Neurokinin B induces c-fos transcription via protein kinase C and activation of serum response factor and Elk-1 in immortalized GnRH neurons

Abstract

Mutations in neurokinin B (NKB) and its receptor, NK3R, were identified in human patients with hypogonadotropic hypogonadism, a disorder characterized by lack of puberty and infertility. Further studies have suggested that NKB acts at the level of the hypothalamus to control GnRH neuron activity, either directly or indirectly. We recently reported that treatment with senktide, a NK3R agonist, induced GnRH secretion and expression of c-fos mRNA in GT1-7 cells. Here, we map the responsive region in the murine c-fos promoter to between -400 and -200 bp, identify the signal transducer and activator of transcription (STAT) (-345) and serum response element (-310) sites as required for induction, a modulatory role for the Ets site (-318), and show that induction is protein kinase C dependent. Using gel shift and Gal4 assays, we further show that phosphorylation of Elk-1 leads to binding to DNA in complex with serum response factor at serum response element and Ets sites within the c-fos promoter. Thus, we determine molecular mechanisms involved in NKB regulation of c-fos induction, which may play a role in modulation of GnRH neuron activation.

Figure 1.

The induction of c-fos promoter activity by senktide requires NK3R and is time and dose dependent. A, GT1-7 cells were transiently transfected with 100 ng TK β-galactosidase, 400 ng of a −1000-bp mouse c-fos luciferase reporter, and either 200 ng of a rat NK3R expression vector or pcDNA3.1 vector control. Cells were treated for 3.5 hours with media change, vehicle, or 30nM senktide. B, GT1-7 cells were transiently transfected with 100-ng TK β-galactosidase, 400-ng −1000-bp c-fos luciferase reporter, and 200 ng of NK3R and then treated with 30nM senktide for 2, 3, 4, 5, or 8 hours. In both A and B, luciferase RLU data were normalized using a β-galactosidase internal control and are represented as fold vs cells treated with media change ± SEM (n = 3–4); *, senktide treated is different from vehicle-treated control by Student's t test (P < .05). C, GT1-7 cells were transiently transfected with 100-ng TK β-galactosidase, 400-ng rat NK3R, and 400 ng of a −1000-bp c-fos luciferase reporter and treated for 3.5 hours with vehicle or senktide at the indicated doses. Luciferase RLU data were normalized using a β-galactosidase internal control and are represented as fold vs cells treated with media change ± SEM (n = 4). A representative dose-response curve fitted by nonlinear regression using a 4-parameter logistic equation is shown. The mean EC₅₀ was 0.037 ± 0.01nM.

Figure 2.

The induction of c-fos gene transcription in response to NK3R activation maps to STAT, SRE, and Ets sites located between −200 and −400 bp in the murine promoter. A, GT1-7 cells were transiently transfected with 100-ng TK β-galactosidase, 200-ng NK3R, and 400 ng of full-length or one of the indicated truncated murine c-fos promoter luciferase reporters (−600, −400, or −200 bp). Cells were treated for 3.5 hours with vehicle or 30nM senktide (n = 3). B, A schematic representation of the location of previously characterized binding sites for STAT, Ets, SRE, and AP-1 between −400 and −200 bp upstream from the transcription start site in the murine c-fos promoter. C, GT1-7 cells were transiently transfected with 100-ng TK β-galactosidase, 200-ng NK3R, and 400 ng of either wild-type or a −1000-bp c-fos luciferase reporter with a cis-mutation in the STAT-, Ets-, SRE-, or AP-1-binding site. Luciferase data are normalized to β-galactosidase and represented as fold vs vehicle-treated, wild-type −1000-bp c-fos luc ± SEM (n = 3). D, GT1-7 cells were transiently transfected with 400-ng NK3R and 400-ng luciferase reporters containing either −1000-bp c-Fos or 4× multimers of STAT- or SRE-binding sites cloned upstream from a TK promoter. RLU Luciferase/β-galactosidase data are normalized to vehicle-treated pGL3 vector control and represented as mean ± SEM (n = 3). Bars labeled with different letters are significantly different by two-way ANOVA, Tukey HSD.

Figure 3.

Basal and c-fos promoter activity induced in response to NK3R activation is reduced by the blockade of SRF and STAT signaling. GT1-7 cells were transiently transfected with 100-ng TK β-galactosidase, either 300 ng of NK3R or pcDNA3.1, 300 ng of a −400-bp c-fos promoter luciferase reporter, and either 200 ng of a truncation mutant (1-266) SRF, a dominant-negative to SRF, or the SOCS3, a negative regulator of STAT signaling. The total amount of DNA used for transfection was balanced using the empty expression vector pCGN (control for SRF 1-266) or phEF-neo (control for SOCS3). Cells were treated for 3.5 hours with either vehicle or 0.4nM senktide. RLU Luciferase/β-galactosidase data are normalized to vehicle-treated pcDNA3.1 and represented as mean ± SEM (n = 4). Bars labeled with different letters are significantly different by two-way ANOVA, Tukey HSD.

Figure 4.

SRF and c-fos promoter activity induced by NK3R are PKC dependent. GT1-7 cells were transfected with 100-ng TK β-galactosidase, 400 ng of either NK3R or an empty pcDNA3.1 expression vector control, and 400 ng of (A) the SRE multimer or (B) the −400-bp c-fos luciferase (Luc) reporter. Cells were pretreated for 30 minutes with either PKC inhibitor Go6983 or DMSO control and then cotreated for 3.5 hours with either vehicle or 0.4nM senktide. RLU Luciferase/β-galactosidase data are normalized to DMSO and vehicle-treated pcDNA3.1 and represented as mean ± SEM (n = 4). Bars labeled with different letters are significantly different by two-way ANOVA, Tukey HSD.

Figure 5.

Activation of PKC increases the amount of phosphorylated Elk-1 binding in complex with SRF to the Ets site within the c-fos promoter. A, Gel shift electromobility assay is performed using preincubation of nuclear extracts from GT1-7 cells treated for times indicated with 100nM TPA, an activator of PKC, with a double-stranded oligonucleotide probe with the SRE-binding site from the c-fos promoter. The complex bound to the probe (lanes 1–5) was shown to contain both phosphorylated and nonphosphorylated forms of SRF protein by supershift (SS) assay using preincubation with antibodies specific for SRF (α-SRF, lanes 13–17) and pSRF (α-pSRF, lanes 6–10). The complex was not supershifted by IgG control (lane 11) and was competed away by unlabeled probe (Self, lane 12). Film was exposed for 4 days with an intensifying screen at room temperature. B, Gel shift electromobility assay was performed using preincubation of nuclear extracts from GT1-7 cells treated for times indicated with 100nM TPA, an activator of PKC, with a double-stranded oligonucleotide probe with the SRE and the adjacent Ets-binding sites from the c-fos promoter (lanes 1–5). The supershift assay was performed with antibodies specific for either pSRF (lanes 6–10) or phosphorylated Elk-1 (lanes 16–20) and the IgG control (lane 11). Competition assay was performed using an oligonucleotide probe with wild-type (WT) Ets + SRF sites (lane 12), mutation (μ) in either Ets (lane 13) or SRF (lane 14), or mutation in both SRE and Ets sites (lane 15). Film was exposed overnight with an intensifying screen at −80°C.

Figure 6.

Elk-1 protein association with DNA is increased by activation of either PKC or NK3R. GT1-7 cells were transfected with 100-ng TK β-galactosidase, 300 ng of either pcDNA3.1 expression vector control or NK3R, 300 ng of a Gal4-Elk-1 fusion protein, and 300 ng of a Gal4 UAS promoter luciferase reporter. Cells were treated for 3.5 hours with vehicle control, 100nM TPA, or 30nM senktide. RLU Luciferase/β-galactosidase data are normalized to control cells transfected with pcDNA3.1 and represented as mean ± SEM (n = 6). Bars labeled with different letters are significantly different by two-way ANOVA, Tukey HSD.