Interaction of calmodulin with the serotonin 5-hydroxytryptamine2A receptor. A putative regulator of G protein coupling and receptor phosphorylation by protein kinase C.

The 5-hydroxytryptamine2A (5-HT2A) receptor is a G(q/11)-coupled serotonin receptor that activates phospholipase C and increases diacylglycerol formation. In this report, we demonstrated that calmodulin (CaM) co-immunoprecipitates with the 5-HT2A receptor in NIH-3T3 fibroblasts in an agonist-dependent manner and that the receptor contains two putative CaM binding regions. The putative CaM binding regions of the 5-HT2A receptor are localized to the second intracellular loop and carboxyl terminus. In an in vitro binding assay peptides encompassing the putative second intracellular loop (i2) and carboxyl-terminal (ct) CaM binding regions bound CaM in a Ca2+-dependent manner. The i2 peptide bound with apparent higher affinity and shifted the mobility of CaM in a nondenaturing gel shift assay. Fluorescence emission spectral analyses of dansyl-CaM showed apparent K(D) values of 65 +/- 30 nM for the i2 peptide and 168 +/- 38 nM for the ct peptide. The ct CaM-binding domain overlaps with a putative protein kinase C (PKC) site, which was readily phosphorylated by PKC in vitro. CaM binding and phosphorylation of the ct peptide were found to be antagonistic, suggesting a putative role for CaM in the regulation of 5-HT2A receptor phosphorylation and desensitization. Finally, we showed that CaM decreases 5-HT2A receptor-mediated [35S]GTPgammaS binding to NIH-3T3 cell membranes, supporting a possible role for CaM in regulating receptor-G protein coupling. These data indicate that the serotonin 5-HT2A receptor contains two high affinity CaM-binding domains that may play important roles in signaling and function.

verse roles in both the central nervous system and peripheral vasculature. In the central nervous system these receptors are widely distributed, being expressed in the neocortex, claustrum, mammilary nuclei, basal ganglia, and anterior cingulate cortex (1). 5-HT 2A receptors are also highly expressed in vascular smooth muscle and renal mesangial cells, where they mediate contraction and proliferation (2)(3)(4), and platelets, where they contribute to aggregation and adherence (5), as well as in kidney (6) and skeletal muscle (7,8). The heterogeneous expression of the 5-HT 2A receptor is accompanied by a diverse array of pathophysiological implications for 5-HT 2A receptor signaling, including roles in sleep, hallucinogenesis, schizophrenia, appetite control, neuroendocrine secretions, hypertension, and depression (9 -12). The 5-HT 2A receptor is involved in the mechanism of action of hallucinogens, atypical neuroleptics, antidepressants, and other psychoactive drugs.
5-HT 2A receptors signal primarily through heterotrimeric proteins of the G q/11 subfamily to the activation of phospholipase C, and the subsequent formation of diacylglycerol and activation of protein kinase C (13)(14)(15). Other second messengers and effectors regulated by the 5-HT 2A receptor include phospholipase A 2 (16 -18), phospholipase D (19), Ca 2ϩ channels (20 -22), reactive oxygen and nitrogen species (23,24), and Na ϩ /H ϩ exchange (25,26). In essentially all cases, activation of downstream signaling molecules has been shown to be mediated by heterotrimeric G proteins.
Calmodulin (CaM) is a small (148 amino acids, ϳ17 kDa), soluble protein, which functions as the major calcium sensor in most cells (27). As a prototypical member of the EF-hand family of Ca 2ϩ -binding proteins, CaM can bind up to four Ca 2ϩ ions, which subsequently extend the protein to expose hydrophobic patches capable of binding cellular targets (28,29). These targets number well over one hundred and include enzymes, ion channels, transcription factors, and cytoskeletal proteins. Recently, CaM has been shown to bind to several plasma membrane receptors, including the epidermal growth factor receptor, platelet glycoprotein VI, and some GPCRs (30,31). The first GPCR that was shown to interact with CaM was the metabotropic glutamate subtype 5 receptor, which contains a CaM-binding site in a region of the extended carboxyl terminus of the receptor also known to bind G protein ␤␥ subunits (32). Subsequently, the interaction of CaM with the third intracellular loop of D 2 -dopamine and -opioid receptors was shown to regulate receptor coupling to Pertussis toxin-sensitive heterotrimeric G proteins (33,34), whereas Nickols et al. (35) showed that the interaction of CaM with the V 2 -vasopressin receptor modulates ligand-induced elevations in intracellular calcium. Our group recently showed that CaM interacts with the G i/ocoupled 5-HT 1A receptor at two distinct sites in the receptor third intracellular loop in regions that overlap with protein kinase C phosphorylation sites (36). Interestingly, binding of CaM to synthetic peptides corresponding to these regions prevented phosphorylation of the peptides by protein kinase C. Furthermore, binding of CaM to the 5-HT 1A receptor decreases G protein coupling as assayed by GTP␥S binding to crude membrane preparations. 2 These examples indicate that CaM interactions may play important and diverse roles in GPCR signaling.
CaM has previously been shown to be a major target for 5-HT 2A receptor signaling. Agonist-mediated up-regulation of the 5-HT 2A receptor is dependent upon both CaM and CaM-dependent kinase 2 (37). Berg et al. (38) showed that CaM is required for 5-HT 2A receptor-mediated formation of cAMP in A1A1 cells. Likewise, the CaM-dependent enzymes CaM-dependent kinase 2 and calcineurin (a CaM-dependent phosphatase) play roles in 5-HT-induced cyclooxygenase 2 mRNA expression in renal mesangial cells (39,40). Finally, the 5-HT 2A receptor activates extracellular signal-regulated mitogen-activated protein kinases through the intermediate action of Ca 2ϩ / CaM (1).
Our group previously reported that CaM interacts with the G i/o -coupled 5-HT 1A receptor third intracellular loop at two distinct sites and that the interaction of CaM with those sites may play a role in regulating receptor phosphorylation and desensitization induced by protein kinase C (36). We were interested in establishing whether the G q/11 -coupled 5-HT 2A receptor could also interact with CaM, and if so, what consequences this binding might have on receptor function. In the current work, we report that, in NIH-3T3 fibroblasts, CaM co-immunoprecipitates in an agonist-dependent manner with the 5-HT 2A receptor. A search of the primary sequence revealed that the 5-HT 2A receptor contains two novel CaM-binding motifs, located in the second intracellular loop and the juxtamembrane region of the carboxyl terminus of the receptor. Both motifs contain consensus phosphorylation sites and are important for G protein coupling, indicating that interaction of CaM with those sites could play roles in regulating receptor function. In this report, we sought to characterize the putative CaMbinding sites in the 5-HT 2A receptor and to determine their functional significance.
Immunoprecipitation-Quiescent NIH-3T3 cell monolayers grown on 100-mm dishes were treated with agonist (1 M 5-HT) or vehicle for the appropriate time then lysed in 500 l of modified radioimmune precipitation assay buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 1% Nonidet P-40, 0.5% Triton X-100, 10% glycerol, 1 mM NaF, 1 mM Na 3 VO 4 , 1 mM phenylmethanesulfonyl fluoride, and 1 g/ml each of aprotinin, leupeptin, and pepstatin). Lysates were homogenized, clarified by centrifugation at 14,000 ϫ g for 15 min, then pre-cleared by incubation with protein A/G-agarose for 30 min at 4°C. Pre-cleared lysates were then incubated with commercial anti-5-HT 2A receptor antibody overnight at 4°C. Immunocomplexes were captured by incubation with protein A/G-agarose for 2 h at 4°C, collected by centrifugation, and washed three times with radioimmune precipitation assay buffer and twice with phosphate-buffered saline. Immunoprecipitates were then resuspended in 2ϫ Laemmli sample buffer, boiled for 5 min, and subjected to SDS-PAGE. Gels were analyzed by immunoblot with either anti-CaM antibody or a second anti-5-HT 2A receptor antibody.
Gel Shift Assays-Gel shift analyses of CaM-peptide complexes were performed using urea-polyacrylamide gel electrophoresis, as described by Erickson-Viitanen and Delgrado (41). Reactions (30 l of total volume) containing 300 pmol of CaM (ϳ5 g) and increasing amounts of peptide (0 -3000 pmol) were incubated in 100 mM Tris-HCl, pH 7.5, 4 M urea and either 0.1 mM CaCl 2 or 1 mM EGTA at 22°C for 30 min. 15 l of a 50% glycerol/0.1% bromphenol blue loading buffer was added to each reaction, and the samples were resolved on 14% polyacrylamide gels containing 4 M urea and either 0.1 mM CaCl 2 or 1 mM EGTA in the running buffer. Protein was visualized by staining with Gel-code blue (Pierce) staining reagent.
Blot Overlay Assays-Peptides (1-100 nmol) were immobilized to PVDF membranes by slot blot and washed twice with 100 mM Tris-HCl, pH 7.5. The membranes were blocked with 5% bovine serum albumin in 100 mM Tris-HCl, pH 7.5, containing 0.1% Tween 20 for 1 h at room temperature, and then were incubated with 0.5 g/ml biotinylated CaM in the presence of either 0.1 mM CaCl 2 or 1 mM EGTA overnight at 4°C. The PVDF membranes were then washed 3ϫ in the same buffer without CaM, followed by incubation with alkaline phosphatase-conjugated avidin for 1 h at room temperature. Detection was with a chemiluminescent reagent.
Fluorometric Measurements with Dansyl-CaM-Dansyl-CaM was synthesized according to the method of Bertrand et al. (42). Briefly, 10 mg of CaM was incubated with ϳ1 mg of dansyl chloride for 1 h at 4°C. Dansyl-CaM was purified from unincorporated dye using a Centricon TM concentrator with a molecular mass cutoff of 10,000 Da. Measurement of absorbance at 340 nm (molar extinction coefficient, 3,400 M Ϫ1 cm Ϫ1 ) gave an incorporation of ϳ1.3 dansyl units per CaM molecule. Fluorescence emission spectra of dansyl-CaM were measured from 400 -600 nm using a SLM 8000 TM C spectrofluorometer (AMINCO-Bowman) with an excitation wavelength of 340 nm. Test peptides (0 -2 M) were incubated with dansyl-CaM in 100 mM Tris-HCl, pH 7.5, supplemented with 0.1 mM CaCl 2 for 2 h at room temperature. The concentration of dansyl-CaM (0.1-0.5 M) was varied, and concentration-response curves were generated for fluorescence enhancement at each dansyl-CaM concentration. The apparent K D values for each concentration of dansyl-CaM were fit to the Hill equation by linear regression to calculate true affinities.
In Vitro Kinase Assays-Thirty-five ng (ϳ0.06 unit) of purified rat brain PKC was incubated with increasing concentrations of ct peptide (0 -6 M) in kinase buffer (20 mM MOPS, pH 7.2, 25 mM ␤-glycerolphosphate, 1 mM dithiothreitol, 1 mM CaCl 2 , 0.1 mg/ml phosphatidylserine, 0.01 mg/ml diacylglycerol, 100 M [␥-32 P]ATP) in a total volume of 25 l. In some cases, purified bovine brain calmodulin (0 -20 M) was added to reactions. Assays were started by the addition of 5 Ci of [␥-32 P]ATP, 2 J. H. Turner and J. R. Raymond, unpublished data. and then incubated at 30°C for 1 h. Reactions were then transferred to nitrocellulose squares, washed several times with 0.75% phosphoric acid, followed by a final wash with 100% acetone. Bound radioactivity was measured using liquid scintillation counting.
Preparation of NIH-3T3/5-HT 2A Receptor Cell Membranes-NIH-3T3 cells overexpressing the human 5-HT 2A receptor (8 pmol/mg) were kindly provided by Dr. Elaine Sanders-Bush (Nashville, TN). Cells were grown to confluence in 100-mm dishes, incubated in serum-free medium for 16 -24 h, scraped into lysis buffer (100 mM Tris-HCl, pH 7.4, 5 mM EDTA, 1 g/ml each of aprotinin, leupeptin, and pepstatin), then subjected to homogenization by twenty strokes in a Dounce homogenizer. Lysed cells were centrifuged at 1,000 ϫ g for 10 min to remove whole cells and nuclear debris, and the supernatant was centrifuged at 37,000 ϫ g for 20 min. The resulting pellet was washed twice in lysis buffer and resuspended in 50 mM Tris HCl, pH 7.4, 2.5 mM MgCl 2 at a final protein concentration of ϳ4 mg/ml. Membranes were frozen by immersion in liquid nitrogen, and stored at Ϫ80°C for future use.

Interaction of CaM with the 5-HT 2A Receptor in NIH-3T3
Fibroblasts-Several groups, including ours, have previously shown that CaM plays a role in numerous 5-HT 2A receptor signaling pathways, including the activation of mitogen-activated protein kinases (1) and Na ϩ /H ϩ exchange (26,45). As a receptor that couples to G q/ll type proteins, the 5-HT 2A receptor is capable of stimulating phosphoinositide turnover and to subsequently increase intracellular Ca 2ϩ levels. We wondered whether the 5-HT 2A receptor might exert some of its Ca 2ϩsensitive and/or -insensitive intracellular effects by directly interacting with CaM. Other GPCRs, including the D 2 dopamine (32), -opioid receptors (34), and 5-HT 1A receptor (36), have been shown to directly interact with CaM through their third intracellular loops. We treated wild-type NIH-3T3 cells and NIH-3T3 cells overexpressing the 5-HT 2A receptor, with 1 M serotonin for 5 min, and immunoprecipitated the receptor with a commercial 5-HT 2A receptor antibody. As shown in Fig. 1, a CaM-specific mouse monoclonal antibody detected a band at ϳ19 kDa in NIH-3T3-5HT 2A R cell immunoprecipitates, which corresponds to the Ca 2ϩ -bound form of CaM. This immunoreactive band was only moderately detectable in immunoprecipitates from untreated NIH-3T3/5-HT 2A R cells, but was significantly increased in immunoprecipitates from cells treated with 5-HT. A second anti-5-HT 2A R antibody detected an immunoreactive band at ϳ60 kDa in immunoprecipitates from receptorexpressing NIH-3T3 cells that was unchanged after treatment with 5-HT. No immunoreactive bands were detected with either antibody in immunoprecipitates from wild-type NIH-3T3 cells. These data suggest that the 5-HT 2A receptor can complex with CaM in an agonist-dependent manner in NIH-3T3 fibroblasts.
Identification of Putative CaM Binding Regions in the 5-HT 2A Receptor-NMR studies have shown that CaM binds target peptides via hydrophobic interactions and via salt bridges typically involving glutamate residues in the EF-hand regions (47,48). Predictably, identified CaM binding regions conform to short peptides that form amphipathic ␣-helices composed of hydrophobic and positively charged amino acid residues (49). Using a web-based computer search program, which identifies such regions based on evaluation criteria such as hydropathy, ␣-helical propensity, residue charge, helical class, residue weight, and hydrophobic residue content (50) To illustrate the amphipathic nature of the putative CaMbinding sequences of the 5-HT 2A receptor, we created helical wheel diagrams using a web-based modeling program (50). As expected, like most CaM-binding sites, both helical wheels showed clusters of positively charged amino acids on one side of the ␣-helix, with mostly hydrophobic amino acids concentrated on the opposite side (data not shown).
CaM Binds to Peptides Derived from the Second Intracellular Loop and Carboxyl Terminus of the 5-HT 2A Receptor-Using synthetic peptides encompassing amino acids 183-202 (i2) and 377-396 (ct) of the 5-HT 2A receptor, we tested the ability of either or both regions to interact with CaM with a modified blot overlay technique. Increasing amounts of each peptide (1-100 nmol) were slot-blotted to PVDF membranes and incubated with biotinylated CaM. To separate specific from nonspecific binding, we used a negative control peptide corresponding to the 17-amino acid CaM binding region of myosin light chain kinase, which contains a point mutation that renders it unable to bind CaM. Both the i2 and ct peptide bound biotinylated CaM in buffer containing 0.1 mM Ca 2ϩ , whereas the myosin light chain kinase control peptide showed no binding (Fig. 3). Binding was significantly reduced when Ca 2ϩ was removed from the buffer and replaced with 1 mM EGTA. Interestingly, the i2 peptide appeared to bind more readily to CaM than did the ct peptide, indicating that, although both peptides bind CaM in a Ca 2ϩ -sensitive manner, they may do so with differing affinities.
We next analyzed the peptide-CaM complexes using polyacrylamide gel electrophoresis in the presence of 4 M urea. This technique eliminates lower affinity and nonspecific proteinprotein interactions (K D Ͼ 100 nM). Increasing amounts of peptide (75-3000 pmol) were incubated with a constant amount of CaM (300 pmol) and subsequently analyzed by nondenaturing gel electrophoresis. In the presence of Ca 2ϩ , the addition of the 5-HT 2A i2 peptide produced an upward shift in the migration of CaM (Fig. 4A). In contrast, no peptide-CaM complexes were formed when Ca 2ϩ was chelated with EGTA. The ct peptide produced no readily apparent shift in the mobility of CaM (Fig. 4B). This suggests that the ct peptide probably binds CaM with an affinity weaker than 100 nM, however the density of the CaM band decreased consistently with increasing peptide concentration, indicating that some shift may have occurred without detection. The incomplete gel shift induced by the i2 peptide, even at a peptide-CaM molar ratio of 10:1, is in line with published reports of other CaM binding regions that utilized this method under the same conditions (33,36,52). Likewise, incubation with a peptide derived from the CaM binding region of myosin light chain kinase, which binds CaM with very high affinity (ϳ6 pM), was also unable to induce a complete shift in CaM mobility in our hands (Fig. 4C).
Interaction of 5-HT 2A i2 and ct Peptides with Dansyl-CaM-Precise affinities are often difficult to calculate using gel-shift assays or in vitro binding techniques. Thus, we evaluated the binding affinities of the i2 and ct peptides for CaM by measuring changes in the fluorescence emission spectrum of dansyl-CaM. Ligand binding to dansyl-CaM enhances the fluorescence emission and blue-shifts the emission peak to a lower wavelength. As shown in Fig. 5A, dansyl-CaM displays weak fluorescence in the absence of Ca 2ϩ , with an emission peak of ϳ520 nm. In the presence of Ca 2ϩ , the fluorescence emission is en- hanced and shifted to a lower wavelength. As expected, both i2 and ct peptides further enhanced dansyl-CaM fluorescence and shifted the emission peak to ϳ495 nm, whereas the negative control myosin light chain kinase mutant peptide was completely ineffective (data not shown). The amounts of i2 and ct peptides were then varied to generate typical concentration-de-pendent saturation curves. The fluorescence emission spectra of dansyl-CaM (0.1, 0.2, 0.3, 0.4, and 0.5 M) were measured in the presence of increasing concentrations of i2 or ct peptides (0 -2 M), and resulting data points were fit by nonlinear least squares analysis to the Scatchard one-site binding equation (Fig. 5, B and C). As expected, higher concentrations of dansyl-CaM required increased concentrations of peptide to enhance fluorescence and reach saturation, resulting in variations in EC 50 measurements. This results from a depletion of free peptide, a phenomenon which is most apparent at low concentrations of dansyl-CaM. By calculating EC 50 values over a range of dansyl-CaM concentrations it was possible to extrapolate to an infinitely low dansyl-CaM concentration, at which point depletion is essentially nonexistent. As shown in Fig. 5D and 5E, EC 50 measurements derived from peptide binding curves were plotted against the concentration of dansyl-CaM. These values fell onto straight lines with slopes of close to one, indicative of a 1:1 binding stoichiometry. Extrapolating the lines to the y-axis gave "true affinities" of 65 Ϯ 30 nM and 168 Ϯ 38 nM for the i2 and ct peptides, respectively.

Functional Overlap of CaM-binding Domain and PKC Phosphorylation Site within the 5-HT 2A Receptor ct Peptide-
Most GPCRs can be modulated by kinase-directed phosphorylation of threonine and serine residues, with the second messenger-dependent kinases (PKA, PKC, and others) predominating at low agonist concentrations, and the G proteincoupled receptor kinases predominating at high agonist concentrations and over extended periods of time (53). Although direct phosphorylation of the 5-HT 2A receptor has yet to be reported, both 5-HT 2A receptor internalization and desensitization appear to be highly dependent on PKC activation (54,55). The 5-HT 2A receptor contains six putative PKC phosphorylation sites, five of which are localized to the receptor third intracellular loop. Interestingly, the only putative PKC phosphorylation site in the carboxyl terminus ( 384 NKTYR 385 ) is centrally located within the ct CaM binding region (Fig. 6A). We consequently wondered whether binding of CaM to this region might regulate phosphorylation of this residue by PKC and, conversely, whether phosphorylation of this residue might inhibit CaM binding. Because typical CaM binding regions contain few or no acidic (negatively charged) amino acids, phosphorylation of the target threonine residue in the ct CaM binding region would add a negative charge and would consequently be expected to perturb CaM binding. We tested this hypothesis by evaluating the fluorescence emission spectrum of dansyl-CaM in the presence of a synthetic peptide identical to the ct peptide, but which contains a phosphorylated threonine at residue 386 (ct-P). As shown in Fig. 6B, the increase in fluorescence induced by the phosphorylated peptide was highly diminished as compared with the shift induced by the unphosphorylated peptide, suggesting that phosphorylation of the ct peptide inhibits CaM binding.
We next assessed whether the peptide could actually be phosphorylated by PKC in vitro, and if so, whether this phosphorylation could be inhibited by CaM. Purified rat brain PKC (composed primarily of PKC-␣ and PKC-␤ isozymes) readily phosphorylated the ct peptide (K m ϭ 8.83 Ϯ 2.31 M, V max ϭ 48.85 Ϯ 7.94 nmol/min/mg). In contrast, phosphorylation of the 5-HT 2A receptor i2 peptide was not observed under identical conditions (data not shown). Interestingly, phosphorylation of the ct peptide was dose-dependently decreased in the presence of CaM, with a 20 M concentration of CaM able to inhibit ct phosphorylation by ϳ80%. These data suggest that phosphorylation of the ct peptide by PKC and peptide binding to CaM are likely to be mutually inhibitory. the 5-HT 2A receptor. The use of an artificial receptor expression system was essential to achieve an adequate signal-tonoise ratio, due largely to the fact that G␣ q/ll proteins are relatively low in abundance, and have much lower rates of nucleotide exchange than G␣ i/o proteins. Stimulation of NIH-3T3 membranes with 10 M 5-HT produced an ϳ25% increase in [ 35 S]GTP␥S binding (Fig. 7A). Interestingly, CaM dose-dependently decreased [ 35 S]GTP␥S binding, with concentrations of 1-10 M reducing [ 35 S]GTP␥S binding to basal levels or lower. This effect was not due to nonspecific effects of the CaM protein, in that membranes incubated with the Ca 2ϩ -binding protein S-100 (which has a similar molecular weight to CaM but different target binding proteins) did not attenuate the 5-HT-induced increase in [ 35 S]GTP␥S binding (data not shown). These data suggest that interaction of CaM with the 5-HT 2A receptor may regulate receptor coupling to heterotrimeric G proteins. However, it remained unclear whether this effect was via CaM binding to the second intracellular loop, carboxyl terminus, or both regions of the receptor.
Previous studies have shown that synthetic peptides derived from specific receptor-G protein interface sites can mimic the activated receptors themselves. This methodology has been used extensively to characterize G protein interaction domains for multiple receptors, including those for angiotensin II (60), glucagon-like peptide (57), and dopamine (61). We investigated the ability of the 5-HT 2A receptor CaM-binding domains to stimulate G protein coupling by measuring [ 35 S]GTP␥S binding to NIH-3T3 cell membranes in the presence of increasing concentrations of i2 or ct peptides. As shown in Fig. 7B, the i2 peptide strongly stimulated [ 35 S]GTP␥S binding, whereas the ct peptide had no effect. These data suggest that the second intracellular loop of the 5-HT 2A receptor is largely responsible for receptor coupling to heterotrimeric G proteins. Likewise, it suggests that the ability of CaM to inhibit G protein coupling is likely mediated by binding to the high affinity CaM-binding domain in the second intracellular loop, rather than to the lower affinity site in the carboxyl terminus.

DISCUSSION
What is new about this work is that we have characterized two putative CaM-binding domains in the 5-HT 2A receptor, and have shown that the interaction of CaM with these domains likely attenuates G protein coupling to the second intracellular loop and phosphorylation of the carboxyl terminus of the receptor by PKC. To our knowledge, this is the first demonstration that CaM attenuates G protein coupling of a G␣ q/11 -preferring receptor. We showed that CaM co-immunoprecipitates with the 5-HT 2A receptor from NIH-3T3 fibroblasts and that the amount of CaM co-immunoprecipitating with the receptor was increased after treatment with 5-HT, suggesting that CaM association is dependent on receptor activation. We subsequently identified two putative CaM-binding sites located within the second intracellular loop and carboxyl terminus of the receptor. These sites have properties typical of CaM-binding domains, including putative amphipathic ␣-helical structures composed of basic and hydrophobic amino acid residues. In addition, both sites could be aligned with well established CaM-binding motifs based on the locations of key hydrophobic residues. Synthetic peptides derived from the i2 and ct regions bound to CaM in a Ca 2ϩ -sensitive manner with average K D values of ϳ65 and ϳ168 nM, respectively. Interestingly, many GPCRs appear to contain one or more putative CaM-binding sites, suggesting that interaction of CaM with different receptor regions may represent a common mechanism through which GPCR-dependent signaling pathways may be regulated ( Table I).
The functions of the 5-HT 2A receptor CaM-binding domains have not been completely elucidated. However, these regions have previously been implicated in numerous signaling pathways and processes. The importance of the second intracellular loop in G protein coupling has been experimentally confirmed for other GPCRs, including the bradykinin B 2 (62), ␣ 2A adrenergic (63), and type B endothelin receptors (64). Using molecular modeling, Prado et al. (62) produced evidence that the distal i2 loop and proximal carboxyl terminus of the bradykinin B 2 receptor may interact with each other to mediate receptor internalization and resensitization, in addition to G protein coupling. Berg et al. (65) showed that the i2 loop of the 5-HT 2C receptor is important for G␣ q coupling, and that different isoforms of the receptor produced via RNA editing in regions of the second intracellular loop varied in their ability to induce second messengers.
Previous work has demonstrated that CaM binding to the receptor third intracellular loops can attenuate the interactions between G␣ i and several receptors, including the 5-HT 1A receptor (36), -opioid receptor (34), D 2 dopamine receptor (33), and the group III metabotropic glutamate receptor, mGluR7a (66,67). New evidence in the current report suggests that CaM binding to the second intracellular loop of the 5-HT 2A receptor likely modulates coupling to heterotrimeric G proteins, most likely G␣ q/11 and/or G␣ 12/13 subunits. The affinity of the second intracellular loop of the 5-HT 2A receptor for CaM (65 Ϯ 30 nM; Fig. 6) is consistent with the ability of 1-10 M CaM to completely attenuate 5-HT 2A receptor-mediated increases in [ 35 S]GTP␥S binding (Fig. 7) and supports the possibility that bona fide interactions of CaM with the 5-HT 2A receptor can occur within the context of the cell. Thus, CaM is likely to play a role in attenuating signal initiation from the 5-HT 2A receptor second intracellular loop by dampening interactions with heterotrimeric G proteins.
One general mechanism of receptor desensitization involves phosphorylation of intracellular residues by second messengers such as PKC, PKA, and the G protein-coupled receptor kinases. Our group previously showed that CaM binding to synthetic The Swiss-Prot protein sequences for members of the human serotonin, adrenergic, and muscarinic GPCR families were entered into a CaM binding site search engine (calcium.uhnres.utoronto.ca/ctdb/pub_pages/search/index.htm). Putative CaM-binding motifs were identified in juxtamembrane regions of the second intracellular loops, third intracellular loops, and/or carboxyl termini of the receptors. As evidenced in the table, the presence and location of putative CaM-binding motifs is a common characteristic of GPCRs and is largely receptor subtype-dependent. peptides derived from the 5-HT 1A receptor third intracellular loop could inhibit phosphorylation by protein kinase C (36 (55) showed that agonist-induced desensitization of the 5-HT 2A receptor was dependent on the activity of PKC and could be modulated by pharmacological inhibitors of CaM. Likewise, activation of PKC in HEK293 cells induced the internalization and recycling of a 5-HT 2A receptor green fluorescent protein fusion protein (54). The current work also supports a potential role for CaM in modulating phosphorylation of the 5-HT 2A receptor. We showed for the first time that the putative PKC phosphorylation site in the 5-HT 2A receptor carboxyl terminus is an excellent PKC substrate in vitro. Furthermore, CaM was capable of concentration-dependently decreasing the ability of PKC to phosphorylate a peptide derived from the ct region. The apparent agonist-dependent binding of CaM to the 5-HT 2A receptor carboxyl terminus suggests a possible mechanism by which this receptor may be partially protected from PKC-dependent phosphorylation and perhaps subsequent internalization and desensitization. Furthermore, we speculate that the failure of several groups to demonstrate PKC-mediated phosphorylation of the 5-HT 2A receptor in vivo may be due, in part, to steric blockade of PKC sites by CaM in live cells. This suggests that a complex contribution of both CaM and PKC may regulate 5-HT 2A receptor internalization and desensitization.
Mutation of Ser 188 in the second intracellular loop has been shown to attenuate 5-HT 2A receptor desensitization (68). Interestingly, this residue is predicted to be a phosphorylation site for the dual-specificity kinase Clk-2. The S188A mutation also caused a 3-fold increase in the EC 50 and a 40% reduction in 5-HT efficacy, suggesting that post-translational modification and/or accessory protein binding to this region might directly disrupt G protein coupling, leading to internalization-independent desensitization. The second intracellular loop of the 5-HT 2A receptor also contains a DRYXX(I/V)XXP motif similar to that identified in the human muscarinic cholinergic receptor, which has been proposed as an important regulator of receptor internalization, whereas the membrane-proximal region of the 5-HT 2A receptor carboxyl terminus has tyrosine-containing motifs previously shown to mediate internalization of GPCRs for neurokinin 1 (69), parathyroid hormone (70), and angiotensin II (71). Such tyrosine-based motifs have been shown to form clathrin-associated protein complexes. CaM has been found to play a role in the general process of receptor endocytosis and trafficking in yeast (72) and human epithelial cells (73). Our group reported a role for CaM in 5-HT 1A receptor internalization, a required step for MEK and subsequent ERK activation (74). Conversely, pharmacological inhibitors of CaM were found to inhibit recycling and degradation of the EGF receptor without affecting internalization, resulting in the accumulation of receptors in enlarged endosomal structures (75). Thus, it is conceivable that CaM could regulate 5-HT 2A receptor internalization and/or trafficking through direct binding to the receptor.
The current work has significance that extends beyond the 5-HT receptor family in that many GPCRs possess putative CaM binding domains (Table I) in intracellular juxtamembrane regions of the second and third intracellular loops and carboxyl termini. Thus far, these motifs have been implicated in G protein coupling and PKC-induced phosphorylation of both G␣ i -and G␣ q/11 -coupled receptors, for which CaM appears to dampen both processes. Because CaM has been implicated in various other GPCR processes such as ERK activation, stimulation of Na ϩ /H ϩ exchange, and receptor trafficking (1,26,(72)(73)(74)(75)(76)(77), direct binding of CaM to GPCRs could modulate any number of receptor processes. Additionally, the presence of putative CaM binding motifs in G s -coupled receptors such as the ␤ 1 -, ␤ 2 -, and ␤ 3 -adrenergic and 5-HT 4 and 5-HT 7 receptors (Table I) raises the possibility that CaM might influence cAMP accumulation by directly binding to those receptors, although that possibility remains to be verified experimentally.
In conclusion, we have identified for the first time the presence of CaM-binding sites in the 5-HT 2A receptor. These sites reside in juxtamembrane regions of the second intracellular loop and carboxyl terminus of the receptor. To our knowledge, this represents the first example of a GPCR with a CaMbinding site in the second intracellular loop and the first demonstration that CaM binding can attenuate G protein coupling to a G␣ q/11 -preferring receptor. In addition, the work extends the concept that CaM binding to GPCR can inhibit receptor phosphorylation by PKC. These results suggest that CaM may serve as a valuable regulator and/or second messenger of 5-HT 2A receptor function.