Gonadotropin-Releasing Hormone-Expressing Neurons Immortalized Conditionally Are Activated by Insulin: Implication of the Mitogen-
http://www.100md.com
《内分泌学杂志》
Service of Endocrinology, Diabetology
Metabolism (R.S., E.C., M.-J.V., M.G., J.-P.R., R.C.G., F.P.P.) and Service of Internal Medicine (P.V.), Department of Medicine, University Hospital, 1011 Lausanne, Switzerland
Service of Endocrinology (F.P.P.), Diabetology and Metabolism, University Hospital, 1211 Geneva, Switzerland
Abstract
Energy balance exerts a critical influence on reproduction via changes in the circulating levels of hormones such as insulin. This modulation of the neuroendocrine reproductive axis ultimately involves variations in the activity of hypothalamic neurons expressing GnRH. Here we studied the effects of insulin in primary hypothalamic cell cultures as well as a GnRH neuronal cell line that we generated by conditional immortalization of adult hypothalamic neurons. These cells, which represent the first successful conditional immortalization of GnRH neurons, retain many of their mature phenotypic characteristics. In addition, we show that they express the insulin receptor. Consistently, their stimulation with insulin activates both the phosphatidylinositol 3-kinase and the Erk1/2 MAPK signaling pathways and stimulates a rapid increase in the expression of c-fos, demonstrating their responsiveness to this hormone. Further work performed in parallel in immortalized GnRH-expressing cells and primary neuronal cultures containing non-GnRH-expressing neurons shows that insulin induces the expression of GnRH in both models. In primary cultures, inhibition of the Erk1/2 pathway abolishes the stimulation of GnRH expression by insulin, whereas blockade of the phosphatidylinositol 3-kinase pathway has no effect. In conclusion, these data strongly suggest that GnRH neurons are directly sensitive to insulin and implicate for the first time the MAPK Erk1/2 signaling pathway in the central effects of insulin on the neuroendocrine reproductive axis.
Introduction
HYPOTHALAMIC GnRH neurons exert a biological function crucial to species survival: they represent the center of integration of the neuroendocrine reproductive axis (1, 2). Their coordinate activation is a prerequisite to the awakening of that axis at puberty as well as continuing reproductive competence in adulthood. Such activation is critically dependent on adequate body energy stores (3), as exemplified by the observation that unfavorable metabolic conditions impede on reproductive activity (4). This inhibition is likely mediated at the level of the hypothalamus via mechanisms implicating the existence of factors signaling the nutritional status of an individual to the central nervous system (5).
Leptin is primarily a central inhibitor of food intake, which also participates in the regulation of reproductive activity by metabolic changes (6, 7). These effects are mediated, at least partially, via hypothalamic neuropeptide Y (NPY)-ergic neurons and pathways (8, 9, 10, 11, 12). In recent years evidence that insulin can play very similar roles has accumulated: like leptin, insulin is transported actively across the blood-brain barrier (13), and has the potential to function as a satiety signal in the hypothalamus (14, 15). In addition, mice harboring a neuron-specific deletion of the insulin receptor also suffer from central hypogonadism (14). This newly discovered role of insulin in the regulation of reproduction is further underlined by the observations, in human and animal models, that insulin-dependent diabetes mellitus is associated with reproductive abnormalities resulting from impaired LH secretion (16, 17). Consistent with this observation, we have demonstrated that insulin activates the neuroendocrine reproductive axis of normal mice in vivo and that this effect involves the stimulation of hypothalamic GnRH expression and secretion (18). However, the precise hypothalamic target neurons of insulin, or the cellular and molecular mechanisms mediating its reproductive effects, remain largely unknown.
We describe here the successful generation of GnRH-expressing cells by conditional immortalization of adult hypothalamic neurons. We then used these immortalized GnRH-expressing cells in parallel with primary cultures of hypothalamic neurons (19) to ascertain the potential direct effects of insulin on GnRH-expressing neurons. Taken together, our data suggest that GnRH neurons themselves may be directly modulated by insulin. If confirmed in vivo, this could provide a novel mechanism underlying the permissive effects exerted by insulin on the gonadotrope axis (14, 18). In addition, we have demonstrated that the intracellular mechanisms involved in the stimulation of GnRH gene expression by insulin implicate the MAPK Erk1/2 signaling pathway. This latter finding is contrasting to data regarding the action of insulin on feeding, which has been linked to the phosphatidylinositol 3-kinase (PI3K) pathway (20). We suggest that the cell lines described here will constitute a useful novel tool for the study of the activation of hypothalamic GnRH neurons at the cellular and molecular levels.
Materials and Methods
All these experiments were approved by both the institutional and state ethical committees on animal research.
Conditional immortalization of GnRH neurons
Twelve different clonal cell lines were derived from primary cultures of adult rat hypothalamic neurons. Hypothalami were obtained from 10- to 12-wk-old female Wistar rats: after killing by decapitation, whole brains were rapidly removed from the skull, and the hypothalamus dissected out by two coronal cuts (anteriorly at the anterior border of the optic chiasma and posteriorly at the level of the mammillary bodies) followed by two parasagittal cuts along the hypothalamic sulci and a final cut dorsally, at a depth of approximately 3 mm from the ventral surface of the tissue block. Cells were dispersed mechanically in PBS buffer (PBS without calcium or magnesium; Gibco, Gaithersburg, MD) supplemented with 0.06% glucose (Fluka Chemie GmbH, Buchs, Switzerland) and 100 U per 100 μg/ml penicillin/streptomycin (Seromed, Flakola, Basel, Switzerland). Freshly excised whole hypothalami were gently passed several times through Pasteur pipettes flamed in the middle of the procedure to decrease the diameter of their opening. Nondispersed tissue was allowed to settle for 5 min and supernatant transferred to a clean tube. The remaining pellet was resuspended in 4 ml of PBS buffer and mechanical dispersion repeated. Supernatants from first and second dispersion were mixed, centrifuged for 5 min at 100 x g and the cellular pellet gently resuspended in 5 ml of neurobasal medium (Gibco) and 0.04% B27 supplement (Gibco BRL, Basel, Switzerland) containing 500 μM glutamine and 25 μM glutamate (Sigma, Buchs, Switzerland). After dispersion, cells were plated at a density of 400,000 live cells/well in six-well plates (Corning Costar Corp., Cambridge, MA) coated with 5 μg/ml poly-D-lysine (Sigma, Fluka Chemie) and grown in neurobasal A medium (Invitrogen AG, Basel, Switzerland) with B27 supplement (Invitrogen) containing 500 μM glutamine and 25 μM glutamate (Sigma, Fluka Chemie) as described (18, 19). Half of the medium was changed every fourth day. Cells were immortalized 10 d after dispersion by the transduction of two different lentiviral vectors (Fig. 1A).
By fusing a 2-kb fragment of the rat GnRH promoter to the reverse tetracycline transactivator (rtTA) gene, the first vector was designed to direct the expression of the rtTA in GnRH-expressing neurons. In the second vector, the tetracycline-responsive element (TRE) (21) was fused to the v-myc oncogene, immediately followed by the minigene encoding resistance to puromycine. With this design, the addition of low concentrations of tetracycline in the culture medium was expected to drive the transcriptional activation of v-myc in GnRH-expressing cells infected with both lentiviral vectors.
The first vector (pSIN-GnRHprom-rtTA-WHV), carrying a modified Tet-On construct (21) in which the cytomegalovirus (CMV) promoter is deleted and replaced by the rat GnRH promoter, was constructed in three phases. First, the proximal 2-kb fragment of the rat GnRH promoter was generated by PCR from genomic DNA, cloned in pGEM-T Easy (Promega, Catalys AG, Wallisellen, Switzerland), and verified by sequencing. Primer pairs used were: 5'-ATACTAGTGCTTATTAGCCAGTGAGCAGCAG-3' (sense) and 5'-AATTGATCACAAGCTTCTGTCAGTAGGTTGAG-3' (antisense, bold letters stand for the restrictions sites SpeI and BclI). Then this fragment was removed by digestion with SpeI and HindIII, the latter site blunted, and subcloned into the pTet-ON plasmid (21) at the blunted EcoRI and SpeI sites in place of the CMV promoter, in the appropriate direction in front of the rtTA element. A fragment containing the 2-kb GnRH promoter and the rtTA element was finally removed from the pTet-ON plasmid by digestion with SpeI and BamHI and inserted at the same restriction sites in the plasmid pSIN-MCS-WHV (22).
For the second vector, the tetracycline response element (TRE) and the MC29 v-myc allele located downstream in the plasmid pUHD10–3myc (23) were removed by digestion with XhoI (blunted) and BamHI and inserted into the second plasmid, pSIN-MCS-IRES-PURO (22), at the sites NotI (blunted) and BamHI. As a result, the TRE is cloned in front of the avian v-myc used as the immortalizing gene and the gene coding for resistance to puromycine.
Recombinant self-inactivating (SIN) lentiviruses (24) were produced by transient cotransfection of 293T cells with a three-plasmid system. Briefly, 60–70% confluent 293T cells, cultured on 10-cm dishes, were cotransfected with the following plasmids: 5 μg of the packaging construct pCMVR8.91, 5 μg of the vesicular stomatitis virus G protein envelope pMD.G, and 15 μg of either of the transfer vectors pSIN-GnRHprom-rtTA or pSIN-TRE-v-myc-IRES-PURO, by using the calcium phosphate technique as described (25). Media were replaced after 16 h and conditioned media collected 48 h after transfection, filtered on 0.45-μm pore size filters (Millipore AG, Volketswil, Switzerland), and either immediately used to infect the hypothalamic neurons or snap frozen in LN2 for later use. To infect the neuronal cultures, equal amounts in volume (1:1) of conditioned media derived from the two lentiviral vectors pSIN-GnRHprom-rtTA and pSIN-TRE-v-myc-IRES-PURO were added to the culture, in the presence of polybrene (8 μg/ml, Sigma, Fluka).
After 48 h of transduction, the culture medium was changed, and stimulation with doxycycline (5 μg/ml, Sigma, Fluka) as well as selection with puromycine (0.5 μg/ml, Sigma, Fluka) was started. Selection with puromycine was carried out for 8 wk before stopping, and stimulation with doxycycline was continued thereafter. After the end of the selection process, 12 different clones were isolated by dilution and following the soft agar method (26). These clones were named GnRH/v-myc (Gnv) cells and numbered from 1 to 12. GnRH expression at the mRNA level was found maximal in clones 3 and 4. Subsequent work reported here was performed using clone 3 (Gnv-3 cells), and clone 4 (Gnv-4 cells) is described in the companion manuscript of this paper (27).
Culture and characterization of Gnv-3 cells
Gnv-3 cells are expanded in a proliferation medium composed as follows: neurobasal A medium with B27 supplement and containing 1% FBS (Invitrogen), 5 μg/ml doxycycline, penicillin and 5 ng/ml bFGF (Invitrogen). They are passaged by trypsin digestion (Invitrogen) every 4–5 days, and several aliquots of the same passage (passage 3 or 4) are frozen and conserved. This procedure allows to use cells at the same early passage for all experiments. For each experiment, one of these aliquots is rewarmed, plated in proliferation medium and allowed 4–5 days before passage. After passage, the cells are plated in 6-well plates or in 10 cm Petri dishes in differentiation medium containing: neurobasal A medium alone, with glutamine (Invitrogen) and bFGF (Invitrogen). This medium was named differentiation medium because under these culture conditions, the shape of the cells changes: their bodies become thinner and elongated, and they develop many intercellular connections. Because of the high glucose content of neurobasal A (25 mM), the medium was changed again, 16–24 h before the experiments, for DMEM low glucose (5 mM glucose, Invitrogen). All experiments were performed using these culture conditions.
Cell growth
Cell growth in the proliferation medium (presence of doxycycline) and the differentiation medium (absence of doxycycline) was evaluated following a modification of the tetrazolium colorimetric assay (28), which is based on the ability of living cells to reduce the yellow salt 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide to formazan. Cells were plated at precisely 50,000 cells/well in 6-well plates. On d 0 of the experiment and every 24 h thereafter for 6 consecutive days, 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide (0.5 mg/ml, Sigma, Fluka) was added to the medium in triplicate wells. Four hours later, the supernatants of these triplicate wells were gently aspirated and the newly formed formazan crystals dissolved by the addition of 1.5 mL of dimethylsulfoxide [100%, Merck (Suisse) SA, Geneva, Switzerland] with 25 μl of glycine buffer [0.1 M glycine (pH 10.5); Merck]. Absorbance was then measured at 540 nm in a spectrophotometer (Uvicon 940, Kontron, Eching, Germany). This experiment was performed on two separate occasions.
RT-PCR and quantitative RT-PCR
For these experiments, cells in the differentiation medium were grown either in 10-cm petri dishes (RT-PCR) or 6-well plates (quantitative RT-PCR). At the end of the experiment, total RNA was extracted using commercially available reagents (TriPure reagent, Roche Diagnostics, Rotkreuz, Switzerland). For RT-PCR experiments, reverse transcription with SuperScript II (Invitrogen) was performed with random primers, and the primer pairs used in the various PCRs are listed in Table 1, together with gene accession numbers. PCRs were performed on a GeneAmp PCR system 9700 thermocycler (PE Applied Biosystems, Rotkreuz, Switzerland).
Quantitative RT-PCR experiments are performed using the LightCycler technology (Roche Diagnostics) with SYBR green detection as described (29). Reverse transcription was done with random primers. Then a standard curve was created with serial dilutions of RNA extracts from whole hypothalami. Different dilutions of the samples were always tested in preliminary experiments to ensure that quantification would be performed within the linear part of this standard curve. After this test, all samples were quantified in at least three different runs. For quantification purposes, the mRNA levels of the gene of interest were always reported to the levels of 2-microglobulin, a constitutively expressed gene. A value of 100% was then attributed to the control sample in each separate experiment, and all measures were reported as a percent of this control.
Immunohistochemistry
Immunohistochemistry was performed at room temperature as described (19), using an avidin-biotinylated horseradish peroxidase macromolecular complex kit (ABC Elite, Vector, Burlingame, CA). Cells were fixed for 40 min in 4% paraformaldehyde (Merck) and rinsed three times in PBS and one time in 0.05 M Tris (Sigma) (pH 7.4), 0.6% NaCl, 0.3% Triton X-100 (Merck), and 10% goat serum (Gibco). After quenching of endogenous peroxidase activity, nonspecific binding was blocked by incubating 1 h in 0.05 M Tris (Sigma) (pH 7.4), 0.6% NaCl, and 0.3% Triton X-100 (Merck). Incubations with the primary antiserum (polyclonal GnRH antiserum, Bio-Yeda Ltd., Rehovot, Israel) were performed overnight at a final dilution of 1:400, and then cells were washed and further incubated 1 h with a biotinylated secondary antibody (Jackson, West Grove, PA). After three washes in 0.1 M Tris-HCl (pH 7.7), cells were incubated in the ABC reagent, washed again, and finally incubated in diaminobenzidine peroxidase substrate solution [50 mg per 100 ml 3–3' diaminobenzidine (Sigma) in 0.1 M Tris-HCl, 0.018% H2O2 (pH 7.7)]. Negative controls were obtained by incubating the primary antiserum with excess GnRH before use.
Western blotting
The expression of v-myc and the insulin receptor (IR) was also assessed by Western blot from cells plated at a density of 40,000 cells/well in 6-well plates. On the day of the experiment, cells were rinsed with ice-cold PBS and lysed with the radioimmunoprecipitation assay buffer [10 mM Tris (pH 7.5), 100 mM NaCl, 1 mM EDTA, 0.5% Na-deoxycholate, 0.1% sodium dodecyl sulfate, 1% Triton X-100] supplemented with complete protease inhibitor cocktail (Roche Biochemicals, Rotkreuz, Switzerland). After this procedure, lysates were homogenized by sonication during 30–60 sec. Cells debris was removed by centrifugation at 14,000 rpm for 15 min at 4 C. The protein concentration of the supernatant was determined using commercially available reagents (Bio-Rad protein assay, Bio-Rad, Reinach, Switzerland). Proteins were resolved by 10% SDS-PAGE electrophoresis and transferred to a polyvinylidine difluoride membrane (Hybond-P, Amersham Biosciences, Otelfingen, Switzerland). Blots were blocked for 2 h at room temperature in PBS containing 5% nonfat dried milk and 0.05% Tween 20. Then they were incubated overnight at 4 C in the same buffer containing the primary antibody.
For v-myc, experiments were performed either with proliferating cells or after the change for the differentiation medium, using a monoclonal antibody (Oncogene Research Products, La Jolla, CA) at a final dilution of 1:1000. For the IR, experiments were performed only with cells in the differentiation medium, using a monoclonal antibody directed against its -subunit (Oncogene Research Products) at a final dilution of 1:1000. Equal amounts of proteins were resolved by 10% SDS-PAGE electrophoresis and transferred to a PDVF membrane (Hybond-P, Amersham Biosciences) and antibody complexes visualized by the enhanced chemiluminescence system (Amersham, Arlington Heights, IL).
Secretion experiments
Static secretion experiments were performed in 6-well plates, and perifusion experiments were performed with cells grown onto plastic coverslips (Assistant, Karl Hecht BmgH & Co. KG, Altnau, Switzerland) coated with 5 μg/ml poly-D-lysine (Sigma, Fluka Chemie) as described (18, 19). We used two coverslips in each perifusion chamber, with approximately 400,000 cells/coverslip. The perifusion rate was set at 150 μl/min, and fractions were collected every 4 min. During the experiment, the perifusion medium was constantly gassed with a mixture of 95% O2 and 5% CO2. GnRH was directly assayed in the medium, or in the effluent, by RIA as previously described (18), using the antiserum used for immunohistochemistry. The limit of detection of this assay is 2 pg/ml.
Experimental procedures
The responsiveness of Gnv-3 cells to N-methyl D-aspartate (NMDA) and nicotine, both recognized activators of GnRH secretion in adult GnRH neurons (30, 31, 32), was assessed in static cultures and, for nicotine, in perifusion experiments. Cells in static culture were exposed for 2 h to either NMDA or nicotine at concentrations ranging from 10–6 to 10–3 M. At the end of the incubation period, GnRH was measured in the medium by RIA. In addition, the effect of NMDA on GnRH gene expression was also studied by incubating the cells for 6 h with NMDA (5 x 10–5 M) and measuring mRNA levels of GnRH in total RNA extracts by real-time quantitative RT-PCR.
To investigate the intracellular signaling pathway(s) potentially activated by insulin, Gnv-3 cells in the differentiation medium (see above) were stimulated with insulin at a concentration of 4 x 10–5 M for various time periods up to 30 min. The use of this insulin concentration was based on preliminary experiments as well as our previous work in primary neuronal cell cultures (18). The activation of the MAPK Erk1/2 pathway or the PI3K pathway was assessed by Western blot. For Erk1/2, membranes were first incubated as described above with a monoclonal antibody against phosphospecific Erk1/2 (Cell Signaling Technology, Inc., Beverly, MA). Data were quantified by densitometric analysis of autoradiograms, using a computerized densitometer (Typhoon system, Molecular Dynamics, Inc., Sunnyvale, CA). Then the membranes were stripped and hybridized again with a monoclonal antibody directed against total Erk1/2 (Cell Signaling Technology) to allow normalization. The activation of the PI3K-dependent pathway was evaluated similarly by measuring the phosphorylation of its downstream substrate Akt. Immunodetection was first performed with antibodies specific for phospho-Akt (Ser 473) (New England Biolabs, Beverly, MA). After quantification, membranes were stripped and incubated with antibodies against total Akt forms (Santa Cruz Biotechnology, Santa Cruz, CA) to allow normalization. For all Western blot data, the ratio of phosphorylated over total enzyme was calculated at each time point and expressed as a percent of baseline. Time 0 was arbitrarily considered as 100%.
The expression of suppressor of cytokine signaling (SOCS)-3 was then evaluated in Gnv-3 cells at different time points between 30 and 120 min after stimulation with the same concentration of insulin (4 x 10–5 M). Similar stimulations with insulin (4 x 10–5 M) were performed in parallel at the same time points to study its effects on the expression of the immediate early genes c-fos and Egr-1, two markers of neuronal activation. Finally, the effects of insulin on the expression of GnRH were also evaluated in Gnv-3 cells. These effects were tested at various time courses between 2 and 24 h, with insulin concentrations ranging between 4 x 10–5 and 4 x 10–8 M. Because insulin was found to stimulate the expression of GnRH at relatively high concentrations, similar experiments were then performed with IGF-I at concentrations between 10–7 and 10–9 M. Effects of IGF-I were checked at 2 h, the time point at which insulin had the highest effect. mRNA levels of SOCS-3, c-fos, Egr-1, and GnRH were measured by semiquantitative, real-time RT-PCR of total mRNA extracts as described above.
Next, the role in the stimulation of GnRH expression by insulin played by the two pathways identified in Gnv-3 cells (PI3K and Erk1/2) was further evaluated in primary hypothalamic neuronal cell cultures (19). First, their stimulation by insulin (4 x 10–4 M) was confirmed in primary neurons. Erk1/2 phosphorylation was assessed by Western blot, and PI3K activity determined as previously described (33) was used as a measure of PI3K stimulation. Then we used enzymatic inhibition of either the MAPK or PI3K pathway to evaluate their respective role in the stimulation of GnRH gene expression by insulin. The effect of insulin (4 x 10–4 M) on GnRH mRNA levels was studied in cells either pretreated for 30 min or not with the PI3K inhibitor wortmannin (1 μM, Sigma, Fluka) or pretreated for 30 min or not with the Erk1/2 inhibitor PD98059 (25 μM, Sigma, Fluka).
Data analysis
Pulse detection in perifusion experiments was performed with Cluster 8, a model-free computerized pulse analysis algorithm to identify statistically significant pulses in relation to dose-dependent measurement error in a hormone time series (34). We used the software package Pulse_xp (2004 version; University of Virginia, Charlottesville, VA) running under Windows XP. Criteria used for pulse identification were the following: 1) individual test cluster sizes of three consecutive points for both the nadir and the peak width; 2) a minimum and maximum intraseries coefficient of variation of 10 and 15%, respectively; and 3) a t statistic to identify a significant increase and a t statistic to define a significant decrease.
All results are expressed as mean ± SEM, and a minimum of three independent experiments were always performed. Statistical significance of all results was assessed by ANOVA, with P < 0.05 considered a significant difference.
Results
Conditional immortalization and characterization of Gnv-3 cells
Figure 1, B and C, summarize the initial characterization of the 12 clones obtained with this approach. Figure 1B represents the results of PCR experiments performed on their genomic DNA (labeled 1–12). These data demonstrate that the rtTA as well as the GAG lentiviral gene are both integrated in the genome of all 12 clones but that the v-myc oncogene is not found in clone 8. Figure 1C displays the results of RT-PCR experiments performed on total RNA samples of the same 12 clones. Consistent with the above results, clone 8 does not express neuron-specific enolase (NSE), a neuronal marker (35). The rest of the work presented here was performed with clone 3 (called Gnv-3 cells, clone 3).
Table 2 summarizes the characterization that was further performed on Gnv-3 cells by RT-PCR. As shown, these cells express the light chain of neurofilaments (NFL), which is another marker of well-differentiated neurons (36), as well as several markers of neurosecretory neurons such as synaptobrevin, synapsin 1, synaptophysin, synaptosomal-associated protein of 25 kDa (SNAP25) vesicle-associated membrane protein (VAMP)1 and VAMP2, or chromogranin B (37, 38, 39, 40, 41). In contrast, a marker of poorly differentiated neurons, the microtubule-associated protein (MAP)-2b isoform of MAP-2 (42), is not expressed by Gnv-3 cells. Similarly, we found no expression of glial cell markers in this clone, further confirming its well-differentiated neuronal phenotype. Finally, among the various neuroendocrine markers that were tested, we found that Gnv-3 cells express exclusively GnRH, both at the mRNA and the protein levels.
This is further exemplified in Fig. 2A, which demonstrates the uniform positive immunostaining for GnRH observed in these cells. Their GnRH secretion was then assessed in perifusion experiments. Figure 2B illustrates the pattern of basal GnRH secretion observed. Significant pulses identified by Cluster 8 are indicated by a solid line on top of the plot. When summarizing the data from three independent perifusion studies, the mean basal frequency of GnRH secretion from Gnv-3 cells was 2.2 ± 0.68 pulses/h, with a mean interpulse interval of 24.69 ± 5.69 min. The mean increase over calculated baseline of these peaks was 3.17 ± 0.1 pg/ml.
Figure 2C illustrates the significant increase in GnRH mRNA levels observed after 6 in Gnv-3 cells exposed to NMDA (5 x 10–5 M). This increase in GnRH gene expression after exposure to NMDA (5 x 10–5 M) is also accompanied by an increase in GnRH secretion (data not shown). Similarly, stimulation of Gnv-3 cells with nicotine (10–6 M), either in static cultures or perifusion experiments, was observed to stimulate GnRH secretion acutely (data not shown). Finally, Fig. 2D demonstrates that the proliferation of Gnv-3 cells is dependent on the presence of doxycycline in the culture medium, which consistently stimulates the expression of v-myc (Fig. 2D, inset). Without doxycycline, v-myc is not detected by Western blot (Fig. 2D, inset) and the proliferation of the cells is markedly lower.
IRs and leptin receptors
Figure 3A demonstrates that the IR is expressed at the mRNA level (RT-PCR) and the protein level (Western blot). However, we failed to demonstrate consistently the presence in this clone of the long isoform of the leptin receptor (Ob-Rb) by RT-PCR. The level of expression of the IR in Gnv-3 cells, as assessed by Western blot, is comparable with that observed in fresh hypothalamic extracts and lower than that observed in muscle or brain. We then investigated the intracellular signaling pathway(s) that may be linked with the IR in these cells. Figure 3B illustrates the rapid, strong, and long-lasting stimulation of the phosphorylation of the MAPKs Erk1/2 that follows exposure of Gnv-3 cells to insulin. Similarly, the phosphorylation of AKT, downstream of PI3K, is rapidly and strongly stimulated by insulin in the same cells (Fig. 3C). Finally, SOCS-3 mRNA levels are increased after exposure to insulin, as reported for other insulin-sensitive tissues (43) (Fig. 3D).
Insulin effects in Gnv-3 cells
The results illustrated in Figs. 4, A and B, show that insulin (4 x 10–5 M) induces a rapid and significant induction of the immediate early genes c-fos and Egr-1, as reflected by the 2- to 3-fold increases measured in their respective mRNA levels. The induction of c-fos is stronger than that observed with Egr-1. In contrast, the stimulation of Egr-1 expression by insulin is of a longer duration: Egr-1 mRNA levels are still significantly higher than baseline 2 h after the stimulation. Finally, Fig. 4C displays the effects of insulin on GnRH expression in these cells. These data demonstrate the significant and dose-dependent increases in GnRH mRNA levels observed after insulin stimulation. This effect is rapid, already maximal at 2 h (as shown in the figure). A similar stimulation was still observed after 4 h of incubation with insulin, but after 24 h, no effect of insulin was noted in Gnv-3 cells (data not shown). In contrast, IGF-I at the concentrations tested did not induce any significant change in GnRH mRNA levels in our model: after 2 h of incubation with IGF-I at the concentration of 10–7 M, these levels were 161.2 ± 25.7% of control (n = 4), and they were 139.0 ± 19.0% of control after 2 h in the presence of IGF-I at the concentration of 10–8 M.
Insulin signaling in primary hypothalamic neurons
Previous work from our group had demonstrated that insulin stimulates GnRH gene expression in primary hypothalamic neuronal cell cultures (18). Therefore, we next examined in primary neurons the respective importance of the two signaling pathways activated in Gnv-3 cells. Figure 5 summarizes our findings. Figure 5A illustrates the very strong and long-lasting increase of PI3K activity observed in primary hypothalamic neurons stimulated by insulin: this activity is still significantly higher than control after 2 h. Results reported in this graph were obtained after immunoprecipitation of the cell extracts with an antiinsulin receptor substrate (IRS)1 antibody, but very similar results were observed after immunoprecipitation with an anti-IRS2 antibody (data not shown). Despite this PI3K activation, its inhibition by pretreatment with wortmannin did not block the GnRH mRNA levels increase by insulin seen in this model (Fig. 5B). PI3K activity inhibition in these conditions was demonstrated by showing complete inhibition of the downstream Akt serine/threonine kinase (data not shown). Therefore, the potential role of the MAPK Erk1/2 pathway was then evaluated in these neurons, using a similar approach. As anticipated from results obtained in Gnv-3 cells, data reported in Fig. 5C show that this pathway is also stimulated by insulin in primary hypothalamic neurons. However, and in sharp contrast to the observation made with wortmannin, inhibition of the phosphorylation of Erk1/2 by pretreatment of the cells with the enzyme inhibitor PD 98059 completely abolished the effect of insulin on GnRH expression.
Discussion
Metabolic changes are known to alter the central nervous system control of reproduction through variations in circulating levels of peripheral signals. Leptin as well as insulin is among the metabolic factors involved (6, 14, 18), acting via a modulation of the activity of highly specialized subpopulations of neurons. The two hormones have in common to modulate feeding behavior, acting as peripheral satiety factors (44), and to work as permissive signals for the activation of the neuroendocrine gonadotrope axis (14, 45). Because the control of feeding and that of reproduction are two vital phenomena controlled by the hypothalamus, important target neurons of leptin and insulin in this setting are likely located in that area of the brain. And indeed, NPY- or proopiomelanocortin (POMC)-expressing neurons of the arcuate nucleus represent such targets for their feeding effects (44, 46, 47, 48). The actions of leptin on the reproductive neuroendocrine axis are also, at least partially, mediated by NPY-ergic neurons and pathways (9, 10, 49). In contrast, the hypothalamic targets of insulin in this respect remain to be elucidated. Here we report the generation of novel cell lines that exhibit many phenotypic characteristics of mature neurons and express GnRH. We used one of these cell lines (Gnv-3) in parallel with primary hypothalamic neuronal cell cultures to evaluate the potential effects and mechanisms of action of insulin in GnRH-expressing neurons.
The use of a cell line was dictated by the difficulty, in primary hypothalamic cultures, to identify the precise type of neurons involved in any measured effect. Indeed, primary cultures consist of different cell types besides GnRH neurons. In such a system, an effect observed on GnRH expression or secretion can result from either a direct modulation of GnRH neurons or local interactions between other cell types directly sensitive to the stimulus applied and the GnRH neurons themselves. In the present case, the insulin-induced increase of GnRH gene expression could be mediated via an inhibition of NPY-expressing neurons. Therefore, we reasoned that the study of a neuronal cell line in parallel with primary cultures would allow to better ascertain the direct sensitivity of GnRH-expressing cells to insulin.
The two different GnRH neuronal cell lines readily available (50, 51, 52) are widely used and have proven extremely useful in the study of GnRH physiology (53, 54). However, we took the option to generate another cell line to test the possibility to render the immortalization process conditional. We used a modified Tet-On system (55) and lentiviral vectors derived from HIV (22) because they can transduce nondividing, postmitotic neurons. The 11 clones expressing NSE obtained after the infection of primary adult hypothalamic neuronal cell cultures represent the first successful conditional immortalization of hypothalamic neurons, demonstrating the validity of our experimental approach. Data reported here were obtained in one of these clones, the Gnv-3 cell line. In addition to the neuronal marker NSE, Gnv-3 cells exhibit different markers of well-differentiated neurons and stain uniformly positive for GnRH. The Tet-On system controlling the expression of v-myc is tightly regulated in these cells, as evidenced by Western blot experiments showing no expression of the oncogene in the absence of doxycycline. Furthermore, the presence of doxycycline in the medium induces a very significant stimulation of cell proliferation.
As anticipated from data obtained in GT1–7 cells (50, 56, 57), the Gnv-3 cells described here secrete GnRH in a pulsatile manner at baseline, confirming that pulsatile GnRH secretion is a feature of this novel cell line. Overall, the results of our phenotypic and functional characterization demonstrate that Gnv-3 cells exhibit many features of adult, mature GnRH neurons, including responsiveness to NMDA (30, 31) and nicotine (32) stimulation. Therefore, these cells probably represent a valuable new tool to study the physiology of GnRH-expressing neurons at the cellular and molecular levels.
To assess potential responsiveness of Gnv-3 cells to leptin or insulin, we first evaluated the expression of both the long isoform of the leptin receptor and that of the IR. We failed to demonstrate any significant expression of the leptin receptor by RT-PCR. This negative result is reminiscent of previous in situ hybridization data that failed to demonstrate any expression of the leptin receptor by GnRH-expressing neurons in vivo (48). Thus, it supports the hypothesis that Gnv-3 cells closely resemble wild-type GnRH neurons.
In contrast, the IR is expressed at both the mRNA and protein levels in Gnv-3 cells, suggesting that GnRH neurons may be directly modulated by insulin. This finding prompted us to investigate the effects of insulin on these GnRH-expressing cells. We first examined the downstream intracellular signaling pathways potentially involved in our model. We show here that in Gnv-3 cells, the IR is coupled to the two classical signaling pathways of insulin: a PI3K-dependent pathway and a MAPK Erk1/2 pathway (43). Moreover, the expression of SOCS-3 is also rapidly stimulated by insulin in these cells, as is true for other insulin-sensitive tissues in which SOCS-3 plays an important physiological role as a negative regulator of insulin signaling (58, 59). These results should be confirmed in vivo, but they suggest that hypothalamic GnRH neurons could be an insulin-sensitive tissue. Consistently, our functional data show that stimulation of Gnv-3 cells by insulin induces a rapid stimulation of the expression of the immediate early genes c-fos and EGR-1. This is followed after 2 h by significant increases in GnRH mRNA levels, providing the first indirect evidence that insulin might activate hypothalamic GnRH neurons. It should be stressed, however, that these effects were observed at relatively high concentrations of insulin. The need to use such high concentrations of peptides could be specific to these in vitro models (18, 19), although very little is known about the actual concentrations of neuropeptides present within synaptic connections in vivo.
We then took advantage of our parallel use of primary cultures of hypothalamic neurons to evaluate the role of these two signaling pathways in nontransformed cells. Given the newly recognized importance of the PI3K-dependent intracellular signaling pathway for the feeding effects of insulin in hypothalamic neurons (20), its potential involvement in the stimulation of GnRH gene expression by insulin was first investigated. We show here that consistent with data obtained in Gnv-3 cells, insulin activates the PI3K pathway in hypothalamic neurons in culture. However, pretreatment of the cells with the PI3K inhibitor wortmannin had no effect on the insulin-induced increase in GnRH mRNA levels observed, suggesting that this pathway is not primarily implicated. Because we could also demonstrate that the MAPK Erk1/2 pathway is also activated after exposure of primary neuronal cell cultures to insulin, we then studied further the potential implication of this pathway. By using an enzymatic inhibitor of the MAPKs, Erk1/2, we show here that inhibition of Erk1/2 phosphorylation before insulin stimulation completely blocks the effect of insulin on GnRH mRNA levels. These results are the first demonstration of an implication of this pathway in the central effects of insulin. Although we cannot ascertain in this model whether GnRH neurons are implicated, these data are consistent with our findings in Gnv-3 cells, suggesting that Erk1/2 may be involved in an activation of GnRH neurons by insulin. This hypothesis is also corroborated by previous results showing that the stimulation of the expression of c-fos and Egr-1 by insulin, as seen in Gnv-3 cells, is dependent on MAPK activation (60, 61). Unfortunately, the effects of an inhibition of the MAPK Erk1/2 pathway could not be elucidated in Gnv-3 cells because of a toxicity of dimethylsulfoxide in that model.
In summary, the GnRH neuronal cell lines described here represent the first conditional immortalization of adult rat hypothalamic neurons. The tight control of the expression of the v-myc oncogene achieved by the TET-ON system may account for their high level of differentiation. Using these cells, which express all the cellular and molecular apparatus characteristic of insulin-sensitive tissues, we provide data suggesting that hypothalamic GnRH neurons can be directly modulated by insulin. Finally, we have identified a novel role of the MAPK Erk1/2 signaling pathway in mediating some important actions of insulin within the central nervous system, a crucial observation to understand the regulation of reproductive activity by this hormone. Further work will be necessary to demonstrate the physiological importance of this pathway in the activation hypothalamic GnRH neurons in vivo.
Acknowledgments
The authors thank Drs. Michel Aubert and Bernard Thorens for their helpful comments and suggestions during the writing of the manuscript.
Footnotes
This work was supported by Grants 32-59465.99 and 32-00B0-100 858/1 (to F.P.P.) from the Swiss National Science Foundation and a grant from the Novartis Foundation (to F.P.P.).
The authors have no conflict of interest.
First Published Online November 17, 2005
1 R.S. and E.C. contributed equally to the work
Abbreviations: CMV, Cytomegalovirus; Gnv, GnRH/v-myc; IR, insulin receptor; IRS, insulin receptor substrate; MAP, microtubule-associated protein; NFL, light chain of neurofilaments; NMDA, N-methyl D-aspartate; NPY, neuropeptide Y; NSE, neuron-specific enolase; Ob-Rb, long isoform of the leptin receptor; PI3K, phosphatidylinositol 3-kinase; POMC, proopiomelanocortin; rtTA, reverse tetracycline transactivator; SIN, self-inactivating; SNAP25, synaptosomal-associated protein of 25 kDa; SOCS, suppressor of cytokine signaling; TRE, tetracycline-responsive element; VAMP, vesicle-associated membrane protein.
Accepted for publication November 4, 2005.
References
Knobil E 1992 Remembrance: the discovery of the hypothalamic gonadotropin-releasing hormone pulse generator and of its physiological significance. Endocrinology 131:1005–1006
Marshall JC, Kelch RP 1986 Gonadotropin-releasing hormone: role of pulsatile secretion in the regulation of reproduction. N Engl J Med 315:1459–1468
Bronson FH 1998 Energy balance and ovulation: small cages versus natural habitats. Reprod Fertil Dev 10:127–137
Bronson FH 1986 Food-restricted, prepubertal female rats: rapid recovery of luteinizing hormone pulsing with excess food, and full recovery of pubertal development with gonadotropin-releasing hormone. Endocrinology 118:2483–2487
Frisch RE, McArthur JW 1974 Menstrual cycles: fatness as a determinant of minimum weight for height necessary for their maintenance or onset. Science 185:949–951
Foster DL, Nagatani S 1999 Physiological perspectives on leptin as a regulator of reproduction: role in timing puberty. Biol Reprod 60:205–215
Cunningham MJ, Clifton DK, Steiner RA 1999 Leptin’s actions on the reproductive axis: perspectives and mechanisms. Biol Reprod 60:216–222
Gonzales C, Voirol MJ, Giacomini M, Gaillard RC, Pedrazzini T, Pralong FP 2004 The neuropeptide Y Y1 receptor mediates NPY-induced inhibition of the gonadotrope axis under poor metabolic conditions. FASEB J 18:137–139
Pralong FP, Gonzales C, Voirol MJ, Palmiter RD, Brunner HR, Gaillard RC, Seydoux J, Pedrazzini T 2002 The neuropeptide Y Y1 receptor regulates leptin-mediated control of energy homeostasis and reproductive functions. FASEB J 16:712–714
Ahima RS, Prabakaran D, Mantzoros C, Qu D, Lowell B, Maratos-Flier E, Flier JS 1996 Role of leptin in the neuroendocrine response to fasting. Nature 382:250–252
Stephens TW, Basinski M, Bristow PK, Bue-Valleskey JM, Burgett SG, Craft L, Hale J, Hoffmann J, Hsiung HM, Kriauciunas A, Mackellar W, Rosteck PR, Schoner B, Smith D, Tinsley FC, Zhang XY, Heiman M 1995 The role of neuropeptide Y in the antiobesity action of the obese gene product. Nature 377:530–532
Gruaz NM, Lalaoui M, Pierroz DD, Englaro P, Sizonenko PC, Blum WF, Aubert ML 1998 Chronic administration of leptin into the lateral ventricle induces sexual maturation in severely food-restricted female rats. J Neuroendocrinol 10:627–633
Baura GD, Foster DM, Porte Jr D, Kahn SE, Bergman RN, Cobelli C, Schwartz MW 1993 Saturable transport of insulin from plasma into the central nervous system of dogs in vivo. A mechanism for regulated insulin delivery to the brain. J Clin Invest 92:1824–1830
Bruning JC, Gautam D, Burks DJ, Gillette J, Schubert M, Orban PC, Klein R, Krone W, Muller-Wieland D, Kahn CR 2000 Role of brain insulin receptor in control of body weight and reproduction. Science 289:2122–2125
Schwartz MW, Porte Jr D 2005 Diabetes, obesity, and the brain. Science 307:375–379
Griffin ML, South SA, Yankov VI, Booth Jr RA, Asplin CM, Veldhuis JD, Evans WS 1994 Insulin-dependent diabetes mellitus and menstrual dysfunction. Ann Med 26:331–340
Steger RW, Kienast SG, Pillai S, Rabe M 1993 Effects of streptozotocin-induced diabetes on neuroendocrine responses to ovariectomy and estrogen replacement in female rats. Neuroendocrinology 57:525–531
Burcelin R, Thorens B, Glauser M, Gaillard RC, Pralong FP 2003 Gonadotropin-releasing hormone secretion from hypothalamic neurons: stimulation by insulin and potentiation by leptin. Endocrinology 144:4484–4491
Bergonzelli GE, Pralong FP, Glauser M, Cavadas C, Grouzmann E, Gaillard RC 2001 Interplay between galanin and leptin in the hypothalamic control of feeding via corticotropin-releasing hormone and neuropeptide Y. Diabetes 50:2666–2672
Niswender KD, Morrison CD, Clegg DJ, Olson R, Baskin DG, Myers Jr MG, Seeley RJ, Schwartz MW 2003 Insulin activation of phosphatidylinositol 3-kinase in the hypothalamic arcuate nucleus: a key mediator of insulin-induced anorexia. Diabetes 52:227–231
Gossen M, Freundlieb S, Bender G, Muller G, Hillen W, Bujard H 1995 Transcriptional activation by tetracyclines in mammalian cells. Science 268:1766–1769
Deglon N, Tseng JL, Bensadoun JC, Zurn AD, Arsenijevic Y, Pereira de Almeida L, Zufferey R, Trono D, Aebischer P 2000 Self-inactivating lentiviral vectors with enhanced transgene expression as potential gene transfer system in Parkinson’s disease. Hum Gene Ther 11:179–190
Oberst C, Hartl M, Weiskirchen R, Bister K 1999 Conditional cell transformation by doxycycline-controlled expression of the MC29 v-myc allele. Virology 253:193–207
Zufferey R, Dull T, Mandel RJ, Bukovsky A, Quiroz D, Naldini L, Trono D 1998 Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J Virol 72:9873–9880
Jordan M, Schallhorn A, Wurm FM 1996 Transfecting mammalian cells: optimization of critical parameters affecting calcium-phosphate precipitate formation. Nucleic Acids Res 24:596–601
Freshney RI 1991 Cloning and selection of specific cell types. In: Culture of animal cells. A manual of basic techniques. New York: John Wiley, Sons, Inc.; 137–153
Igaz P, Salvi R, Rey J-P, Glauser M, Pralong FP, Gaillard RC2006 Effects of cytokines on gonadotropin-releasing hormone (GnRH) gene expression in primary hypothalamic neurons and in GnRH neurons immortalized conditionally. Endocrinology 147:1037–1043
van de Loosdrecht AA, Beelen RH, Ossenkoppele GJ, Broekhoven MG, Langenhuijsen MM 1994 A tetrazolium-based colorimetric MTT assay to quantitate human monocyte mediated cytotoxicity against leukemic cells from cell lines and patients with acute myeloid leukemia. J Immunol Methods 174:311–320
Giusti V, Verdumo C, Suter M, Gaillard RC, Burckhardt P, Pralong F 2003 Expression of peroxisome proliferator-activated receptor-1 and peroxisome proliferator-activated receptor-2 in visceral and subcutaneous adipose tissue of obese women. Diabetes 52:1673–1676
Bourguignon JP, Gerard A, Alvarez Gonzalez ML, Franchimont P 1993 Control of pulsatile secretion of gonadotrophin releasing hormone from hypothalamic explants. Hum Reprod 8(Suppl 2):18–22
Gore AC 2001 Gonadotropin-releasing hormone neurons, NMDA receptors, and their regulation by steroid hormones across the reproductive life cycle. Brain Res Brain Res Rev 37:235–248
Andersson K, Eneroth P, Fuxe K, Harfstrand A 1986 Nicotine-induced increases in brain luteinizing hormone releasing hormone-like immunoreactivity and in serum luteinizing hormone levels of the male rat. Neurosci Lett 71:289–292
Vollenweider P, Menard B, Nicod P 2002 Insulin resistance, defective insulin receptor substrate 2-associated phosphatidylinositol-3' kinase activation, and impaired atypical protein kinase C (/) activation in myotubes from obese patients with impaired glucose tolerance. Diabetes 51:1052–1059
Veldhuis JD, Johnson ML 1986 Cluster analysis: a simple, versatile, and robust algorithm for endocrine pulse detection. Am J Physiol 250:E486–E493
Cameron HA, Woolley CS, McEwen BS, Gould E 1993 Differentiation of newly born neurons and glia in the dentate gyrus of the adult rat. Neuroscience 56:337–344
Kilty IC, Barraclough R, Schmidt G, Rudland PS 1999 Isolation of a potential neural stem cell line from the internal capsule of an adult transgenic rat brain. J Neurochem 73:1859–1870
Sudhof TC, Czernik AJ, Kao HT, Takei K, Johnston PA, Horiuchi A, Kanazir SD, Wagner MA, Perin MS, De Camilli P 1989 Synapsins: mosaics of shared and individual domains in a family of synaptic vesicle phosphoproteins. Science 245:1474–1480
Mandic R, Trimble WS, Lowe AW 1997 Tissue-specific alternative RNA splicing of rat vesicle-associated membrane protein-1 (VAMP-1). Gene 199:173–179
Elferink LA, Trimble WS, Scheller RH 1989 Two vesicle-associated membrane protein genes are differentially expressed in the rat central nervous system. J Biol Chem 264:11061–11064
Bargou RC, Leube RE 1991 The synaptophysin-encoding gene in rat and man is specifically transcribed in neuroendocrine cells. Gene 99:197–204
Hepp R, Grant NJ, Chasserot-Golaz S, Aunis D, Langley K 2001 The hypophysis controls expression of SNAP-25 and other SNAREs in the adrenal gland. J Neurocytol 30:789–800
Riederer BM, Innocenti GM 1992 MAP2 isoforms in developing cat cerebral cortex and corpus callosum. Eur J Neurosci 4:1376–1386
Pirola L, Johnston AM, Van Obberghen E 2004 Modulation of insulin action. Diabetologia 47:170–184
Schwartz MW, Woods SC, Porte Jr D, Seeley RJ, Baskin DG 2000 Central nervous system control of food intake. Nature 404:661–671
Cheung CC, Thornton JE, Nurani SD, Clifton DK, Steiner RA 2001 A reassessment of leptin’s role in triggering the onset of puberty in the rat and mouse. Neuroendocrinology 74:12–21
Elmquist JK 2001 Hypothalamic pathways underlying the endocrine, autonomic, and behavioral effects of leptin. Int J Obes Relat Metab Disord 25(Suppl 5):S78–S82
Schwartz MW, Sipols AJ, Marks JL, Sanacory G, White JD, Scheurink A, Kahn SE, Baskin DG, Woods SC, Figlewicz DP, Porte Jr D 1992 Inhibition of hypothalamic neuropeptide Y gene expression by insulin. Endocrinology 130:3608–3616
Finn PD, Cunningham MJ, Pau KY, Spies HG, Clifton DK, Steiner RA 1998 The stimulatory effect of leptin on the neuroendocrine reproductive axis of the monkey. Endocrinology 139:4652–4662
Aubert ML, Pierroz DD, Gruaz NM, d’Alleves V, Vuagnat BA, Pralong FP, Blum WF, Sizonenko PC 1998 Metabolic control of sexual function and growth: role of neuropeptide Y and leptin. Mol Cell Endocrinol 140:107–113
Mellon PL, Windle JJ, Goldsmith PC, Padula CA, Roberts JL, Weiner RI 1990 Immortalization of hypothalamic GnRH neurons by genetically targeted tumorigenesis. Neuron 5:1–10
Radovick S, Wray S, Lee E, Nicols DK, Nakayama Y, Weintraub BD, Westphal H, Cutler Jr GB, Wondisford FE 1991 Migratory arrest of gonadotropin-releasing hormone neurons in transgenic mice. Proc Natl Acad Sci USA 88:3402–3406
Radovick S, Wray S, Muglia L, Westphal H, Olsen B, Smith E, Patriquin E, Wondisford FE 1994 Steroid hormone regulation and tissue-specific expression of the human GnRH gene in cell culture and transgenic animals. Horm Behav 28:520–529
Wetsel WC 1995 Immortalized hypothalamic luteinizing hormone-releasing hormone (LHRH) neurons: a new tool for dissecting the molecular and cellular basis of LHRH physiology. Cell Mol Neurobiol 15:43–78
Maggi R, Pimpinelli F, Molteni L, Milani M, Martini L, Piva F, Wetsel WC 2000 Immortalized luteinizing hormone-releasing hormone neurons show a different migratory activity in vitro. Endocrinology 141:2105–2112
Mansuy IM, Bujard H 2000 Tetracycline-regulated gene expression in the brain. Curr Opin Neurobiol 10:593–596
Wetsel WC, Valenca MM, Merchenthaler I, Liposits Z, Lopez FL, Weiner RI, Mellon PL, Negro-Vilar A 1992 Intrinsic pulsatile secretory activity of immortalized luteinizing hormone-releasing hormone-secreting neurons. Proc Natl Acad Sci USA 89:4149–4153
Martinez de la Escalera G, Choi AL, Weiner RI 1992 Generation and synchronization of gonadotropin-releasing hormone (GnRH) pulses: intrinsic properties of the GT1–1 GnRH neuronal cell line. Proc Natl Acad Sci USA 89:1852–1855
Emanuelli B, Glondu M, Filloux C, Peraldi P, Van Obberghen E 2004 The potential role of SOCS-3 in the interleukin-1-induced desensitization of insulin signaling in pancreatic -cells. Diabetes 53(Suppl 3):S97–S103
Emanuelli B, Peraldi P, Filloux C, Chavey C, Freidinger K, Hilton DJ, Hotamisligil GS, Van Obberghen E 2001 SOCS-3 inhibits insulin signaling and is up-regulated in response to tumor necrosis factor- in the adipose tissue of obese mice. J Biol Chem 276:47944–47949
Keeton AB, Bortoff KD, Bennett WL, Franklin JL, Venable DY, Messina JL 2003 Insulin-regulated expression of Egr-1 and Krox20: dependence on ERK1/2 and interaction with p38 and PI3-kinase pathways. Endocrinology 144:5402–5410
Harada S, Smith RM, Smith JA, White MF, Jarett L 1996 Insulin-induced egr-1 and c-fos expression in 32D cells requires insulin receptor, Shc, and mitogen-activated protein kinase, but not insulin receptor substrate-1 and phosphatidylinositol 3-kinase activation. J Biol Chem 271:30222–30226(Roberto Salvi1, Einar Castillo1, Marie-J)
Metabolism (R.S., E.C., M.-J.V., M.G., J.-P.R., R.C.G., F.P.P.) and Service of Internal Medicine (P.V.), Department of Medicine, University Hospital, 1011 Lausanne, Switzerland
Service of Endocrinology (F.P.P.), Diabetology and Metabolism, University Hospital, 1211 Geneva, Switzerland
Abstract
Energy balance exerts a critical influence on reproduction via changes in the circulating levels of hormones such as insulin. This modulation of the neuroendocrine reproductive axis ultimately involves variations in the activity of hypothalamic neurons expressing GnRH. Here we studied the effects of insulin in primary hypothalamic cell cultures as well as a GnRH neuronal cell line that we generated by conditional immortalization of adult hypothalamic neurons. These cells, which represent the first successful conditional immortalization of GnRH neurons, retain many of their mature phenotypic characteristics. In addition, we show that they express the insulin receptor. Consistently, their stimulation with insulin activates both the phosphatidylinositol 3-kinase and the Erk1/2 MAPK signaling pathways and stimulates a rapid increase in the expression of c-fos, demonstrating their responsiveness to this hormone. Further work performed in parallel in immortalized GnRH-expressing cells and primary neuronal cultures containing non-GnRH-expressing neurons shows that insulin induces the expression of GnRH in both models. In primary cultures, inhibition of the Erk1/2 pathway abolishes the stimulation of GnRH expression by insulin, whereas blockade of the phosphatidylinositol 3-kinase pathway has no effect. In conclusion, these data strongly suggest that GnRH neurons are directly sensitive to insulin and implicate for the first time the MAPK Erk1/2 signaling pathway in the central effects of insulin on the neuroendocrine reproductive axis.
Introduction
HYPOTHALAMIC GnRH neurons exert a biological function crucial to species survival: they represent the center of integration of the neuroendocrine reproductive axis (1, 2). Their coordinate activation is a prerequisite to the awakening of that axis at puberty as well as continuing reproductive competence in adulthood. Such activation is critically dependent on adequate body energy stores (3), as exemplified by the observation that unfavorable metabolic conditions impede on reproductive activity (4). This inhibition is likely mediated at the level of the hypothalamus via mechanisms implicating the existence of factors signaling the nutritional status of an individual to the central nervous system (5).
Leptin is primarily a central inhibitor of food intake, which also participates in the regulation of reproductive activity by metabolic changes (6, 7). These effects are mediated, at least partially, via hypothalamic neuropeptide Y (NPY)-ergic neurons and pathways (8, 9, 10, 11, 12). In recent years evidence that insulin can play very similar roles has accumulated: like leptin, insulin is transported actively across the blood-brain barrier (13), and has the potential to function as a satiety signal in the hypothalamus (14, 15). In addition, mice harboring a neuron-specific deletion of the insulin receptor also suffer from central hypogonadism (14). This newly discovered role of insulin in the regulation of reproduction is further underlined by the observations, in human and animal models, that insulin-dependent diabetes mellitus is associated with reproductive abnormalities resulting from impaired LH secretion (16, 17). Consistent with this observation, we have demonstrated that insulin activates the neuroendocrine reproductive axis of normal mice in vivo and that this effect involves the stimulation of hypothalamic GnRH expression and secretion (18). However, the precise hypothalamic target neurons of insulin, or the cellular and molecular mechanisms mediating its reproductive effects, remain largely unknown.
We describe here the successful generation of GnRH-expressing cells by conditional immortalization of adult hypothalamic neurons. We then used these immortalized GnRH-expressing cells in parallel with primary cultures of hypothalamic neurons (19) to ascertain the potential direct effects of insulin on GnRH-expressing neurons. Taken together, our data suggest that GnRH neurons themselves may be directly modulated by insulin. If confirmed in vivo, this could provide a novel mechanism underlying the permissive effects exerted by insulin on the gonadotrope axis (14, 18). In addition, we have demonstrated that the intracellular mechanisms involved in the stimulation of GnRH gene expression by insulin implicate the MAPK Erk1/2 signaling pathway. This latter finding is contrasting to data regarding the action of insulin on feeding, which has been linked to the phosphatidylinositol 3-kinase (PI3K) pathway (20). We suggest that the cell lines described here will constitute a useful novel tool for the study of the activation of hypothalamic GnRH neurons at the cellular and molecular levels.
Materials and Methods
All these experiments were approved by both the institutional and state ethical committees on animal research.
Conditional immortalization of GnRH neurons
Twelve different clonal cell lines were derived from primary cultures of adult rat hypothalamic neurons. Hypothalami were obtained from 10- to 12-wk-old female Wistar rats: after killing by decapitation, whole brains were rapidly removed from the skull, and the hypothalamus dissected out by two coronal cuts (anteriorly at the anterior border of the optic chiasma and posteriorly at the level of the mammillary bodies) followed by two parasagittal cuts along the hypothalamic sulci and a final cut dorsally, at a depth of approximately 3 mm from the ventral surface of the tissue block. Cells were dispersed mechanically in PBS buffer (PBS without calcium or magnesium; Gibco, Gaithersburg, MD) supplemented with 0.06% glucose (Fluka Chemie GmbH, Buchs, Switzerland) and 100 U per 100 μg/ml penicillin/streptomycin (Seromed, Flakola, Basel, Switzerland). Freshly excised whole hypothalami were gently passed several times through Pasteur pipettes flamed in the middle of the procedure to decrease the diameter of their opening. Nondispersed tissue was allowed to settle for 5 min and supernatant transferred to a clean tube. The remaining pellet was resuspended in 4 ml of PBS buffer and mechanical dispersion repeated. Supernatants from first and second dispersion were mixed, centrifuged for 5 min at 100 x g and the cellular pellet gently resuspended in 5 ml of neurobasal medium (Gibco) and 0.04% B27 supplement (Gibco BRL, Basel, Switzerland) containing 500 μM glutamine and 25 μM glutamate (Sigma, Buchs, Switzerland). After dispersion, cells were plated at a density of 400,000 live cells/well in six-well plates (Corning Costar Corp., Cambridge, MA) coated with 5 μg/ml poly-D-lysine (Sigma, Fluka Chemie) and grown in neurobasal A medium (Invitrogen AG, Basel, Switzerland) with B27 supplement (Invitrogen) containing 500 μM glutamine and 25 μM glutamate (Sigma, Fluka Chemie) as described (18, 19). Half of the medium was changed every fourth day. Cells were immortalized 10 d after dispersion by the transduction of two different lentiviral vectors (Fig. 1A).
By fusing a 2-kb fragment of the rat GnRH promoter to the reverse tetracycline transactivator (rtTA) gene, the first vector was designed to direct the expression of the rtTA in GnRH-expressing neurons. In the second vector, the tetracycline-responsive element (TRE) (21) was fused to the v-myc oncogene, immediately followed by the minigene encoding resistance to puromycine. With this design, the addition of low concentrations of tetracycline in the culture medium was expected to drive the transcriptional activation of v-myc in GnRH-expressing cells infected with both lentiviral vectors.
The first vector (pSIN-GnRHprom-rtTA-WHV), carrying a modified Tet-On construct (21) in which the cytomegalovirus (CMV) promoter is deleted and replaced by the rat GnRH promoter, was constructed in three phases. First, the proximal 2-kb fragment of the rat GnRH promoter was generated by PCR from genomic DNA, cloned in pGEM-T Easy (Promega, Catalys AG, Wallisellen, Switzerland), and verified by sequencing. Primer pairs used were: 5'-ATACTAGTGCTTATTAGCCAGTGAGCAGCAG-3' (sense) and 5'-AATTGATCACAAGCTTCTGTCAGTAGGTTGAG-3' (antisense, bold letters stand for the restrictions sites SpeI and BclI). Then this fragment was removed by digestion with SpeI and HindIII, the latter site blunted, and subcloned into the pTet-ON plasmid (21) at the blunted EcoRI and SpeI sites in place of the CMV promoter, in the appropriate direction in front of the rtTA element. A fragment containing the 2-kb GnRH promoter and the rtTA element was finally removed from the pTet-ON plasmid by digestion with SpeI and BamHI and inserted at the same restriction sites in the plasmid pSIN-MCS-WHV (22).
For the second vector, the tetracycline response element (TRE) and the MC29 v-myc allele located downstream in the plasmid pUHD10–3myc (23) were removed by digestion with XhoI (blunted) and BamHI and inserted into the second plasmid, pSIN-MCS-IRES-PURO (22), at the sites NotI (blunted) and BamHI. As a result, the TRE is cloned in front of the avian v-myc used as the immortalizing gene and the gene coding for resistance to puromycine.
Recombinant self-inactivating (SIN) lentiviruses (24) were produced by transient cotransfection of 293T cells with a three-plasmid system. Briefly, 60–70% confluent 293T cells, cultured on 10-cm dishes, were cotransfected with the following plasmids: 5 μg of the packaging construct pCMVR8.91, 5 μg of the vesicular stomatitis virus G protein envelope pMD.G, and 15 μg of either of the transfer vectors pSIN-GnRHprom-rtTA or pSIN-TRE-v-myc-IRES-PURO, by using the calcium phosphate technique as described (25). Media were replaced after 16 h and conditioned media collected 48 h after transfection, filtered on 0.45-μm pore size filters (Millipore AG, Volketswil, Switzerland), and either immediately used to infect the hypothalamic neurons or snap frozen in LN2 for later use. To infect the neuronal cultures, equal amounts in volume (1:1) of conditioned media derived from the two lentiviral vectors pSIN-GnRHprom-rtTA and pSIN-TRE-v-myc-IRES-PURO were added to the culture, in the presence of polybrene (8 μg/ml, Sigma, Fluka).
After 48 h of transduction, the culture medium was changed, and stimulation with doxycycline (5 μg/ml, Sigma, Fluka) as well as selection with puromycine (0.5 μg/ml, Sigma, Fluka) was started. Selection with puromycine was carried out for 8 wk before stopping, and stimulation with doxycycline was continued thereafter. After the end of the selection process, 12 different clones were isolated by dilution and following the soft agar method (26). These clones were named GnRH/v-myc (Gnv) cells and numbered from 1 to 12. GnRH expression at the mRNA level was found maximal in clones 3 and 4. Subsequent work reported here was performed using clone 3 (Gnv-3 cells), and clone 4 (Gnv-4 cells) is described in the companion manuscript of this paper (27).
Culture and characterization of Gnv-3 cells
Gnv-3 cells are expanded in a proliferation medium composed as follows: neurobasal A medium with B27 supplement and containing 1% FBS (Invitrogen), 5 μg/ml doxycycline, penicillin and 5 ng/ml bFGF (Invitrogen). They are passaged by trypsin digestion (Invitrogen) every 4–5 days, and several aliquots of the same passage (passage 3 or 4) are frozen and conserved. This procedure allows to use cells at the same early passage for all experiments. For each experiment, one of these aliquots is rewarmed, plated in proliferation medium and allowed 4–5 days before passage. After passage, the cells are plated in 6-well plates or in 10 cm Petri dishes in differentiation medium containing: neurobasal A medium alone, with glutamine (Invitrogen) and bFGF (Invitrogen). This medium was named differentiation medium because under these culture conditions, the shape of the cells changes: their bodies become thinner and elongated, and they develop many intercellular connections. Because of the high glucose content of neurobasal A (25 mM), the medium was changed again, 16–24 h before the experiments, for DMEM low glucose (5 mM glucose, Invitrogen). All experiments were performed using these culture conditions.
Cell growth
Cell growth in the proliferation medium (presence of doxycycline) and the differentiation medium (absence of doxycycline) was evaluated following a modification of the tetrazolium colorimetric assay (28), which is based on the ability of living cells to reduce the yellow salt 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide to formazan. Cells were plated at precisely 50,000 cells/well in 6-well plates. On d 0 of the experiment and every 24 h thereafter for 6 consecutive days, 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide (0.5 mg/ml, Sigma, Fluka) was added to the medium in triplicate wells. Four hours later, the supernatants of these triplicate wells were gently aspirated and the newly formed formazan crystals dissolved by the addition of 1.5 mL of dimethylsulfoxide [100%, Merck (Suisse) SA, Geneva, Switzerland] with 25 μl of glycine buffer [0.1 M glycine (pH 10.5); Merck]. Absorbance was then measured at 540 nm in a spectrophotometer (Uvicon 940, Kontron, Eching, Germany). This experiment was performed on two separate occasions.
RT-PCR and quantitative RT-PCR
For these experiments, cells in the differentiation medium were grown either in 10-cm petri dishes (RT-PCR) or 6-well plates (quantitative RT-PCR). At the end of the experiment, total RNA was extracted using commercially available reagents (TriPure reagent, Roche Diagnostics, Rotkreuz, Switzerland). For RT-PCR experiments, reverse transcription with SuperScript II (Invitrogen) was performed with random primers, and the primer pairs used in the various PCRs are listed in Table 1, together with gene accession numbers. PCRs were performed on a GeneAmp PCR system 9700 thermocycler (PE Applied Biosystems, Rotkreuz, Switzerland).
Quantitative RT-PCR experiments are performed using the LightCycler technology (Roche Diagnostics) with SYBR green detection as described (29). Reverse transcription was done with random primers. Then a standard curve was created with serial dilutions of RNA extracts from whole hypothalami. Different dilutions of the samples were always tested in preliminary experiments to ensure that quantification would be performed within the linear part of this standard curve. After this test, all samples were quantified in at least three different runs. For quantification purposes, the mRNA levels of the gene of interest were always reported to the levels of 2-microglobulin, a constitutively expressed gene. A value of 100% was then attributed to the control sample in each separate experiment, and all measures were reported as a percent of this control.
Immunohistochemistry
Immunohistochemistry was performed at room temperature as described (19), using an avidin-biotinylated horseradish peroxidase macromolecular complex kit (ABC Elite, Vector, Burlingame, CA). Cells were fixed for 40 min in 4% paraformaldehyde (Merck) and rinsed three times in PBS and one time in 0.05 M Tris (Sigma) (pH 7.4), 0.6% NaCl, 0.3% Triton X-100 (Merck), and 10% goat serum (Gibco). After quenching of endogenous peroxidase activity, nonspecific binding was blocked by incubating 1 h in 0.05 M Tris (Sigma) (pH 7.4), 0.6% NaCl, and 0.3% Triton X-100 (Merck). Incubations with the primary antiserum (polyclonal GnRH antiserum, Bio-Yeda Ltd., Rehovot, Israel) were performed overnight at a final dilution of 1:400, and then cells were washed and further incubated 1 h with a biotinylated secondary antibody (Jackson, West Grove, PA). After three washes in 0.1 M Tris-HCl (pH 7.7), cells were incubated in the ABC reagent, washed again, and finally incubated in diaminobenzidine peroxidase substrate solution [50 mg per 100 ml 3–3' diaminobenzidine (Sigma) in 0.1 M Tris-HCl, 0.018% H2O2 (pH 7.7)]. Negative controls were obtained by incubating the primary antiserum with excess GnRH before use.
Western blotting
The expression of v-myc and the insulin receptor (IR) was also assessed by Western blot from cells plated at a density of 40,000 cells/well in 6-well plates. On the day of the experiment, cells were rinsed with ice-cold PBS and lysed with the radioimmunoprecipitation assay buffer [10 mM Tris (pH 7.5), 100 mM NaCl, 1 mM EDTA, 0.5% Na-deoxycholate, 0.1% sodium dodecyl sulfate, 1% Triton X-100] supplemented with complete protease inhibitor cocktail (Roche Biochemicals, Rotkreuz, Switzerland). After this procedure, lysates were homogenized by sonication during 30–60 sec. Cells debris was removed by centrifugation at 14,000 rpm for 15 min at 4 C. The protein concentration of the supernatant was determined using commercially available reagents (Bio-Rad protein assay, Bio-Rad, Reinach, Switzerland). Proteins were resolved by 10% SDS-PAGE electrophoresis and transferred to a polyvinylidine difluoride membrane (Hybond-P, Amersham Biosciences, Otelfingen, Switzerland). Blots were blocked for 2 h at room temperature in PBS containing 5% nonfat dried milk and 0.05% Tween 20. Then they were incubated overnight at 4 C in the same buffer containing the primary antibody.
For v-myc, experiments were performed either with proliferating cells or after the change for the differentiation medium, using a monoclonal antibody (Oncogene Research Products, La Jolla, CA) at a final dilution of 1:1000. For the IR, experiments were performed only with cells in the differentiation medium, using a monoclonal antibody directed against its -subunit (Oncogene Research Products) at a final dilution of 1:1000. Equal amounts of proteins were resolved by 10% SDS-PAGE electrophoresis and transferred to a PDVF membrane (Hybond-P, Amersham Biosciences) and antibody complexes visualized by the enhanced chemiluminescence system (Amersham, Arlington Heights, IL).
Secretion experiments
Static secretion experiments were performed in 6-well plates, and perifusion experiments were performed with cells grown onto plastic coverslips (Assistant, Karl Hecht BmgH & Co. KG, Altnau, Switzerland) coated with 5 μg/ml poly-D-lysine (Sigma, Fluka Chemie) as described (18, 19). We used two coverslips in each perifusion chamber, with approximately 400,000 cells/coverslip. The perifusion rate was set at 150 μl/min, and fractions were collected every 4 min. During the experiment, the perifusion medium was constantly gassed with a mixture of 95% O2 and 5% CO2. GnRH was directly assayed in the medium, or in the effluent, by RIA as previously described (18), using the antiserum used for immunohistochemistry. The limit of detection of this assay is 2 pg/ml.
Experimental procedures
The responsiveness of Gnv-3 cells to N-methyl D-aspartate (NMDA) and nicotine, both recognized activators of GnRH secretion in adult GnRH neurons (30, 31, 32), was assessed in static cultures and, for nicotine, in perifusion experiments. Cells in static culture were exposed for 2 h to either NMDA or nicotine at concentrations ranging from 10–6 to 10–3 M. At the end of the incubation period, GnRH was measured in the medium by RIA. In addition, the effect of NMDA on GnRH gene expression was also studied by incubating the cells for 6 h with NMDA (5 x 10–5 M) and measuring mRNA levels of GnRH in total RNA extracts by real-time quantitative RT-PCR.
To investigate the intracellular signaling pathway(s) potentially activated by insulin, Gnv-3 cells in the differentiation medium (see above) were stimulated with insulin at a concentration of 4 x 10–5 M for various time periods up to 30 min. The use of this insulin concentration was based on preliminary experiments as well as our previous work in primary neuronal cell cultures (18). The activation of the MAPK Erk1/2 pathway or the PI3K pathway was assessed by Western blot. For Erk1/2, membranes were first incubated as described above with a monoclonal antibody against phosphospecific Erk1/2 (Cell Signaling Technology, Inc., Beverly, MA). Data were quantified by densitometric analysis of autoradiograms, using a computerized densitometer (Typhoon system, Molecular Dynamics, Inc., Sunnyvale, CA). Then the membranes were stripped and hybridized again with a monoclonal antibody directed against total Erk1/2 (Cell Signaling Technology) to allow normalization. The activation of the PI3K-dependent pathway was evaluated similarly by measuring the phosphorylation of its downstream substrate Akt. Immunodetection was first performed with antibodies specific for phospho-Akt (Ser 473) (New England Biolabs, Beverly, MA). After quantification, membranes were stripped and incubated with antibodies against total Akt forms (Santa Cruz Biotechnology, Santa Cruz, CA) to allow normalization. For all Western blot data, the ratio of phosphorylated over total enzyme was calculated at each time point and expressed as a percent of baseline. Time 0 was arbitrarily considered as 100%.
The expression of suppressor of cytokine signaling (SOCS)-3 was then evaluated in Gnv-3 cells at different time points between 30 and 120 min after stimulation with the same concentration of insulin (4 x 10–5 M). Similar stimulations with insulin (4 x 10–5 M) were performed in parallel at the same time points to study its effects on the expression of the immediate early genes c-fos and Egr-1, two markers of neuronal activation. Finally, the effects of insulin on the expression of GnRH were also evaluated in Gnv-3 cells. These effects were tested at various time courses between 2 and 24 h, with insulin concentrations ranging between 4 x 10–5 and 4 x 10–8 M. Because insulin was found to stimulate the expression of GnRH at relatively high concentrations, similar experiments were then performed with IGF-I at concentrations between 10–7 and 10–9 M. Effects of IGF-I were checked at 2 h, the time point at which insulin had the highest effect. mRNA levels of SOCS-3, c-fos, Egr-1, and GnRH were measured by semiquantitative, real-time RT-PCR of total mRNA extracts as described above.
Next, the role in the stimulation of GnRH expression by insulin played by the two pathways identified in Gnv-3 cells (PI3K and Erk1/2) was further evaluated in primary hypothalamic neuronal cell cultures (19). First, their stimulation by insulin (4 x 10–4 M) was confirmed in primary neurons. Erk1/2 phosphorylation was assessed by Western blot, and PI3K activity determined as previously described (33) was used as a measure of PI3K stimulation. Then we used enzymatic inhibition of either the MAPK or PI3K pathway to evaluate their respective role in the stimulation of GnRH gene expression by insulin. The effect of insulin (4 x 10–4 M) on GnRH mRNA levels was studied in cells either pretreated for 30 min or not with the PI3K inhibitor wortmannin (1 μM, Sigma, Fluka) or pretreated for 30 min or not with the Erk1/2 inhibitor PD98059 (25 μM, Sigma, Fluka).
Data analysis
Pulse detection in perifusion experiments was performed with Cluster 8, a model-free computerized pulse analysis algorithm to identify statistically significant pulses in relation to dose-dependent measurement error in a hormone time series (34). We used the software package Pulse_xp (2004 version; University of Virginia, Charlottesville, VA) running under Windows XP. Criteria used for pulse identification were the following: 1) individual test cluster sizes of three consecutive points for both the nadir and the peak width; 2) a minimum and maximum intraseries coefficient of variation of 10 and 15%, respectively; and 3) a t statistic to identify a significant increase and a t statistic to define a significant decrease.
All results are expressed as mean ± SEM, and a minimum of three independent experiments were always performed. Statistical significance of all results was assessed by ANOVA, with P < 0.05 considered a significant difference.
Results
Conditional immortalization and characterization of Gnv-3 cells
Figure 1, B and C, summarize the initial characterization of the 12 clones obtained with this approach. Figure 1B represents the results of PCR experiments performed on their genomic DNA (labeled 1–12). These data demonstrate that the rtTA as well as the GAG lentiviral gene are both integrated in the genome of all 12 clones but that the v-myc oncogene is not found in clone 8. Figure 1C displays the results of RT-PCR experiments performed on total RNA samples of the same 12 clones. Consistent with the above results, clone 8 does not express neuron-specific enolase (NSE), a neuronal marker (35). The rest of the work presented here was performed with clone 3 (called Gnv-3 cells, clone 3).
Table 2 summarizes the characterization that was further performed on Gnv-3 cells by RT-PCR. As shown, these cells express the light chain of neurofilaments (NFL), which is another marker of well-differentiated neurons (36), as well as several markers of neurosecretory neurons such as synaptobrevin, synapsin 1, synaptophysin, synaptosomal-associated protein of 25 kDa (SNAP25) vesicle-associated membrane protein (VAMP)1 and VAMP2, or chromogranin B (37, 38, 39, 40, 41). In contrast, a marker of poorly differentiated neurons, the microtubule-associated protein (MAP)-2b isoform of MAP-2 (42), is not expressed by Gnv-3 cells. Similarly, we found no expression of glial cell markers in this clone, further confirming its well-differentiated neuronal phenotype. Finally, among the various neuroendocrine markers that were tested, we found that Gnv-3 cells express exclusively GnRH, both at the mRNA and the protein levels.
This is further exemplified in Fig. 2A, which demonstrates the uniform positive immunostaining for GnRH observed in these cells. Their GnRH secretion was then assessed in perifusion experiments. Figure 2B illustrates the pattern of basal GnRH secretion observed. Significant pulses identified by Cluster 8 are indicated by a solid line on top of the plot. When summarizing the data from three independent perifusion studies, the mean basal frequency of GnRH secretion from Gnv-3 cells was 2.2 ± 0.68 pulses/h, with a mean interpulse interval of 24.69 ± 5.69 min. The mean increase over calculated baseline of these peaks was 3.17 ± 0.1 pg/ml.
Figure 2C illustrates the significant increase in GnRH mRNA levels observed after 6 in Gnv-3 cells exposed to NMDA (5 x 10–5 M). This increase in GnRH gene expression after exposure to NMDA (5 x 10–5 M) is also accompanied by an increase in GnRH secretion (data not shown). Similarly, stimulation of Gnv-3 cells with nicotine (10–6 M), either in static cultures or perifusion experiments, was observed to stimulate GnRH secretion acutely (data not shown). Finally, Fig. 2D demonstrates that the proliferation of Gnv-3 cells is dependent on the presence of doxycycline in the culture medium, which consistently stimulates the expression of v-myc (Fig. 2D, inset). Without doxycycline, v-myc is not detected by Western blot (Fig. 2D, inset) and the proliferation of the cells is markedly lower.
IRs and leptin receptors
Figure 3A demonstrates that the IR is expressed at the mRNA level (RT-PCR) and the protein level (Western blot). However, we failed to demonstrate consistently the presence in this clone of the long isoform of the leptin receptor (Ob-Rb) by RT-PCR. The level of expression of the IR in Gnv-3 cells, as assessed by Western blot, is comparable with that observed in fresh hypothalamic extracts and lower than that observed in muscle or brain. We then investigated the intracellular signaling pathway(s) that may be linked with the IR in these cells. Figure 3B illustrates the rapid, strong, and long-lasting stimulation of the phosphorylation of the MAPKs Erk1/2 that follows exposure of Gnv-3 cells to insulin. Similarly, the phosphorylation of AKT, downstream of PI3K, is rapidly and strongly stimulated by insulin in the same cells (Fig. 3C). Finally, SOCS-3 mRNA levels are increased after exposure to insulin, as reported for other insulin-sensitive tissues (43) (Fig. 3D).
Insulin effects in Gnv-3 cells
The results illustrated in Figs. 4, A and B, show that insulin (4 x 10–5 M) induces a rapid and significant induction of the immediate early genes c-fos and Egr-1, as reflected by the 2- to 3-fold increases measured in their respective mRNA levels. The induction of c-fos is stronger than that observed with Egr-1. In contrast, the stimulation of Egr-1 expression by insulin is of a longer duration: Egr-1 mRNA levels are still significantly higher than baseline 2 h after the stimulation. Finally, Fig. 4C displays the effects of insulin on GnRH expression in these cells. These data demonstrate the significant and dose-dependent increases in GnRH mRNA levels observed after insulin stimulation. This effect is rapid, already maximal at 2 h (as shown in the figure). A similar stimulation was still observed after 4 h of incubation with insulin, but after 24 h, no effect of insulin was noted in Gnv-3 cells (data not shown). In contrast, IGF-I at the concentrations tested did not induce any significant change in GnRH mRNA levels in our model: after 2 h of incubation with IGF-I at the concentration of 10–7 M, these levels were 161.2 ± 25.7% of control (n = 4), and they were 139.0 ± 19.0% of control after 2 h in the presence of IGF-I at the concentration of 10–8 M.
Insulin signaling in primary hypothalamic neurons
Previous work from our group had demonstrated that insulin stimulates GnRH gene expression in primary hypothalamic neuronal cell cultures (18). Therefore, we next examined in primary neurons the respective importance of the two signaling pathways activated in Gnv-3 cells. Figure 5 summarizes our findings. Figure 5A illustrates the very strong and long-lasting increase of PI3K activity observed in primary hypothalamic neurons stimulated by insulin: this activity is still significantly higher than control after 2 h. Results reported in this graph were obtained after immunoprecipitation of the cell extracts with an antiinsulin receptor substrate (IRS)1 antibody, but very similar results were observed after immunoprecipitation with an anti-IRS2 antibody (data not shown). Despite this PI3K activation, its inhibition by pretreatment with wortmannin did not block the GnRH mRNA levels increase by insulin seen in this model (Fig. 5B). PI3K activity inhibition in these conditions was demonstrated by showing complete inhibition of the downstream Akt serine/threonine kinase (data not shown). Therefore, the potential role of the MAPK Erk1/2 pathway was then evaluated in these neurons, using a similar approach. As anticipated from results obtained in Gnv-3 cells, data reported in Fig. 5C show that this pathway is also stimulated by insulin in primary hypothalamic neurons. However, and in sharp contrast to the observation made with wortmannin, inhibition of the phosphorylation of Erk1/2 by pretreatment of the cells with the enzyme inhibitor PD 98059 completely abolished the effect of insulin on GnRH expression.
Discussion
Metabolic changes are known to alter the central nervous system control of reproduction through variations in circulating levels of peripheral signals. Leptin as well as insulin is among the metabolic factors involved (6, 14, 18), acting via a modulation of the activity of highly specialized subpopulations of neurons. The two hormones have in common to modulate feeding behavior, acting as peripheral satiety factors (44), and to work as permissive signals for the activation of the neuroendocrine gonadotrope axis (14, 45). Because the control of feeding and that of reproduction are two vital phenomena controlled by the hypothalamus, important target neurons of leptin and insulin in this setting are likely located in that area of the brain. And indeed, NPY- or proopiomelanocortin (POMC)-expressing neurons of the arcuate nucleus represent such targets for their feeding effects (44, 46, 47, 48). The actions of leptin on the reproductive neuroendocrine axis are also, at least partially, mediated by NPY-ergic neurons and pathways (9, 10, 49). In contrast, the hypothalamic targets of insulin in this respect remain to be elucidated. Here we report the generation of novel cell lines that exhibit many phenotypic characteristics of mature neurons and express GnRH. We used one of these cell lines (Gnv-3) in parallel with primary hypothalamic neuronal cell cultures to evaluate the potential effects and mechanisms of action of insulin in GnRH-expressing neurons.
The use of a cell line was dictated by the difficulty, in primary hypothalamic cultures, to identify the precise type of neurons involved in any measured effect. Indeed, primary cultures consist of different cell types besides GnRH neurons. In such a system, an effect observed on GnRH expression or secretion can result from either a direct modulation of GnRH neurons or local interactions between other cell types directly sensitive to the stimulus applied and the GnRH neurons themselves. In the present case, the insulin-induced increase of GnRH gene expression could be mediated via an inhibition of NPY-expressing neurons. Therefore, we reasoned that the study of a neuronal cell line in parallel with primary cultures would allow to better ascertain the direct sensitivity of GnRH-expressing cells to insulin.
The two different GnRH neuronal cell lines readily available (50, 51, 52) are widely used and have proven extremely useful in the study of GnRH physiology (53, 54). However, we took the option to generate another cell line to test the possibility to render the immortalization process conditional. We used a modified Tet-On system (55) and lentiviral vectors derived from HIV (22) because they can transduce nondividing, postmitotic neurons. The 11 clones expressing NSE obtained after the infection of primary adult hypothalamic neuronal cell cultures represent the first successful conditional immortalization of hypothalamic neurons, demonstrating the validity of our experimental approach. Data reported here were obtained in one of these clones, the Gnv-3 cell line. In addition to the neuronal marker NSE, Gnv-3 cells exhibit different markers of well-differentiated neurons and stain uniformly positive for GnRH. The Tet-On system controlling the expression of v-myc is tightly regulated in these cells, as evidenced by Western blot experiments showing no expression of the oncogene in the absence of doxycycline. Furthermore, the presence of doxycycline in the medium induces a very significant stimulation of cell proliferation.
As anticipated from data obtained in GT1–7 cells (50, 56, 57), the Gnv-3 cells described here secrete GnRH in a pulsatile manner at baseline, confirming that pulsatile GnRH secretion is a feature of this novel cell line. Overall, the results of our phenotypic and functional characterization demonstrate that Gnv-3 cells exhibit many features of adult, mature GnRH neurons, including responsiveness to NMDA (30, 31) and nicotine (32) stimulation. Therefore, these cells probably represent a valuable new tool to study the physiology of GnRH-expressing neurons at the cellular and molecular levels.
To assess potential responsiveness of Gnv-3 cells to leptin or insulin, we first evaluated the expression of both the long isoform of the leptin receptor and that of the IR. We failed to demonstrate any significant expression of the leptin receptor by RT-PCR. This negative result is reminiscent of previous in situ hybridization data that failed to demonstrate any expression of the leptin receptor by GnRH-expressing neurons in vivo (48). Thus, it supports the hypothesis that Gnv-3 cells closely resemble wild-type GnRH neurons.
In contrast, the IR is expressed at both the mRNA and protein levels in Gnv-3 cells, suggesting that GnRH neurons may be directly modulated by insulin. This finding prompted us to investigate the effects of insulin on these GnRH-expressing cells. We first examined the downstream intracellular signaling pathways potentially involved in our model. We show here that in Gnv-3 cells, the IR is coupled to the two classical signaling pathways of insulin: a PI3K-dependent pathway and a MAPK Erk1/2 pathway (43). Moreover, the expression of SOCS-3 is also rapidly stimulated by insulin in these cells, as is true for other insulin-sensitive tissues in which SOCS-3 plays an important physiological role as a negative regulator of insulin signaling (58, 59). These results should be confirmed in vivo, but they suggest that hypothalamic GnRH neurons could be an insulin-sensitive tissue. Consistently, our functional data show that stimulation of Gnv-3 cells by insulin induces a rapid stimulation of the expression of the immediate early genes c-fos and EGR-1. This is followed after 2 h by significant increases in GnRH mRNA levels, providing the first indirect evidence that insulin might activate hypothalamic GnRH neurons. It should be stressed, however, that these effects were observed at relatively high concentrations of insulin. The need to use such high concentrations of peptides could be specific to these in vitro models (18, 19), although very little is known about the actual concentrations of neuropeptides present within synaptic connections in vivo.
We then took advantage of our parallel use of primary cultures of hypothalamic neurons to evaluate the role of these two signaling pathways in nontransformed cells. Given the newly recognized importance of the PI3K-dependent intracellular signaling pathway for the feeding effects of insulin in hypothalamic neurons (20), its potential involvement in the stimulation of GnRH gene expression by insulin was first investigated. We show here that consistent with data obtained in Gnv-3 cells, insulin activates the PI3K pathway in hypothalamic neurons in culture. However, pretreatment of the cells with the PI3K inhibitor wortmannin had no effect on the insulin-induced increase in GnRH mRNA levels observed, suggesting that this pathway is not primarily implicated. Because we could also demonstrate that the MAPK Erk1/2 pathway is also activated after exposure of primary neuronal cell cultures to insulin, we then studied further the potential implication of this pathway. By using an enzymatic inhibitor of the MAPKs, Erk1/2, we show here that inhibition of Erk1/2 phosphorylation before insulin stimulation completely blocks the effect of insulin on GnRH mRNA levels. These results are the first demonstration of an implication of this pathway in the central effects of insulin. Although we cannot ascertain in this model whether GnRH neurons are implicated, these data are consistent with our findings in Gnv-3 cells, suggesting that Erk1/2 may be involved in an activation of GnRH neurons by insulin. This hypothesis is also corroborated by previous results showing that the stimulation of the expression of c-fos and Egr-1 by insulin, as seen in Gnv-3 cells, is dependent on MAPK activation (60, 61). Unfortunately, the effects of an inhibition of the MAPK Erk1/2 pathway could not be elucidated in Gnv-3 cells because of a toxicity of dimethylsulfoxide in that model.
In summary, the GnRH neuronal cell lines described here represent the first conditional immortalization of adult rat hypothalamic neurons. The tight control of the expression of the v-myc oncogene achieved by the TET-ON system may account for their high level of differentiation. Using these cells, which express all the cellular and molecular apparatus characteristic of insulin-sensitive tissues, we provide data suggesting that hypothalamic GnRH neurons can be directly modulated by insulin. Finally, we have identified a novel role of the MAPK Erk1/2 signaling pathway in mediating some important actions of insulin within the central nervous system, a crucial observation to understand the regulation of reproductive activity by this hormone. Further work will be necessary to demonstrate the physiological importance of this pathway in the activation hypothalamic GnRH neurons in vivo.
Acknowledgments
The authors thank Drs. Michel Aubert and Bernard Thorens for their helpful comments and suggestions during the writing of the manuscript.
Footnotes
This work was supported by Grants 32-59465.99 and 32-00B0-100 858/1 (to F.P.P.) from the Swiss National Science Foundation and a grant from the Novartis Foundation (to F.P.P.).
The authors have no conflict of interest.
First Published Online November 17, 2005
1 R.S. and E.C. contributed equally to the work
Abbreviations: CMV, Cytomegalovirus; Gnv, GnRH/v-myc; IR, insulin receptor; IRS, insulin receptor substrate; MAP, microtubule-associated protein; NFL, light chain of neurofilaments; NMDA, N-methyl D-aspartate; NPY, neuropeptide Y; NSE, neuron-specific enolase; Ob-Rb, long isoform of the leptin receptor; PI3K, phosphatidylinositol 3-kinase; POMC, proopiomelanocortin; rtTA, reverse tetracycline transactivator; SIN, self-inactivating; SNAP25, synaptosomal-associated protein of 25 kDa; SOCS, suppressor of cytokine signaling; TRE, tetracycline-responsive element; VAMP, vesicle-associated membrane protein.
Accepted for publication November 4, 2005.
References
Knobil E 1992 Remembrance: the discovery of the hypothalamic gonadotropin-releasing hormone pulse generator and of its physiological significance. Endocrinology 131:1005–1006
Marshall JC, Kelch RP 1986 Gonadotropin-releasing hormone: role of pulsatile secretion in the regulation of reproduction. N Engl J Med 315:1459–1468
Bronson FH 1998 Energy balance and ovulation: small cages versus natural habitats. Reprod Fertil Dev 10:127–137
Bronson FH 1986 Food-restricted, prepubertal female rats: rapid recovery of luteinizing hormone pulsing with excess food, and full recovery of pubertal development with gonadotropin-releasing hormone. Endocrinology 118:2483–2487
Frisch RE, McArthur JW 1974 Menstrual cycles: fatness as a determinant of minimum weight for height necessary for their maintenance or onset. Science 185:949–951
Foster DL, Nagatani S 1999 Physiological perspectives on leptin as a regulator of reproduction: role in timing puberty. Biol Reprod 60:205–215
Cunningham MJ, Clifton DK, Steiner RA 1999 Leptin’s actions on the reproductive axis: perspectives and mechanisms. Biol Reprod 60:216–222
Gonzales C, Voirol MJ, Giacomini M, Gaillard RC, Pedrazzini T, Pralong FP 2004 The neuropeptide Y Y1 receptor mediates NPY-induced inhibition of the gonadotrope axis under poor metabolic conditions. FASEB J 18:137–139
Pralong FP, Gonzales C, Voirol MJ, Palmiter RD, Brunner HR, Gaillard RC, Seydoux J, Pedrazzini T 2002 The neuropeptide Y Y1 receptor regulates leptin-mediated control of energy homeostasis and reproductive functions. FASEB J 16:712–714
Ahima RS, Prabakaran D, Mantzoros C, Qu D, Lowell B, Maratos-Flier E, Flier JS 1996 Role of leptin in the neuroendocrine response to fasting. Nature 382:250–252
Stephens TW, Basinski M, Bristow PK, Bue-Valleskey JM, Burgett SG, Craft L, Hale J, Hoffmann J, Hsiung HM, Kriauciunas A, Mackellar W, Rosteck PR, Schoner B, Smith D, Tinsley FC, Zhang XY, Heiman M 1995 The role of neuropeptide Y in the antiobesity action of the obese gene product. Nature 377:530–532
Gruaz NM, Lalaoui M, Pierroz DD, Englaro P, Sizonenko PC, Blum WF, Aubert ML 1998 Chronic administration of leptin into the lateral ventricle induces sexual maturation in severely food-restricted female rats. J Neuroendocrinol 10:627–633
Baura GD, Foster DM, Porte Jr D, Kahn SE, Bergman RN, Cobelli C, Schwartz MW 1993 Saturable transport of insulin from plasma into the central nervous system of dogs in vivo. A mechanism for regulated insulin delivery to the brain. J Clin Invest 92:1824–1830
Bruning JC, Gautam D, Burks DJ, Gillette J, Schubert M, Orban PC, Klein R, Krone W, Muller-Wieland D, Kahn CR 2000 Role of brain insulin receptor in control of body weight and reproduction. Science 289:2122–2125
Schwartz MW, Porte Jr D 2005 Diabetes, obesity, and the brain. Science 307:375–379
Griffin ML, South SA, Yankov VI, Booth Jr RA, Asplin CM, Veldhuis JD, Evans WS 1994 Insulin-dependent diabetes mellitus and menstrual dysfunction. Ann Med 26:331–340
Steger RW, Kienast SG, Pillai S, Rabe M 1993 Effects of streptozotocin-induced diabetes on neuroendocrine responses to ovariectomy and estrogen replacement in female rats. Neuroendocrinology 57:525–531
Burcelin R, Thorens B, Glauser M, Gaillard RC, Pralong FP 2003 Gonadotropin-releasing hormone secretion from hypothalamic neurons: stimulation by insulin and potentiation by leptin. Endocrinology 144:4484–4491
Bergonzelli GE, Pralong FP, Glauser M, Cavadas C, Grouzmann E, Gaillard RC 2001 Interplay between galanin and leptin in the hypothalamic control of feeding via corticotropin-releasing hormone and neuropeptide Y. Diabetes 50:2666–2672
Niswender KD, Morrison CD, Clegg DJ, Olson R, Baskin DG, Myers Jr MG, Seeley RJ, Schwartz MW 2003 Insulin activation of phosphatidylinositol 3-kinase in the hypothalamic arcuate nucleus: a key mediator of insulin-induced anorexia. Diabetes 52:227–231
Gossen M, Freundlieb S, Bender G, Muller G, Hillen W, Bujard H 1995 Transcriptional activation by tetracyclines in mammalian cells. Science 268:1766–1769
Deglon N, Tseng JL, Bensadoun JC, Zurn AD, Arsenijevic Y, Pereira de Almeida L, Zufferey R, Trono D, Aebischer P 2000 Self-inactivating lentiviral vectors with enhanced transgene expression as potential gene transfer system in Parkinson’s disease. Hum Gene Ther 11:179–190
Oberst C, Hartl M, Weiskirchen R, Bister K 1999 Conditional cell transformation by doxycycline-controlled expression of the MC29 v-myc allele. Virology 253:193–207
Zufferey R, Dull T, Mandel RJ, Bukovsky A, Quiroz D, Naldini L, Trono D 1998 Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J Virol 72:9873–9880
Jordan M, Schallhorn A, Wurm FM 1996 Transfecting mammalian cells: optimization of critical parameters affecting calcium-phosphate precipitate formation. Nucleic Acids Res 24:596–601
Freshney RI 1991 Cloning and selection of specific cell types. In: Culture of animal cells. A manual of basic techniques. New York: John Wiley, Sons, Inc.; 137–153
Igaz P, Salvi R, Rey J-P, Glauser M, Pralong FP, Gaillard RC2006 Effects of cytokines on gonadotropin-releasing hormone (GnRH) gene expression in primary hypothalamic neurons and in GnRH neurons immortalized conditionally. Endocrinology 147:1037–1043
van de Loosdrecht AA, Beelen RH, Ossenkoppele GJ, Broekhoven MG, Langenhuijsen MM 1994 A tetrazolium-based colorimetric MTT assay to quantitate human monocyte mediated cytotoxicity against leukemic cells from cell lines and patients with acute myeloid leukemia. J Immunol Methods 174:311–320
Giusti V, Verdumo C, Suter M, Gaillard RC, Burckhardt P, Pralong F 2003 Expression of peroxisome proliferator-activated receptor-1 and peroxisome proliferator-activated receptor-2 in visceral and subcutaneous adipose tissue of obese women. Diabetes 52:1673–1676
Bourguignon JP, Gerard A, Alvarez Gonzalez ML, Franchimont P 1993 Control of pulsatile secretion of gonadotrophin releasing hormone from hypothalamic explants. Hum Reprod 8(Suppl 2):18–22
Gore AC 2001 Gonadotropin-releasing hormone neurons, NMDA receptors, and their regulation by steroid hormones across the reproductive life cycle. Brain Res Brain Res Rev 37:235–248
Andersson K, Eneroth P, Fuxe K, Harfstrand A 1986 Nicotine-induced increases in brain luteinizing hormone releasing hormone-like immunoreactivity and in serum luteinizing hormone levels of the male rat. Neurosci Lett 71:289–292
Vollenweider P, Menard B, Nicod P 2002 Insulin resistance, defective insulin receptor substrate 2-associated phosphatidylinositol-3' kinase activation, and impaired atypical protein kinase C (/) activation in myotubes from obese patients with impaired glucose tolerance. Diabetes 51:1052–1059
Veldhuis JD, Johnson ML 1986 Cluster analysis: a simple, versatile, and robust algorithm for endocrine pulse detection. Am J Physiol 250:E486–E493
Cameron HA, Woolley CS, McEwen BS, Gould E 1993 Differentiation of newly born neurons and glia in the dentate gyrus of the adult rat. Neuroscience 56:337–344
Kilty IC, Barraclough R, Schmidt G, Rudland PS 1999 Isolation of a potential neural stem cell line from the internal capsule of an adult transgenic rat brain. J Neurochem 73:1859–1870
Sudhof TC, Czernik AJ, Kao HT, Takei K, Johnston PA, Horiuchi A, Kanazir SD, Wagner MA, Perin MS, De Camilli P 1989 Synapsins: mosaics of shared and individual domains in a family of synaptic vesicle phosphoproteins. Science 245:1474–1480
Mandic R, Trimble WS, Lowe AW 1997 Tissue-specific alternative RNA splicing of rat vesicle-associated membrane protein-1 (VAMP-1). Gene 199:173–179
Elferink LA, Trimble WS, Scheller RH 1989 Two vesicle-associated membrane protein genes are differentially expressed in the rat central nervous system. J Biol Chem 264:11061–11064
Bargou RC, Leube RE 1991 The synaptophysin-encoding gene in rat and man is specifically transcribed in neuroendocrine cells. Gene 99:197–204
Hepp R, Grant NJ, Chasserot-Golaz S, Aunis D, Langley K 2001 The hypophysis controls expression of SNAP-25 and other SNAREs in the adrenal gland. J Neurocytol 30:789–800
Riederer BM, Innocenti GM 1992 MAP2 isoforms in developing cat cerebral cortex and corpus callosum. Eur J Neurosci 4:1376–1386
Pirola L, Johnston AM, Van Obberghen E 2004 Modulation of insulin action. Diabetologia 47:170–184
Schwartz MW, Woods SC, Porte Jr D, Seeley RJ, Baskin DG 2000 Central nervous system control of food intake. Nature 404:661–671
Cheung CC, Thornton JE, Nurani SD, Clifton DK, Steiner RA 2001 A reassessment of leptin’s role in triggering the onset of puberty in the rat and mouse. Neuroendocrinology 74:12–21
Elmquist JK 2001 Hypothalamic pathways underlying the endocrine, autonomic, and behavioral effects of leptin. Int J Obes Relat Metab Disord 25(Suppl 5):S78–S82
Schwartz MW, Sipols AJ, Marks JL, Sanacory G, White JD, Scheurink A, Kahn SE, Baskin DG, Woods SC, Figlewicz DP, Porte Jr D 1992 Inhibition of hypothalamic neuropeptide Y gene expression by insulin. Endocrinology 130:3608–3616
Finn PD, Cunningham MJ, Pau KY, Spies HG, Clifton DK, Steiner RA 1998 The stimulatory effect of leptin on the neuroendocrine reproductive axis of the monkey. Endocrinology 139:4652–4662
Aubert ML, Pierroz DD, Gruaz NM, d’Alleves V, Vuagnat BA, Pralong FP, Blum WF, Sizonenko PC 1998 Metabolic control of sexual function and growth: role of neuropeptide Y and leptin. Mol Cell Endocrinol 140:107–113
Mellon PL, Windle JJ, Goldsmith PC, Padula CA, Roberts JL, Weiner RI 1990 Immortalization of hypothalamic GnRH neurons by genetically targeted tumorigenesis. Neuron 5:1–10
Radovick S, Wray S, Lee E, Nicols DK, Nakayama Y, Weintraub BD, Westphal H, Cutler Jr GB, Wondisford FE 1991 Migratory arrest of gonadotropin-releasing hormone neurons in transgenic mice. Proc Natl Acad Sci USA 88:3402–3406
Radovick S, Wray S, Muglia L, Westphal H, Olsen B, Smith E, Patriquin E, Wondisford FE 1994 Steroid hormone regulation and tissue-specific expression of the human GnRH gene in cell culture and transgenic animals. Horm Behav 28:520–529
Wetsel WC 1995 Immortalized hypothalamic luteinizing hormone-releasing hormone (LHRH) neurons: a new tool for dissecting the molecular and cellular basis of LHRH physiology. Cell Mol Neurobiol 15:43–78
Maggi R, Pimpinelli F, Molteni L, Milani M, Martini L, Piva F, Wetsel WC 2000 Immortalized luteinizing hormone-releasing hormone neurons show a different migratory activity in vitro. Endocrinology 141:2105–2112
Mansuy IM, Bujard H 2000 Tetracycline-regulated gene expression in the brain. Curr Opin Neurobiol 10:593–596
Wetsel WC, Valenca MM, Merchenthaler I, Liposits Z, Lopez FL, Weiner RI, Mellon PL, Negro-Vilar A 1992 Intrinsic pulsatile secretory activity of immortalized luteinizing hormone-releasing hormone-secreting neurons. Proc Natl Acad Sci USA 89:4149–4153
Martinez de la Escalera G, Choi AL, Weiner RI 1992 Generation and synchronization of gonadotropin-releasing hormone (GnRH) pulses: intrinsic properties of the GT1–1 GnRH neuronal cell line. Proc Natl Acad Sci USA 89:1852–1855
Emanuelli B, Glondu M, Filloux C, Peraldi P, Van Obberghen E 2004 The potential role of SOCS-3 in the interleukin-1-induced desensitization of insulin signaling in pancreatic -cells. Diabetes 53(Suppl 3):S97–S103
Emanuelli B, Peraldi P, Filloux C, Chavey C, Freidinger K, Hilton DJ, Hotamisligil GS, Van Obberghen E 2001 SOCS-3 inhibits insulin signaling and is up-regulated in response to tumor necrosis factor- in the adipose tissue of obese mice. J Biol Chem 276:47944–47949
Keeton AB, Bortoff KD, Bennett WL, Franklin JL, Venable DY, Messina JL 2003 Insulin-regulated expression of Egr-1 and Krox20: dependence on ERK1/2 and interaction with p38 and PI3-kinase pathways. Endocrinology 144:5402–5410
Harada S, Smith RM, Smith JA, White MF, Jarett L 1996 Insulin-induced egr-1 and c-fos expression in 32D cells requires insulin receptor, Shc, and mitogen-activated protein kinase, but not insulin receptor substrate-1 and phosphatidylinositol 3-kinase activation. J Biol Chem 271:30222–30226(Roberto Salvi1, Einar Castillo1, Marie-J)