Creation, Expression, andCharacterization of a Constitutively

Active Mutant of the Human Serotonin5-HT6 Receptor

ANIL PUROHIT, KATHARINE HERRICK-DAVIS, AND MILT TEITLER*Center for Neuropharmacology and Neuroscience, Albany Medical College, Albany, New York 12208

KEY WORDS serotonin receptors; constitutive activity; G-protein coupled receptors;antipsychotic; clozapine; inverse agonist

ABSTRACT The serotonin 5-HT6 receptor, a G-protein-coupled receptor, displayshigh affinity for antipsychotic, antidepressant, and psychotropic drugs. We created aconstitutively active form of the human 5-HT6 receptor in order to probe the moleculardomains of receptor activation and to determine if inverse agonist activities of antipsy-chotic drugs contribute to their clinical profile. Previous studies from our laboratorysupport a critical role for the c-terminal region of the third intracellular loop (il3) in theactivation of Gq-coupled serotonin receptors. In the present study, PCR-based mutagen-esis was used to mutate serine 267 (S6.34) in the c-terminal region of il3 to lysine(S267K). The native and S267K 5-HT6 receptors were expressed in COS-7 cells to studythe functional effects of the mutation. The S267K receptor shows 10-fold higher affinityfor serotonin than the native receptor and demonstrates agonist-independent activity.Clozapine decreased the basal activity of the S267K receptor to vector control levels.Therefore, we can conclude that the S267K mutation renders the 5-HT6 receptor con-stitutively active and that clozapine is an inverse agonist at the mutant 5-HT6 receptor.These results indicate that the c-terminal region of il3 of the Gs-coupled 5-HT6 receptoris a key domain for G-protein coupling, similar to the Gq-coupled 5-HT receptors. Theinverse agonist action of clozapine indicates that drugs displaying competitive antago-nist activity at native 5-HT6 receptors may display inverse agonist activity at theconstitutively activated form of the receptor. Synapse 47:218–224, 2003.© 2002 Wiley-Liss, Inc.

INTRODUCTION

The serotonin 5-HT6 receptor couples to Gs and in-creases intracellular cAMP when stimulated (Kohen etal., 1996; Ruat et al., 1993; Monsma et al., 1993). In therat brain, 5-HT6 receptor mRNA is expressed in highlevels in the striatum (putamen), nucleus accumbensand hippocampus (Monsma et al., 1993; Ward andDorsa, 1996; Ward et al., 1995; Gerard et al., 1996).The pattern of 5-HT6 receptor mRNA expression in thehuman brain is similar to the rat brain (Kohen et al.,1996). 5-HT6 receptors have been shown to function aspostsynaptic receptors (Gerard et al., 1996). The phys-iological roles of 5-HT6 receptors are not known, al-though recent studies indicate that the 5-HT6 receptorcould be involved in modulating the activity of braincholinergic systems (Bourson et al., 1998). Interest-ingly, the 5-HT6 receptor has high affinity for severalantipsychotic drugs, including clozapine, the classical

atypical antipsychotic drug (Kohen et al., 1996; Roth etal., 1994).

Classically, G-protein-coupled receptors (GPCR) areconsidered to be inactive in the absence of an agonistand activate signal transduction pathways only in thepresence of an agonist. GPCR have the potential todisplay constitutive activity (receptor activation in theabsence of agonist) and inverse agonists decrease the

Abbreviations: GPCR, G-protein coupled receptors; il3, third intracellularloop; 5-HT, serotonin or 5-hydroxytryptamine; LSD, lysergic acid diethylamide;PCR, polymerase chain reaction.

Contract grant sponsor: PHS; Contract grant numbers: MH56650 (to M.T.),MH57019 (to K.H.D).

*Correspondence to: Dr. Milt Teitler, Center for Neuropharmacology andNeuroscience, Albany Medical College, 47 New Scotland Avenue, Albany, NY12208. E-mail: teitlem@mail.amc.edu

Received 8 July 2002; Accepted 31 August 2002

DOI 10.1002/syn.10157

SYNAPSE 47:218–224 (2003)

© 2002 WILEY-LISS, INC.

basal activity of constitutively active GPCR. Lefkowitzand colleagues (Kjelsberg et al., 1992) demonstratedthat mutation of a single amino acid in the c-terminalregion of the third intracellular loop (il3) could renderthe �1B-adrenergic receptor constitutively active. Ac-cording to the extended ternary complex model theconstitutively active receptors represent the agonist-activated state of the receptor (Samama et al., 1993).Presumably, the activation of a GPCR by agonists ormutations involves a similar conformational change inthe critical domains of the GPCR involved in signaltransduction. In several other GPCR, including 5-HT2A

and 5-HT2C receptors, mutations of analogous aminoacids in the c-terminal region of il3 produce constitu-tively active receptors (Egan et al., 1998; Herrick-Daviset al., 1997). The 5-HT6 receptor, like the 5-HT2A re-ceptor, has a short il3 and a long c-terminal tail, butunlike the 5-HT2A receptor (which couples to Gq), the5-HT6 receptor belongs to the Gs-coupled subfamily ofGPCR. Another interesting aspect of the 5-HT6 recep-tor involves a developing model concerning the essen-tial role of an ionic interaction between a critical glu-tamate residue (E6.30) in il3 and an arginine (R3.50) inthe conserved DRY region in maintaining an inactivestate of monoamine GPCR (Shapiro et al., 2002; Palc-zewski et al., 2000; Ballesteros et al., 2001). The 5-HT6

receptor lacks the critical glutamate and expresses analanine in the analogous position (A6.30). Thus, themodel predicts that the native 5-HT6 receptor shouldbe constitutively active. The model also includes animportant interaction involving a hydrogen bond be-tween R3.50 and a serine (S6.34) or threonine (T6.34)in il3 at the analogous position to S267 in the 5-HT6

receptor. Therefore, the present study was designed todetermine if the native 5-HT6 receptor displays consti-tutive activity and if the c-terminal region of il3 is a keymolecular domain involved in 5-HT6 receptor activa-tion. Based on a review of the literature, we hypothe-sized that disturbance of the predicted hydrogen bondbetween R3.50 and S6.34 should result in constitutiveactivation of the 5-HT6 receptor. We tested this hypoth-esis by mutating S267 (S6.34) in the human 5-HT6

receptor. While mutating R3.50 would seem to be aviable aim, mutating this amino acid can impair recep-tor expression and signaling (Alewijnse et al., 2000;Scheer et al., 2000).

MATERIALS AND METHODSMaterials

COS-7 cells were purchased from ATCC (AmericanType Culture Collection, Rockville, MD). Chemicalsand reagents were purchased as follows: 3H-LSD and3H-5-HT (NEN Life Science Products, Boston, MA);5-HT, Ro-20-1724, ascorbic acid, and Clozapine (SigmaChemical Co., St. Louis, MO); Dulbecco’s modified Ea-gle’s medium (DMEM), fetal bovine serum (FBS), andLipofectamine (Life Technologies, Gaithersburg, MD);

Pfu Turbo (Stratagene, La Jolla, CA); pCDNA3 vector(Invitrogen, La Jolla, CA); Qia-quik miniprep kit (Qia-gen, Chatsworth, CA); Ecoscint cocktail (National Di-agnostics, Manville, NJ). Oligonucleotides were pur-chased from Sigma Genosys. The human 5-HT6

receptor cDNA was generously provided by Dr. BryanRoth (Case Western Reserve University, Cleveland,OH).

Mutagenesis

The GPCR numbering scheme has been previously de-scribed (Ballesteros and Weinstein, 1995). Site-directedmutagenesis was performed on the human 5-HT6 recep-tor (in the pCDNA3 expression vector) to mutate serine267 to lysine. Primers (sense 5�-GGAAGGCCCT-GAAGGCCAAGCTTACGCTGGGCATCCTGC-3� andantisense 5�-GCAGGATGCCCAGCGTAAGCTTGGC-CTTCAGGGCCTTCC-3�) were designed to mutateserine (AGC) to lysine (AAG). A Hind III restriction site(A!AGCTT) was added by changing the wobble base ofthe adjacent leucine from CTG to CTT. PCR was per-formed using Pfu Turbo DNA Polymerase enzyme. ThePCR-amplified DNA was transformed into DH5� ac-cording to the manufacturer’s protocol. Plasmid DNAwas purified from DH5� using the Qia-quick miniprepkit. The presence of the desired mutation was verifiedby restriction digest and then confirmed by DNA se-quencing (Sanger et al., 1977).

Cell culture and transfection

COS-7 cells were cultured in DMEM supplementedwith 10% FBS in 5% CO2 at 37°C. Twenty-four hoursprior to transfection cells were seeded at 80% conflu-ence in 100-mm dishes for radioligand binding assaysor 105 cells/well in 24-well cluster plates for cAMPassays. Cells were transfected with vector DNA(pCDNA3), native or mutant 5-HT6 receptor cDNAsusing 3 �l of Lipofectamine and 0.2 �g plasmid DNAper well for 24-well cluster plates and 30 �l Lipo-fectamine and 2 �g of DNA per 100 mm dish.

Radioligand binding

Thirty-six hours after transfection, membranes wereprepared from COS-7 cells by scraping in 50 mM Tris-HCl/5 mM MgCl2/0.5 mM EDTA, pH 7.4 (tissue buffer),and centrifuged at 12,000g for 30 min. Membraneswere resuspended in tissue buffer, homogenized bypolytron, and centrifuged as above. The membrane pel-let was resuspended in tissue buffer and 0.5 ml ali-quots were added to 0.5 ml of assay buffer (tissuebuffer � 0.2% ascorbic acid) containing 2 nM 3H-LSD.For saturation assays, 3H-LSD concentration was var-ied from 0.25–25 nM and 3H-5-HT concentration wasvaried from 0.5–100 nM. Clozapine (10 �M) was usedto determine nonspecific binding. Specific bindingranged from 80–90%. Samples were incubated for 1 h

CONSTITUTIVELY ACTIVE 5-HT6 RECEPTORS 219

at room temperature, filtered on a Brandel Cell Har-vester, and counted in 5 ml Ecoscint cocktail in a Beck-man liquid scintillation counter at 40% efficiency. Spe-cific activities of 3H-LSD and 3H-5-HT were 83 Ci/mmoland 23.7 Ci/mmol, respectively. Data analyses wereperformed using nonlinear regression analysis of thesaturation data to calculate KD and Bmax values(GraphPad Prism, San Diego, CA).

Intracellular cAMP assays

Twenty-four hours post-transfection cells were incu-bated in serum-free media overnight to eliminate ef-fects of residual 5-HT present in the serum. On the dayof assay cells were washed with PBS and incubated for10 min in serum-free assay media containing 0.25 mMRo-20-1724 (a phosphodiesterase inhibitor). Drugswere added to the assay media at a final concentrationof 10 �M, or various concentrations of 5-HT (0.1 nM to10 �M) and cells were incubated for 30 min at 37°C and5% CO2. Cells were lysed with 200 �l 1N KOH, neu-tralized with 200 �l 0.9N HCl, and 20 �l of cell lysatewas incubated with 2.5 nM 3H-cAMP and cyclic-AMPbinding protein (purified from calf adrenal glands) for2.5 h. In parallel, known amounts of nonradioactivecAMP ranging from 0.1–30 pmol were incubated togenerate a standard curve. The amount of cAMP pro-duced in each sample was calculated from the standardcurve using computer-assisted nonlinear regressionanalysis (GraphPad Prism).

RESULTSCreation and expression of a mutant

S267K 5-HT6 receptor

Using PCR-based mutagenesis, serine 267 (S6.34) inthe human 5-HT6 receptor was mutated to lysine(S267K). COS-7 cells were transfected with native ormutant 5-HT6 receptor cDNA in order to characterizethe functional effect(s) of the mutation. LSD has beenshown to have high affinity for rat and human 5-HT6

receptors and radiolabeled forms (I125-LSD, 3H-LSD)have been used to measure receptor expression(Monsma et al., 1993; Roth et al., 1994). As shown inFig. 1, membranes prepared from COS-7 cells express-ing native and mutant 5-HT6 receptors display a singleclass of high affinity, saturable sites for 3H-LSD. Non-linear regression analyses revealed similar Bmax (na-tive � 8.2 � 0.5 pmol/mg protein; S267K � 6.8 � 0.8pmol/mg protein) and KD values (native � 8.6 � 0.6nM; S267K � 8.9 � 0.4 nM) for native and mutantreceptors. Data represent the mean and SEM of threeindependent transfections, each experimental datapoint determined in triplicate. Our results are in agree-ment with the published values for the native 5-HT6

receptor (Kohen et al., 1996; Roth et al., 1994).3H-5-HT saturation studies performed on mem-

branes prepared from transfected COS-7 cells revealed

saturable 3H-5-HT binding to native and S267K recep-tors (Fig. 2). Nonlinear regression analysis of the 3H-5-HT saturation curves produced Bmax and KD valuesof 1.3 � 0.1 pmol/mg and 35 � 1.3 nM, respectively, fornative receptors, and 3.3 � 0.5 pmol/mg and 14 � 0.9nM, respectively, for S267K receptors. The 3H-5-HTBmax and KD values for S267K receptors were signifi-

Fig. 1. Saturable 3H-LSD binding to native and S267K 5-HT6receptors. Membranes were prepared from COS-7 cells expressingnative or S267K receptors. The graphs shown are from a singlerepresentative experiment, while Bmax and KD values are the meanand SEM of three independent transfections, each experimental datapoint determined in triplicate. Native receptor: Bmax � 8.2 � 0.5pmol/mg; KD � 8.6 � 0.6 nM. S267K receptor: Bmax � 6.8 � 0.8pmol/mg; KD � 8.9 � 0.4 nM. Data analyses were performed usingnonlinear regression analysis of the saturation data to calculate KDand Bmax values.

Fig. 2. Saturable 3H-5-HT binding to native and S267K 5-HT6receptors. Membranes were prepared from COS-7 cells expressingnative or S267K receptors. The graphs shown are from a singlerepresentative experiment, while Bmax and KD values are the meanand SEM of three independent transfections, each experimental datapoint determined in triplicate. Native receptor: 3H-5-HT Bmax � 1.3 �0.1 pmol/mg; KD � 35 � 1.3 nM. S267K receptor: Bmax � 3.3 � 0.5pmol/mg, KD � 14 � 0.9 nM. Data analyses were performed usingnonlinear regression analysis of the saturation data to calculate KDand Bmax values. P � 0.01 for native vs. S267K 3H-5-HT Bmax, andnative vs. S267K 3H-5-HT KD (Student’s t-test).

220 A. PUROHIT ET AL.

cantly different (P � 0.01, Student’s t-test) from theBmax and KD values for native 5-HT6 receptors (TableI). The increase in 3H-5-HT Bmax and the increase in5-HT affinity for S267K receptors are consistent with aconstitutively active form of the receptor.

5-HT competition for 3H-LSD-labeled receptors3H-LSD was used as a radiolabel in 5-HT competi-

tion experiments (Fig. 3). The S267K mutation causeda 10-fold leftward shift in the 5-HT competition curve,indicating a 10-fold increase in 5-HT affinity for themutant receptor. 5-HT Ki values for native and S267Kreceptors were determined to be 33 � 3.5 nM and 3.3 �0.9 nM, respectively.

Stimulation of intracellular cAMPaccumulation

To determine the effect of the S267K mutation on theability to activate adenylate cyclase, cAMP productionwas measured in COS-7 cells expressing native orS267K receptors. In order to measure the amount ofcAMP produced following 5-HT6 receptor activation,endogenous levels of cAMP produced in COS-7 cellstransfected with vector alone (31 � 7 pmol/105 cells)were subtracted from the amount of cAMP measured in

cells expressing native or S267K receptors (Table II).The EC50 values for 5-HT stimulation of cAMP produc-tion were similar for native and S267K receptors (Fig.4). This was somewhat unexpected, as generally theconstitutively active form of the receptor is more po-tently stimulated by the agonist (see Discussion). Bothnative and S267K receptors showed similar maximalresponses to 10 �M 5-HT (Fig. 5). cAMP production inresponse to 10 �M 5-HT was 219 � 19 pmol/105 cellsfor native receptors and 176 � 15 pmol/105 cells forS267K receptors. While the mutation did not alter theability of the receptor to produce a maximal cAMPresponse, 5-HT-stimulated cAMP production was 18-

TABLE I. 3H-LSD and 3H-5-HT saturation studies

Receptor LigandKD

(nM)Bmax

(pmol/mg) % Bmax

Native 5-HT6 receptor 3H-LSD 8.6 � 0.6 8.2 � 0.5 100 � 6Native 5-HT6 receptor 3H-5-HT 35 � 1.3 1.3 � 0.1 16 � 1.2S267K 5-HT6 receptor 3H-LSD 8.9 � 0.4 6.8 � 0.8 100 � 11S267K 5-HT6 receptor 3H-5-HT 14 � 0.9* 3.3 � 0.5** 49 � 7.4

Radioligand binding saturation experiments were performed in COS-7 cells expressing native or S267K receptors. Datarepresent the mean and SEM from three independent experiments (independent transfections), each performed intriplicate. *P � 0.01 vs. native 3H-5-HT KD; **P � 0.01 vs. native 3H-5-HT Bmax (Student’s t-test).

Fig. 3. 5-HT competition isotherms for 3H-LSD-labeled 5-HT6 re-ceptors. Membranes were prepared from COS-7 cells expressing na-tive or S267K receptors. Data are the mean and SEM three indepen-dent transfections, each experimental data point determined intriplicate. Native 5-HT Ki � 33 � 3.5 nM. S267K 5-HT Ki � 3.3 � 0.9nM.

TABLE II. Constitutive activity of the S267K 5-HT6 receptor andinverse agonist activity of clozapine

Receptor anddrug treatment

cAMP(pmol/105 cells)

Fold-increaseover basal n

Native 5-HT6 basal 12 � 9 — 4Native 5-HT6 � 5-HT 219 � 19* 18 4S267K 5-HT6 basal 58 � 17* 4.8 4S267K 5-HT6 � 5-HT 176 � 15** 3.0 3S267K � clozapine �5 � 5** — 3

cAMP production was measured in COS-7 cells expressing native or S267Kreceptors in the absence or presence of 10 �M 5-HT or clozapine. cAMP produc-tion in cells transfected with vector alone (31 � 7 pmol/105 cells) was subtractedfrom each value. n represents the number of independent experiments (indepen-dent transfections) each performed in triplicate. *P � 0.01 vs. native basal;**P � 0.01 vs. S267K basal (ANOVA with post hoc SNK comparisons).

Fig. 4. 5-HT dose–response curves. cAMP production was mea-sured in COS-7 cells expressing native and S267K 5-HT6 receptors.Native 5-HT EC50 � 30 � 1 nM. S267K 5-HT EC50 � 41 � 6 nM. Datarepresent the mean and SEM of three independent transfections, eachexperimental data point determined in triplicate.

CONSTITUTIVELY ACTIVE 5-HT6 RECEPTORS 221

fold over basal levels for native receptors and 3-foldover basal levels for S267K receptors (Table II). Thebasal level of cAMP stimulated by the S267K receptor(in the absence of agonist) was 4.8 fold higher than thebasal level of cAMP stimulated by the native receptor.Clozapine (10 �M) reduced basal cAMP productionstimulated by S267K receptors down to vector controllevels. The same concentration of clozapine did nothave any effect on cAMP production in COS-7 cellstransfected with vector alone or native 5-HT6 receptorcDNA.

DISCUSSION

Previous studies have suggested the importance ofthe c-terminal region of il3 in GPCR activation. In thepresent study, we provide evidence that the nativeform of the human 5-HT6 receptor does not show con-stitutive activity in a heterologous expression system(COS-7 cells) and mutating serine 267 to lysine producesa constitutively active receptor. In addition, clozapineacts as an inverse agonist at the S267K receptor. Anincrease in agonist affinity, agonist-independent basalactivity, and reversal of the basal activity by an inverseagonist are hallmark characteristics of constitutivelyactive 5-HT receptors (for a review, see Teitler et al.,2002). Native and S267K receptors were expressed atsimilar levels, as measured by 3H-LSD saturation, in-dicating that the higher level of basal activity observed

for the mutant receptors is not due to higher expressionlevels of the mutant receptors. The lack of an effect ofthe S267K mutation on the binding affinity for LSDsuggests that the mutation did not cause globalchanges in receptor conformation or the binding site.

3H-5-HT labeled only a fraction of the total nativeand S267K binding sites that were labeled by 3H-LSD(Table I), consistent with the classical ternary complexmodel for GPCR. More importantly, 3H-5-HT labeled ahigher number of 5-HT binding sites in cells expressingS267K receptors than in cells expressing native recep-tors. Generally, agonist radioligands label only thehigh-affinity, active state of the receptor. Therefore,3H-5-HT would be expected to label only a fraction ofthe total receptor population. A greater number of 3H-5-HT binding sites in the S267K receptor population isconsistent with the hypothesis that constitutively acti-vating mutations increase the fraction of receptors inthe high-affinity state. It was observed that the affinityof 5-HT for the S267K receptor was lower when mea-sured in 3H-5-HT saturation studies (14 nM) thanwhen determined in 3H-LSD competition studies (3.3nM). The difference in 5-HT affinity determined by thetwo different methods is unexplainable at this time.Possibly the presence of another drug, 3H-LSD, in thecompetition assays may be altering the dynamics of theagonist–receptor interaction in some unknown man-ner.

According to the extended ternary complex model ahigher proportion of receptors in the active state shouldresult in an increase in the affinity of the receptor foragonist and an increase in basal activity. The S267Kreceptor showed 10-fold higher affinity for 5-HT thanthe native receptor and was constitutively active. Un-expectedly, the potency of 5-HT in stimulating cAMPproduction was unchanged. Agonists generally showhigher potency for constitutively active receptors. Inagreement with our results, an analogous mutation inthe TSH receptor produced a constitutively active re-ceptor without changing the potency of TSH for stim-ulating cAMP production (Parma et al., 1993). Theconstitutive activity of the S267K receptor was alsoobserved in CHO-K1 cells and thus is not an idiosyn-cratic feature of the COS-7 cell-line (data not shown). Itwill be of interest to determine if the equal potency of5-HT at native and S267K receptors is observed in acell line stably expressing these receptors. More uni-form receptor expression level and a more robust cAMPstimulation may reveal a higher 5-HT potency for theS267K receptor in a stable cell line (currently in devel-opment).

Based on the crystal structure of rhodopsin, currentmodels of receptor activation have focused on aminoacid 6.34, analogous to S267 of the 5-HT6 receptor(Palczewski et al., 2000). In the rhodopsin receptor,this amino acid is predicted to form a hydrogen bondwith arginine (R3.50) in the DRY region. The crystal

Fig. 5. Constitutive activity of S267K receptors and inverse ago-nist activity of clozapine. cAMP production was measured in COS-7cells expressing native and S267K 5-HT6 receptors in the absence orpresence of 10 �M 5-HT or clozapine. Data represent the mean andSEM of three independent transfections, each experimental datapoint determined in triplicate. *P � 0.01 vs. native basal; #P � 0.01vs. S267K basal (ANOVA with post-hoc SNK comparisons).

222 A. PUROHIT ET AL.

structure of rhodopsin provides a physical explanationfor the results observed in our study and in previousstudies in which amino acid 6.34 has been mutated.Mutations of amino acids in this position, such as A293in the �1B-adrenergic receptor (Kjelsberg et al., 1992),T373 in the �2A-adrenergic receptor (Ren et al., 1993),L322 in the 1-adrenergic receptor (Lattion et al.,1999), C322 in the 5-HT2A receptor (Egan et al., 1998),S312 in the 5-HT2C receptor (Herrick-Davis et al.,1997), T313 in the 5-HT1B receptor (Pauwels et al.,1999), and T279 in the �-opioid receptor (Huang et al.,2001) have produced constitutively active receptors.Taken together, these studies support the conclusionthat the domains involved in receptor activation arefairly conserved among GPCR.

During the completion of the work presented herein,a report containing a different result was published.Kohen et al. (2001) recently cloned the mouse 5-HT6

receptor, which shows 89% identity to its human coun-terpart. Mutagenesis studies were attempted to ex-plore the role of the c-terminal region of il3 in receptoractivation. The data were interpreted as demonstrat-ing that the wild-type mouse 5-HT6 receptor was con-stitutively active. The mutations K264I, K267A, andA268R decreased receptor basal activity, whereas themutations A265R and L266R had no effect on the basalactivity of the mouse 5-HT6 receptor expressed inCOS-7 cells. In the present study, we mutated S267 inthe human 5-HT6 receptor, which is analogous to S269of the mouse 5-HT6 receptor. It could be possible thatspecies differences in the constitutive activity of the5-HT6 receptors exist which could explain the discrep-ancies with the data presented herein. However, Ko-hen et al. (2001) were unable to detect receptor expres-sion by radioligand binding in their experiments andintracellular cAMP was not determined directly, butrather with an indirect reporter gene under the controlof a cAMP response element. Therefore, in their studyartifacts due to the measurement of downstream re-porter gene activation cannot be ruled out. The lack ofdetectable receptor expression also makes interpretingthe data in the study difficult.

The phenomenon of GPCR constitutive activity hasreceived considerable attention over the past few yearsand many native and mutant receptors displaying con-stitutive activity have been identified (for a review, seePauwels and Wurch, 1998). Morriset et al. (2000) dem-onstrated a physiological role for constitutively activehistamine H3 autoreceptors in regulating histaminerelease in the rat brain in vivo. Constitutive activity ofmutant GPCR has been linked to several human dis-eases, including congenital night blindness (Rao et al.,1994), thyroid adenomas (Parma et al., 1993), malefamilial precocious puberty (Shenker et al., 1993), andhypercalcemia (Parfitt et al., 1996).

Based on the distribution of 5-HT6 receptors in cog-nitive areas of the brain and high affinity for antipsy-

chotic drugs, the 5-HT6 receptor is a potential target fortreatment of psychiatric illnesses. Interestingly, wediscovered that clozapine acts as an inverse agonist atthe constitutively active form of the human 5-HT6 re-ceptor. Clozapine is also known to act as an inverseagonist at human 5-HT2A (Egan et al., 1998) and5-HT2C receptors (Herrick-Davis et al., 2000). As of yetthere is no evidence for the presence of receptor acti-vating mutations in schizophrenia. However, constitu-tive activity of native receptors in vivo is a distinctpossibility. In such a scenario, an inverse agonist mightbe a more effective therapeutic agent than a neutralantagonist. The S267K receptor provides a new tool toscreen drugs for 5-HT6 receptor inverse agonist activ-ity. Currently, in our laboratory, studies are inprogress to explore the inverse agonist activity of atyp-ical and typical antipsychotic drugs at the 5-HT6 recep-tor. It should be pointed out that there is no indicationthat the human 5-HT6 receptor shows constitutive ac-tivity in vivo nor is there any clinical data indicatingthat an inverse agonist demonstrates superior clinicalefficacy relative to a classical competitive antagonist.The possible role of constitutive activation of GPCRand the role of inverse agonist activity in therapeuticsis still a relatively new field of endeavor.

REFERENCES

Alewijnse AE, Timmerman H, Jacobs EH, Smit MJ, Roovers E, Cotec-chia S, Leurs R. 2000. The effect of mutations in the DRY motif onthe constitutive activity and structural instability of the histamineH(2) receptor. Mol Pharmacol 57:890–898.

Ballesteros JA, Weinstein H. 1995. Integrated methods for modelingG-protein coupled receptors. Methods Neurosci 366–428.

Ballesteros JA, Jensen AD, Liapakis G, Rasmussen SG, Shi L, GetherU, Javitch JA. 2001. Activation of the beta 2-adrenergic receptorinvolves disruption of an ionic lock between the cytoplasmic ends oftransmembrane segments 3 and 6. J Biol Chem 276:29171–29177.

Bourson A, Boess FG, Bos M, Sleight AJ. 1998. Involvement of 5-HT6receptors in nigro-striatal function in rodents. Br J Pharmacol125:1562–1566.

Egan CT, Herrick-Davis K, Teitler M. 1998. Creation of a constitu-tively activated state of the 5-hydroxytryptamine2A receptor bysite-directed mutagenesis: inverse agonist activity of antipsychoticdrugs. J Pharmacol Exp Ther 286:85–90.

Gerard C, el Mestikawy S, Lebrand C, Adrien J, Ruat M, Traiffort E,Hamon M, Martres MP. 1996. Quantitative RT-PCR distribution ofserotonin 5-HT6 receptor mRNA in the central nervous system ofcontrol or 5,7-dihydroxytryptamine-treated rats. Synapse 23:164–173.

Herrick-Davis K, Egan C, Teitler M. 1997. Activating mutations of theserotonin 5-HT2C receptor. J Neurochem 69:1138–1144.

Herrick-Davis K, Grinde E, Teitler M. 2000. Inverse agonist activityof atypical antipsychotic drugs at human 5HT2C receptors. J Phar-macol Exp Ther 295:226–232.

Huang P, Li J, Chen C, Visiers I, Weinstein H, Liu-Chen LY. 2001.Functional role of a conserved motif in TM6 of the rat mu opioidreceptor: constitutively active and inactive receptors result fromsubstitutions of Thr6.34(279) with Lys and Asp. Biochemistry 40:13501–13509.

Kjelsberg MA, Cotecchia S, Ostrowski J, Caron MG, Lefkowitz RJ.1992. Constitutive activation of the alpha 1B-adrenergic receptor byall amino acid substitutions at a single site. Evidence for a regionwhich constrains receptor activation. J Biol Chem 267:1430–1433.

Kohen R, Metcalf MA, Khan N, Druck T, Huebner K, Lachowicz JE,Meltzer HY, Sibley DR, Roth BL, Hamblin MW. 1996. Cloning,characterization, and chromosomal localization of a human 5-HT6serotonin receptor. J Neurochem 66:47–56.

Kohen R, Fashingbauer LA, Heidmann DE, Guthrie CR, HamblinMW. 2001. Cloning of the mouse 5-HT6 serotonin receptor and

CONSTITUTIVELY ACTIVE 5-HT6 RECEPTORS 223

mutagenesis studies of the third cytoplasmic loop. Brain Res MolBrain Res 90:110–117.

Lattion A, Abuin L, Nenniger-Tosato M, Cotecchia S. 1999. Constitu-tively active mutants of the beta1-adrenergic receptor. FEBS Lett457:302–306.

Monsma Jr FJ, Shen Y, Ward RP, Hamblin MW, Sibley DR. 1993.Cloning and expression of a novel serotonin receptor with highaffinity for tricyclic psychotropic drugs. Mol Pharmacol 43:320–327.

Morriset S, Rouleau A, Ligneau X, Gbahou F, Tardivel-Lacombe J,Stark H, Schunack W, Ganellin CR, Schwartz JC, Arrang JM. 2000.High constitutive activity of native H3 receptors regulates hista-mine neurons in brain. Nature 408:860–864.

Palczewski K, Kumasaka T, Hori T, Behnke CA, Motoshima H, FoxBA, Le TI, Teller DC, Okada T, Stenkamp RE, Yamamoto M,Miyano M. 2000. Crystal structure of rhodopsin: a G protein-cou-pled receptor. Science 289:739–745.

Parfitt AM, Schipani E, Rao DS, Kupin W, Han ZH, Juppner H. 1996.Hypercalcemia due to constitutive activity of the parathyroid hormone(PTH)/PTH-related peptide receptor: comparison with primary hyper-parathyroidism. J Clin Endocrinol Metab 81:3584–3588.

Parma J, Duprez L, Van Sande J, Cochaux P, Gervy C, Mockel J,Dumont J, Vassart G. 1993. Somatic mutations in the thyrotropinreceptor gene cause hyperfunctioning thyroid adenomas. Nature365:649–651.

Pauwels PJ, Wurch T. 1998. Review: amino acid domains involved inconstitutive activation of G-protein-coupled receptors. Mol Neuro-biol 17:109–135.

Pauwels PJ, Gouble A, Wurch T. 1999. Activation of constitutive5-hydroxytryptamine(1B) receptor by a series of mutations in theBBXXB motif: positioning of the third intracellular loop distal junc-tion and its G(o)alpha protein interactions. Biochem J 343(Pt 2):435–442.

Rao VR, Cohen GB, Oprian DD. 1994. Rhodopsin mutation G90D anda molecular mechanism for congenital night blindness. Nature 367:639–642.

Ren Q, Kurose H, Lefkowitz RJ, Cotecchia S. 1993. Constitutivelyactive mutants of the alpha 2-adrenergic receptor. J Biol Chem268:16483–16487.

Roth BL, Craigo SC, Choudhary MS, Uluer A, Monsma Jr FJ, Shen Y,Meltzer HY, Sibley DR. 1994. Binding of typical and atypical anti-psychotic agents to 5-hydroxytryptamine-6 and 5-hydroxytrypta-mine-7 receptors. J Pharmacol Exp Ther 268:1403–1410.

Ruat M, Traiffort E, Arrang JM, Tardivel-Lacombe J, Diaz J, Leurs R,Schwartz JC. 1993. A novel rat serotonin (5-HT6) receptor: molec-ular cloning, localization and stimulation of cAMP accumulation.Biochem Biophys Res Commun 193:268–276.

Samama P, Cotecchia S, Costa T, Lefkowitz RJ. 1993. A mutation-induced activated state of the beta 2-adrenergic receptor. Extendingthe ternary complex model. J Biol Chem 268:4625–4636.

Sanger F, Nicklen S, Coulson AR. 1977. DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci USA 74:5463–5467.

Scheer A, Costa T, Fanelli F, De Benedetti PG, Mhaouty-Kodja S,Abuin L, Nenniger-Tosato M, Cotecchia S. 2000. Mutational anal-ysis of the highly conserved arginine within the Glu/Asp-Arg-Tyrmotif of the alpha(1b)-adrenergic receptor: effects on receptorisomerization and activation. Mol Pharmacol 57:219–231.

Shapiro DA, Kristiansen K, Weiner DM, Kroeze WK, Roth BL. 2002.Evidence for a model of agonist-induced activation of 5-hydroxy-tryptamine 2A serotonin receptors that involves the disruption of astrong ionic interaction between helices 3 and 6. J Biol Chem277(13):11441–11449.

Shenker A, Laue L, Kosugi S, Merendino Jr JJ, Minegishi T, Cutler JrGB. 1993. A constitutively activating mutation of the luteinizinghormone receptor in familial male precocious puberty. Nature 365:652–654.

Teitler M, Herrick-Davis K, Purohit A. 2002. Constitutive activity ofg-protein coupled receptors: emphasis on serotonin receptors. CurrTop Med Chem 2:529–538.

Ward RP, Dorsa DM. 1996. Colocalization of serotonin receptor sub-types 5-HT2A, 5-HT2C, and 5-HT6 with neuropeptides in rat stri-atum. J Comp Neurol 370:405–414.

Ward RP, Hamblin MW, Lachowicz JE, Hoffman BJ, Sibley DR,Dorsa DM. 1995. Localization of serotonin subtype 6 receptor mes-senger RNA in the rat brain by in situ hybridization histochemistry.Neuroscience 64:1105–1111.

224 A. PUROHIT ET AL.