, suggesting the involvement of TrkA receptors as the main mediator of the anti-apoptotic action of DHEA ( 2b). Moreover, inhibition of p75NTR by a specific antibody (MAB365R, Millipore) against its extracellular domain, strongly induced a DHEA- or NGF-mediated anti-apoptotic effect, clearly indicating that p75NTR receptor serves a pro-apoptotic role in SCGs also, effect which is apparent only in the absence of TrkA receptor, as it was also shown in nnr5 cells.

[3H]DHEA binds with high affinity to HEK293TrkA and HEK293p75NTR cell membranes.

We have previously shown the presence of specific DHEA binding sites to membranes isolated from PC12, primary human sympathoadrenal, and primary rat hippocampal cells, with Kd 0.9, 0.1 and 0.06 nM, respectively [9]. The presence of DHEA-specific membrane binding sites on PC12 cells has been confirmed by flow cytometry and confocal laser microscopy of cells stained with the membrane impermeable DHEA-BSA-FITC conjugate. In contrast to estrogens, glucocorticoids and androgens displaced [3H]DHEA from its membrane binding sites, acting as pure antagonists by blocking theanti-apoptotic effect of DHEA in serum deprived PC12 cells9. In the present study, we repeated this series of experiments using membranes isolated from HEK293 cells transfected with the plasmid cDNAs of TrkA or p75NTR receptors. Homologous [3H]DHEA displacement experiments with unlabeled DHEA showed the presence of specific DHEA binding on membranes from both HEK293TrkA and HEK293p75NTR cells with IC50 0.9 and 5.6 nM, respectively ( 3a). No specific binding of [3H]DHEA was observed on membranes isolated from non-transfected HEK293 cells or from HEK293 cells transfected with the empty vectors.

The selectivity of DHEA binding on HEK293TrkA and HEK293p75NTR cell membranes was examined by performing heterologous [3H]DHEA displacement experiments using a number of non-labeled steroids or NGF. Binding of [3H]DHEA on membranes isolated from both HEK293TrkA and HEK293p75NTR cells was effectively displaced by NGF (IC50: 0.08 and 1.1 nM, respectively) ( 3a). NGF was also effectively displacing [3H]DHEA binding on membranes isolated from PC12 cells (IC50: 0.8 nM, data not shown). Estradiol failed to displace [3H]DHEA from its binding on membranes from HEK293TrkA and HEK293p75NTR cells at concentrations ranging from 0.1 to 1000 nM. In contrast, displacement of [3H]DHEA binding on membranes from both HEK293TrkA and HEK293p75NTR cells was shown by testosterone (Testo) (IC50: 3.3 and 7.4 nM, respectively). Glucocorticoid dexamethasone (Dex) effectively competed [3H]DHEA binding on membranes from HEK293TrkA (IC50: 10.5 nM) but was ineffective in displacing DHEA binding on membranes from HEK293p75NTR cells. Homologous [125I]NGF displacement experiments with unlabeled NGF confirmed the presence of specific NGF binding on membranes from both HEK293TrkA and HEK293p75NTR cells with IC50 0.03 and 1.7 nM respectively (data not shown). It is of note that in contrast to unlabeled NGF, DHEA was unable to displace binding of [125I]NGF on membranes isolated from HEK293TrkA and HEK293p75NTR transfectants.

DHEA-BSA-FITC conjugate stains HEK293TrkA and HEK293p75NTR cell membranes.

Incubation of PC12 cells with the membrane impermeable, fluorescent DHEA-BSA-fluorescein conjugate results in a specific spot-like membrane fluorescent staining [9]. In the present study, we have tested the ability of DHEA-BSA-FITC conjugate to stain HEK293TrkA and HEK293p75NTR transfectants. Fluorescence microscopy analysis revealed that DHEA-BSA-FITC clearly stained the membranes of HEK293TrkA and HEK293p75NTR cells ( 3b). No such staining was found in non-transfected HEK293 cells (data not shown) or in HEK293 cells transfected with the vectors empty of TrkA and p75NTR cDNAs ( 3b). Furthermore, BSA-FITC conjugate was ineffective in staining both transfectants (data not shown). We have further confirmed the presence of membrane DHEA-BSA-FITC staining of HEK293TrkA and HEK293p75NTR cells with flow cytometry (FACS) analysis ( 3c). Specific staining was noted in both transfectants. No such staining was seen in non-transfected HEK293 cells (data not shown) or in HEK293 cells transfected with the empty vectors ( 3c). In both fluorescence microscopy and FACS experiments membrane staining of TrkA or p75NTR proteins in HEK293TrkA and HEK293p75NTR cells was also shown using specific antibodies for each protein (s 3b, 3c).

Immobilised DHEA pulls down TrkA and p75NTR receptors.

Our binding assays with radiolabeled DHEA suggest that DHEA physically interacts with NGF receptors. To test this hypothesis we covalently linked DHEA-7-O-(carboxymethyl) oxime DHEA-7-CMO) to polyethylene glycol amino resin (NovaPEG amino resin) and we tested the ability of immobilized DHEA to pull down TrkA and p75NTR proteins. Precipitation experiments and western blot analysis of precipitates with specific antibodies against TrkA and p75NTR proteins ( 4) showed that immobilized DHEA effectively precipitated recombinant TrkA and p75NTR proteins. Similar results were obtained when cell extracts isolated from HEK293 cells transfected with TrkA and p75NTR cDNAs, PC12 cells and whole rat brain were treated with immobilised DHEA ( 4, panels marked with A). No precipitation of TrkA and p75NTR proteins was shown with polymer-supported DHEA-7-CMO incubated with cell extracts from untransfected HEK293 cells or HEK293 cells transfected with the empty vectors. A control experiment was performed with NovaPeg amino resin (no DHEA-7-CMO present) which was found ineffective in precipitating TrkA and p75NTR proteins ( 4). The presence of TrkA and p75NTR receptors in HEK293TrkA and HEK293p75NTR transfectants and in PC12 and fresh rat brain was confirmed with western blot analysis using specific antibodies against TrkA and p75NTR proteins and GAPDH as reference standard ( 4, panels marked with B).

DHEA induces TrkA- and p75NTR-mediated signaling.

Previous findings have shown that NGF controls the responsiveness of sensitive cells through induction of TrkA phosphorylation and regulation of the levels of each own receptors [17]. We compared the ability of NGF and DHEA to induce phosphorylation of TrkA in HEK293 cells transfected with the cDNAs of TrkA receptors. HEK293TrkA transfectants were exposed for 10 and 20 min to 100 nM of DHEA or 100 ng/ml of NGF, and cell lysates were immunoprecipitated with anti-tyrosine antibodies and analyzed by western blotting, using specific antibodies against TrkA receptors. Both NGF and DHEA strongly increased phosphorylation of TrkA as early as 10 min, effect which was also maintained at 20 min ( 5a). We also tested the effects of DHEA and NGF in PC12 cells, endogenously expressing TrkA receptors. Naive or siRNATrkA transfected PC12 cells were incubated for 10 min with DHEA or NGF, and cell lysates were analyzed with western blotting, using specific antibodies against Tyr490-phosphorylated TrkA and total TrkA. Both NGF and DHEA strongly induced the phosphorylation of TrkA in naive PC12 cells, effects which were diminished in siRNATrkA transfected PC12 cells ( 5a). The stimulatory effect of DHEA on TrkA phosphorylation might be due to an increase of NGF production. To test this hypothesis, we measured with ELISA the levels of NGF in culture media of HEK293 and PC12 cells exposed for 5 to 30 min to 100 nM of DHEA. NGF levels in culture media of control and DHEA-treated HEK293 and PC12 cells were undetectable, indicating that DHEA-induced TrkA phosphorylation was independent of NGF production.

We compared the ability of NGF and DHEA to induce phosphorylation of TrkA-sensitive Shc, ERK1/2 and Akt kinases. Serum deprived naive or siRNATrkA transfected PC12 cells were incubated for 10 min with 100 nM DHEA or 100 ng/ml NGF and cell lysates were analyzed with western blotting, using specific antibodies against the phosphorylated and total forms of kinases mentioned above. Both DHEA and NGF strongly increased phosphorylation of Shc, ERK1/2 and Akt kinases in naive PC12 cells, effects which were almost absent in siRNATrkA transfected PC12 cells, suggesting that both DHEA and NGF induce Shc, ERK1/2 and Akt phosphorylation via TrkA receptors ( 5a).

The effectiveness of DHEA to promote the interaction of p75NTR receptors with its effector proteins TRAF6, RIP2 and RhoGDI was also assessed. It is well established that NGF induces the association of p75NTR receptors with TNF receptor-associated factor 6 (TRAF6), thus, facilitating nuclear translocation of transcription factor NFκB [18]. Furthermore, p75NTR receptors associate with receptor-interacting protein 2 (RIP2) in a NGF-dependent manner [19]. RIP2 binds to the death domain of p75NTR via its caspase recruitment domain (CARD), conferring nuclear translocation of NFκB. Finally, naive p75NTR interacts with RhoGDP dissociation inhibitor (RhoGDI), activating small GTPase RhoA [20]. In that case, NGF binding abolishes the interaction of p75NTR receptors with RhoGDI, thus, inactivating RhoA. We co-transfected HEK293 cells with the plasmid cDNAs of p75NTR and of each one of the effectors TRAF6, RIP2 or RhoGDI, tagged with the flag (TRAF6) or myc (RIP2, RhoGDI) epitopes. Transfectants were exposed to 100 nM DHEA or 100 ng/ml NGF, and lysates were immunoprecipitated with antibodies against flag or myc, followed by immunoblotting with p75NTR specific antibodies. Both DHEA and NGF efficiently induced the association of p75NTR with effectors TRAF6 and RIP2, while facilitated the dissociation of RhoGDI from p75NTR receptors ( 5b).

DHEA reverses the apoptotic loss of TrkA positive sensory neurons in dorsal root ganglia of NGF null mouse embryos.

NGF null mice have less sensory neurons in dorsal root ganglia due to their apoptotic loss [13]. Heterozygous mice for the NGF deletion were interbred to obtain mice homozygous for the NGF gene disruption. The mothers were treated daily with an intraperitoneal injection of DHEA (2 mg) or vehicle (4.5% ethanol in 0.9% saline). Embryos were collected at E14 day of pregnancy and sections were stained for Caspase 3 and Fluoro jade C, markers of apoptotic and degenerative neurons, respectively. ngf-/- embryos at E14 showed a dramatic increase in the number of Fluoro Jade C and Caspase 3 positive neurons in the DRG compared to the ngf+/- embryos (s 6a, 6b). DHEA treatment significantly reduced Fluoro Jade C and Caspase 3 positive neurons in the DRG to levels of ngf+/- embryos. Furthermore, TrkA and TUNEL double staining of DRGs has shown that in ngf+/- embryos, numbers of TUNEL-positive apoptotic neurons were minimal, while TrkA positive staining was present in a large number of neuronal cell bodies of the DRG and their collaterals were extended within the marginal zone to the most dorsomedial region of the spinal cord. On the contrary, in DRG of ngf-/- embryos levels of TUNEL-positive apoptotic neurons were dramatically increased while TrkA neuronal staining was considerably decreased and DRG collaterals of the dorsal funiculus were restricted in the dorsal root entry zone ( 6c). DHEA treatment resulted in a significant increase of TrkA positive staining and the extension of TrkA staining within the marginal zone to the most dorsomedial region of the spinal cord similarly to the ngf+/- embryos ( 6d), while staining of TUNEL-positive apoptotic neurons was decreased to levels shown in ngf+/- embryos.

DISCUSSION

DHEA exerts multiple actions in the central and peripheral nervous system, however no specific receptor has been reported to date for this neurosteroid. Most of its actions in the nervous tissue were shown to be mediated via modulation, at micromolar concentrations, of membrane neurotransmitter receptors, such as NMDA, GABAA and sigma1 receptors. DHEA may also influence brain function by direct binding at micromolar concentrations to dendritic brain microtubule-associated protein MAP2C [21]. In the present study we provide evidence that DHEA binds with high affinity to NGF receptors. This is the first report showing a highly affine, direct binding of a steroid to neurotrophin receptors. Displacement binding assays of [3H]DHEA on membranes isolated from HEK293 cells transfected with the cDNAs of TrkA and p75NTR receptors showed that DHEA binds with high affinity to both membranes (IC50 0.9 and 5.6 nM, respectively). Non-radioactive NGF effectively displaced [3H]DHEA binding to both membrane preparations, with IC50 values 0.08 nM for HEK293TrkA cells and at 1.1 nM for HEK293p75NTR cells, respectively. Furthermore, pull down experiments using DHEA covalently immobilized on NovaPEG amino resin suggest that DHEA binds directly to TrkA and p75NTR proteins. Indeed, polymer-supported DHEA-7-CMO effectively pulled down recombinant TrkA and p75NTR proteins, and precipitated both proteins from extracts prepared from cells expressing both receptors (HEK293TrkA, HEK293p75NTR and PC12 cells and freshly isolated rat brain). Interestingly, DHEA was unable to effectively displace binding of [125I]NGF on membranes isolated from HEK293TrkA and HEK293p75NTR transfectants. It is possible that dissociation of binding of peptidic NGF from its receptors lasts longer due to the multiple sites of interaction within the binding cleft of this large peptidic molecule compared to smaller in volume steroid. The domains of TrkA and p75NTR proteins involved in DHEA binding were not defined in the present study. Mutagenesis assays combined with NMR spectroscopy are planned to map the domains of both receptors related to DHEA binding. However, our findings that DHEA mimics NGF in binding to both TrkA and p75NTR receptors and that NGF effectively displaces DHEA binding to both receptors, support the hypothesis that NGF and DHEA share the same binding sites. Other small molecules, like antidepressant amitriptyline and gamboge's natural extract gambogic amide bind, although with much lower affinity compared to DHEA (Kd 3mM and 75 nM, respectively), in the extracellular and the cytoplasmic juxtamembrane domains of TrkA receptor [22,23].

Our findings suggest that binding of DHEA to NGF receptors is functional, mediating its anti-apoptotic effects. Indeed, blocking of TrkA expression by RNAi almost completely reversed the ability of DHEA to protect PC12 cells from serum deprivation-induced apoptosis and to maintain elevated levels of the anti-apoptotic Bcl-2 protein. Additionally, in dispersed primary sympathetic neurons in culture, DHEA effectively compensated NGF deprivation by decreasing the levels of apoptotic neurons, effect which was reversed by a specific TrkA inhibitor further supporting the involvement of TrkA receptors in the anti-apoptotic action of DHEA. Finally, DHEA effectively rescued from apoptosis TrkA-positive dorsal root ganglia sensory neurons of NGF null mouse embryos.

It appears that the decision between survival and death among DHEA-responsive cells is determined by the ratio of TrkA and p75NTR receptors. In fact, DHEA and NGF induced apoptosis of nnr5 cells, a clone of PC12 cells expressing only pro-death p75NTR receptors. The pro-death effects of both agents were completely blocked by p75NTR shRNA and were remarkably restored after transfection of nnr5 cells with the TrkA cDNA. It is of note that during brain development the ratio of TrkA to p75NTR varies tempospatially [24]. Thus, the ability of DHEA to act in a positive or negative manner on neuronal cell survival may depend upon the levels of the two receptors during different stages of neuronal development.

Binding of DHEA on both TrkA and p75NTR receptors was effectively competed by testosterone (IC50: 3.3 and 7.4 nM, respectively) while synthetic glucocorticoid dexamethasone displaced DHEA binding only to pro-survival TrkA receptors (IC50: 10.5 nM). In a previous study we had shown that both steroids effectively displaced DHEA from its specific membrane binding sites of sympathoadrenal cells, acting as DHEA antagonists by blocking its anti-apoptotic effect and the induction of anti-apoptotic Bcl-2 proteins [9]. Our findings suggest that testosterone and glucocorticoids may act as neurotoxic factors by antagonizing endogenous DHEA and NGF for their binding to NGF receptors [25,26]. Glucocorticoids show a bimodal effect on hippocampal neurons causing acutely an increase in performance of spatial memory tasks, while chronic exposure has been associated with decreased cognitive performance, and neuronal atrophy [27]. Acute administration of glucocorticoids results in a glucocorticoid receptor-mediated phosphorylation and activation of hippocampal TrkB receptors, exerting trophic effects on dentate gyrus hippocampal neurons [28], via an increase in the sensitivity of hippocampal cells to neurotrophin BDNF, the endogenous TrkB ligand known to promote memory and learning [29]. However, overexposure to glucocorticoids during prolonged periods of stress is detrimental to central nervous system neurons, especially in aged animals, affecting mainly the hippocampus. It is possible that part of neurotoxic effects of glucocorticoids may be due to their antagonistic effect on the neuroprotective effect of endogenous DHEA and NGF, via TrkA receptor antagonism. The decline of brain DHEA and NGF levels during aging and in Alzheimer's disease [27] might exacerbate this phenomenon, rendering neurons more vulnerable to glucocorticoid toxicity. Indeed, glucocorticoid neurotoxicity becomes more pronounced in aged subjects since cortisol levels in the CSF increase in the course of normal aging, as well as in relatively early stages of Alzheimer's disease [27].

A number of neurodegenerative conditions are associated with lower production or action of both DHEA and NGF [30, 31]. Animal studies suggest that NGF may reverse, or slow down the progression of Alzheimer's related cholinergic basal forebrain atrophy [31]. Furthermore, the neurotrophic effects of NGF in experimental animal models of neurodegenerative conditions, like MPTP (Parkinson's disease), experimental allergic encephalomyelitis (multiple sclerosis) or ischemic retina degeneration mice [32-34] support its potential as a promising neuroprotective agent. However, the use of NGF in the treatment of these conditions is limited because of its poor brain blood barrier permeability. It is of interest that DHEA also exerts neuroprotective properties in some of these animal models [7,35]. These findings suggest that synthetic DHEA analogs, deprived of endocrine effects, may represent a new class of brain blood barrier permeable NGF receptor agonists with neuroprotective properties. We have recently reported the synthesis of 17-spiro-analogs of DHEA, with strong anti-apoptotic and neuroprotective properties, deprived of endocrine effects [36], which are now tested for their ability to bind and activate NGF receptors.

We have previously defined the pro-survival signaling pathways that are initiated by DHEA at the membrane level [3]. These pathways include MEK1/2/ERK1/2, and PI3K/Akt pro-survival kinases. We now provide experimental evidence that DHEA activates these kinases via TrkA receptors ( 7). Downregulation of TrkA receptors using siRNAs, resulted in an almost complete reversal of the ability of DHEA to increase the phosphorylation of kinases Shc, Akt and ERK1/2. In addition to TrkA receptors, binding of DHEA to the low affinity NGF receptor was also functional, affording the activation of p75NTR receptors. Unlike TrkA receptors, p75NTR lacks any enzymatic activity. Signal transduction by p75NTR proceeds via ligand-dependent recruitment and release of cytoplasmic effectors to and from the receptor. Indeed, DHEA like NGF facilitated the recruitment of two major cytoplasmic interactors of p75NTR, TRAF6 and RIP2 proteins. Additionally, DHEA-mediated activation of p75NTR led to the dissociation of bound RhoGDI, a protein belonging to small GTPases and interacting with RhoA [20].

It is worth noticing that the interaction of DHEA with the NGF system was first suggested fifteen years ago by Compagnone et al, showing co-localized staining of CYP17, the rate limiting enzyme of DHEA biosynthesis, and NGF receptors in mouse embryonic DRGs [37]. About one fifth of CYP17-immunopositive DRG neurons in the mouse were found to be also TrkA-immunopositive. Among the TrkA-expressing cells, about one third also expresses CYP17, while p75NTR-expressing neurons represent only 13% of the cells in the DRG. Thus, about one fifth of CYP17-immunopositive neurons may be able to respond to both DHEA and NGF stimulation.

Recent studies have shown the expression of CYP17 in invertebrate cephalochordata Amphioxus [38]. Amphioxus is also expressing TrkA receptor homologous AmphiTrk, which effectively transduces signals mediated by NGF [39]. Phylogenetic analysis of neurotrophins revealed that they emerged with the appearance of vertebrates (530-550 million years ago), when complexity of neural tissue increased [40]. Invertebrate cephalochordata like Amphioxus are positioned on the phylogenetic boundary with vertebrates (600 million years ago). It is thus tempting to hypothesize that DHEA contributed as one of the “prehistoric” neurotrophic factors in an ancestral, simpler in structure invertebrate nervous system [41], then when a strict tempospatial regulation of evolving nervous system of vertebrates was needed peptidic neurotrophins emerged to afford rigorous and cell specific neurodevelopmental processes.

In conclusion, our findings suggest that DHEA and NGF cross-talk via their binding to NGF receptors to afford brain shaping and maintenance during development. During aging, the decline of both factors may leave the brain unprotected against neurotoxic challenges. This may also be the case in neurodegenerative conditions associated with lower production or action of both factors. DHEA analogs may represent lead molecules for designing non-endocrine, neuroprotective and neurogenic micromolecular NGF receptor agonists.

REFERENCES

1. Baulieu EE, Robel P (1995) Dehydroepiandrosterone (DHEA) and DHEA sulfate as neuroactive neurosteroids Proc Natl Acad Sci U. S. A. 95:4089-4091

2. Compagnone NA, Mellon SH (2000) Neurosteroids: biosynthesis and function of these novel neuromodulators. Front Neuroendocrinol 21:1-56.

3. Charalampopoulos I, Remboutsika E, Margioris AN, Gravanis A (2008) Neurosteroids as modulators of neurogenesis and neuronal survival. Trends Endocrinol Metab 19:300-307

4. Kimonides VG, Khatibi NH, Svendsen CN, Sofroniew MV, Herbert J (1998) Dehydroepiandrosterone (DHEA) and DHEA-sulfate (DHEAS) protect hippocampal neurons against excitatory amino acid-induced neurotoxicity. Proc Natl Acad Sci USA 95:1852-1857.

5. Bastianetto S, Ramassamy C, Poirier J, Quirion R (1999) Dehydroepiandrosterone (DHEA) protects hippocampal cells from oxidative stress-induced damage. Brain Res Mol Brain Res 66:35-41.

6. Lapchak PA, Chapman DF, Nunez SY, Zivin JA (2000) Dehydroepiandrosterone sulfate is neuroprotective in a reversible spinal cord ischemia model: possible involvement of GABA(A) receptors. Stroke 31:1953-1956

7. Belanger N, Gregoire L, Bedard PJ, Di Paolo T (2006) DHEA improves symptomatic treatment of moderately and severely impaired MPTP monkeys. Neurobiol Aging 27:1684-1693.

8. Charalampopoulos I, Tsatsanis C, Dermitzaki E, Alexaki VI, Castanas E, Margioris AN, Gravanis A (2004) Dehydroepiandrosterone and allopregnanolone protect sympathoadrenal medulla cells against apoptosis via antiapoptotic Bcl-2 proteins. Proc Natl Acad Sci USA 101:8209-8214.

9. Charalampopoulos I, Alexaki VI, Lazaridis I, Dermitzaki E, Avlonitis N, Tsatsanis C, Calogeropoulou T, Margioris AN, Castanas E, Gravanis A (2006) G protein-associated, specific membrane binding sites mediate the neuroprotective effect of dehydroepiandrosterone. FASEB J 20:577-579.

10. Charalampopoulos I, Margioris AN, Gravanis A (2008) Neurosteroid dehydroepiandrosterone exerts anti-apoptotic effects by membrane-mediated, integrated genomic and non-genomic pro-survival signaling pathways. J Neurochem 107:1457-1469

11. Reichardt LF (2006) Neurotrophin-regulated signaling pathways. Phil Trans R Soc B 361:1545-1564

12. Riccio A, Ahn S, Davenport M, Blendy J, Ginty D (1999) Mediation by a CREB family transcription factor of NGF-dependent survival of sympathetic neurons. Science 286:2358-2361

13. Crowley G, Spencer S, Nishimura M, Chen K, Pitts-Meek S, Armanini M, Ling L, McMahon S, Shelton D, Levinson A, Phillips H. (1994). Mice lacking nerve growth factor display perinatal loss of sensory and sympathetic neurons yet develop basal fore brain cholinergic neurons. Cell 78:1001-1011

14. Vilar M, Charalampopoulos I, Kenchappa RS, Simi A, Karaca E, Reversi A, Choi S, Bothwell M, Mingarro I, Friedman WJ, Schiavo G, Bastiaens PI, Verveer PJ, Carter BD, Ibáñez CF (2009) Activation of the p75 neurotrophin receptor through conformational rearrangement of disulphide-linked receptor dimers. Neuron. 62:72-83.

15. Smeyne RJ, Klein R, Schnapp A, Long LK, Bryant S, Lewin A, Lira S, Barbacid M (1994) Severe sensory and sympathetic neuropathies in mice carrying a disrupted Trk/NGF receptor gene. Nature 368:246 –249.

16. Fagan AM, Zhang H, Landis S, Smeyne RJ, Silos-Santiago I, Barbacid M. (1996) TrkA, but not TrkC, receptors are essential for survival of sympathetic neurons in vivo. J Neurosci. 16:6208-6218.

17. Zhou J, Vallena JS, Grimes ML, Mobley W (1995) Multiple levels for regulation of TrkA in PC12 cells by nerve growth factor. J Neurochem 65:1146-1156

18. Khursigara G, Ortinick JR, Chao MV (1999) Association of the p75 neurotrophin receptor with TRAF6. J Biol Chem 274:2597-2600

19. Khursigara G, Bertin J, Yano H, Moffett H, DiStefano PS, Chao MV (2001) A prosurvival function for the p75 receptor death domain mediated via the caspase recruitment domain receptor-interacting protein 2. J Neurosci 21:5854-5863

20. Yamashita T, Toyama M (2003) The p75 receptor acts as a displacement factor that releases Rho from Rho-GDI. Nat Neurosci 6:461-467

21. Laurine E, Lafitte D, Grégoire C, Sérée E, Loret E, Douillard S, Michel B, Briand C, Verdier JM. (2003) Specific binding of dehydroepiandrosterone to the N-terminus of the microtubule-associated protein MAP2. J Biol Chem. 278:29979-29986.

22. Jang SW, Okada M, Sayeed I, Xiao G, Stein D, Jin P, Ye K (2007) Gambogic amide a selective agonist for TrkA receptor that possesses robust neurotrophic activity, prevent neuronal cell death. Proc Natl Acad Sci USA 104:16329-16334

23. Jang SW, Liu X, Chan CB, Weinshenker D, Hall RA, Xiao G, Ye K (2009) Amitriptyline is a TrkA and TrkB receptor agonist that promotes TrkA/TrkB heterodimerization and has potent neurotrophic activity. Chem Biol. 16:644-656.

24. Barrett GL, Bartlett PF (1994) The p75 nerve growth factor receptor mediates survival or death depending on the stage of sensory neuron development. Proc Natl Acad Sci U S A. 91:6501-6505.

25. Yang SH, Perez E, Cutright J, Liu R, He Z, Day AL, Simpkins JW (2002) Testosterone increases neurotoxicity of glutamate in vitro and ischemia-reperfusion injury in an animal model. J. Appl. Physiol. 92:195–201.

26. Nuñez JL, McCarthy MM (2008). Androgens predispose males to GABAA-mediated excitotoxicity in the developing hippocampus. Exp. Neurol. 210:699–708.

27. Lupien SJ, McEwen BS, Gunnar MR, Heim C (2009) Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nat Rev Neurosci. 10:434-45.

28. Jeanneteau F, Garabedian M, Chao M (2008). Activation of Trk neurotrophin receptors by glucocorticoids provides a neuroprotective effect. Proc Nat Acad Sci USA 105:4862-4867

29. Lee JK, Everitt BJ, Thomas KL (2004) Independent cellular processes for hippocampal memory consolidation and reconsolidation. Science 304:839-843

30. Weill-Engerer S, David JP, Sazdovitch V, Liere P, Eychenne B, Pianos A,Schumacher M, Delacourte A, Baulieu EE, Akwa Y (2002). Neurosteroid quantification in human brain regions: comparison between Alzheimer's and nondemented patients. J Clin Endocrinol Metab. 87:5138-5143

31. Capsoni S, Ugolini G, Comparini A, Ruberti F, Berardi N, Cattaneo A (2000) Alzheimer-like neurodegeneration in aged antinerve growth factor transgenic mice. Proc Natl Acad Sci U S A. 97:6826-6831.

32. Villoslada P, Hauser SL, Bartke I, Unger J, Heald N, Rosenberg D, Cheung SW, Mobley WC, Fisher S, Genain CP (2000) Human nerve growth factor protects common marmosets against autoimmune encephalomyelitis by switching the balance of T helper cell type 1 and 2 cytokines within the central nervous system. J Exp Med. 191:1799-806.

33. Shimoke K, Chiba H (2001) Nerve growth factor prevents 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-induced cell death via the Akt pathway by suppressing caspase-3-like activity using PC12 cells: relevance to therapeutical application for Parkinson's disease. J Neurosci Res. 63:402-409.

34. Sivilia S, Giuliani A, Fernández M, Turba ME, Forni M, Massella A, De Sordi N, Giardino L, Calzà L (2009) Intravitreal NGF administration counteracts retina degeneration after permanent carotid artery occlusion in rat. BMC Neurosci. 10:52

35. Du C, Khalil MW, Sriram S (2001) Administration of dehydroepiandrosterone suppresses experimental allergic encephalomyelitis in SJL/J mice. J Immunol. 167:7094-7101.

36. Calogeropoulou T, Avlonitis N, Minas V, Pantzou A, Alexi X, Charalampopoulos I, Zervou M, Vergou V, Lazaridis I, Alexis MN, Gravanis A (2009) Novel dehydroepiandrosterone derivatives with anti-apoptotic, neuroprotective activity. J Med Chem, 52:6569-6587

37. Compagnone N, Bulfone A, Rubenstein J, Mellon S (1995) Steroidogenic enzyme P450c17 is expressed in the embryonic central nervous system. Endocrinology 136:5213-5223

38. Mizuta T, Kubokawa K (2007) Presence of sex steroids and cytochrome P450 genes in Amphioxus. Endocrinology 148:3554-3505

39. Benito-Gutierrez E, Nake C, Llovera M, Comella J, Garcia-Fernadez J (2005) The single AphiTrk receptor highlights increased complexity of neurotrophin signaling in vertebrates and suggests an early role in developing sensory neuroepidermal cells. Development 132:2191-2202

40. Hallbook F (1999) Evolution of the vertebrate neurotrophin and Trk receptor gene families. Curr Opin Neurobiol 9:616-621

41. Miller G (2009) On the origin of the nervous system. Science 325:24-26

MATERIALS AND METHODS

si/shRNAs, plasmids and antibodies.

PC12 cells were transfected with specific si/shRNAs for blocking the expression of TrkA and/or p75NTR receptors. More specific, three siRNAs and two shRNAs for TrkA and p75NTR respectively were obtained. The sequences for TrkA siRNAs (Ambion) were: GCCUAACCAUCGUGAAGAG (siRNA ID 191894), GCAUCCAUCAUAAUAGCAA (siRNA ID 191895) and CCUGACGGAGCUCUAUGUG (siRNA ID 191893). Sequences for p75NTR (Qiagen) were: GACCUAUCUGAGCUGAAA (CatNo SI00251090) and GCGUGACUUUCAGGGAAA (CatNo SI00251083).

Rat TrkA was expressed from the pHA vector backbone and rat p75NTR was expressed from the pCDNA3 vector backbone (InVitrogen) using a full length coding sequence flanked by an N-terminal hemagglutinin (HA) epitope tag. Plasmids to express RIP2 [19] and RhoGDI [36] were myc-tagged, while TRAF6 [19] was FLAG-tagged, as previously described.

The origin of antibodies was as follow: Bcl-2 (Cat.No. C-2, sc-7382, Santa Cruz Biotechnology Inc.), phospho TrkA (Cat.No. 9141, Cell Signaling), TrkA (CatNo. 2505, Cell Signaling, was used for Western Blotting and Cat.No. 06-574, Upstate, was used for immunostainings), p75NTR (Cat.No. MAB365R, Millipore), c-myc (CatNo 9E10, sc-40, Santa Cruz Biotechnology Inc.), phospho ERK1/2 (Cat.No. 9106, Cell Signaling), Erk1/2 (Cat.No. 9102, Cell Signaling), phospho-Shc (Tyr239/240) Antibody (Cat.No. 2434, Cell Signaling), Shc (Cat.No. 2432, Cell Signaling), phospho-Akt (Ser473) (Cat.No. 9271, Cell Signaling), Akt (Cat.No 9272, Cell Signaling), anti-FLAG (M2) mouse monoclonal (Cat.No. F1804, Sigma), pTyr (Cat.No. sc-508, Santa Cruz Biotechnology Inc.), active Caspase-3 (Cat. No. ab13847, Abcam), Tyrosine Hydroxylase (Cat. No. ab6211, Abcam), anti-rabbit-R-phycoerythrin conjugated (Cat.No. P9537, Sigma), anti-mouse-fluorescein conjugated (Cat.No. AP124F, Millipore), anti-rabbit Alexa Fluor 488 (Cat. No. A21206), anti-rabbit Alexa Fluor 546 (Cat.No. A10040), and GAPDH (Cat.No. 2118, Cell Signaling).

Cell cultures.

PC12 cells were obtained from LGC Promochem (LGC Standards GmbH, Germany), and nnr5 cells from Dr C.F. Ibáñez (Karolinska Institute). Both cell types were grown in RPMI 1640 containing 2mM L-glutamine, 15mM HEPES, 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 10% horse serum, 5% fetal calf serum (both charcoal-stripped for removing endogenous steroids) at 5% CO2 and 37°C. HEK-293 cells were obtained from LGC Promochem. Cells were grown in DMEM medium containing 10% Fetal Bovine Serum (charcoal-stripped for removing endogenous steroids), 100 units/ml penicillin and 0.1 mg/ml streptomycin, at 5% CO2 and 37°C. HEK-293 and PC12 cells were transfected with Lipofectamine 2000 (InVitrogen) according to manufacturer's instructions. Transfected cells were typically used on the 2nd day after transfection.

Measurement of apoptosis

PC12 cells were cultured in 6-well plates, 24h later they were transfected with the si/shRNAs for TrkA and/or p75NTR. Twenty hour hours later the medium was aspirated and replaced either with complete medium (serum supplemented) or serum free medium in the absence or the presence of DHEA or DHEA-BSA conjugate at 10 nM. Apoptosis was quantified 24h later with annexin V-FITC and PI (BD Pharminogen) according to our protocol [8].

[3H]DHEA binding assays

Membranes isolated from HEK293 cells transfected with the cDNAs of TrkA and p75NTR receptors were cultured in 225 cm2 flasks. After two washes with PBS, they were detached by vigorous shaking. After centrifugation at 1200 rpm, transfectants were homogenized in a 50 mM Tris–HCl buffer, pH 7.4 (at 4°C), containing freshly added protease inhibitors (1 mM PMSF and 1 μg/ml aprotinin). Crude membrane fractions were isolated by differential centrifugation at 2500×g (10 min at 4°C, to remove unbroken cells and nuclei) and 102,000×g (1 h, at 4°C). Membranes were washed once with ice-cold 50 mM Tris–HCl buffer, pH 7.4 and re-suspended in the same buffer. Membranes at 2mg protein/ml were incubated with 1 nM of [3H]DHEA in the absence (for total binding) or the presence of various unlabeled competitors (DHEA, NGF, testosterone (Testo), estradiol (E2), and dexamethasone (DEX)) at concentrations varying from 10−12 to 10−6 M. Binding was performed at a final volume of 100 μl at 37°C in a water bath for 30 min for HEK293TrkA and 60 min for HEK293p75NTR cell membranes. The reaction was stopped by rapid filtration through GF/B filters, pre-wetted in a 0.5% PEI solution for 30 min at 4°C. Filters were washed three times with ice-cold 50 mM Tris–HCl pH7.4, air-dried and counted in a β-scintillation counter (Perkin Elmer, Foster City, CA) with 60% efficiency for Tritium.

Fluorescence microscopy

HEK293 cells were allowed to grow on gelatin coated glass coverslips for 24hr in culture medium, and 24h later they were transfected with the cDNAs of TrkA, and p75NTR receptors or the vector alone. Staining was performed 48h after transfection. Culture medium was aspirated and transfectants were washed twice with PBS buffer. Primary antibodies against TrkA (rabbit, Upstate, No. 06-574, diluted 1:100) or p75NTR (mouse monoclonal ab, MAB365R, Millipore, dilution 1:500) were added for 30min at 37°C. Transfectants were also incubated with the DHEA-BSA-FITC or the BSA-FITC conjugates (10-7M) for 15min at 37°C in the dark, then they were washed with serum free culture medium and incubated for another 15min in serum free culture medium containing 4% BSA. Secondary antibodies, anti-rabbit-R-phycoerythrin conjugated (Sigma, No. P9537) and anti-mouse-fluorescein conjugated (No. AP124F, Millipore) were added at 1:100 dilution and transfectants were incubated for 30min at 37°C, then they were washed three times with PBS, and counterstained with Hoechst nuclear stain (Molecular Probes) for 5min. Coverslips were mounted to slides with 90% glycerin and were analyzed with a confocal laser scanning microscope (Leica TCS-NT, Leica Microsystems GmbH, Heidelberg, Germany).

Flow cytometry

HEK293 cells were cultured in 12-well plates, 24 hours later they were transfected with the cDNAs of TrkA and/or p75NTR receptors, or the vector alone. Staining was performed 48h later. Transfectants (5x105 cells) were peleted and incubated with 20μl of the primary antibodies against TrkA (rabbit, Upstate, No. 06-574, dilution 1:100) or p75NTR receptors (mouse, Millipore MAB365R, dilution 1:500), or with 20μl of DHEA-BSA-FITC (10-7M) for 30min at 4°C. Afterwards, transfectants were washed three times with PBS and 20μl of the secondary antibodies, anti-rabbit-R-phycoerythrin conjugated and anti-mouse-fluorescein conjugated were added at 1/100 dilution for 30min at 4°C. Transfectants were washed twice with PBS, resuspended in 500μl of PBS, and were analyzed by a Beckton-Dickinson FACSArray apparatus and the CELLQuest software (Beckton-Dickinson, Franklin Lakes, NJ).

Synthesis of immobilised DHEA-7-CMO

NovaPEG amino resin (loading value 0.78 mmol/g) was purchased from Novabiochem. NMR spectra were recorded on a Varian 300 spectrometer operating at 300 MHz for 1H and 75.43 MHz for 13C or on a Varian 600 operating at 600 MHz for 1H. 1H NMR spectra are reported in units of d relative to the internal standard of signals of the remaining protons of deuterated chloroform, at 7.24 ppm. 13C NMR shifts are expressed in units of δ relative to CDCl3 at 77.0 ppm. 13C NMR spectra were proton noise decoupled. IR spectra was recorded at Bruker Tensor 27. Absorption maxima are reported in wavenumbers (cm-1).

3β-Acetoxy-17,17-ethylenedioxyandrost-5-ene (0.74 g, 1.98 mmol) and N-hydroxy phthalimide (0.71 g, 2.2 mmol) were dissolved in acetone (39 mL) containing 1 mL of pyridine. The mixture was stirred vigorously at room temperature and sodium dichromate dihydrate (0.89 g, 3 mmol) was added. Additional portions of solid sodium dichromate dihydrate (0.89 g, 3 mmol) were added after 10 and 20 hours stirring at room temperature. After reaction completion (48hrs), the mixture was diluted with dichloromethane, filtered through a bed of celite and filtrate washed with water, saturated sodium bicarbonate solution and brine. The organic layer was dried over anhydrous sodium sulfate, the solvent evaporated in vacuo and the residue purified by flash column chromatography using hexane/acetone/ 25% NH4OH (85:15:0.1 mL) as eluent to afford 3β-acetoxy-17,17-ethylenedioxyandrost-5-ene-7-one ( 0.6 g, yield: 78%). 1H NMR (CDCl3, 300MHz) δ: 0.87 (s, 3H), 1.21 (s, 3H), 1.26-2.00 (m, 14H), 2.05 (s, 3H), 2.20-2.51 (m, 3H), 3.84-3.92 (m, 4H), 4.68-4.76 (m, 1H), 5.70 (d, J = 1.58 Hz, 1H).

To a solution of 3β-acetoxy-17,17-ethylenedioxyandrost-5-en-7-one (0.1 g, 0.26 mmol) in pyridine (1.9 mL) was added O-(carboxymethyl)hydroxylamine hemihydrochloride (0.11 g, 0.52 mmol) and the reaction mixture was stirred overnight under argon. After completion of the reaction, the solvent was evaporated and the residue was diluted with ethyl acetate. The organic layer was washed with water and brine, dried over anhydrous sodium sulfate and the solvent was evaporated in vacuo to afford 3β-acetoxy-17,17-ethylenedioxyandrost-5-en-7-one7-(O-carboxymethy1) oxime as a white foam (0.12 g, yield: 100 %). 1H NMR (CDCl3, 300MHz) δ: 0.88 (s, 3H), 1.13 (s, 3H), 1.16-1.95 (m, 12H), 2.04 (s, 3H), 2.25-2.59 (m, 5H), 3.84-3.95 (m, 4H), 4.59 (d, J = 2.29 Hz, 2H), 4.62-4.73 (m, 1H), 6.51 (d, J = 1.47 Hz, 1H).

To a solution of 3β-acetoxy-17,17-ethylenedioxyandrost-5-en-7-one-7-(O-carboxymethy1) oxime (0.12 g, 0.26 mmol) in a mixture of acetone/water (5:1, 6.3 mL) was added p-toluenesulfonic acid monohydrate (0.019 g, 0.10 mmol) and the reaction mixture was stirred until the starting material was consumed (48 hrs). The solvent was evaporated in vacuo and the residue was diluted with ethyl acetate. The organic layer was washed with water and brine, dried over anhydrous sodium sulfate and the solvent was evaporated in vacuo to afford 3β-acetoxy-androst-5-en-7,17-dione 7-(O-carboxymethy1) oxime as a white foam (0.11 g, yield: 100 %). 1H NMR (CDCl3, 600MHz) δ: 0.90 (s, 3H), 1.15 (s, 3H), 1.20-1.95 (m, 12H), 2.05 (s, 3H), 2.09-2.68 (m, 5H), 4.63 (d, J = 4.18 Hz, 2H), 4.65-4.71 (m, 1H), 6.56 (d, J = 1.39 Hz, 1H).

To a solution of 3β-acetoxy-androst-5-en-7,17-dione 7-(O-carboxymethy1) oxime (0.11 g, 0.26 mmol) in methanol (3.9 mL) was added LiOH (1.5mL, 1.5 mmol, 1N solution) and the reaction mixture was stirred until the starting material was consumed (4 hrs). The solvent was evaporated in vacuo and the residue was diluted with water. The solution was acidified with 10% hydrochloric acid and DHEA-7-CMO precipitated as a white solid, which was isolated by filtration (0.097 g, yield: 100 %). 1H NMR (CDCl3/CD3OD, 600MHz) δ: 0.90 (s, 3H), 1.14 (s, 3H), 1.20-2.75 (m, 17H), 3.49-3.54 (m, 1H), 4.54 (s, 2H), 6.54 (s, 1H).

3β-Hydroxy-17-oxoandrost-5-en-7-O-(carboxymethyl)oxime (DHEA-7-CMO) (192 mg, 0.511 mmol) in DMF (5 mL) was treated with HOBt (69 mg, 0.511 mmol) and DIC (0.08 mL, 0.511 mmol) and the resulting mixture was stirred at room temperature for 30 min. This solution was added to NovaPEG amino resin (130 mg, 0.102 mmol, 0.78 mmol/gr) (pre-swollen with DMF for 1 h) and the slurry was shaken at room temperature overnight. The mixture was filtered, the resin was sequentially washed with dichloromethane (3x), methanol (3x), and diethyl ether (3x) and was dried in vacuo overnight. Yield 175 mg (100%), loading value 0.61 mmol/gr. 13C NMR (gel phase, CDCl3) δ: 220.66, 170.15, 157.10, 154.15, 113.11, 72.57, 66.59, 49.92, 47.86, 42.15, 38.46, 37.08, 36.53, 35.49, 31.20, 30.71, 24.96, 20.15, 18.05, 13.95; IR: νmax/cm-1 2865 (s), 1735 (m), 1669 (w), 1653 (w), 1637 (w), 1456 (m), 1348 (w), 1289 (w), 1247 (w), 1093 (s), 946 (w).

Co-immunoprecipitation and pull-down assays

HEK293 cells were transfected with the appropriate plasmids (TrkA, p75NTR, RIP2, TRAF-6 and RhoGDI) by using Lipofectamine 2000 (Invitrogen). Cells were harvested 48 hours after transfection, suspended in lysis buffer (50 mM Tris-HCl, 0.15 M NaCl, 1% Triton-X100, pH 7.4) supplemented with protease inhibitors. Lysates were precleared for 1h with Protein A-Sepharose beads (Amersham) and immunoprecipitated with the appropriate antibody (pTyr, Flag or c-myc) overnight at 4°C. Protein A Sepharose beads were incubated with the lysates for 4h at 4°C with gentle shaking. In the case of immobilised DHEA-7-CMO, HEK293 or PC12 cells lysates or purified receptors (both from R&D Systems, Recombinant Mouse NGF R/TNFRSF16/Fc Chimera, Cat.No.: 1157-NR and Recombinant Rat Trk A/Fc Chimera, Cat.No.: 1056-TK), were incubated onernight at 4oC with the NovaPEG amino resin alone or conjugated with DHEA. Beads were collected by centrifugation, washed four times with lysis buffer, and resuspended in SDS loading buffer. Proteins were separated by SDS/PAGE, followed by immunoblotting with specific antibodies.

Western Blot Analysis

PC12 or HEK293 cells lysates were electrophoresed through a 12% SDS-polyacrylamide gel, then proteins were transferred to nitrocellulose membranes, which were processed according to standard Western blotting procedures, as previously described [8]. To detect protein levels, membranes were incubated with the appropriate antibodies: Bcl-2 (dilution 1:500), phospho TrkA (dilution 1:500), total TrkA (dilution1:500), p75NTR (dilution 1:500), phospho Shc (dilution 1:1000), total Shc (dilution 1:1000), phospho Akt (dilution 1:500), total Akt (dilution 1:500), phospho ERK1/2 (dilution 1:500) and total ERK1/2 (dilution 1:500). Proteins were visualized using the ECL Western blotting kit (ECL Amersham Biosciences, UK) and blots were exposed to Kodak X-Omat AR films. A PC-based Image Analysis program was used to quantify the intensity of each band (Image Analysis, Inc., Ontario, Canada).

To normalize for protein content the blots were stripped and stained with GAPDH antibody (dilution 1:1000); the concentration of each target protein was normalized versus GAPDH. Where phosphorylation of TrkA or kinases was measured, membranes were first probed for the phosphorylated form of the protein, then stripped and probed for the total protein.

SCG neuronal cultures

Superior cervical ganglia were removed from newborn (P0-P1) rat pups and dissociated in 0.25% trypsin (Gibco, 15090) for 30 minutes at 37oc. SCG neurons were pre-plated for 2 hours, in order to extract attached non-neuronal cells, the supernatant was gently aspirated and after a brief centrifugation neurons were re-suspended in culture medium (Gibco, Neurobasal Cat.No. 21103) containing 1% FBS, 100 units/ml penicillin, 0.1 mg/ml streptomycin, anti-mitotic 100nM and 100ng/ml NGF (Millipore, 01-125). Cells were plated on collagen coated 96 well plates and cultured for 5 days prior to use. For NGF withdrawal experiments, cells were washed twice with Neurobasal containing 1% FBS, and fresh culture medium lacking NGF and containing anti-NGF antibody at 1:50 dilution (Millipore, AB1526). DHEA, TrkA-inhibitor (Calbiochem, 648450) and anti-p75NTR (mouse, MAB365R Millipore) were used at 100nM, 100nM and 1:50 respectively.

In vivo experiments with the NGF null mice

NGF+/- mice [13] were obtained from the Jackson Laboratory and maintained on C57BL/6 background. All procedures described below were approved by the Animal Care Committee of the University of Crete School of Medicine. Animals were housed in cages maintained under a constant 12 h light–dark cycle at 21–23°C, with free access to food and tap water. Genotyping was performed on tail DNA using the following primers: NGFKOU2 (5'CCG TGA TAT TGC TGA AGA GC3'), NGFU6 (5'CAG AAC CGT ACA CAG ATA GC3') and NGFD1 (5'TGT GTC TAT CCG GAT GAA CC3'). Genomic PCR reactions containing the 3 primers were incubated for 32 cycles at 95°C (30 seconds)/59°C (30 seconds)/72°C (1 minute).

Mice heterozygous for the NGF null mutation were interbred to obtain mice homozygous for the NGF gene disruption and the first day of gestation determined by the discovery of a copulation plug. The mothers were treated daily with a subcutaneous injection of DHEA (2 mg/day) or vehicle (4.5% ethanol in 0.9% saline) starting from the third day after gestation. Animals were collected at E14. At the day of collection the mothers were deeply anesthetized with sodium pentobarbital (Dolethal 0,7 ml/kg i.p) and were perfused transcardially with saline solution containing heparin for about 15min and then undergone perfusion in 4% PFA, 15% Picric Acid, 0.05% GA in PB 0.1M, for another 15min. After the perfusion the embryos were collected and maintained in the same fixative over night at 4°C. Embryos were then washed in 0.1M PB and cryoprotected by using 10% sucrose followed by 20% sucrose over night at 4°C. Finally embryos were frozen in OCT in iso-pentane over liquid nitrogen for five minutes and the frozen tissues were stored for later use at -80oC. The samples were sectioned (20μm) and mounted onto Superfrost plus slides (Menzel-Glaser J1800AMNZ). Slides left to air-dry overnight at room temperature (RT), and were then either used immediately or were fixed in cold acetone for 1 minute and stored at -80oC for later use.

Stored or fresh slides fixed for 15 minutes in cold acetone at 4oC and left to dry for 10 minutes at RT. Slides were washed in PB 0.1 M then in TBS-T 0.1% and incubated for 45 min with 10% hors serum in TBS-T 0.1%. The normal serum was tipped off and the primary antibodies (anti-TrkA diluted 1:400 and active Caspase-3 diluted 1:50), diluted in TBS-T 0.1% with 1% hors serum, were added. Sections were incubated for 4h at RT and over night at 4oC and were then washed in TBS-T 0.1% and the secondary antibodies (Alexa Fluor 488 and Alexa Fluor 546, both anti-rabbit and diluted 1:1000) diluted in TBS-T 0.1%, were added and sections were incubated for 6h at RT. Sections washed in TBS-T, TBS and in PB 0.1 M and were coverslipped with Vectashield (Vector, H-1400) and analyzed in a confocal microscope. TUNEL (Roche, Cat.No 12156792910) and Fluoro-Jade C (Millipore, Cat.No. AG325) staining of apoptotic and degenerating neurons, respectively, were performed according to manufacturer instructions.

LEGENDS

1. RNA interference against NGF receptors affects the anti-apoptotic effect of DHEA. PC12 or PC12nnr5 cells were transfected with si/shRNAs of TrkA and/or p75NTR (panels a and b) and/or expressing vectors of TrkA (panel c). Twenty four hours later the medium was replaced either with complete medium (serum supplemented) or serum free medium in the absence or the presence of DHEA, DHEA-BSA (100 nM) or NGF (100 ng/ml). Apoptosis was quantified 24h later by FACS using Annexin V-FITC and PI. a) Upper panel: levels of apoptosis expressed as % of difference from serum supplemented cells [* P<0.01 versus control (serum free conditions), n:8]. Lower panel: representative FACS analysis of Annexin V-FITC and PI staining. b) Levels of Bcl-2 protein in serum deprived PC12 cells with or without DHEA treatment. Cellular extracts containing total proteins were collected and levels of Bcl-2 protein were measured by western blot, and normalized per GAPDH protein content. Upper panel: mean±SE of Bcl-2 levels, normalized against GAPDH (*P<0.01 versus control, n:4), lower panel: representative western blots of Bcl-2, TrkA, p75NTR and GAPDH proteins. c) Upper panel: levels of apoptosis in PC12nnr5 cells expressed as % of difference from serum deprivation condition. (*P<0.01 versus control-naive cells, n:4). Lower panel: western blots of TrkA, p75NTR and GAPDH proteins for each condition. (d), Serum deprived PC12 cells were exposed to 100 nM of DHEA or 100 ng/ml of NGF for 12, 14 and 48 hours, TrkA and p75NTR protein levels were measured in cell lysates with immunoblotting, using specific antibodies against TrkA and p75NTR proteins and were normalized against GAPDH (* P<0.01 versus control-Serum Free, n:5).

2. DHEA rescues TrkA positive primary sympathetic neurons from NGF deprivation-induced apoptosis, in a TrkA dependent manner. (a) Light microscopy photographs of organotypic cultures of rat SCG at P1 incubated for 48 hours in the presence of 100 ng/ml NGF, or in the same medium but lacking NGF and containing a polyclonal goat anti-rat NGF-neutralizing antiserum in the absence or the presence of 100 nM DHEA. (b) Light and Annexin V-FITC stained fluorescence microscopy photographs of dispersed primary sympathetic neurons in culture, isolated from rat SCG at P1. (c) Sympathetic neurons were incubated in the presence of 100 ng/ml NGF, or in the same medium but lacking NGF and containing a polyclonal rabbit anti-NGF-neutralizing antiserum and or 100 nM DHEA in the absence or the presence of TrkA-inhibitor or a mouse anti-p75NTR-neutralizing antibody. The results shown are the mean±SE from 3 separate experiments were over 300 neurons were counted in 6 to 7 randomly selected optical fields (*P<0.01 versus anti-NGF condition). Photograph (d) depicts tyrosine hydroxylase (TH) staining of sympathetic neurons.

3. DHEA binds with high affinity to HEK293TrkA and HEK293p75NTR cell membranes. (a) Competition binding experiments were performed using isolated membranes (at a final concentration 2 mg protein/ml) from HEK293 cells transfected with the plasmid cDNAs of TrkA and p75NTR receptors, incubated for 30 min with 5 nM [3H]DHEA in the absence (for total binding) or the presence of various unlabeled steroids (DHEA, testosterone (Testo), estradiol (E2) dexamethasone (Dex) or NGF at concentrations varying from 0.01 to 1000 nM. (b, c) Fluorescence localization of membrane binding of DHEA on HEK293 cells transfected with the plasmid cDNAs of TrkA and p75NTR receptors. Transfectants were incubated with either the membrane impermeable DHEA-BSA-FITC conjugate (100 nM), BSA-FITC (100 nM) or with specific antibodies against TrkA and p75NTR proteins. Transfectants were analyzed under the confocal laser scanning microscope (b) or by FACS analysis (c). Blue staining in (b) photographs depicts Hoechst nuclear staining.

4. Immobilised DHEA pulls down TrkA and p75NTR receptors. Covalently linked DHEA-7-O-(carboxymethyl) oxime DHEA-7-CMO) to polyethylene glycol amino resin (NovaPEG amino resin) was incubated with recombinant TrkA and p75NTR proteins (a) or with cell extracts isolated from HEK293TrkA HEK293p75NTR transfectants (b), PC12 cells and whole rat brain (c). Precipitation experiments (panels marked with A) and western blot analysis (panels marked with B) of precipitates with specific antibodies against TrkA and p75NTR proteins were performed as described in Materials and Methods. DB: DHEA-7-O-(carboxymethyl) oxime DHEA-7-CMO) polyethylene glycol amino resin, B: polyethylene glycol amino resin, P: pellet, S: supernatant.

5. DHEA induces TrkA- and p75NTR-mediated signaling. (a) HEK293TrkA transfectants were exposed for 10 and 20 min to 100 nM of DHEA or 100 ng/ml of NG