Abstract
Serotonin 5-HT2C receptors (5-HT2CR) activate Gq proteins and are expressed in the central nervous system (CNS). 5-HT2CR regulate emotion, feeding, reward, or cognition and may serve as promising drug targets to treat psychiatric disorders or obesity. Owing to technical difficulties in isolating cells from the CNS and the lack of suitable cell lines endogenously expressing 5-HT2CR, our knowledge about this receptor subtype in native environments is rather limited. The hypothalamic mHypoA-2/10 cell line was recently established and resembles appetite-regulating hypothalamic neurons of the paraventricular nucleus (PVN), where 5-HT2CR have been detected in vivo. Therefore, we tested mHypoA-2/10 cells for endogenous 5-HT2CR expression. Serotonin or the 5-HT2CR preferential agonist WAY-161,503 initiated cAMP response element (CRE)-dependent gene transcription with EC50 values of 15.5 ± 9.8 and 1.1 ± 0.9 nM, respectively. Both responses were blocked by two unrelated 5-HT2CR-selective antagonists (SB-242,084, RS-102,221) but not by a 5-HT2AR (EMD-281,014) or 5-HT2BR (RS-127,455) antagonists. By single-cell calcium imaging, we found that serotonin and WAY-161,503 induced robust calcium transients, which were also blunted by both 5-HT2CR antagonists. Additionally we revealed, first, that 5-HT2CR induced CRE activation via protein kinase C (PKC)-mediated engagement of extracellular-regulated kinases-1/2 and, second, that intrinsic activity of WAY-161,503 was in the range of 0.3–0.5 compared with serotonin, defining the frequently used 5-HT2CR agonist as a partial agonist of endogenous 5-HT2CR. In conclusion, we have shown that hypothalamic mHypoA-2/10 cells endogenously express 5-HT2CR and thus are the first cell line in which to analyze 5-HT2CR pharmacology, signaling, and regulation in its natural environment.
Introduction
The neurotransmitter 5-hydroxytryptamine (5-HT) or serotonin exerts its physiologic effects via at least 14 distinct receptor subtypes, including 5-HT1 receptors that couple to G proteins of the Gi/o family and 5-HT2 receptors coupling to Gq proteins (Hoyer et al., 1994; Chang et al., 2000). The 5-HT2C receptor (5-HT2CR) subtype is predominantly expressed in the central nervous system (CNS), mostly in brain regions associated with emotion, feeding, reward, or cognition, and thus might serve as a drug target for the treatment of CNS disorders (Bhatnagar et al., 2001; Di Giovanni and De Deurwaerdere, 2016). In fact, 5-HT2CR activity–controlling substances have been suggested as therapeutics against schizophrenia, depression, Parkinson’s disease, anxiety, sleeping disorders, drug abuse, and obesity (Tecott et al., 1995; Frank et al., 2002; Rocha et al., 2002; Meltzer et al., 2003; Isaac, 2005; Millan, 2005; Di Giovanni et al., 2006; Thomsen et al., 2008). A major obstacle hampering the development of 5-HT2CR-targeting drugs are technical difficulties in isolating significant numbers of primary cells from the CNS and the current lack of suitable CNS-derived cell lines that endogenously express 5-HT2CR. Hence, discovering reliable CNS-derived cell lines that endogenously express the functional 5-HT2CR protein will be an important asset for 5-HT2CR-related research in general and for the development 5-HT2CR-targeting drugs in particular.
5-HT-producing neurons of the raphe nuclei project to different hypothalamic regions, including the nucleus arcuatus (ARC) and the paraventricular nucleus (PVN) (Lam et al., 2010). 5-HT released to hypothalamic neurons reduces food intake by at least three distinct mechanisms: 1) 5-HT2CR-induced excitation of the anorexigenic melanocyte-stimulating hormone (MSH)–producing ARC neurons, 2) 5-HT1BR-mediated inhibition of the orexigenic neuropeptide Y (NPY)–producing ARC neurons, and 3) action upon MSH- and NPY-sensitive PVN neurons via 5-HT1BR and 5-HT2CR (Currie et al., 1999, 2002; Sohn et al., 2011; Doslikova et al., 2013; Burke et al., 2014). Consequently, mice deficient for the 5-HT2CR protein exhibit severe hyperphagia leading to diabetes and concomitant hyperglycemia and insulin resistance (Tecott et al., 1995; Heisler et al., 1998; Tecott and Abdallah, 2003). Furthermore, administration of a 5-HT2CR (WAY-161,503) or a 5-HT1BR (CP-94,253) preferential agonists inhibited food intake by reducing meal size and duration (Halford and Blundell, 1996; Lee and Simansky, 1997; Rosenzweig-Lipson et al., 2006; Thomsen et al., 2008; Fletcher et al., 2009; Doslikova et al., 2013). Overall, it appears that PVN neurons play an important role in the central effects of 5-HT on appetite control and express 5-HT1BR and HT2CR subtypes.
The hypothalamic cell line mHypoA-2/10 has recently been introduced and characterized as a suitable PVN-derived cell model with regard to sensitivity toward MSH and NPY, but its sensitivity toward 5-HT has not been investigated yet (Belsham et al., 2009; Dalvi et al., 2012; Nazarians-Armavil et al., 2013; Breit et al., 2015). In the present study we dissected the effects of 5-HT, WAY-161,503, and CP-94,253 on classic G protein-signaling cascades that involve cAMP, calcium, extracellular-regulated kinases-1/2 (ERK-1/2), and CRE binding protein (CREB)/cAMP response element (CRE). We found that 5-HT and CP-94,253 inhibited forskolin-induced cAMP production, indicating expression of Gi/o-coupled 5-HT1BR in mHypoA-2/10 cells. 5-HT and WAY-161,503 elicited calcium transients, extracellular-regulated kinases-1/2 phosphorylation, and CREB/CRE activation, which was blocked by two distinct selective 5-HT2CR but not by a 5-HT2AR or 5-HT2BR antagonist, providing strong pharmacological evidence of functional 5-HT2CR expression in mHypoA-2/10 cells. Hence, mHypoA-2/10 cells are a unique cell line that now allows the analysis of the 5-HT2CR protein in its natural environment and that could serve as a 5-HT-sensitive PVN-derived cell model permitting the characterization of HT2CR- and 5-HT1BR-induced signaling when activated individually or in concert.
Materials and Methods
Materials
Invitrogen cell culture reagents were obtained from ThermoFisher Scientific (Darmstadt, Germany) and TurboFect from ThermoScientific Fermentas (Schwerte, Germany). The anti-pERK-1/2 (E-4) antiserum was from Santa Cruz Biotechnology (Heidelberg, Germany). The pCREB-Ser-133 antibody was purchased from Cell Signaling Technology (Leiden, The Netherlands), and RS-127,455 or the peroxidase-conjugated anti-mouse or anti-rabbit antibody, both raised in goat, from Sigma-Aldrich (Deisenhofen, Germany). The firefly luciferase substrate was from Promega (Mannheim, Germany). BIM-X, 5-HT, WAY-161,503, CP-94,253, EMD-281,014, and NPY were from Tocris Bioscience (Wiesbaden-Nordenstadt, Germany), and MSH from Biotrend Chemicals (Cologne, Germany). PD-184,352 was from Enzo Life Sciences (Lörrach, Germany). The signal transducer and activator of transcription-3 (STAT-3) and the nuclear factor of activated T cells (NFAT) reporter genes were purchased from Promega. The CRE reporter, containing six 5′-TGACCTCAC-3′sites, has been reported previously (Damm et al., 2012).
Cell Culture and Transfections
The adult mouse hypothalamic cell line mHypoA-2/10 (Clu-176) was purchased from Cedarlane Labs (Burlington, Canada) and cultured in Dulbecco’s modified Eagle’s medium (DMEM; 10% fetal bovine serum, 2 mM l-glutamine, penicillin/streptomycin) at 37°C and 5% CO2. To obtain mHypoA-2/10 cells, hypothalamic neurons from 2-month-old male C57Bl/6 mice were isolated, ciliary neurotrophic factor was used to trigger neurogenesis of hypothalamic primary cultures, and immortalization was induced by retroviral transfer of SV40 T-Ag (Belsham et al., 2009). To obtain mHypoA-2/10 cells stably expressing the CRE reporter, cells were cotransfected with the CRE-reporter construct and an empty pcDNA4 vector containing the resistance gene for zeocin (Breit et al., 2015). Transfected cells were then selected with 600 μg/ml zeocin for 4 weeks and positively tested cell pools were used for further studies.
Fura-2-Based Single-Cell Calcium Imaging
Seventy-two hours prior to the experiment mHypoA-2/10 cells were seeded at a density of 2–4 × 104 cells on glass cover slips in six-well plates coated with 0.1% poly-l-lysine. Twenty-four hours before the experiment cells were serum-starved and at the day of the experiment loaded with 10 μM Fura 2-AM in HEPES-buffered Iscove’s modified Dulbecco’s medium (without phenol red) supplemented with 0.02% pluronic F-127. Cover slips were washed twice with Iscove’s modified Dulbecco’s medium and placed in a recording chamber. Cells were analyzed using a Polychrome 5000 monochromator (Till-Photonics, Gräfelfing, Germany) and an Andor charge-coupled device camera coupled to an inverted microscope (IX71, Olympus, Hamburg, Germany). Calcium binding to Fura-2 was determined by alternate excitation of the cells with 340 nm or 380 nm every 0.5 seconds and concomitant detection of fluorescence emission at 520 nm. Fura-2 ratios (340/380) were normalized by defining the first ratio measured as 100%.
Western Blot
Cells (∼100,000/well) were seeded on six-well plates, cultured for 1 day, serum-starved for 24 hours the next day, stimulated for the indicated period of time with 5-HT, WAY-161,503, or MSH, and then lysed in Laemmli buffer. Lysates were subjected to SDS-PAGE and protein transferred to nitrocellulose by Western blotting. For detection of ERK-1/2 or CREB, phosphorylation blots were separated by a horizontal cut at 25 kDa. The upper part was used to analyze ERK-1/2 or CREB phosphorylation with phospho-specific antisera and the lower part to detect total histone proteins as a loading control. Immune-reactivity was quantified by densitometry, ratios between p-CREB or p-ERK-1/2 and histone signals were calculated, and ligand-induced protein phosphorylation normalized to not-stimulated cells.
ERK-1/2 pThr202/Tyr204 Enzyme-Linked Immunosorbent Assay
Phosphorylated ERK-1/2 was also detected using the PhosphoTracer ELISA Kit from abcam (ab119659) according to the manufacturer’s protocol.
Whole-Cell ELISA.
Cells were seeded in 24-well plates (∼50,000/well) 2 days prior to the experiment. After 20 hours of serum-starvation, cells were stimulated with increasing concentrations of 5-HT or WAY-161,503 for 3 minutes, placed on ice, and washed with ice-cold phosphate-buffered saline (PBS). After fixing the cells for 15 minutes with 4% paraformaldehyde, they were permeabilized with ice-cold methanol/acetone (50:50) for 5 minutes. Cells were then washed with PBS and then incubated with 1.0% of bovine serum albumin (BSA) in PBS for 15 minutes at room temperature (RT) to block unspecific binding sites. After an additional washing step, cells were incubated for 30 minutes at RT with anti-p-ERK-1/2 antiserum (1:2500) in PBS/BSA. First, antibodies were removed, cells washed, and incubated with the HRP-conjugated goat anti-rabbit secondary antibody (1:4,000) in PBS/BSA for 30 minutes at RT. After several washing steps 200 μl of the 1-Step-Ultra-TMB-ELISA Substrate (ThermoFisher Scientific, Waltham, MA) was added to the cells and incubated for 15 minutes at RT. The reaction was stopped by adding 50 ml of 1 M H2SO4, and 200 μl transferred to 96-well plates with clear bottom and OD450 measured in a FLUOstar Omega plate reader (BMG LABTECH, Inc., Cary, NC).
Reporter Gene Assays
mHypoA-2/10 cells (∼20,000/well) were seeded on 12-well plates and 24 hours later transfected with various luciferase reporter gene constructs using the TurboFect reagent from ThermoFisher Scientific according to the manufacturer’s protocol. mHypoA-2/10-CRE cells (∼20,000/well) were seeded on 24-well plates. After serum-starvation of 24 hours, cells were stimulated the following day for 4–6 hours in serum-free medium. After stimulation cells were lysed (25 mM Tris/HCl, pH 7.4, 4 mM EGTA, 8 mM MgCl2, 1 mM dithiothreitol, and 1% Triton-X-100), and luciferase activity measured in white-bottom 96-well plates after automatic injection of luciferase substrate. Resulting total light emission was detected every second for 10 seconds postinjection in a FLUOstar Omega plate reader. Average light emission during this period was recorded and is indicated relative to unstimulated cells.
cAMP Accumulation
To determine agonist-induced cAMP accumulation, ∼100,000 cells were seeded in 12-well dishes 48 hours prior to the experiment and labeled in serum-free DMEM containing 2 μCi/ml of [3H]adenine for 24 hours. Cells were stimulated for 30 minutes at 37°C in DMEM containing 1 μM IBMX with MSH alone or along with forskolin and NPY, 5-HT, CP-94,253, or WAY-161,503. The reaction was terminated by removal of the medium and addition of ice-cold 5% trichloroacetic acid to the cells. [3H]ATP and [3H]cAMP were then purified by sequential chromatography (DOWEX-resin/aluminum oxide columns), and the accumulation of [3H]cAMP was expressed as the ratio of [3H]cAMP/([3H]cAMP + [3H]ATP).
Data Analysis
Data were analyzed using Prism5.0 (GraphPad Software Inc., San Diego, CA). Statistical significance of differences was assessed by the one-test indicated by hash signs (###P < 0.001, ##P < 0.01, #P < 0.05) or by a two-sample Student’s t test indicated by asterisks (***P < 0.001, **P < 0.01, *P < 0.05). In Fig. 1B Kolmogorov-Smirnov test was first used to ensure that the data follow a normal distribution. Next, one-way two-way analysis of variance (ANOVA) was performed to analyze the variance of the data. Results were indicated as F[degree of freedom (Df) of the sample/Df of the residual (error)] = F-value, and the P value. Afterward Tukey’s multiple-comparison test was used to indicate a significant difference to forskolin-treated cells by dagger signs (†††P < 0.001, ††P < 0.01, †P < 0.05). In Figs. 3C and 6D, Kolmogorov-Smirnov test was also used to ensure that the data follow a normal distribution. Next, ANOVA was performed to analyze the variance of the data. Results were indicated as F[Df of the sample/Df of the residual (error)] = F-value, and the P value. Afterward Bonferroni post-test was used to indicate a significant differences to basal by dagger signs (†††P < 0.001, ††P < 0.01, †P < 0.05).
mHypoA-2/10 cells respond to MSH, NPY, and 5-HT on the level of cAMP. (A, B) cAMP accumulation was assessed after labeling of serum-starved (24 hours) mHypoA-2/10 cells with [3H]adenine followed by the purification of [3H]cAMP by chromatography. (A) Cells were stimulated with 1 μM MSH and (B) with 1 μM forskolin (FSK) alone or along with 1 μM 5-HT, 500 nM CP-94,253, 500 nM WAY-161,503, or 100 nM NPY for 30 minutes at 37°C. (A) Data of five independent experiments performed in quadruplicate are presented as the mean ± S.E.M. Asterisks indicate a significant difference between basal and MSH (two-sample t test). (B) Upper panel, one representative experiment performed in quadruplicate. Lower panel, data of five independent experiments were normalized and presented as the mean ± S.E.M. Hash signs (one-sample t test) indicate a significant difference to zero. One-way ANOVA test results are F(4/18) = 7.891, P < 0.01. Dagger signs indicate significant differences (Tukey’s test) to FSK alone.
Results
mHypoA-2/10 Cells Respond to 5-HT with Decreased cAMP Production.
mHypoA-2/10 cells express a wide range of hypothalamic markers and, as with to PVN neurons, have been shown to respond to MSH and NPY (Belsham et al., 2009; Dalvi et al., 2012; Nazarians-Armavil et al., 2013; Breit et al., 2015). In line with these previous reports, mHypoA-2/10 cells used herein strongly reacted to MSH with cAMP accumulation, indicating expression of Gs-coupled MSH receptors (Fig. 1A). NPY significantly inhibited forskolin-induced cAMP production (Fig. 1B), illustrating that mHypoA-2/10 cells respond to MSH and NPY on the level of cAMP. It has been shown that PVN neurons respond to 5-HT either via the Gi/o-coupled 5-HT1BR or the Gq-coupled 5-HT2CR subtype. In line with these observations, 5-HT- and the 5-HT1BR-selective agonist CP-94,253 inhibited forskolin-induced cAMP production to the same extent as did NPY, indicating expression of the 5-HT1BR subtype in mHypoA-2/10 cells.
5-HT2CR Activate CREB-Dependent Gene Expression in mHypoA-2/10 Cells.
Having shown that mHypoA-2/10 cells respond to 5-HT on the level of cAMP, we next sought to analyze putative effects of 5-HT on gene expression in hypothalamic cells. To this end, we stimulated mHypoA-2/10 cells expressing either a SRE/TCF-, STAT-, CREB-, or NFAT-dependent reporter gene construct with 5-HT for 6 hours. As shown in Fig. 2, 5-HT enhanced CREB-dependent gene expression, providing further evidence for endogenous 5-HTR expression in mHypoA-2/10 cells. Thus, we took advantage of previously established mHypoA-2/10-CRE cells, which stably express the CREB reporter, to determine which 5-HTR subtype is involved in 5-HT-induced CRE activation. The 5-HT and the 5-HT2CR preferential agonist WAY-161,503, but not the 5-HT1BR agonist CP-94,253, activated the CREB reporter (Fig. 3A), suggesting that endogenous 5-HT2CR enhance CRE-controlled gene expression in the PVN-derived cell model. However, WAY-161,503 has been reported to activate 5-HT2AR, 5-HT2BR, and 5-HT2CR, but only the latter with a potency in the low nanomolar range. Therefore, we determined concentration-response curves of 5-HT and WAY-161,503 in mHypoA-2/10-CRE cells (Di Giovanni and De Deurwaerdere, 2016). As shown in Fig. 3B, WAY-161,503 induced CREB-dependent reporter activity with a potency of 1.1 ± 0.9 nM, strongly suggesting that the 5-HT2CR subtype is the target of WAY-161,503 in mHypoA-2/10 cells. Furthermore, the efficacy of WAY-161,503 to activate the CRE reporter was significantly lower compared with 5-HT (∼45%), suggesting that WAY-161,503 is a partial 5-HT2CR agonist or that 5-HT interacts with other CRE-activating 5-HTRs besides HT2CR. To further evaluate the role of distinct 5-HTR in hypothalamic CRE activation, we took advantage of the 5-HT2CR antagonists SB-242,084 or RS-102,221, the 5-HT2AR antagonist EMD-281,014, and the 5-HT2BR antagonist RS-127,455 (Cussac et al., 2000; Walker et al., 2007). Both 5-HT2CR, but none of the other antagonists, blunted WAY-161,503-induced CRE-dependent gene expression (Fig. 3C), indicating that 5-HT2CR are indeed activated by WAY-161,503 in mHypoA-2/10 cells. 5-HT-induced CRE activation (Fig. 3C) and CREB phosphorylation (Fig. 3D) was also fully inhibited by 5-HT2CR, but not by 5-HT2AR or 5-HT2BR antagonists, providing further evidence for 5-HT2CR expression and suggesting that WAY-161,503 and 5-HT engage the same receptor to activate CRE in mHypoA-2/10 cells, albeit with distinct intrinsic activities.
5-HT activates a CREB-dependent reporter in mHypoA-2/10 cells. mHypoA-2/10 cells were transfected with (A) serum response factor/ternary complex factor (SRF/TCF-), (B) STAT-3-, (C) CREB-, or (D) NFAT-dependent reporter constructs, serum-starved for 24 hours, stimulated with 5-HT (1 μM) for 6 hours, and then luciferase expression in cell lysates determined after injection of luciferin at time point 0 seconds. One representative experiment performed in quadruplicate is shown. Data of five independent experiments are compiled in (E) by setting the corresponding basal reporter activation to 1.0 and are given as the mean ± S.E.M. Hash sign (one-sample t test) indicates a significant difference to 1.0.
5-HT2CR phosphorylate CREB and activate CRE in mHypoA-2/10 cells. Ligand-induced reporter activation was measured in serum-starved (24 hours) mHypoA-2/10-CRE cells. (A) Cells were stimulated with 1 μM 5-HT, 500 nM WAY-161,053, or 500 nM CP-94,253 for 6 hours and luciferase expression determined. Data from two representative experiments performed in quadruplicate were compiled and expressed as the mean ± S.E.M. One-way ANOVA test results are F(2/12) = 4.623, P < 0.01. Dagger signs indicate significant differences (Tukey’s test) to basal. (B) Cells were stimulated with increasing concentrations of 5-HT or WAY-161,053. Compared with (A), the substrate concentrations were increased to enhance the signal. Data of four independent experiments performed in quadruplicate were compiled by subtracting the corresponding basal value and expressed as the mean ± S.E.M. (C) dimethyl sulfoxide (DMSO) (0.01%), SB-242,084 (100 nM), RS-102,221 (500 nM), RS-127,455 (500 nM), or EMD-281,014 (500 nM) pre-treated cells were stimulated or not with 1 μM 5-HT or 500 nM WAY-161,503. No significant differences between DMSO and cells treated with any antagonist were observed. Data from 10–20 experiments performed in quadruplicate were compiled by setting basal (DMSO) to 1.0 and expressed as the mean ± S.E.M. Hash signs indicate a significant difference (one-sample t test) to 1.0, asterisks (two-sample t test) between 5-HT and WAY-161,503. Two-way ANOVA test results are F(2/133) = 7.501, P < 0.01. Dagger signs indicate significant differences (Bonferroni post-test) to basal (DMSO). (D) 5-HT-induced CREB phosphorylation was measured in serum-starved mHypoA-2/10 cells stimulated for the indicated period of time with 5-HT (1 μM). Cells were pretreated with DMSO (0.01%), SB-242,084 (100 nM; left), RS-127,455 (500 nM; middle), or EMD-281,014 (500 nM; right) and lysates subjected to Western analysis using either a phospho-specific antiserum against CREB (p-CREB) or against the total histone protein to control for the total protein amount. One representative blot is shown. Data of 8–12 lysates were quantified by densitometry, ratios between p-CREB and histone signals calculated, and 5-HT-induced CREB phosphorylation normalized to not-stimulated cells and expressed as the mean ± S.E.M. Asterisks (two-sample t test) indicate a significant difference between DMSO and SB-242,084 pre-treated cells.
5-HT2CR Phosphorylate ERK-1/2 in mHypoA-2/10 Cells.
In nonhypothalamic cells, 5-HT phosphorylates ERK-1/2 via 5-HT2CR and almost all other 5-HTR subtypes. Hence, we next tested whether 5-HT would induce ERK-1/2 phosphorylation in mHypoA-2/10 cells. 5-HT stimulated significant ERK-1/2 phosphorylation after 2.5 and 5.0 minutes of stimulation, but as with our results in the CRE assay, the effects of WAY-161,503 were much weaker and did not reach statistical differences with unstimulated cells (Fig. 4A). Therefore, we additionally analyzed ERK-1/2 phosphorylation with a highly sensitive enzyme-linked immunosorbent assay (ELISA) (Fig. 4B) and a whole-cell ELISA (Fig. 4C). By both approaches, we detected significant increases in ERK-1/2 phosphorylation in response to 5-HT and WAY-161,503. The 5-HT2CR preferential agonist exhibited a potency in the low nanomolar range (EC50 value: 1.0 ± 0.5 nM) and significantly reduced efficacy compared with 5-HT (∼47%), further supporting our findings that mHypoA-2/10 cells endogenously express 5-HT2CR and that WAY-161,503 acts as a partial agonist. In line with this notion and as with results obtained on the level of CREB and CRE, 5-HT-induced ERK-1/2 phosphorylation was fully inhibited by a 5-HT2CR antagonist and not by 5-HT2AR or 5-HT2BR antagonists (Fig. 4D).
5-HT2CR phosphorylate ERK-1/2 in mHypoA-2/10 cells. (A) Ligand-induced ERK-1/2 phosphorylation was measured in serum-starved mHypoA-2/10 cells stimulated for the indicated period of time with 5-HT (1 μM) or WAY-161,503 (500 nM). Cell lysates were subjected to Western analysis using either a phospho-specific antiserum against ERK-1/2 (p-ERK-1/2) or against the total histone protein to control for the total protein amount. One representative blot is shown. Data of six lysates were quantified by densitometry, ratios between p-ERK-1/2 and histone signals calculated, ligand-induced ERK-1/2 phosphorylation normalized to not-stimulated cells and expressed as the mean ± S.E.M. Hash signs indicate a significant difference (one-sample t test) to 100. (B) Ligand-induced ERK-1/2 phosphorylation was measured in serum-starved mHypoA-2/10 cells stimulated for the indicated period of time with 5-HT (1 μM) or WAY-161,503 (500 nM). Cell lysates were subjected to ELISA analysis using a phospho Tracer ELISA Kit. Data of four (5-HT) or six (WAY-163,503) independent experiments are shown after subtracting the corresponding basal. Asterisks indicate a significant (two-sample t test) difference between ligand stimulation and time point zero. (C) ERK-1/2 phosphorylation in mHypoA-2/10 cells was measured by whole-cell ELISA after stimulation with increasing concentration of 5-HT or WAY-161,503 for 3 minutes. Data of four experiments performed in quadruplicate were compiled by subtracting the corresponding basal and expressed as the mean ± S.E.M. (D) ERK-1/2 phosphorylation was measured in serum-starved mHypoA-2/10 cells pretreated with DMSO (0.01%) or SB-242,084 (100 nM; left), RS-127,455 (500 nM; center), or EMD-281,014 (500 nM; right) and stimulated for the indicated period of time with 5-HT (1 μM). Cell lysates were subjected to Western analysis using either a phospho-specific antiserum against ERK-1/2 (p-ERK-1/2) or against the total histone protein to control for the total protein amount. One representative blot is shown. Data of 8–12 lysates were quantified by densitometry, ratios between p-ERK-1/2 and histone signals calculated, and 5-HT-induced ERK-1/2 phosphorylation normalized to unstimulated cells and expressed as the mean ± S.E.M. Asterisks indicate a significant difference (two-sample t test) between DMSO and SB-242,084 pretreated cells.
5-HT2CR Induce Calcium Signaling in mHypoA-2/10 Cells.
To further characterize 5-HT2CR signaling in mHypoA-2/10 cells, we next tested 5-HT- and WAY-161,503-induced calcium signaling by performing Fura-2-based single-cell calcium imaging. As seen in Fig. 5A, 5-HT induced robust calcium signals that peaked at ∼20% above basal. Effects of 5-HT on calcium signaling were significantly reduced by SB-242,084 (∼83%) and RS-102,222 (∼86%), suggesting a significant role for the 5-HT2CR subtype in 5-HT-induced calcium signaling. WAY-161,503 also significantly induced calcium signals, which were strongly inhibited by both 5-HT2CR antagonists and, in line with our data obtained on the level of CRE and ERK-1/2, exhibited reduced efficacy (∼5% above basal) compared with 5-HT (Fig. 5B).
Ligand-induced calcium signals in single, Fura-2-loaded, serum-starved (24 hours) mHypoA-2/10 cells were monitored. (A) 5-HT (1 μM) was manually added at time point 25 seconds either to DMSO (0.01%), SB-242,084 (100 nM), or RS-102,221 (500 nM) pretreated (5.0 minutes) cells. Data of 1037 (DMSO), 401 (SB-242,084), or 249 (RS-102,221) cells were compiled and expressed as the mean ± S.E.M. (B) WAY-161,503 (500 nM) was manually added at time point 25 seconds either to DMSO (0.01%), SB-242,084 (100 nM), or RS-102,221 (500 nM) pretreated (5.0 minutes) cells. Data of 1044 (DMSO), 410 (SB-242,084), or 423 (RS-102,221) cells were compiled and expressed as the mean ± S.E.M
5-HT2CR Activate CRE-Dependent Reporter Activity via Protein Kinase C and ERK-1/2 in mHypoA-2/10 Cells.
Having established endogenous 5-HT2CR expression in mHypoA-2/10 cells, we next sought to define the signaling components leading to ERK-1/2 and CREB phosphorylation. Because Gq-coupled receptors have frequently been shown to phosphorylate ERK-1/2 via protein kinase C (PKC), we used the PKC inhibitor BIM-X (Fig. 6A). BIM-X blunted 5-HT-induced ERK-1/2 phosphorylation but did not affect phosphorylation by MSH. Thus, we provide evidence that 5-HT2CR expressed in hypothalamic cells activate the ERK-1/2 pathway via PKC. Finally, we tested a putative link between PKC, ERK-1/2, and CREB by analyzing effects of BIM-X and the ERK-1/2 inhibitor PD-184,352 on 5-HT-induced phosphorylation of CREB. Both inhibitors completely diminished 5-HT-induced CREB phosphorylation, indicating that endogenous 5-HT2CR activate CREB/CRE via PKC-induced ERK-1/2 activation in a PVN-derived cell model (Fig. 6, B and C). To support this model we also tested effects of BIM-X, PD-184,352, and the phospholipase C (PLC) inhibitor U-73122 in the reporter assay. All inhibitors blunted 5-HT-induced CRE reporter activation, providing the first evidence that 5-HT2CR mediate CREB/CRE signaling via PLC, PKC, and ERK-1/2 in hypothalamic cells.
5-HT2CR phosphorylate CREB via PKC and ERK-1/2 in mHypoA-2/10 cells. (A) ERK-1/2 phosphorylation was measured in serum-starved mHypoA-2/10 cells pretreated with DMSO (0.2%) or BIM-X (5 μM) and stimulated for the indicated period of time with 5-HT (1 μM) or MSH (1 μM). Cell lysates were subjected to Western analysis using either a phospho-specific antiserum against ERK-1/2 (p-ERK-1/2) or against the total histone protein to control for the total protein amount. One representative blot is shown. 5-HT data of six lysates were quantified by densitometry, ratios between p-ERK-1/2 and histone signals calculated, 5-HT-induced ERK-1/2 phosphorylation normalized to not-stimulated cells and expressed as the mean ± S.E.M. Asterisks indicate a significant difference (two-sample t test) between DMSO and BIM-X pretreated cells. Ligand-induced CREB phosphorylation was measured in serum-starved mHypoA-2/10 cells pretreated with DMSO (0.2%) or BIM-X (5 μM) in (B) or with DMSO (0.1%) or PD-184,352 (10 μM) in (C) and stimulated for the indicated period of time with 5-HT (1 μM) or MSH (1 μM) in (A) or only with 5-HT (1 μM) in (B). Cell lysates were subjected to Western analysis using either a phospho-specific antiserum against CREB (p-CREB) or against the total histone protein to control for the total protein amount. One representative blot is shown. Data of six independent experiments were quantified by densitometry, ratios between p-CREB and histone signals calculated, 5-HT-induced CREB phosphorylation normalized to not-stimulated cells and expressed as the mean ± S.E.M. Asterisks indicate a significant difference (two-sample t test) between DMSO- or inhibitor-pretreated cells. (D) 5-HT-induced reporter activation was measured in serum-starved (24 hours) mHypoA-2/10-CRE cells stably expressing the CRE-dependent reporter. DMSO (0.2%), BIM-X (5 μM), PD-184,352 (10 μM), U-73122 (10 μM), or its inactive analog U-73433 (10 μM) pretreated cells were stimulated or not with 1 μM 5-HT. No significant differences between DMSO and inhibitor-treated cells were observed. Data from four to six experiments performed in quadruplicate were compiled by setting basal to 1.0 and expressed as the mean ± S.E.M. Two-way ANOVA test results are F(4/27) = 7.535, P < 0.01. Dagger signs indicate significant differences (Bonferroni post-test) to basal (DMSO) and hash signs (one-sample t test) to 1.0.
Discussion
In this work we took advantage of the recently introduced mHypoA-2/10 cell line that has been shown to be sensitive toward MSH, NPY, and T3 hormone and thus resembles 5-HT1BR- and 5-HT2CR-expressing neurons of the hypothalamic PVN (Belsham et al., 2009; Dalvi et al., 2012; Nazarians-Armavil et al., 2013; Breit et al., 2015). We provide compelling pharmacological evidence on the basis of selective 5-HTR antagonists and agonists that mHypoA-2/10 cells express functional 5-HT1BR and, of particular interest, 5-HT2CR. Thus mHypoA-2/10 cells represent the first CNS-derived cell model known so far that allows investigation of 5-HT2CR biology in its natural environment. Because of the outstanding role of 5-HT2CR in many CNS disorders or pathophysiological states, mHypoA-2/10 cells represent a novel experimental tool for several 5-HT2CR-based research areas, such as emotion, cognition, reward, and feeding.
It has been established that 5-HT exerts central anorexigenic effects by excitation of MSH-producing and inhibition of NPY-releasing neurons of the ARC (Currie et al., 1999; Currie et al., 2002; Sohn et al., 2011; Doslikova et al., 2013; Burke et al., 2014). Furthermore, activation of the 5-HT2CR in the PVN has been shown to decrease food-intake and to enhance the hypothalamic-pituitary-adrenal axis by releasing corticotrophin-releasing hormone and adrenocorticotropic hormone (Heisler et al., 2007). However, the underlying signaling events have remained elusive. Analyzing mHypoA-2/10 cells, we first obtained data about gene expression–regulating signaling pathways induced by 5-HT2CR in a PVN-like cell line. We showed that 5-HT2CR activated CRE-dependent gene expression via PKC-induced and ERK-1/2-mediated phosphorylation of CREB. These findings are of particular interest because: 1) they provide the first piece of evidence for hypothalamic CRE activation by a Gq-coupled receptor in general and by the 5-HT2CR in particular; 2) the corticotrophin-releasing hormone -promoter contains a binding site for CREB, suggesting that the PKC/ERK-1/2/CREB pathway might be responsible for the effects of the 5-HT2CR on the hypothalamic-pituitary-adrenal axis; and 3) on the basis of data obtained in overexpression systems, it has been posited that 5-HT2CR-induced ERK-1/2 activation is independent of PKC, thus highlighting the fact that endogenous 5-HT2CR engage distinct signaling pathways compared with recombinant 5-HT2CR (Legradi et al., 1997; Labasque et al., 2008).
The transcription factor CREB has traditionally been assigned to the PKA pathway and so far CREB/CRE activation in hypothalamic cells has been attributed to α-MSH-induced Gs activation (Shaywitz and Greenberg, 1999; Harris et al., 2001). In nonhypothalamic cells Gq-controlled kinases such as calcium/calmodulin-dependent kinase II (CamKII), PKC, and ERK-1/2 have also been proposed to phosphorylate CREB, but the subsequent effects on CRE activation are not clear. Gq activation via the gastrin-releasing peptide leads to CREB phosphorylation via PKC/ERK-1/2 in overexpression systems without subsequent CRE activation (Fitzgerald et al., 1999). Carbachol-induced activation of the muscarinic 3 receptor has been shown to phosphorylate CREB via CamKII and ERK-1/2, but only CamKII-induced CREB phosphorylation led to CRE activation (Chan et al., 2005; Rosethorne et al., 2008). Likewise, PKC-mediated ERK-1/2 activation has been shown to phosphorylate CREB via bradykinin-B2 receptors, calcium-sensing receptors, and orexin-2 receptors, but no CRE activation was found (Rosethorne et al., 2008; Guo and Feng, 2012; Avlani et al., 2013). Furthermore, CREB phosphorylation by the gonadotropin-releasing hormone required epidermal growth-factor receptor transactivation in hypothalamic cells (Neithardt et al., 2006). Interestingly, as with our results showing that 5-HT2CR-induced signaling differs between recombinant and hypothalamic cell models, orexin-2 receptor–induced CREB phosphorylation varied in Chinese hamster ovary and hypothalamic cells (Guo and Feng, 2012). It appears that Gq-induced CREB/CRE activation is regulated by multilayer signaling networks strongly dependent on the stimulus and the cellular background. Thus, it will be important to analyze whether PKC/ERK-1/2-mediated CREB phosphorylation and subsequent CRE activation in hypothalamic cells is restricted to the 5-HT2CR or also applies to other Gq-activation stimuli.
Because of the Gq coupling of the 5-HT2CR, one might also expect 5-HT-induced NFAT activation. Activation of PLC via Gq leads to a transient release of calcium from intracellular stores, which is not necessarily sufficient to activate NFAT, because calcium influx through calcium release–activated calcium channels is required to activate NFAT (Serafini et al., 1995; Crabtree and Olson, 2002). Enhanced calcium signaling by this “inside-out” pathway might be particularly important for Gq-induced NFAT activation in endogenous expression systems, when the initial calcium signaling from intracellular stores is rather low.
Interestingly, compared with WAY-161,503, 5-HT signaling on the level of CRE, ERK-1/2, and calcium was increased, suggesting that WAY-161,503 might be a partial 5-HT2CR agonist or that 5-HT interacts with other 5-HTR besides the HT2CR to activate CRE, ERK-1/2, or calcium signaling in mHypoA-2/10 cells. We can exclude a role for the 5-HT1BR in 5-HT-induced CRE, ERK-1/2, or calcium signaling because the 5-HT1BR agonist CP-94,253 did not activate either pathway (Fig. 3A, data not shown). We can further exclude the 5-HT2AR or 5-HT2BR subtypes, because antagonists of either receptor did not affect 5-HT-induced CREB and ERK-1/2 phosphorylation or CRE activation. In clear contrast, both 5-HT2CR antagonists blunted 5-HT signaling, indicating that 5-HT exclusively employs the 5-HT2CR to activate CREB/CRE and ERK-1/2, thus showing that WAY-161,503 and 5-HT engage the same receptor to activate CRE in mHypoA-2/10 cells, albeit with distinct intrinsic activities (Fig. 7). Previous reports using overexpression systems reported that WAY-161,503 reached maximal receptor activation in Chinese hamster ovary cells and ∼80% activation levels in human embryonic kidney-293 cells, indicating that the efficacy of WAY-161,503 is generally overestimated in overexpression systems, probably owing to spare receptors (Schlag et al., 2004; Rosenzweig-Lipson et al., 2006). Partial 5-HT2CR activation by WAY-161,503 in endogenous cells suggests that the anorexigenic effects of WAY-161,503 observed in animal studies might not reflect the full therapeutic potential of 5-HT2CR activation on feeding. It is noteworthy in this context that the 5-HT2CR-selective agonist lorcaserin (Belvique) decreases body weight in obese individuals and was touted as one of the most promising drugs available to treat obesity (Thomsen et al., 2008). However, the clinical use of lorcaserin has been limited because of severe cardiac side effects, most probably because of its binding to the 5-HT2AR subtype expressed in cardiomyocytes (Miller and Wacker, 2010; Miller, 2013; Comerma-Steffensen et al., 2014). Thus, lorcaserin-based therapy could be improved either by increasing the selectivity toward the 5-HT2CR subtype or by identifying the exact signaling events induced by lorcaserin via 5-HT2CR in appetite-regulating neurons. mHypoA-2/10 cells might be quite useful in this regard: first, to determine the intrinsic activity of lorcaserin to activate endogenous 5-HT2CR, and second, to identify pertinent signaling pathways in a hypothalamic cell model.
WAY-163,503 is a partial agonist of endogenous 5-HT2CR in mHypoA-2/10 cells. Data from Figs. 3C, 4B, and 5 were analyzed by setting the average signal induced by 5-HT (1 μM) in the absence of any antagonist to 1.0. Intrinsic activity of WAY-163,503 (500 nM), or of the SB-242,084-sensitive portion of the 5-HT signal, were calculated for each experiment. Asterisks indicate a significant difference between WAY-163,503 and 5-HT.
Herein, we not only provide data about 5-HT2CR expression in mHypoA-2/10 cells, we also report that these cells endogenously express the 5-HT1BR subtype. 5-HT1BR are expressed in the hypothalamus and are involved in 5-HT-induced regulation of feeding. Furthermore, in vivo experiments strongly suggested that simultaneous 5-HT1BR and 5-HT2CR activation depressed food intake in a synergistic fashion. However, it still remains unclear whether both receptor subtypes co-operate in the same cell and, if so, which signaling pathways might be involved (Doslikova et al., 2013). Thus, mHypoA-2/10 cells may be a useful cell model to decipher complex functional interactions between 5-HT1BR and 5-HT2CR in a native environment.
In conclusion, we report endogenous 5-HT2CR receptor expression in hypothalamic mHypoA-2/10 cells. Analysis of 5-HT2CR signaling in an endogenous cell system rapidly revealed so-far unknown features of the hypothalamic 5-HT system, such as partial agonism of a selective 5-HT2CR agonist and PKC/ERK-1/2-promoted CREB/CRE activation, implying that detailed dissection of the 5-HT2CR biology in mHypoA-2/10 cells could significantly increase our knowledge about this intriguing receptor subtype.
Authorship Contributions
Participated in research design: Lauffer, Gudermann, Breit.
Conducted experiments: Lauffer, Glas, Breit.
Contributed new reagents or analytic tools: Breit.
Performed data analysis: Lauffer, Breit.
Wrote or contributed to the writing of the manuscript: Gudermann, Breit.
Footnotes
- Received January 22, 2016.
- Accepted April 25, 2016.
Abbreviations
- ANOVA
- analysis of variance
- ARC
- nucleus arcuatus
- BIM-X
- bisindolylmaleimide X
- BSA
- bovine serum albumin
- CNS
- central nervous system
- CP-94,253
- 5-propoxy-3-(1,2,3,6-tetrahydro-4-pyridinyl)-1H-pyrrolo[3,2-b]pyridine
- CRE
- cAMP response element
- CREB
- CRE binding protein
- Df
- degree of freedom
- DMEM
- Dulbecco’s modified Eagle’s medium
- DMSO
- dimethyl sulfoxide
- ELISA
- enzyme-linked immunosorbent assay
- EMD-281,014
- 7-[[4-[2-(4-fluorophenyl)ethyl]-1-piperazinyl]carbonyl]-1H-indole-3-carbonitrile hydrochloride
- ERK-1/2
- extracellular-regulated kinases-1/2
- 5-HT
- 5-Hydroxytryptamine
- 5-HTR
- 5-HT receptor(s)
- IBMX
- 3-isobutyl-1-methylxanthine
- MSH
- melanocyte-stimulating hormone
- NFAT
- nuclear factor of activated T cells
- NPY
- neuropeptide Y
- PBS
- phosphate-buffered saline
- PD-184,352
- 2-(2-chloro-4-iodophenylamino)-N-cyclopropylmethoxy-3,4-difluorobenzamide
- PKC
- protein kinase C
- PVN
- paraventricular nucleus
- RS-102,221
- 8-[5-(2,4-dimethoxy-5-(4-trifluoromethylphenylsulphonamido)phenyl-5-oxopentyl]-1,3,8-triazaspiro[4.5]decane-2,4-dione
- RS-127,455
- 1′-[2-[4-(trifluoromethyl)phenyl]ethyl]-spiro[4H-3,1-benzoxazine-4,4′-peperidin]-2(1H)-one
- RT
- room temperature
- SB-242,084
- 6-chloro-2,3-dihydro-5-methyl-N-[6-[(2-methyl-3-pyridinyl)oxy]-3-pyridinyl]-1H-indole-1-carboxyamide
- STAT-3
- signal transducer and activator of transcription-3
- WAY-161,503
- 8,9-dichloro-2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]quinoxalin-5(6H)-one
- Copyright © 2016 by The American Society for Pharmacology and Experimental Therapeutics
References
In this issue
5-HT2CR Expression in mHypoA-2/10 Cells
