| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,2
* Service de Biochimie, IFR 139, Hôpital Lariboisière and Laboratoire de Biologie Cellulaire, EA3621, Faculté de Pharmacie, Université Paris V, Paris, France;
Différenciation Cellulaire et Prions, CNRS UPR 1983, Institut André Lwoff, Villejuif, and Institut Pasteur, Département de Biologie Cellulaire et Infections, Paris, France; and
Pharma Research Department, Hoffmann-La Roche A.G., Basel, Switzerland
1Correspondence: Service de Biochimie, Hôpital Lariboisière, 75475 Paris Cedex 10, France. E-mail: jean-marie.launay{at}lrb.aphp.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-subunit, impairing the electrochemical gradient necessary to 5-HT uptake. SERT hyperphosphorylation also affects antidepressant recognition. Finally, such 5-HT2B receptor-mediated control of SERT activity operates in primary neurons from raphe nuclei. Altogether, our data shed new light on the 5-HT-driven post-translational modifications involved in the control of SERT activity.—Launay, J-M., Schneider, B., Loric, S., Da Prada, M., Kellermann, O. Serotonin transport and serotonin transporter-mediated antidepressant recognition are controlled by 5-HT2B receptor signaling in serotonergic neuronal cells.
Key Words: SERT • neuronal differentiation • transporter(s)
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The maintenance of a robust neurotransmission largely depends on the precise control of extracellular 5-HT levels. This process is driven by the 5-HT transporter (SERT), which belongs to the Na+/Cl–-dependent transporter family (2
, 3)
. SERT assumes the uptake of extracellular 5-HT across the cell membrane of neuronal and non-neuronal cells, such as serotonergic neurons of raphe nuclei, platelets, or gut enterochromaffin cells. Dysregulation of the tightly controlled external concentration of 5-HT is at the origin of a host of metabolic and psychiatric disorders, including depressive, anxious, and obsessive-compulsive disorders. For instance, many therapeutic compounds efficient in the treatment of such pathologies target SERT, act as reuptake inhibitors (tricyclic antidepressants, e.g., imipramine; SSRIs, e.g., fluoxetine, paroxetine) (4
5
6)
, or substrate-type releasers (e.g., amphetamines) (7)
and optimize 5-HT concentration at the neuronal synapse. The benefit of antidepressants may also be related to a potentiation of 5-HT autoreceptor signaling function as a consequence of a prolonged trafficking of 5-HT receptors to the plasma membrane (8)
. In any case, according to the monoamine theory of affective disorders, these drugs interfere with some central serotonergic dysfunctions (8
9
10
11)
. Reductions of 5-HT transport and imipramine binding sites in brain often coincide with the diagnosis of major depression or obsessive-compulsive disorder, as well as with suicide. Such relations between 5-HT transport and neuropsychiatric diseases (12
13
14
15
16)
strengthen the interest of studying the molecular and regulatory features of SERT functions.
Progress in the field of 5-HT transport was allowed by the cloning of cDNAs encoding SERTs (17
18
19)
and purification of the SERT itself (20
, 21)
. Analysis of the amino acid sequence of SERT indicates the presence of several consensus sites for phosphorylation by PKA, PKC, and Ca2+/calmodulin-dependent protein kinase. In addition, in a non-neuronal cell context, SERT uptake activity was reported to be under the control of histamine receptors and of adenosine receptor-associated PKG signaling pathways (22
23
24)
. However, the exact regulatory mechanisms that affect the function of the transporter in neurons remain largely unknown (see ref. 25
for review).
The 1C11 cell line, endowed with the capacity to differentiate into neuronal serotonergic cells (26
, 27)
, is a useful tool to investigate the regulation of SERT function. Upon 4 days of induction, nearly 100% of 1C11 precursor cells, which lack neuron-associated functions, convert into serotonergic cells. The resulting 1C115-HTd4 cells express neuron-associated markers and have acquired the ability to synthesize, store, catabolize, and take up 5-HT. Three serotonergic receptors are selectively induced during the differentiation of 1C11 cells. 5-HT1B/1D and 5-HT2B receptors become functional on day 2, whereas the onset of 5-HT2A receptors occurs on day 4, simultaneously with that of a 5-HT active transport (26
, 28)
. The inducible 1C11 clone therefore implements a functional SERT in an integrated serotonergic phenotypic context.
In the present work we demonstrate that 5-HT2B receptors, through their couplings to the PKC and PKG/NO pathways, govern the overall 5-HT transport system by promoting phosphorylation of both the SERT and its energy source, the Na+,K+-ATPase, in 1C11-derived serotonergic cells as well as in primary raphe neurons. In addition, our results highlight that SERT hyperphosphorylation mediated by agonist-dependent 5-HT2B receptor stimulation impairs the binding of antidepressants to the serotonin transporter.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-32P]-ATP (2590 Ci/mmol) were from NEN Life Science Products (Boston, MA, USA). Desipramine and CGP 6085A, citalopram, paroxetine, indalpine, fluoxetine, and LY266097 were kindly provided by Sanofi-Synthélabo (Paris, France), Novartis (Basel, Switzerland), H. Lundbeck (Copenhagen, Denmark), Ferrosan-Novo (Copenhagen, Denmark), Aventis (Paris, France), and Elli Lilly (Indianapolis, IN, USA), respectively. All other chemicals were reagent grade purchased from the usual commercial sources. Rabbit polyclonal antibodies toward SERT or 5-HT2B receptors were from Calbiochem (Darmstadt, Germany) or BD Pharmingen (San Diego, CA, USA), respectively. Mouse monoclonal anti--tubulin was purchased from Sigma-Aldrich (Saint Louis, MO, USA) and provided as ascites fluid (Western blot working dilution is 1:500).
1C11 cell culture
1C11 cells were grown and induced to differentiate along the serotonergic pathway by the addition of 1 mM Bt2cAMP and 0.05% CCA in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% FCS (27)
. The presence of 5-HT in animal sera used in cell culture (29)
led us to remove this biogenic amine from the FCS by dialysis through a MWCO 500 cellulose ester sterile membrane (Spectrum, Houston, TX, USA) and made up identical to undialyzed FCS for most cytokines and growth factors. The residual 5-HT concentration of such a reconstituted FCS was systematically checked, using a highly specific and sensitive radio-enzymatic assay (30)
, to be lower than 1 nM.
Mouse synaptosomes
Adult male C57BL/6By mice (11±3 wk old) were rapidly decapitated, brains were removed, and the frontal cortex was dissected on ice. Tissues were prepared at 4°C as a P2 pellet (31)
. The protein concentration was determined with the bicinchoninic acid (BCA) reagent (Pierce, Chester, UK). Uptake assays were performed according to (32)
with 0.5 mg protein · ml–1.
Primary raphe neuronal-enriched cell cultures and 5-HT2B receptor expression
Neuronal cell cultures were obtained from the ventromedial medullary (VMM) of E-10 129PAS mice following a protocol described for serum-free cultures (33)
, which was adapted to raphe neurons. Briefly, a wedge of tissue from the ventromedial portion of the rostral one-half to two-thirds of the medulla was removed that contained the raphe pallidus, raphe magnus, raphe obscurus, and tissue immediately adjacent to these nuclei. Dissected tissue was dissociated and plated on poly-L-ornithine- and laminin-coated CELLocate coverslips at a density of 8.5 x 104 cells/ml in 10% fetal calf serum (FCS), 54% modified Eagle’s medium, and 36% neurobasal medium with B27 supplement (Gibco BRL, Grand Island, NY, USA). Medium was conditioned for 1 day before use by glial cultures obtained from the VMM. Basic fibroblast growth factor (bFGF) (0.5 ng·ml–1) and FGF-5 (2.5 ng/ml) (Gibco BRL) were added to the culture medium to enhance survival and cytosine-ß-D-arabino-furanoside hydrochloride (3 µM) was added to inhibit glial growth.
Lysates of primary neurons from embryonic raphe nuclei were prepared by incubating cells for 15 min at room temperature in radio-immuno-precipitation assay (RIPA) lysis buffer (25 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 0.5% deoxycholate, 0.5% Nonidet P-40, 1% SDS, 1 mM Na3VO4, and a cocktail of protease inhibitors; Roche, Nutley, NJ, USA) after extensive washing in cold PBS. Extracts were centrifuged at 14,000 g for 15 min. Protein concentration in the supernatant was measured by using the BCA protein assay (Pierce). Assessment of 5-HT2B receptor expression in primary neurons from embryonic raphe nuclei was performed by SDS/10% PAGE (20 µg protein lysate) and transfer to Immobilon membranes (Millipore, Billerica, MA, USA). Membranes were blocked in 5% nonfat dried milk and Tris-buffered saline with 0.1% Tween 20 for 1 h at room temperature, then incubated overnight at 4°C with 0.5 µg·ml–1 5-HT2BR primary antibody (Ab). Bound Ab was revealed by enhanced chemiluminescence (ECL) detection (Amersham Pharmacia, Arlington Heights, IL, USA).
Cell membrane preparations, radioligand binding studies, 5-HT uptake, data analysis, and statistics
All these experiments were performed as already reported in (20
, 26
, 34)
. To prepare crude membrane for radioligand binding assays, 1C115-HT cells were harvested in PBS containing a cocktail of protease inhibitors (Roche). Cells were then pelleted by centifugation and resuspended in cold buffer A containing 4 mM EDTA, 1 mM EGTA, 0.1 mM PMSF, 10 mM imidazole, pH 7.3. After centrifugation, the supernatant was poured onto a 20% sucrose cushion, then centrifuged at 100,000 g for 90 min. The membrane-containing pellet was resuspended in buffer B (75 mM KCl, 5 mM MgCl2, 1 mM EGTA, 10 mM imidazole, pH 7.3). Protein contents were determined by the BCA protein assay (Pierce).
Radioligand binding experiments were performed using either intact cells or crude membranes. Briefly, with intact cells, binding assays were initiated by the addition of DMEM containing the radiolabeled ligand and the appropriate competing ligands. After a 30 min incubation at 37°C, cells were washed twice with cold PBS and 1N HClO4 was added. Radioactivity was counted in a scintillation counter (Packard, Issy-les-Moulineaux, France). The specific binding was defined as the binding that was inhibited by 1 µM of homologous unlabeled ligands. With crude membranes (20 µg of proteins), the binding was initiated at 37°C by the addition of radiolabeled ligand and appropriate competing ligands into 50 mM Tris, pH 7.4. A 30 min incubation period was followed by the addition of ice-cold 10 mM Tris, pH 7.4. Samples were filtered on PEI-treated filters and radioactivity retained on filters was counted.
Binding data were analyzed by the iterative nonlinear fitting software Prism (GraphPad, San Diego, CA, USA). This allowed the calculation of dissociation equilibrium constants (KD) for saturation experiments, as well as inhibition constants (KI) for displacement studies. The statistical analysis on small groups used nonparametric tests and the Instat software (GraphPad). The chosen significant criterion was P < 0.01. All values are given as means ± SE or 95% confidence intervals.
[3H]-5-HT (5-HT) binoxalate was used as a sensitive probe to assess SERT uptake activity. The uptake of [3H]-5-HT binoxalate into 1C115-HT cells was initiated in serotonin free medium in the presence or absence of 5-HT uptake inhibitors. [3H]-5-HT binoxalate uptake was measured within a short incubation time (60 s) at 37°C. Cell protein content was measured using the bicinchoninic method (Pierce). Apparent KM and Vmax values were determined by a nonlinear regression using the EZ-FIT software (35)
according to ref. 26
.
RNA analysis of 1C115-HT differentiating cells
Total RNA from 1C11 precursor and 1C115-HT cells was prepared and Northern analysis was performed as described (36)
. Each RNA sample was treated with deoxyribonuclease, purified, and subjected to PAT (poly(A) test, a polymerase chain reaction (PCR) -based assay) cDNA synthesis (37)
. All transcript-specific amplification were performed on the same cDNA preparation. Because the amount of RNA was not determined for each time point, quantitative estimates cannot be made. Poly(A) tail lengths are reported as maximally elongated species. Similar conclusions were drawn if the poly(A) tail lengths were estimated as the average elongation.
RNA isolation and reverse transcriptase-polymerase chain reaction analyses with embryonic raphe nuclei-derived neuronal cells.
Total RNA isolation from embryonic raphe nuclei-enriched neuronal cells was performed using the RNeasy Maxi Kit (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA). Semiquantitative RT-polymerase chain reaction (RT-PCR) was performed using the following specific primers: 5-HT2B receptor 5' AGGAATCGAGACTGATGTGAT 3' and 5' CTTAGGAAAACTGTGGGCACA 3' (230 bp); GAPDH 5' TGAAGGTCGGTGTGAACGGATTTGGC 3' and 5' CATGTAGGCCATGAGGTCCACCAC 3'(1100 bp).
[32P]-labeling and pharmacological treatments
Cells were incubated for at 30°C in 2 µl of Krebs-bicarbonate buffer [124 mM NaCl, 4 mM KCl, 26 mM NaHCO3, 1.5 mM CaC12, 1.5 mM MgSO4, 0.25 mM KH2PO4, 10 mM glucose], oxygenated with 95% O2/5% CO2 (v/v). After 15 min this incubation buffer was replaced with the same amount of fresh buffer containing 2.5 mCi of [32P]-orthophosphoric acid and cells were incubated for 60 min. The radioactive buffer was then removed, and [32P]-labeled cells were washed twice with 2 µl of fresh buffer and incubated for an additional 2–60 min in the presence or absence of drugs. After drug treatment, the buffer was removed and cells were rapidly frozen in liquid nitrogen and stored at –70°C until assayed.
Determination of the Na+,K+-ATPase activity
Cells were incubated for 30 min at 22°C in the presence or absence of drugs. The Na+,K+-ATPase activity was probed by the hydrolysis of ATP with crude membranes of 1C115-HT cells as previously reported in (38
, 39)
. In a typical assay, crude membranes of 1C115-HT cells were preincubated for 10 min in a medium containing 50 mM NaCl, 10 mM MgCl2, 1 mM EGTA, 100 mM Tris-HCl pH 7.40, 10 mM ATP, and [-32P]-ATP in tracer amounts (
5 nCi/µl). The reaction was started by adding 5 mM KCl and incubated for 15 min at 37°C. The amount of proteins was selected precisely to hydrolyze <20% of the total ATP. In this range, the time course of ATP hydrolysis was linear. The reaction was stopped by the addition of activated charcoal and sample were centrifuged at 13,000 g. Phosphate liberated by the hydrolysis of [-32P]-ATP was counted in the supernatant. Determination of Mg2+-dependent ATPase activity was performed by incubating membranes with 3 mM ouabain in a reacting buffer lacking NaCl and KCl. Specific Na+,K+-ATPase activity was calculated as the difference between total and Mg-dependent activities.
Immunoprecipitation of Na+,K+-ATPase
[32P]-labeled cells were sonicated in 1 µl of lysis buffer (20 mM Tris/HCl, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 0.2% BSA, pH 8.00) containing 50 mM NaF to block phosphatase activity and the following protease inhibitors: 1 mM EGTA, 25 mM benzamidine, 100 mM phenylmethyl-sulfoxide, 20 mg/ml chymostatin, 20 mg/ml pepstatin A, 5 mg/ml leupeptin, and 5 mg/ml antipain. Aliquots of the homogenate (20 µl) were used for determination of total [32P]-incorporation into trichloroacetic acid-precipitated proteins. Ten milligrams of protein A-Sepharose CL-4B was added to each tube and the samples were mixed for 30 min at 4°C. The Sepharose beads and associated, nonspecifically adsorbed, proteins were removed by centrifugation. The supernatants were mixed for 2 h at 4°C with 15 µl of a mouse monoclonal antibody specific for the 1-isoform of Na+,K+-ATPase (Sigma-Aldrich), followed by a 1 h incubation with 15 µl of affinity-purified rabbit antimouse Ab. The samples were then incubated for 1 h at 4°C with 10 mg of protein A-Sepharose beads. The beads were collected by centrifugation, washed extensively with 1 µl of lysis buffer containing 0.1% SDS, resuspended in 50 µl of SDS-PAGE sample buffer, vortexed, and centrifuged. The recovered proteins were separated by SDS-PAGE on 7.5% acrylamide gels. Gels were dried and [32P]-incorporation into Na+,K+-ATPase was quantified using a Phosphor-Imager 400B. Individual values of [32P]-incorporation into Na+,K+-ATPase were corrected for the total [32P]-incorporation measured for each sample.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Here, we carefully compare the functional and pharmacological properties of SERT expressed in 1C115-HT cells grown in either standard culture conditions [10% FCS containing 0.1 to 1 µM 5-HT (29)
] or in a dialyzed culture medium containing less than 1 nM 5-HT (Table 1
).
|
At day 4 of serotonergic differentiation, KD values of either [3H]-paroxetine or [3H]-cyanopramine measured with 1C115-HT cells are very similar to those measured with synaptosomes or platelets. These binding constants remained unchanged whatever the external 5-HT concentration during differentiation. Upon 5-HT exposure, the KM for 5-HT transport varied significantly by a factor of 30% (0.91±0.05x10–7 M vs. 1.43±0.16x10–7 M in 5-HT-depleted medium) with 1C115-HT cells. Nevertheless, these KM values remain in the same range of magnitude as those obtained with platelets or synaptosomes (Table 1)
.
Nontricyclic and tricyclic antidepressants behave as competitive inhibitors of the binding of [3H]-paroxetine to SERT. A highly significant correlation (P<0.001) was obtained between the KI values of these compounds measured with either 1C115-HTd4 cells or mouse synaptosomes (Fig. 1
). Varying the 5-HT concentration in the culture medium did not change this correlation (not shown).
|
With 1C115-HT d4 cells grown in 10% FCS medium, both the apparent Vmax of 5-HT transport (8-fold) and the Bmax for both [3H]-paroxetine (3-fold) and [3H]-cyanopramine (3-fold) are decreased when compared with the corresponding values monitored with cells cultured under 5-HT starvation (Table 1)
. Upon addition of 0.5 µM 5-HT to the dialyzed medium at the beginning of the serotonergic program, the above kinetics parameters reached values close to those obtained in 10% FCS conditions. Therefore, the observed down-regulation of 5-HT transport and of the number of antidepressant binding sites are attributable to 5-HT itself.
By assuming that the number of paroxetine binding sites reflects the number of active sites for 5-HT transport, a mean period of translocation cycle at each transport site can be calculated. With 1C115-HTd4 cells underexposed to 5-HT, a value of 0.14 s is deduced, instead of 0.35 s under standard culture conditions (Table 1)
. The former value is close to those measured with mouse synaptosomes (0.09 s) or blood platelets (0.14 s).
We may therefore conclude that, during serotonergic differentiation, extracellular 5-HT exerts a negative feedback on the onset of functional SERT. This feedback reduces not only the 5-HT transport velocity and the number of antidepressant binding sites, but also the translocation rate associated with each SERT molecule.
SERT mRNA translation coincides with poly(A) addition
The observation that 5-HT transport varies with external 5-HT concentration prompted us to investigate the onset of SERT during 1C11 differentiation. 1C11 precursor cells were induced to differentiate either in standard conditions or in a 5-HT-depleted medium, and SERT-encoding mRNAs were analyzed by Northern blots. SERT transcripts were detected as soon as day 0, indicating that, at the committed stem cell state, 1C11 precursor cells already transcribe the SERT gene.
Whatever the external 5-HT concentration, the steady-state level of SERT transcripts remained roughly constant along the 4 days of differentiation (Fig. 2
A). We therefore focused our experiments on a possible variation of the poly(A) tail length of SERT mRNA. Indeed, in most animal species, the length of the poly(A) tail at the 3'-end of specific mRNAs can control translational activation during early embryogenesis (see refs. 40
, 41
for reviews). As shown in Fig. 2B
, in 1C11 progenitor cells, SERT mRNAs display a short poly(A) extension of 40 nucleotides (nt). The length of this tail rapidly increases upon induction of differentiation in 10% FCS medium, with a peak at 200 nt in 1C115-HTd3 cells. Further shortening of the tail (125 nt) occurs on day 4. With the dialyzed culture medium, similar variations were observed upon differentiation (not shown). Clearly, the SERT mRNA gains or loses poly(A) at precise times during the course of the serotonergic program irrespective of the external 5-HT concentration.
|
Anti-SERT antibodies were used to follow de novo SERT mRNA translation. As shown in Fig. 2C
, the SERT mRNA is dormant at the stem cell stage (day 0). At day 2, whatever the external 5-HT concentration, an expression of the SERT (68 kDa) becomes detectable in cell homogenates by Western blot. However, day 4 must be awaited to observe [3H]-paroxetine or [3H]-cyanopramine binding with 1C115-HT homogenates or plasma membranes (Table 2
).
|
Cocaine analog binding precedes antidepressant recognition
The lag of 2 days between the onset of SERT expression and the first detection of antidepressant binding (Table 2)
led us to examine whether other ligands could be recognized by the transporter beforehand. 1C115-HT cells lack the DA and NE transporters (27)
. Moreover, the cocaine binding sites of SERT are likely to be distinct from those of antidepressants (4)
. Consequently, among the nonselective monoamine transport inhibitors available, we selected RTI-55, a cocaine-related molecule, to trace the SERT protein.
The cocaine congener [125I]-RTI-55 started binding 1C115-HT homogenates on day 2 (Table 2)
. The extent of RTI binding further increased up to day 3 and remained constant until day 4. If plasma membranes were used instead of cell homogenates, RTI binding became measurable on day 3 only, i.e., 1 day before the first detection of paroxetine and cyanopramine binding and the onset of 5-HT uptake. Of note, with homogenates or plasma membranes of 1C115-HTd4 cells, [125I]-RTI-55 binding was completely displaced in competition experiments by unlabeled paroxetine (KI=0.38±0.03 nM) or cyanopramine (KI=1.72±0.09 nM) (Fig. 3
). These overall data most certainly demonstrate that RTI-55 cocaine congener specifically targets SERT protein in 1C115-HT differentiating cells.
|
In summary, SERT translation starts on day 2. Translation closely follows polyadenylation of the SERT mRNA. The newly synthesized SERT molecules display cocaine binding properties. SERT insertion at the plasma membrane occurs on day 3. Between day 3 and day 4, the transporter becomes competent for 5-HT transport. In parallel, it acquires the capacity to recognize antidepressants.
5-HT regulates the number of SERT molecules functional for 5-HT transport
To go deeper into the 5-HT-mediated regulation of SERT function, we evaluated the threshold concentration of extracellular 5-HT causing inhibition of 5-HT transport. 1C11 cells were left to differentiate in a dialyzed culture medium supplemented with various 5-HT concentrations from day 0. RTI and paroxetine bindings were measured in cell homogenates on day 4 (Fig. 4
). RTI binding remained unchanged (220 sites per cell) whatever the 5-HT concentration in the culture medium. Below 10 nM 5-HT, the paroxetine binding did not differ from that of RTI. By contrast, beyond 10 nM 5-HT, the number of paroxetine binding sites started to decrease. Above 500 nM 5-HT, the numbers of paroxetine and cyanopramine binding sites were reduced by up to two-thirds compared with that of RTI (75 sites/cell vs. 220). Identical results were obtained if the culture medium was supplemented with 5-HT on day 2 instead of day 0 (not shown).
|
While external 5-HT had no effect on RTI binding measured with 1C115-HTd4 homogenates (Table 2)
, it reduced by nearly one-third the number of RTI binding sites associated with the plasma membranes (Table 2
, 1273±94 vs. 1835±132 fmol/mg protein). This difference likely reflects internalization of 30% of the SERT molecules caused by an excess of exogenous 5-HT. As shown in Table 2
, in the presence of 5-HT the ratio between total SERT proteins (i.e., the molecules that bind RTI) and functional SERT (i.e., the paroxetine/cyanopramine-recognizing molecules) is the same whether in homogenates (708 vs. 208/271 fmol/mg protein) or plasma membranes (1273 vs. 355/381 fmol/mg protein). In homogenates or at the cell surface, only one-third of total SERT molecules are therefore competent for antidepressant binding.
Thus, our results support the view that an excess of external 5-HT causes at the same time 1) a down-regulation of the number of antidepressant binding sites and of the velocity of 5-HT transport, 2) a reduction of the translocation rate sustained by each functional SERT molecule, and 3) some SERT internalization that accounts for 10% of the reduction of 5-HT uptake. Our data thus provide evidence for the existence of distinct subpopulations of SERT differing by their drug binding properties, but equally sensitive to endocytosis.
Since the capacity of 1C115-HTd4 cell homogenates to bind the cocaine analog does not depend on external 5-HT, it is likely to reflect total SERT synthesis. At the lowest serotonin concentration assayed, close to the concentrations present in cerebrospinal fluid (CSF) (0.11–1.43 nM (42)
) or in plasma (0.27–1.49 nM, (43)
), nearly all SERT molecules show full functionality on day 4 of differentiation. By contrast, under overexposure to 5-HT, both the number of SERT antidepressant binding sites and the period of the SERT translocation cycle are reduced by 30–40%. These combined reductions largely account for the 8-fold decrease of the Vmax for 5-HT transport (Table 1)
.
5-HT2B receptor activation reduces the Na+,K+-ATPase activity
Variation in the efficiency of 5-HT transport may result from post-translational modification(s) of either the transporter itself and/or an indirect effect on the electrochemical Na+ gradient via control of the Na+,K+-ATPase activity. We compared the ATPase activity on day 4 in 1C115-HT cells left to grow in the presence or absence of 1 µM 5-HT from day 0 of the serotonergic program. Upon 5-HT addition, the Na+,K+-ATPase activity was decreased by 18 ± 6% (n=4). Addition of 1 µM 5-HT on day 2 rather than on day 0 of serotonergic differentiation gave similar results.
5-HT1B/1D and/or 5-HT2B receptors are likely mediators of the reduction in Na+,K+-ATPase activity of 1C115-HT cells since these receptors become functional on day 2 of differentiation (28
, 34)
. If added on day 2, agonists of 5-HT1B/1D receptors (5-CT or GR127935 up to 100 nM) had no effect on Na+,K+-ATPase activity. In contrast, addition of 100 nM DOI or BW723C86, two 5-HT2B receptor (5-HT2BR) agonists, decreased overall Na+,K+-ATPase activity by up to 18%. These agonist-mediated inhibitions were cancelled in the presence of 10 nM LY266097, a selective 5-HT2BR antagonist (44)
. A functional relationship between the 5-HT2BR and the ATPase could be further established by monitoring the effects of various 5-HT2 agonists and antagonists on both the binding of DOI and the Na+,K+-pump activity. A significant correlation (P<0.001) was evidenced between the apparent affinities of the drugs for the receptor, on the one hand, and their effects on the pump activity on the other. Such a correlation indicates a direct involvement of the 5-HT2BR in the control of the Na+,K+-ATPase activity (Fig. 5
A).
|
The 5-HT2B receptor promotes 5-HT-dependent phosphorylation of the Na+,K+-ATPase
In vitro experiments have already shown that PKC and cAMP-dependent protein kinases modulate the ion pump through phosphorylation of its catalytic -subunit (see refs. 45
46
47
for reviews). The previous observation that the 5-HT2B receptor is coupled to the PLC-PKC pathway in 1C11 cells (34)
prompted us to examine the phosphorylation of the Na+,K+-ATPase in response to 5-HT. All experiments were carried out with 1C115-HTd2 cells, i.e., at a stage when 1C11 cells become responsive to 5-HT and start implementing the 5-HT transport system. Cells grown in a dialyzed culture medium were exposed to [32P]-orthophosphoric acid for 1 h prior to the addition of 10 µM 5-HT for 10 min. Cell homogenates were immunoprecipitated with an Ab specific of the 95 kDa 1-isoform of the Na+,K+-ATPase. This isoform is the only one that is expressed in 1C11 cells (not shown). SDS-PAGE analysis and autoradiogram of the immunoprecipitated proteins demonstrated a marked reinforcement (6-fold) of the [32P]-labeling of the Na+,K+-ATPase 1 subunit upon exposure to 5-HT (Fig. 5B
). Similar intensities of phosphorylation were observed with either 1 or 10 µM 5-hydroxytryptamine. The increase in the phosphorylation level of the 1 subunit was completed within 10 min of 5-HT addition.
The effect of phorbol-12,13-dibutyrate (PDBu), a direct activator of PKC, on the activity and the phosphorylation of the Na+,K+-ATPase was also investigated. Treatment of 1C115-HTd2 cells with 5 µM PDBu for 10 min resulted in a 31 ± 5% (n=5) decrease in the pump activity. In parallel, phosphorylation of the Na+,K+-ATPase 1 subunit increased by a factor of 14 ± 6% (n=5) (Fig. 5B
).
The basic function of Na+,K+-ATPase is to maintain the high Na+ and K+ gradients across the plasma membrane. Since phosphorylation of Na+,K+-ATPase 1 subunit is linked to a reduction of its enzymatic activity (45
, 46)
, the electrochemical Na+ gradient is impaired (47)
. As SERT assumes the transport of 5-HT across cell membrane using the Na+ gradient as its energy source (2)
, we may conclude that a slower SERT translocation cycle, such as that evidenced with 1C11 cells exposed to 5-HT, may be related at least partly to a reduction in the rate of recycling of Na+ ions by SERT. Moreover, our data highlight that the 5-HT-dependent control of the phosphorylation level of the Na+,K+-ATPase catalytic subunit is mediated by the 5-HT2BR receptor.
5-HT transport velocity is tightly controlled by 5-HT2B receptor signaling activity
The 5-HT transport system depends both on the activity of the Na+,K+-ATPase and the number of functional SERT expressed at the cell membrane. To investigate whether the 5-HT2BR was also responsible for regulating the number of active SERT molecules, the following set of experiments was designed. First, cells were left to grow from day 2 in the presence of various concentrations of 5-HT2BR agonists. Like 5-HT itself (Table 1)
, these drugs lowered the maximal velocity (Vmax) of 5-HT transport measured on day 4. Second, similar experiments were performed with antagonists. In this case, the effect of each drug with respect to 5-HT transport velocity was defined as the concentration of this drug that relieved 50% of the inhibition of serotonin transport caused by the presence of 100 nM DOI. The concentration values obtained were compared with the binding constants of the same set of drugs toward 5-HT2BR (Fig. 6
A). The correlation observed (P<0.001) strongly suggests a pivotal role for 5-HT2BR in the control of serotonin transport beyond the regulation of the Na+,K+-ATPase activity.
|
5-HT2B receptor-mediated phosphorylation of SERT regulates 5-HT transport
To assess whether phosphorylation of the SERT molecule could be involved in the control of 5-HT transport, [32P]-treated 1C115-HTd2 cells, grown in the absence of 5-HT, were exposed to 100 nM DOI or BW723C86. Cell extracts were immunoprecipitated on day 4 with anti-SERT antibodies and [32P] incorporation was measured (Fig. 6B
). SERT phosphorylation was increased by 2.1 or 1.9-fold (n=5) upon treatment with DOI (Fig. 6B
) or with BW723C86 (not shown), respectively. Addition of 10 nM LY266097 fully prevented the effect of both agonists. We can therefore conclude that the HT2BR is directly involved in the control of SERT phosphorylation. Addition of 5 µM PDBu to the [32P]-labeled 1C115-HTd2 cells also resulted in a 2.1 ± 0.4-fold increase (n=5) in the incorporation of radioactivity by SERT (Fig. 6B
). Thus, 5-HT2BR coupling to PKC appears to mediate SERT phosphorylation in response to 5-HT or DOI.
Actually, the SERT is already phosphorylated in 1C115-HT cells prior to the addition of DOI (Fig. 6B
). Such a basal level of phosphorylation, associated with a maximal SERT transport capacity (Vmax=280 pmol/mg protein/min), is likely to result from the 5-HT2BR intrinsic activity (34
, 48)
. We next searched for 5-HT2BR-dependent transduction mechanisms, distinct from the PKC pathway, that might be at the origin of the SERT basal phosphorylation. SERT phosphorylation was cancelled upon exposure of 1C115-HTd2 cells to 50 nM ritanserin, an inverse agonist of 5-HT2B receptors (Fig. 6B
). In addition, incubation of 1C115-HTd2 cells with 1 µM KT58231, a selective inhibitor of PKG activity, fully abrogated basal phosphorylation of SERT (Fig. 6B
). Conversely, direct PKG activation by treatment of 1C115-HTd2 cells with 1 µM 8-Br-cGMP slightly enhanced incorporation of radioactivity by SERT (Fig. 6B
). Altogether, the 5-HT2BR-mediated NO/PKG pathway (49)
thus appears to account for the basal phosphorylation of SERT. Noteworthy, KT58231 also cancelled the DOI-induced SERT phosphorylation (Fig. 6B
). Therefore, the 5-HT2BR/NO-dependent phosphorylation of SERT appears to be a prerequisite for further phosphorylation through the PKC pathway.
Agonist stimulation of the 5-HT2B receptor or PKG activation also trigger SERT phosphorylation in raphe-derived serotonergic neurons
To draw a link between our findings and the in vivo situation, we established neuronal-enriched cultures from embryonic raphe nuclei. We first investigated whether 5-HT2B receptors are expressed in these primary neurons through reverse transcription and Western blot analyses. As shown in Fig. 7
A, embryonic raphe nuclei-derived neuronal cells contain transcripts encoding 5-HT2BR and expressed this receptor subtype at the protein level. This set of experiments clearly establishes the presence of 5-HT2B receptors in embryonic raphe-derived neurons, in line with the reported abundant expression of 5-HT2BR transcripts in adult raphe nuclei (50)
.
|
As previously observed using 1C115-HT cells, exposure of the raphe nuclei-derived neuronal cells to 100 nM BW723C86 for 15 min increased the phosphorylation of the SERT (Fig. 7B
). Our experiment thus provides strong evidence for a 5-HT2BR-dependent phosphorylation of the SERT in serotonergic neurons. As for 1C115-HT cells, blockade of the 5-HT2BR intrinsic activity by 50 nM ritanserin cancelled basal phosphorylation of SERT (Fig. 7B
). PKG stimulation by 1 µM 8-Br-cGMP also promoted an up-regulation of SERT phosphorylation in primary neurons from raphe nuclei (Fig. 7B
). This additional observation underlines the importance of the NO-cGMP-PKG pathway in vivo. Hence, despite the heterogeneity of raphe primary cultures, our positive data confirm the critical role of 5-HT2BR in the post-translational control of 5-HT transport.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this study, the use of the cocaine congener RTI-55 to trace the full number of SERT proteins during 1C11 differentiation enables us to clarify various aspects of SERT regulation and trafficking. 1) 5-HT has no effect on the transcriptional and translational events sustaining SERT synthesis, as deduced from Northern blot and poly(A)-tail mRNA analyses as well as from cocaine analog binding measurements. Detection of mRNAs encoding SERT in 1C11 precursor cells, which do not express any neuronal markers, is in line with the isolation of transcripts of several neuronal marker in sympatho-adrenal progenitors prior to their cell fate decision (51)
. 2) Anti-SERT antibodies and RTI binding indicate that after the initiation of SERT translation, 1 day is needed to detect SERT molecules competent for the recognition of RTI at the cell plasma membrane; another day is needed for SERT molecules to acquire the capacity to bind antidepressants and to take up 5-HT. 3) Upon 5-HT exposure, the number of SERT proteins at the cell surface slightly decreases (10%). This decrease can be explained by internalization of a fraction of the transporter population. However, the internalized fraction of SERT is too small to account for the down-regulation of 5-HT uptake by 5-HT itself.
Actually, the overall 5-HT transport is under the control of 5-HT2B receptors. Two controls are seen that act on the phosphorylation levels of the SERT itself and its energy source. Pharmacological correlations between 5-HT2BR binding properties and either the Vmax of 5-HT transport or the Na+,K+-ATPase activity give credit to this conclusion. In line with data supporting the idea that signaling pathways, including PKG and PKC, are involved in the regulation of SERT phosphorylation and activity (22
23
24
, 52
, 53)
, we show here that two distinct 5-HT2BR-dependent transduction mechanisms control 5-HT transport. First, a 5-HT2BR intrinsic activity for its NO/PKG coupling governs basal phosphorylation of the SERT. This phosphorylation, observed in 1C115-HTd4 cells grown in the absence of external 5-hydroxytryptamine, coincides with maximal antidepressant binding and 5-HT uptake capacities. In raphe serotonergic neurons as well, we demonstrate that 5-HT2BR targets the SERT, most likely through the NO/PKG pathway. Second, upon addition of 5-HT, 5-HT2BR-dependent PKC activation promotes additional phosphorylation of SERT. The Na+,K+-ATPase is phosphorylated in parallel. These modifications add their effects to reduce the 5-HT transport efficiency. We also observe that inhibition of the PKG pathway by KT58231 impairs DOI-induced SERT phosphorylation. Therefore, basal phosphorylation, depending on the PKG pathway, is a prerequisite for the SERT to be recruited as a target by the PKC pathway. Altogether, our results are consistent with a view where the intrinsic 5-HT2BR-NO coupling is sufficient to allow full implementation of the 5-HT transport system, with maximal antidepressant binding and uptake capacities. Further phosphorylation of the SERT through 5-HT2BR stimulation and subsequent PKC activation would reduce 5-HT transport and antidepressant binding capacities without any change in cocaine recognition. Accordingly, engineered constitutive phosphorylation of SERT protein was shown to affect the inhibitory power of typical antidepressants (i.e., imipramine, fluoxetine, or sertraline) on SERT transport activity, while cocaine inhibition of SERT-mediated 5-HT translocation was insensitive to SERT chronic phosphorylation (54)
. In our study, we also establish that recruitment of the 5-HT2BR-PKC pathway affects phosphorylation of the Na+,K+-ATPase. By reducing both the SERT velocity and Na+,K+-ATPase activity, the phosphorylations induced by 5-HT2BR stimulation all contribute to a down-regulation of the overall 5-HT transport system.
SERT sensitivity to antidepressant and repression of 5-HT uptake depend on autoreceptor signaling. Such a control likely represents an intrinsic homeostatic loop that links transporter capacity to external 5-HT availability. Such a link may contribute to the adaptive mechanisms that sustain the long-lasting response of 5-HT neurons to antidepressant treatment (see refs. 10
,55
for reviews). Along the same idea, fluctuating levels of 5-HT modulate the 5-HT transporter function over a long period. Post-translational autoregulation of 5-HT transport by 5-HT autoreceptor-coupled signaling pathways may be a key to understand the time-delayed therapeutic efficiency of antidepressants, contrasting with the quasi-immediate effect of cocaine assessed by its historical use as a local anesthetic (56)
. In agreement, Ramamoorthy and Blakely provide evidence that antidepressants interfere with kinase-linked post-translational modifications of SERT protein, which would contribute to (re)sensitization of 5-HT neurons to psychostimulants (52)
.
Serotonin synthesis and uptake become active at early stages during embryogenesis (see refs. 57
,58
for reviews). Noteworthy, pharmacological inhibition of 5-HT uptake causes both neural tube defects and malformations of the skull, brain, and spinal cord (59)
. In the adult, increasing evidence points to a connection between depression and a slowdown in new neuron growth (60
61
62)
. Recently, the blockade of neurogenesis in adult mice hippocampus was shown to render antidepressant treatment ineffective (62)
. However, the mechanisms sustaining the relationship between neurogenesis and the behavioral effects of antidepressant still remain unclear. These overall observations argue for the occurrence of a complex link between 5-HT transport and neurogenesis and/or morphogenesis. In keeping with this notion is our demonstration that 5-HT2BR take part in the control of the SERT function in differentiating serotonergic cells. Indeed, the 5-HT2B receptor, which is detectable in forebrain and neural tube during early embryogenesis, appears to mediate numerous 5-HT developmental functions (see refs 58
for a review). Here, we provide evidence that this receptor is present on embryonic raphe neurons where it may ensure autocrine function(s) during the ontogeny of the serotonergic system. We also detected 5-HT2B binding sites in neuronal-enriched primary cultures of adult raphe nuclei (not shown). This may indicate a role of the 5-HT2B receptor in the adult central nervous system. By identifying the SERT as a target of a 5-HT receptor, our work has implications as to the homeostasis of the serotonergic system in pathophysiological situations such as neurogenesis and neuropsychiatric diseases.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication January 20, 2006. Accepted for publication April 10, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, KI=0.38±0.03 nM) or cyanopramine (•, KI=1.72±0.09 nM). Each curve represents the mean of triplicates from a single experiment representative of two others.
) and [3H]-paroxetine (
) binding sites were measured in cell homogenates on day 4. Values were normalized with respect to total protein content in cell extracts. Means ± SE of seven independent experiments performed in triplicate are indicated.
