Enhanced activation of Akt and extracellular-regulated kinase pathways by simultaneous occupancy of Gq-coupled 5-HT2A receptors and Gs-coupled 5-HT7A receptors in PC12 cells

Authors

  • Nadine N. Johnson-Farley,

    1. Department of Psychiatry, University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School, Piscataway, New Jersey, USA
    Search for more papers by this author
  • Sylvia B. Kertesy,

    1. Department of Physiology and Biophysics, Case School of Medicine, Case Western Reserve University, Cleveland, Ohio, USA
    Search for more papers by this author
  • George R. Dubyak,

    1. Department of Physiology and Biophysics, Case School of Medicine, Case Western Reserve University, Cleveland, Ohio, USA
    Search for more papers by this author
  • Daniel S. Cowen

    1. Department of Psychiatry, University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School, Piscataway, New Jersey, USA
    Search for more papers by this author

Address correspondence and reprints requests to Daniel S. Cowen, Department of Psychiatry, UMDNJ-Robert Wood Johnson Medical School, 125 Paterson Street, Suite 2200, New Brunswick, NJ 08901, USA. E-mail: cowends@umdnj.edu

Abstract

The most commonly prescribed antidepressants, the serotonin (5-HT) selective reuptake inhibitors, increase 5-HT without targeting specific receptors. Yet, little is known about the interaction of multiple receptor subtypes expressed by individual neurons. Specifically, the effect of increases in cAMP induced by Gs-coupled 5-HT receptor subtypes on the signaling pathways modulated by other receptor subtypes has not been studied. We have, therefore, examined the activation of the extracellular-regulated kinase (ERK) and Akt pathways by Gs-coupled 5-HT7A receptors and Gq-coupled 5-HT2A receptors, which are co-expressed in discrete brain regions. Agonists for both receptors were found to activate ERK and Akt in transfected PC12 cells. 5-HT2A receptor-mediated activation of the two pathways was found to be Ca2+-dependent. In contrast, 5-HT7A receptor-mediated activation of Akt required increases in both [cAMP] and intracellular [Ca2+], while activation of ERK was inhibited by Ca2+. The activation of ERK and Akt stimulated by simultaneous treatment of cells with 5-HT2A and 5-HT7A receptor agonists was found to be at least additive. Cell-permeable cAMP analogs mimicked 5-HT7A receptor agonists in enhancing 5-HT2A receptor-mediated activation of ERK and Akt. A role was identified for the cAMP–guanine exchange factor, Epac, in this augmentation of ERK, but not Akt, activation. Our finding of enhanced activation of neuroprotective Akt and ERK pathways by simultaneous occupancy of 5-HT2A and 5-HT7A receptors may also be relevant to the interaction of other neuronally expressed Gq- and Gs-coupled receptors.

Abbreviations used
8-CPT-Me-cAMP

8-(4-chlorophenylthio)-2′-O-methyladenosine 3′,5′-cyclic monophosphate

5-CT

5-carboxamidotryptamine maleate

DOI

R(–)-2,5-dimethoxy-4-iodoamphetamine hydrochloride

ERK

extracellular signal-regulated kinase

GEF

guanine nucleotide exchange factor

H-89

N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide

5-HT

serotonin

pCPT-cAMP

8-(4-chlorophenylthio)-adenosine 3′,5′-cyclic monophosphate

PDK

3-phosphoinositide-dependent protein kinase

PKA

protein kinase A

The selective serotonin reuptake inhibitors represent the most commonly prescribed class of antidepressants. These drugs, as their name implies, selectively increase serotonin (5-HT), for which there are at least 14 receptors. These receptors have been classified into seven families in which: 5-HT1 receptors and 5-HT5 receptors couple to inhibition of adenylate cyclase; 5-HT2 receptors couple to Gq-type G proteins, activating phosphoinositide hydrolysis; and 5-HT4, 5-HT6 and 5-HT7 receptors couple to Gs, stimulating adenylate cyclase. In contrast, 5-HT3 receptors are not G-protein-coupled receptors, but are ligand-gated ion channels. Although it is known that increases in [5-HT] can effectively induce remission of depression, it is not known which of the 14 or more receptors for 5-HT mediate this action. In fact, selective serotonin reuptake inhibitor-induced increases in 5-HT act at all 5-HT receptors exposed to the higher [5-HT].

The identities of the cellular pathways utilized by 5-HT receptors in the treatment of depression also remain unknown. However, it is hypothesized that the pathways may be similar to those mediating antidepressant-induced neuroprotective changes in the hippocampus and other brain regions. Extracellular-regulated kinase (ERK) MAPK and Akt (protein kinase B) are thought to be relevant, as they have been found to confer neuroprotection in several models of apoptosis (Tamatani et al. 1998; Hetman et al. 1999; Matsuzaki et al. 1999; Yamaguchi et al. 2001). ERK1 and ERK2 are serine/threonine protein kinases that phosphorylate a number of transcription factors (see Denton and Tavare 1995 for review). Activation requires phosphorylation at Thr202/Tyr204 by MEK (MAPK kinase). Akt is also a serine/threonine protein kinase that is known to stimulate neuroprotective changes. It is activated by phosphorylation at Thr308 by 3-phosphoinositide-dependent protein kinase (PDK-1) and at Ser473 by an unidentified PDK-2 (Alessi et al. 1996, 1997).

Given the large number of 5-HT receptor subtypes, it is perhaps not surprising that the expression of individual receptors exhibits overlap. However, little is currently known about the cellular effects of multiple subtypes of 5-HT receptors interacting in individual neurons. The pattern of expression of 5-HT2 and 5-HT7 receptors supports the possibility that those two receptors may interact. Both are expressed in the cortex, hypothalamus, thalamus, and the CA3 pyramidal region of the hippocampus (Pompeiano et al. 1994; Gustafson et al. 1996; Heidmann et al. 1998). Gq-coupled 5-HT2 receptors have been found to couple to activation of ERK in vascular and tracheal smooth muscle cells (Hershenson et al. 1995; Watts et al. 1996), mesangial cells (Greene et al. 2000) and PC12 cells (Quinn et al. 2002). In contrast, agonists for Gs-coupled receptors cause inhibition of ERK activation in most cell types. Activation is seen only in specific types of cells, such as neurons, Chinese hamster ovary cells, and HEK293 cells. In fact, Gs-coupled 5-HT7 receptors have been shown to activate the MAPK in HEK293 cells, PC12 cells, and cultured hippocampal neurons (Errico et al. 2001; Lin et al. 2003; Norum et al. 2003).

In this study we used a PC12 cell line expressing endogenous 5-HT2A receptors and transfected 5-HT7A receptors to directly compare the coupling of the two receptor subtypes to ERK and Akt. The effect of agonists for Gs-coupled 5-HT receptor subtypes on the activation of cellular signals stimulated by agonists for receptor subtypes coupling to other G proteins has not been studied. We, therefore, also studied the interaction of 5-HT2A receptor-mediated increases in intracellular [Ca2+] and 5-HT7A receptor-mediated increases in [cAMP] in the activation of ERK and Akt. Our identification of a positive interaction between 5-HT2A and 5-HT7A receptors suggests that other Gq- and Gs-coupled receptors may similarly positively interact in neurons and neuronal cell types.

Experimental procedures

Materials

R-(+)-Alpha-(2,3-dimethoxyphenyl)-1-[2-(4-fluorophenyl)ethyl]-4-piperidinemethanol (MDL100907) was kindly provided by Aventis Pharmaceuticals (Bridgewater, NJ, USA). Thapsigargin was obtained from Calbiochem and from Alomone Laboratories (Jerusalem, Israel). 8-Bromoadenosine 3′,5′-cyclic monophosphate (8-Br-cAMP), 5-HT, N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinolinesulfonamide (H-89), 8-(4-chlorophenylthio)-adenosine 3′,5′-cyclic monophosphate (pCPT-cAMP), [R]-3-[2-(2-[4-methyl-piperidin-1-yl]ethyl)pyrrolidine-1-sulfonyl]phenol (SB269970), and 5-carboxamidotryptamine maleate (5-CT) were obtained from Sigma (St. Louis, MO, USA). 8-(4-Chlorophenylthio)-2′-O-methyladenosine 3′,5′-cyclic monophosphate (8-CPT-Me-cAMP) was obtained from Axxora LLC (San Diego, CA, USA).

PC12 cell culture

PC12 cells were obtained from American Type Culture Collection (Rockville, MD, USA), and were routinely cultured in Dulbecco's modified Eagle's medium supplemented with l-glutamine, minimal essential medium non-essential amino acids, 15% dialyzed fetal bovine serum (dialyzed in membranes with 1000 Da cut-offs against a 100-fold greater volume of 150 mm NaCl to remove endogenous 5-HT), and 100 U penicillin−100 µg streptomycin/mL at 37°C (95% air, 5% CO2). A stable, tightly adherent cell population was obtained after several cycles of washing off loosely adherent cells (Quinn et al. 2002). The gene for the human 5-HT7A receptor was isolated by PCR from Quick Clone human brain cDNA (Clonetech, Palo Alto, CA, USA). The gene was then cloned into a murine leukemia retroviral packaging vector and transduced into PC12 cells. Cells stably expressing receptors were selected in the presence of 500 µg/mL geneticin. Cells represent mixed cultures, as individual clones were not selected. This avoided the potential problem of selecting clones not representative of PC12 cells.

Transient transfections of cells

A cDNA encoding HA epitope-tagged Epac1 was kindly provided by Dr Johannes L. Bos. Transient transfections of PC12 cells with 9 µg plasmid cDNA per 60 mm cell culture dish were performed 48 h prior to cellular studies with Lipofectamine 2000 according to the manufacturer's suggestions (Gibco BRL, Rockville, MD, USA).

Immunoblots

Monoclonal anti-phospho-ERK1/ERK2 (Thr202/Tyr204), rabbit polyclonal anti-phospho-Akt (Ser473), and rabbit polyclonal anti-total Akt were obtained from Cell Signaling (Beverly, MA, USA). Rabbit polyclonal anti-total ERK1/ERK2 and horseradish peroxidase-conjugated secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Cells were washed the day prior to use with phosphate-buffered saline and cultured overnight under low-serum (0.5%) conditions. Reagents were then added directly to the culture media. Treated cells were washed with phosphate-buffered saline, and routinely lysed with a 26-gauge needle in 25 mm HEPES (pH 7.4), 150 mm NaCl, 1% Triton X-100, 1 mmβ-glycerol phosphate, 50 mm NaF, 5 mm EDTA, 1 mm sodium orthovanadate, 250 µm 4-(2-aminoethyl)-benzene-sulfonylfluoride hydrochloride, 0.1% aprotinin, and 10 µg/mL leupeptin. After 20 min on ice, the lysate was centrifuged at 10 000 g for 10 min at 4°C. Supernatant proteins were separated on 10% resolving gels (Bio-Rad Laboratories, Hercules, CA, USA) and transferred to 0.45 µm Immobolin-P polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA, USA). Bound antibodies were visualized using Enhanced Luminol Chemiluminescence Reagent (Perkin–Elmer Life Sciences, Boston, MA, USA) and direct exposure to a Kodak Image Station 440CF with a cooled, full-frame-capture CCD camera (Kodak). Net intensity of bands was calculated directly from stored images using Kodak Digital Science 1D Image Analysis Software (version 3.5) on defined regions of interest.

Fluorimetric analysis of cytosolic Ca2+

One day before fluorimetric analysis of agonist-induced changes in cytosolic Ca2+ concentration, 100 mm dishes with cells at ˜ 80% confluence were washed with phosphate-buffered saline and cultured overnight under low-serum conditions in Dulbecco's modified Eagle's medium supplemented with antibiotics and 0.1% bovine serum albumin. Cells were removed from the culture dish by brief trypsinization, washed, and resuspended to ˜ 106 cells/mL in a balanced saline solution (140 mm NaCl, 5 mm KCl, 1.5 mm CaCl2, 1 mm MgCl2, 10 mm d-glucose, 1 mg/mL BSA, and 20 mm NaHEPES pH 7.5). The cell suspension was supplemented with 1 µm fura-2AM (Molecular Probes) and incubated at room temperature for 45 min. The cells were pelleted, washed once, and resuspended in fresh balanced saline solution to a density of 106 cells/mL. Cytosolic Ca2+ levels in 1.5 mL stirred aliquots of cell suspension were assayed fluorimetrically at 37°C using equipment and calibration protocols that have been described previously (Cowen et al. 1989). The cells were treated with various 5-HT receptor agonists or antagonists (as indicated in specific figure legends) added as 0.5–5 µL aliquots from concentrated stocks. The cells were then permeabilized with 50 µg/mL digitonin to facilitate calibration of fura-2AM fluorescence as a function of extracellular Ca2+ levels.

Statistics

Results are expressed as the means ± SEM of three or more experiments, performed in duplicate. Experimental groups were compared by anova followed by post-hoc Bonferroni tests.

Results

5-HT2A receptors couple to activation of ERK and Akt in PC12 cells via Ca2+-dependent pathways

We have previously reported that 5-HT stimulates a large activation of ERK in non-transfected PC12 cells through endogenous Gq-coupled 5-HT2A receptors (Quinn et al. 2002). We now find that 5-HT also stimulates activation (phosphorylation) of the neuroprotective Akt pathway in non-transfected PC12 cells (Fig. 1b). Activation of both ERK and Akt was found to be inhibited by 10 nm concentrations of MDL100907, a highly selective 5-HT2A receptor antagonist (Fig. 1). Activation of both pathways was also found to be inhibited when receptor-mediated increases in intracellular Ca2+ were attenuated. Activation of ERK was found to be inhibited by 48% when the extracellular [Ca2+] was reduced from 2 mm to 100 nm (a concentration, as shown in Figs 5 and 6, similar to the intracellular [Ca2+] found in resting PC12 cells) by pre-treatment with EGTA (Fig. 2a). Pre-treatment of cells over 30 min with a low (30 nm) concentration of the endoplasmic reticular Ca2+ ATPase inhibitor, thapsigargin, to slowly deplete intracellular stores of Ca2+ prior to treatment with 5-HT, caused a 71% reduction in activated ERK. Combined treatment with EGTA and thapsigargin caused complete inhibition, demonstrating a requirement for increases in intracellular [Ca2+] originating from both influx of extracellular Ca2+ and release from intracellular stores. Activation of Akt was also found to be sensitive to Ca2+ (Fig. 2b). In fact, the pathway for activation of Akt was more sensitive to changes in Ca2+ than was the pathway for activation of ERK. Reduction of either extracellular [Ca2+] or intracellular stores caused complete inhibition of receptor-mediated activation of Akt.

Figure 1.

Serotonin (5-HT) stimulates activation of ERK and Akt in non-transfected PC12 cells through endogenous 5-HT2A receptors. Non-transfected PC12 cells were treated for 5 min with 1 µm 5-HT in the absence or presence of 10 nm MDL100907 (MDL), or were treated with 300 nm 5-CT, and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a) phospho-ERK1/ERK2 (p-ERK) and (b) phospho-Akt (p-Akt). Membranes were then stripped and analyzed with antibody to (a) total ERK1/ERK2 (Total ERK) and (b) total Akt (Total Akt). ***p < 0.001 versus 5-HT. Representative immunoblots from three separate experiments performed in duplicate are shown.

Figure 5.

300 nm 5-CT elicits 5-HT7A receptor-dependent increases in intracellular [Ca2+]. PC12 cells stably expressing human 5-HT7A receptors were trypsinized and loaded with fura-2AM. Aliquots (1.5 mL) of fura-2AM-loaded cells (106/mL) were assayed for Ca2+-dependent changes in fluorescence in a stirred, thermostatted (37°C) cuvette in response to 10 µm 5-HT, 3 µm 5-CT, or 300 nm 5-CT. Where indicated, the cells were pre-treated for 2 min with 3 µm SB269970 before addition of the 5-CT agonist. At the conclusion of each assay, the cells were permeabilized by addition of 50 µg/mL digitonin (Dig) to permit calibration of the released fura-2 fluorescence as a function of [Ca2+]. The illustrated traces are representative of results from three separate experiments.

Figure 6.

SB269970 selectively antagonizes the increase in intracellular [Ca2+] mediated by 5-HT7A receptors, but not that mediated by 5-HT2A receptors. PC12 cells stably expressing human 5-HT7A receptors were trypsinized and loaded with fura-2AM. Aliquots (1.5 mL) of fura-2AM-loaded cells (106/mL) were assayed for Ca2+-dependent changes in fluorescence in a stirred, thermostatted (37°C) cuvette in response to 1 µm DOI or 600 nm 5-CT. Where indicated, cells were pre-treated for 2 min with either 3 µm SB269970 or 10 nm MDL100907. At the conclusion of each assay, the cells were permeabilized by addition of 50 µg/mL digitonin (Dig) to permit calibration of the released fura-2 fluorescence as a function of [Ca2+]. The illustrated traces are representative of results from three separate experiments.

Figure 2.

5-HT2A receptor-mediated activation of ERK and Akt requires Ca2+. Non-transfected PC12 cells were pre-treated with EGTA (to reduce extracellular [Ca2+] from 2 mm to 100 nm) and/or with 30 nm thapsigargin for 30 min prior to treatment with 1 µm 5-HT for 5 min, and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a) phospho-ERK1/ERK2 (p-ERK) and (b) phospho-Akt (p-Akt). Membranes were then stripped and analyzed with antibody to (a) total ERK1/ERK2 (Total ERK) and (b) total Akt (Total Akt). ***p < 0.001 versus 5-HT. Representative immunoblots from three separate experiments performed in duplicate are shown.

5-HT7 receptors couple to activation of both ERK and Akt

We have previously reported that agonists for the Gs-coupled 5-HT7A receptor stimulate activation of ERK in cultured hippocampal neurons and transfected PC12 cells (Errico et al. 2001; Lin et al. 2003). We now find that 5-HT7A receptors similarly couple to activation of Akt. PC12 cells stably expressing the receptor exhibited > 2-fold increases in the level of activated Akt when treated with the 5-HT1/5-HT7 receptor agonist 5-CT (Fig. 3b). Pre-treatment with the selective 5-HT7 receptor antagonist SB269970 completely inhibited activation of both ERK and Akt. Also, no activation of ERK or Akt was seen when non-transfected cells were treated with the same (300 nm) concentration of 5-CT (Fig. 1). Higher concentrations of 5-CT were not used in our studies so as to maintain 5-HT7 receptor-selectivity. Significantly, no inhibition of 5-HT7 receptor agonist-stimulated activation of ERK and Akt was seen when transfected cells were treated with 3 µm MDL100907 (Fig. 3), a concentration 300-fold higher than that found to inhibit the activation mediated by 5-HT2A receptors (Fig. 1). Neither MDL100907 nor SB269970 induced any change in basal ERK or Akt levels (not shown).

Figure 3.

5-CT stimulates activation of ERK and Akt in PC12 cells stably expressing 5-HT7A receptors. PC12 cells stably expressing human 5-HT7A receptors were treated for 5 min with 300 nm 5-CT in the absence or presence of 3 µm SB269970 (SB) or 3 µm MDL100907 (MDL) and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a) phospho-ERK1/ERK2 (p-ERK) and (b) phospho-Akt (p-Akt). Membranes were then stripped and analyzed with antibody to (a) total ERK1/ERK2 (Total ERK) and (b) total Akt (Total Akt). ***p < 0.001 versus 5-CT. Representative immunoblots from three separate experiments performed in duplicate are shown.

Complete activation of Akt by 5-HT7 receptor agonists requires both Ca2+ and cAMP

We have previously found that 5-HT7 receptors couple to activation of adenylate cyclase in transfected PC12 cells, such that 5-CT induces an 80-fold increase in intracellular [cAMP] (Lin et al. 2003). We therefore sought to determine the role of cAMP in activating Akt. We found that the cell-permeable cAMP analogs, 8-bromoadenosine 3′,5′-cyclic monophosphate (8-Br-cAMP) and 8-(4-chlorophenylthio)-adenosine 3′,5′-cyclic monophosphate (pCPT-cAMP) both stimulate activation of Akt (Fig. 4b). However, the increase in phosphorylated Akt was only ˜ 40% of that stimulated by 5-CT. Both cAMP analogs were similar to 5-CT in also stimulating activation of ERK. However, the activation stimulated by pCPT-cAMP was larger than that stimulated by 8-Br-cAMP (Fig. 4a). The smaller, 2-fold, activation stimulated by 8-Br-cAMP was consistently observed across multiple experiments. However, it did not reach statistical significance when anova was performed on data in which the larger 6- to 7-fold activations by pCPT-cAMP and 5-CT were included. In contrast, statistical significance (p < 0.05) was achieved when pERK levels for 8-Br-cAMP-treated and vehicle-treated cells were compared directly in two-tailed paired t-tests.

Figure 4.

cAMP analogs stimulate a smaller magnitude of activation of Akt than does 5-CT. PC12 cells stably expressing human 5-HT7A receptors were treated for 5 min with 300 nm 5-CT, 200 µm pCPT-cAMP, or 200 µm 8-Br-cAMP, and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a) phospho-ERK1/ERK2 (p-ERK) and (b) phospho-Akt (p-Akt). Membranes were then stripped and analyzed with antibody to (a) total ERK1/ERK2 (Total ERK) and (b) total Akt (Total Akt). *p < 0.05; **p < 0.01; ***p < 0.001 versus control. Representative immunoblots from four separate experiments performed in duplicate are shown.

Agonists for 5-HT7 receptors have been reported to stimulate increases in intracellular [Ca2+] in glomerulosa cells and transfected HEK293 cells (Baker et al. 1998; Lenglet et al. 2002). We similarly found agonist-induced increases in intracellular [Ca2+] in PC12 cells. Treatment of cells with 300 nm 5-CT caused intracellular [Ca2+] to increase from the resting level of 165 nm to a peak of 220 nm, followed by a prolonged plateau in the 200 nm range (Fig. 5). These changes in intracellular [Ca2+] were completely inhibited in the presence of 3 µm SB269970, confirming that 300 nm 5-CT selectively activates 5-HT7A receptors in the absence of activating 5-HT2A receptors (Table 1). The Ca2+ signals induced by 300 nm 5-CT were small compared with the much larger increases (520 nm peak followed a 250 nm plateau) in intracellular [Ca2+] triggered by 10 µm 5-HT, which activates both endogenous 5-HT2A receptors and transfected 5-HT7A receptors. However, at 5-CT concentrations > 300 nm, larger increases in [Ca2+] could be also stimulated, reflecting a non-selective activation of 5-HT2A receptors on top of the expected 5-HT7A receptor activation. For example, a 2-fold peak increase in [Ca2+] (from 160 to 320 nm) was triggered by 3 µm 5-CT, and only partially antagonized by SB269970.

Table 1.  5-CT and DOI can be used as selective 5-HT7 and 5-HT2A agonists, respectively, in Ca2+ studies of cells expressing both receptors
TreatmentMean Δ[Ca2+] ± SEMp
  1. PC12 cells stably expressing human 5-HT7A receptors were trypsinized and loaded with fura-2AM. Aliquots (1.5 mL) of fura-2AM-loaded cells (106/mL) were assayed for Ca2+-dependent changes in fluorescence in a stirred, thermostatted (37°C) cuvette in response to 1 µm DOI or 300 nm 5-CT. Where indicated, the cells were pre-treated for 2 min with either 3 µm SB269970 or 10 nm MDL100907. Results are expressed as the mean Δ[Ca2+] ± SEM of three separate experiments where Δ[Ca2+] = peak [Ca2+] subsequent to treatment – basal [Ca2+] prior to treatment. p-value versus 5-CT or DOI in the absence of inhibitor; n.s., not significant (p > 0.05).

300 nm 5-CT59 ± 16
300 nm 5-CT + 3 µm SB26697013 ± 7p < 0.05
300 nm 5-CT + 10 nm MDL10090765 ± 21n.s.
1 µm DOI113 ± 19
1 µm DOI + 3 µm SB266970121 ± 20n.s.
1 µm DOI + 10 nm MDL1009072.5 ± 2.5< 0.01

5-HT2A receptor-mediated pathways can be studied in the 5-HT7A receptor-transfected PC12 cell line by treating cells with R(–)-2,5-dimethoxy-4-iodoamphetamine hydrochloride (DOI), a partial agonist selective for 5-HT2 receptors. One micromolar DOI stimulated changes in [Ca2+]˜ 2-fold larger than those stimulated by 300 nm 5-CT (Figs 5, 6). As expected, the ability of 1 µm DOI to trigger increased intracellular [Ca2+] was completely inhibited in the presence of 10 nm MDL100907 (Table 1 and Fig. 6). In contrast, the DOI-induced change in Ca2+ was not affected by the presence of 3 µm SB269970. The Ca2+ transient elicited by 600 nm 5-CT is shown in Fig. 6 for comparison. Although the magnitude is similar to that stimulated by 1 µm DOI, it was partially inhibited by both MDL100907 and SB269970, reflecting activation of both 5-HT2A and 5-HT7 receptors.

The increase in Ca2+ induced by 300 nm 5-CT appeared to be required for maximal 5-HT7A receptor-mediated activation of Akt. 5-CT-stimulated activation was inhibited by 53% when extracellular [Ca2+] was reduced from 2 mm to 100 nm by pre-treatment with EGTA (Fig. 7b). A similar 64% inhibition was seen when intracellular stores of Ca2+ were slowly depleted by pre-treatment with 30 nm thapsigargin for 30 min prior to treatment with 5-CT. Complete inhibition of Akt activation occurred when both extracellular and intracellular Ca2+ were reduced following pre-treatment with both EGTA and thapsigargin. These effects were not the result of cell toxicity as both thapsigargin and EGTA potentiated 5-CT-stimulated activation of ERK (Fig. 7a), a finding that was in contrast to 5-HT2A receptor-mediated activation of ERK (Fig. 2a). Treatment with EGTA and/or thapsigargin, alone, induced a small decrease in basal Akt activity and a small increase in ERK activity (not shown). In summary, maximal 5-HT7A receptor-mediated activation of Akt, but not ERK, showed a requirement for increases in intracellular Ca2+. However, the sensitivity of 5-CT-stimulated Akt activation to Ca2+ was less than that exhibited by 5-HT2A receptor-mediated activation (Fig. 2b).

Figure 7.

Maximal 5-HT7A receptor-mediated activation of Akt, but not ERK, requires increases in intracellular [Ca2+]. PC12 cells stably expressing human 5-HT7A receptors were pre-treated with EGTA (to reduce extracellular [Ca2+] to 100 nm) and/or 30 nm thapsigargin to deplete intracellular stores for 30 min prior to treatment with 5-CT for 5 min. Supernatant was analyzed by immunoblotting with antibody to (a) phospho-ERK1/ERK2 (p-ERK) and (b) phospho-Akt (p-Akt). Membranes were then stripped and analyzed with antibody to (a) total ERK1/ERK2 (Total ERK) and (b) total Akt (Total Akt). *p < 0.05; **p < 0.01; ***p < 0.001 versus 5-CT. Representative immunoblots from three separate experiments performed in duplicate are shown.

5-HT7A receptor-mediated increases in cAMP and 5-HT2A receptor-mediated increases in Ca2+ positively interact to increase activation of Akt

Our finding that 5-HT7A receptors utilized a pathway for activation of Akt requiring increases in both cAMP and Ca2+ suggested that each second messenger could enhance the activation of Akt stimulated by the other. We therefore tested the effect of concomitant treatment with agonists for 5-HT7A and 5-HT2A receptors. Our hypothesis was that the large increase in cAMP stimulated by 5-CT would enhance the activation of Akt stimulated by 5-HT2A receptor agonist-induced increases in intracellular [Ca2+]. DOI, as expected, stimulated activation of both ERK and Akt (Fig. 8). The activations of ERK and Akt stimulated by simultaneous treatment with 5-CT and DOI were found to be slightly more than additive (Fig. 8a,c). Significantly, the activation of both pathways by concomitant treatment was to approximately the same magnitude observed when cells were treated with 1 µm 5-HT, which acts at both receptor subtypes. Consistent with the 5-CT-stimulated enhancement of DOI activity being mediated by cAMP, we found that pCPT-cAMP similarly enhanced the activation of ERK and Akt stimulated by DOI (Fig. 8b,d).

Figure 8.

5-CT and the cAMP analog pCPT-cAMP enhance 5-HT2A receptor-mediated ERK and Akt activation. (a, c) PC12 cells stably expressing human 5-HT7A receptors were treated for 5 min with 300 nm 5-CT, 1 µm DOI, 300 nm 5-CT plus 1 µm DOI, or 1 µm 5-HT, and then lysed. (b, d) PC12 cells stably expressing human 5-HT7A receptors were treated for 5 min with 200 µm pCPT-cAMP and/or 1 µm DOI, and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a, b) phospho-ERK1/ERK2 (p-ERK) and (c, d) phospho-Akt (p-Akt). (a, c) ***p < 0.001 versus 5-CT. (b, d) ***p < 0.001 versus DOI. Data were obtained from three separate experiments performed in duplicate.

We next examined the role of protein kinase A (PKA) in mediating the effects of cAMP. We have previously found that 5-CT stimulates a 3-fold activation of PKA in 5-HT7 receptor-expressing PC12 cells. This activation is completely inhibited by pre-treatment of cells with the PKA inhibitor H-89, at concentrations as low as 0.1 µm (Lin et al. 2003). In contrast, 10-fold higher concentrations of H-89 (1 µm) caused no inhibition of 5-CT-stimulated activation of ERK and Akt (Fig. 9). The best characterized pathway for activation of Akt requires phosphatidylinositol 3-kinase-dependent phosphorylation at Thr308/Ser473 by PDK-1 and -2 (Alessi et al. 1996, 1997). In contrast, PKA has been reported to activate Akt via an apparently different pathway, independent of phosphatidylinositol 3-kinase (Filippa et al. 1999). Consistent with our finding that PKA was not required for 5-CT-stimulated activation of Akt, we found a requirement for phosphatidylinositol 3-kinase. Pre-treatment with either 30 nm wortmannin or 25 µm LY294002, two chemically distinct phosphatidylinositol 3-kinase inhibitors, completely inhibited activation phosphorylation of Akt (Fig. 9c).

Figure 9.

5-HT7 receptor-mediated activation of Akt is independent of PKA and sensitive to phosphatidylinositol 3-kinase inhibitors. PC12 cells stably expressing human 5-HT7A receptors were pre-treated for 30 min with (a, b) 1 µm H-89 or (c) either 30 nm wortmannin or 25 µm LY294002 (LY) prior to treatment with 300 nm 5-CT for 5 min, and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a) phospho-ERK1/ERK2 (p-ERK) and to (b, c) phospho-Akt (p-Akt). Membranes were then stripped and analyzed with antibody to (a) total ERK1/ERK2 (Total ERK) and (b, c) total Akt (Total Akt). ***p < 0.001 versus 5-CT in the absence of inhibitors. Representative immunoblots from three separate experiments performed in duplicate are shown.

We therefore sought to determine whether the cAMP-activated guanine-nucleotide exchange factor, Epac, plays a role in 5-HT7A receptor-mediated activation of Akt, similar to the role we have previously identified in the activation of ERK (Lin et al. 2003). We first used transiently transfected PC12 cells to examine whether over-expression of Epac could potentiate the activation of Akt mediated by 5-HT7A receptors. Cells transiently transfected with cDNA for Epac1 did exhibit a larger activation of ERK in response to 5-CT treatment (Fig. 10). However, over-expression of Epac1 did not potentiate 5-CT- stimulated activation of Akt.

Figure 10.

5-HT7A receptor-mediated activation of ERK, but not Akt, can utilize Epac. (a, c) PC12 cells stably expressing human 5-HT7A receptors were transiently transfected with cDNA for Epac1 or a control plasmid. Cells were then treated with 30 nm 5-CT for 5 min prior to lysis. (b, d) PC12 cells stably expressing human 5-HT7A receptors were treated for 5 min with 1 µm DOI and either 200 µm pCPT-cAMP or 200 µm 8-CPT-Me-cAMP, and then lysed. Supernatant was analyzed by immunoblotting with antibody to (a, b) phospho-ERK1/ERK2 (p-ERK) and (c, d) phospho-Akt (p-Akt). (a, c) ***p < 0.001 versus control plasmid-transfected cells treated with 5-CT. (b, d) ***p < 0.001 versus DOI in the absence of cAMP analogs. Data were obtained from four separate experiments performed in duplicate.

We next sought to determine whether Epac was responsible for mediating cAMP-induced augmentation of DOI-stimulated ERK and Akt activation. The cell-permeable cAMP analog, 8-CPT-Me-cAMP, has recently been shown to activate Epacs 1 and 2 without activating PKA (Enserink et al. 2002; Christensen et al. 2003). Cells treated with 8-CPT-Me-cAMP, did in fact exhibit the same magnitude of enhancement of DOI-stimulated ERK activation (Fig. 10b) seen with pCPT-cAMP, which activates both PKA and Epac (Christensen et al. 2003). However, selective activation of Epac did not enhance DOI-stimulated Akt activation (Fig. 10d), suggesting a role for a cAMP-dependent protein other than Epac.

Discussion

Agonists for Gs-coupled receptors have been reported to inhibit ERK activity in most cell types. However, such agonists conversely stimulate activation of ERK in neurons and cell lines of neuronal lineage, including PC12 cells. The effect of cAMP on activation of Akt has been found to be similarly cell dependent. For example, direct activation of adenylate cyclase with forskolin inhibits phosphorylation and activation of Akt in COS monkey kidney cells, Swiss 3T3 cells, HEK293 cells and Rat2 fibroblasts (Kim et al. 2001; Mei et al. 2002). This inhibition may be secondary to the observed effect of cAMP decreasing phosphatidylinositol 3-kinase activity, as phosphatidylinositol 3,4,5-triphosphate modulates the membrane localization of PDK1 and Akt (Kim et al. 2001). However, in some cell types cAMP has been found to stimulate activation of Akt. For example, forskolin and agonists for Gs-coupled pituitary adenylate cyclase-activating polypeptide receptors have been reported to stimulate activation of Akt in PC12 cells (Lee et al. 2002; Piiper et al. 2002). Interestingly, pituitary adenylate cyclase-activating polypeptide is similar to agonists for the 5-HT7A receptor in that it has been shown in some systems to increase Ca2+ in addition to increasing cAMP.

Until recently, it was assumed that PKA mediates all cellular responses to cAMP. However, it is now known that elevations in cAMP, such as those induced by Gs-coupled receptors, can also elicit effects independent of PKA. The best characterized proteins mediating PKA-independent activity are Epac1 and Epac2, cAMP-activated guanine-nucleotide exchange factors (cAMP-GEFs) for Rap1 and Rap2 (de Rooij et al. 1998; Kawasaki et al. 1998; Pham et al. 2000; Kashima et al. 2001; Laroche-Joubert et al. 2002; Mei et al. 2002). Evidence for a role for Epac in the activation of Akt has been reported for some cell types. For example, HEK293 cells exhibit forskolin-induced activation of Akt only after transfection with cDNA for Epac (Mei et al. 2002). Similarly, a constitutively activated form of Epac has been reported to independently stimulate activation of Akt in transfected C6 glioma cells (Wang et al. 2001).

We have previously reported a role for Epac in 5-HT7A receptor-mediated activation of ERK (Lin et al. 2003). We found that over-expression of Epac1 causes a potentiation of agonist-stimulated activation of ERK in PC12 cells. The larger magnitude of ERK activation observed in our current studies by pCPT-cAMP relative to 8-Br-cAMP is also consistent with mediation by Epac. Though both are cell-permeable analogs of cAMP, pCPT-cAMP has been shown to be a more potent activator of Epac (Christensen et al. 2003). We also found in this study that the selective activator of Epac, 8-CPT-Me-cAMP, mimics 5-CT and pCPT-cAMP in enhancing the activation of ERK stimulated by the 5-HT2A receptor agonist DOI.

In contrast, we found no role for Epac in 5-HT7A receptor-mediated activation of Akt. Over-expression of Epac1 did not potentiate agonist-stimulated activation of Akt. Nor did we observe a difference in the magnitude of Akt activation stimulated by pCPT-cAMP relative to 8-Br-cAMP. Moreover, direct activation of Epac with 8-CPT-Me-cAMP did not mimic the actions of 5-CT and pCPT-cAMP in enhancing 5-HT2A receptor-mediated activation of Akt. However, we also found that activation of Akt does not require PKA. 5-CT-stimulated activity was found to be insensitive to H-89 at a concentration 10-fold higher than that required to completely inhibit PKA in PC12 cells (Lin et al. 2003). The observed sensitivity of Akt activation to phosphatidylinositol 3-kinase inhibitors was also consistent with a PKA-independent process (Filippa et al. 1999; Mei et al. 2002). We can therefore conclude that 5-HT7A receptor-mediated activation of Akt is independent of both PKA and Epac. Possibly, another cAMP-GEF, such as CNrasGEF (Pham et al. 2000), could be required.

We found that PC12 cells are similar to glomerulosa cells and transfected HEK293 cells in exhibiting 5-HT7A receptor-mediated increases in intracellular [Ca2+] (Baker et al. 1998; Lenglet et al. 2002). A possible mechanism is via phospholipase C epsilon, which has been reported to mediate other Gs-coupled receptor-stimulated increases in [Ca2+]. The phospholipase is expressed in brain, and is activated both by Rap and Ras (Schmidt et al. 2001; Song et al. 2002; Wu et al. 2003). Interestingly, potentiation of 5-HT7 receptor-mediated ERK activation was observed when the small, 5-CT-stimulated increases in [Ca2+] were prevented. This inhibitory effect of Ca2+ might be the result of inhibition of cAMP synthesis. Type VI adenylate cyclase is expressed in PC12 cells (Chern et al. 1995) and is known to be inhibited by Ca2+. In contrast, the activation of ERK stimulated by simultaneous treatment with DOI and 5-CT was found to be at least additive. It would appear that the Ca2+-dependent inhibition of 5-CT-stimulated ERK activation is compensated for by a cAMP-dependent augmentation of DOI-stimulated ERK activation. Alternatively, the increases in [Ca2+] stimulated by DOI might be compartmentalized such that they do not modulate the coupling of 5-HT7 to activation of ERK. In fact, compartmentalization of the MAP kinase pathway through scaffold proteins and docking sites has been described (Pouyssegur et al. 2002).

Little has been reported previously on the interactions between subtypes of 5-HT receptors. However, Berg et al. (1994, 1996), have shown that Gq-coupled 5-HT2C receptors inhibit Gi-coupled 5-HT1B receptor-mediated inhibition of adenylate cyclase. Our finding that 5-HT2A and 5-HT7A receptors positively interact to activate ERK and Akt suggests that other Gs- and Gq-coupled receptors will be found to similarly do so in neuronal cell types. This could be of relevance to understanding the actions of 5-HT at other receptor subtypes exposed to SSRI-induced increases in synaptic 5-HT. It may also be of relevance to understanding the mechanism of action of other classes of antidepressants, such as the tricyclic antidepressants, that increase both norepinephrine and 5-HT. Interactions between various Gs- and Gq-coupled receptor subtypes for each of the monoamines could be hypothesized.

Interestingly, the specific interaction of 5-HT2A and 5-HT7A receptors may be of clinical significance with regard to antipsychotic medications. A characteristic of the newer, ‘atypical’ anti-psychotics is their antagonism at 5-HT2 receptors. A subset of these compounds has been found to also bind with relatively high affinity to the 5-HT7 receptors (Roth et al. 1994). Our findings suggest that such dual inhibition would cause greater inhibition of ERK and Akt in neurons co-expressing both receptors than would atypical anti-psychotics exhibiting antagonism solely at 5-HT2 receptors. The role of ERK and Akt, if any, in mediating psychosis is not currently known. Therefore, it is not clear whether increased inhibition would be expected to confer increased anti-psychotic efficacy.

Acknowledgements

These studies were supported by NIMH grant MH60100 to DSC and NIGMS grant GM36387 to GRD.

Ancillary