Proc. Nati. Acad. Sci. USAVol. 89, pp. 3630-3634, April 1992Pharmacology

Human serotonin iD receptor is encoded by a subfamily of twodistinct genes: 5-HT1Da and 5-HT1DPRICHARD L. WEINSHANK*, JOHN M. ZGOMBICK, MARY J. MACCHI, THERESA A. BRANCHEK,AND PAUL R. HARTIGSynaptic Pharmaceutical Corporation,t Paramus, NJ 07652

Communicated by Eric R. Kandel, January 17, 1992 (received for review September 16, 1991)

ABSTRACT The serotonin iD (5-HT1D) receptor is a phar-macologicaily defied binding site and functional receptor site.Observed variations in the properties of 5-HTID receptors indifferent tissues have led to the speculation that multiplereceptor proteins with slightly different properties may exist.We report here the cloning, deduced amino acid sequences,pharmacological properties, and second-messenger couplingof a pair of human 5-HTID receptor genes, which we havedesignated 5-HTID, and 5-HTDp due to their strong similar-ities in sequence, pharmacological properties, and second-messenger coupling. Both genes are free of introns in theircoding regions, are expressed in the human cerebral cortex,and can couple to inhibition of adenylate cyclase activity. Thepharmacological binding properties of these two human recep-tors are very similar, and match closely the pharmacologicalproperties ofhuman, bovine, and guinea pig 5-HTID sites. Bothreceptors exhibit high-affinity binding of sumatriptan, a newanti-mIgae medication, and thus are candidates for thepharmacological site of action of this drug.

The serotonin iD (5-HT1D) receptor was initially character-ized by radioligand binding procedures using membranesderived from the bovine caudate nucleus (1). In physiologicalassays, the 5-HTID receptor subtype has been shown tomodulate vascular tone (2) and to inhibit neurotransmitterrelease through its role as a presynaptic heteroreceptor (3, 4)or as a terminal autoreceptor (5). The 5-HT1D receptor isknown to be a G protein-coupled receptor based upon itsfunctional coupling to adenylate cyclase inhibition (6) and thesensitivity of high-affinity agonist ([3H]5-HT) binding toguanine nucleotides (7). Sumatriptan, the only 5-HTlD-selective ligand yet identified (8, 9), has been reported toeffectively treat acute migraine (10).We have recently identified a previously cloned G protein-

coupled receptor that displayed high homology to the 5-HT1Areceptor, RDC4 (11), as a canine 5-HT1D receptor subtype(12). In the current study, we have used this RDC4 gene asa probe for the isolation of its human homolog from a humanhippocampus cDNA library. As described below, this humanclone has been sequenced, transfected, and characterizedand has been shown to encode a human 5-HT1D receptorsubtype. A preliminary report describing some properties ofthis clone has been published (13). Subsequently, a reportappeared describing a human 5-HT1D receptor clone that isvery similar in pharmacological properties (14). We havedesignated this first human 5-HT1D receptor gene, which isgene RDC4 in the dog, as 5-HTlDa. In addition, we havecloned a highly homologous, yet distinct, 5-HT1D receptor(5-HTlDp) from a human genomic library. These two 5-HT1Dsubtypes represent two closely related members of a sub-family of 5-HT1D receptors.t

MATERIALS AND METHODSCloning and Sequencing. A human placental genomic li-

brary (Stratagene) was screened with the 1.3-kilobase (kb)HindIII-Sph I genomicDNA fragment from the canine RDC4clone (12). The probe was labeled with 32P by random priming(15). Hybridization and Southern blot analysis were per-formed as described (16). For subcloning and further South-ern blot analysis, DNA was inserted into plasmid pUC18.cDNA clones were isolated by screening a human hippoc-ampal AZAPII cDNA library (Stratagene) with an oligonu-cleotide probe derived from the canine RDC4 sequence (17)(corresponding to amino acids 165-188 in transmembraneregion IV). Overlapping oligomers were labeled with[32P]dATP and [32PjdCTP by synthesis with the large frag-ment of DNA polymerase. Hybridization was performed asdescribed (16). DNA fragments cloned in the AZAPII vectorwere excised by helper phage and recircularized to generatesubclones in the pBluescript (Stratagene) phagemid vector.Nucleotide sequence analysis was done by the dideoxynu-cleotide chain-termination method (18) on denatured double-stranded plasmid templates with Sequenase (United StatesBiochemical).

Expression. The entire coding region of the 5-HTID, or5-HTDf3 receptor gene was subcloned into the expressionvector pcEXV3 (16) or pSVL (Pharmacia). Stable cell lineswere generated by cotransfection with pcEXV3-5-HT1,D orpSVL-5-HTD3 and pGCcos3neo, using the calcium phos-phate method. Cells were grown as monolayers and selectedfor antibiotic resistance (16). Stable transfectants werescreened to specifically bind [3H]5-HT, and two clones,5-HTD1 and 5-HTlDp, were selected for pharmacologicalcharacterization. No specific binding of [3H]5-HT to mem-branes from sham-transfected cells was detected. Cells weregrown as monolayers to 100% confluency before membraneswere harvested for binding assays (16). Freshly preparedmembranes for radioligand binding assays were preparedfrom stably transfected cells (12).

Radioligand Bing Studies. Binding studies were con-ducted with [3H]5-HT (24-30 Ci/mmol; DuPont-NEN; 1 Ci =37 GBq) as the radioligand, and 5-HTlD assay conditions inthe absence of masking ligands were as described (12).Occupancy studies were performed with [3H]5-HT concen-trations ranging from 0.25 to 50 nM, and competition studieswere performed with 5 nM [3H]5-HT. Binding data wereanalyzed by computer-assisted nonlinear regression analysis(ACCUFIT and ACCUCOMP, Lundon Software, Chagrin Falls,OH). IC50 values were converted to Ki values by using theCheng-Prusoff equation (19). Drugs were obtained as de-scribed (12).

Abbreviation: 5-HT, serotonin (5-hydroxytryptamine).*To whom reprint requests should be addressed.tFormerly Neurogenetic Corporation.tThe sequences reported in this paper have been deposited in theGenBank data base (accession nos. M81589 and M81590).

3630

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

, 202

0

Pharmacology: Weinshank et al.

Functional Studies. Intracellular cAMP levels were deter-mined (12) by radioimmunoassay (Advanced Magnetics,Cambridge, MA). Dose-response curves for 5-HT and suma-triptan were obtained by using Y-1 cells stably transfectedwith the 5-HT1D9 receptor gene. Functional response curveswere fit to a four-parameter logistic equation to obtain EC50and Emau values (GraphPAD InPlot, San Diego, CA).

RESULTSWe have screened a human genomic placental library (Strat-agene) with a 1.3-kb Hind III-Sph I restriction fragmentderived from the canine genomic clone RDC4 (12). Two oftheclones obtained, hp8 and hp84, were nonoverlapping partialclones that represented human homologs of the canine RDC4receptor. A human hippocampus cDNA library was subse-quently screened with an oligonucleotide probe derived fromtransmembrane region IV of the canine clone RDC4. Twooverlapping clones were isolated and characterized by DNA

Proc. Natl. Acad. Sci. USA 89 (1992) 3631

sequence analysis. Complementary portions of the cloneswere then ligated at an internal Bgl II restriction site (corre-sponding to nucleotides 624-629) to obtain a full-lengthclone, designated 5-HTlDa. Overall, 88% amino acid se-quence conservation was observed between the dog RDC4sequence and the human receptor over 377 amino acids. Anadditional genomic clone, hpll, was also isolated and char-acterized. A comparison ofthe deduced amino acid sequenceof clone hpll with previously characterized neurotransmitterreceptors indicates that the protein product of hpll is amemberofthe G protein-coupled receptor family. Clone hpllcontains an uninterrupted open reading frame that encodes aprotein of 390 amino acids.Amino acid sequence comparison shows that hpll is

closely related to, but distinct from, the 5-HT1,D receptor(Fig. 1). The overall amino acid identity between 5-HT1,D andhpll is 63%, with 77% amino acid identity within the trans-membrane regions alone. This high amino acid homology,particularly in the transmembrane regions, is characteristic of

liT,,,. M s L N sikMPA E G L PE.P :A5-HT M.EEP..Aac A P.- P P PA S E T P.5-HT^, M D V L S P.G a G N N I T S P P A P5-HTc M V N L G N A V R S L L M H L I G L L V w Q F D S I S P V A A5-HT2 M E I L C E D N S L S S I P N S L MO L G D G P R L Y H N D F N S R D A N T S E A S N W T I D A E

1523183250

5-%....SN RSCN A T ET SE A W DPRT L SALK I SL AV VL sV.IITL.ATl .SNA'-VV'L T-T.: 625-HT,2# NL S':S'A P.S ON':C S'A KY YODDS S LP VK V'LLVM L L AL -T.L.A-T T L SNAl VIAT:'1 735-AT1A FE T G GN T T'G I.:SV TVSV S ........VI T SLLLS.GT L I FC''A V L:GN: A C:V:VA-A 605-HT2c. I VTD T FNS S'D GG RL F FP D GVSN W P AL SI V VI. I IMT IdGONI. LVIMN A 785-AT, Nt R T N L S C E G Y L P P T C L S L H L S E K N W S A L L 'T T V V II I .T. I A G N I L V.I.:4MA 98

5-HT,,, L L T ft K L -YPA L I G'S L A T J J) L IL 11VSAL V NM L .C 11

5~T,,[, V T ft ~Tft K LU T P ~A N Y L L A S L A V -T: D L: L VX S I L. V N4 P 1. 5 T M:YV-T. V. .TGS NV 0 V V C- 1225-HTI2 I ALERSL N VANWY:L- I G.S.L AV T L M'VS:V L:VL P MA AL:OQ'V LN .KW.T.:LSGS-VTJ~C- 1095-HT~cVL A A M L V. 0 L L V 14 P'L S 'L L A L V D YVWV P L'PRfYVL.C 125-AT2 VSL E kKI0N A TN YFFLM S LA AbM:L L GF.LVTM P V S'MLT LY.SV RW PLPSKL. 4

IIII -911~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...5 ,,AT 0.. I L S S S I T C ACT AS J L H L C V'A A L::S fR:F W.A 4T S A: L' E T§I RK ft .: T 'A GK A AA I N.L: ::A.I 161OAT0 F V L S S 5 T C C ~:T: A S~IL H L C V 0 A L`0 RYtVVA I T'0 A.v EY:5VS:A:: R.: T P 'K ft A. A -V: M A:-LV 172

5OAT, AL F IA L ~V L::C':CTJ~S -.SILLHLC'A I-A IORVYVW:A'I .T 0-P I D:Y-:V KRT P:RPRA .LIO:S~L` 1585AHTn P VWI''S L V L FST-A S.I MHLC A IS L'DRY V A I RkP IE HSR FNSRTKAIKII 1785-AT A IVWI YLDYL FST ASIM HLC A:'IS L.SNYVVA IS0N P I'H FN S RTK AFLKI. IA198

I IT TIV-4

5A-HT,;, VWAI SI CIS I-P.P L-F .AWR'.QA kA QE E MS DCL V N TS0. S'YT.IYSTCGA F'Y' 1208OAT, WVVVF SI:S I S.L:P. P'FPF.WER.SA KA E E E V SS~C V V.'NTD.ILVY.T VYSTVS: .-A.F'Y:F219OAT2VVLISFL1SI~~~~~~~~P.PMLGW.P SDDCIKH.SGT' STFAF 205

OAHT VVWA I Si GTSV 'VPIT-P'VIG LRDE SK'V FV NNTT CV L NDP N ..FL'I G'S FVA:F':-IF226OAHT, VVWT S VSGI SNP I P-V-FG L D DDS0KVF .K E GS CL LA D0N .F-VOL I SSFVAPI24

5OAT,. P S V LL II Y G.R IYRA ANRWRI L .N P PS LYGK R FTTA'HLVT G.SAGSSL 2535-HT.[ P .-T LL LIA L Y Y VEA RS RI L K. T P NRT GK RL TR-A LITEDSPGSTS'SV 267OAHT P L LL M LVL GRIF AARF RI R KTV K KVE K TG ADTRH GA SPAP PKKSVN 254OA .P LTIMVITOLLLIRVLRRSTLMTLRD.ATEEEANMS LN FLNC CCKKNGS270

OAHT P LT'I M VITYF LTIK SLSKEAT LC V S DL ST RA KLA S.F S FLPSSSLS.S'EK 293

5 HT ~ .CS LNS S LEGMISHAS AGS PL FF A V KI KL A0SAL-E.............2875T T I NSR V PD-VP SES .GDS PV YV.WO V KV RV S0ALLE.............300

OAHT_ G E SGS RNAWERL .G VE SK A GGA L CA N GAVRGD DDGAAL:E.VI E VARVGNSKEH 3035OATI E EE NA PNP NP DQKP RRKKK E KRP R GT M.......................303O AT, L F 5 Rt S I ARft P 0GSYVA GRft ........T M 0 ......................313

.H-. ................R KfRtIDSAARE 'k A'T KI' .'GIO GA' F 310

.,. ................K K KL MA ARE 'K A T K TLGIIL A: F 323O H LPL PS E A GPT PC A PA S F EARKAE R NAAEAK R KMA LA RE RKTV KT L G I I M T'' 353

. ................AlI NNEK KA SKXVL G IVFFVF 322.OT................SI SNES KACK VL GI'VFFL F 332

OA Tj. II CWLP F FV V SLV L:P IC RISC W.I HP ALFD FFTWLGSOLNW85LINWPIIYTTV 308OAHT IV C WL PPFFI I SLVMKP I CK DAC.WVH L A IPFDPFFTWILGYL NSLI NP.I1TIN371OHTI, IL C WLPPFF IV ALV LPFPCE S S C .M PTL L0GA II NWLGYSVWSNSLL NPVL-YAY 401OAHT, L IMWC P FF I T NIL SVL C GKA C N DKLM E KLL NV F VWIGYV CSGINWP LV..T'L 3725HT, V V MVWCPPFF IT NIM A VIC K ES C E NV IGA L L V FVWIGYL S SAV NPLVYTL 382

OAHT1,,F N E F RDA FSK IVP FR KAS...............................377OAHT,, S NE DFK SA FAK LIRPFK CTS...............................390OAHT, F NK 0 F NA FK.KIIK CNFCR0..............................421SAHT, PNK IYfRRAFS KTYL RC DYK P D K KP .POVERS1PNROVAA T AL S GfRE L NOVNI YfRT 421OAHT, PNKTOYR SA FS RY.:ISCQYKE NR K PL L I LV T I PALAYVKS S QLDV GDK K NS432

5Tj. 377T.,? 3900

OAHT,, N E RVA RK A ND PE P GI EA V E NL EL PV NP SAVVS ER I SS V 4605-HT 0 E D A ft 5 T V 0 D C S M V T L G K 4 Q S E E N C T D N A T V N A K V S C V 471

FIG. 1. Comparison of the 5-HTD. and 5-HTDg receptor deduced amino acid sequences with those of other serotonin receptors. The sevenputative membrane-spanning domains (I-VII) are indicated by overbars. Homologies between the 5-HTDD receptor and other receptors arenoted by shading. In the amino-terminal region, consensus sequences for N-linked glycosylation sites are underlined. Selected invariant residuesare highlighted by arrows.

Dow

nloa

ded

by g

uest

on

Aug

ust 1

, 202

0

3632 Pharmacology: Weinshank et al.

r , . .2 4 6

Bound, pmnol/mg

O 10 20 30 40 50 60

2.0

1.5-06

I-.~~~~~~~~02

0 1 2

Bound, pmol/mg

Free fHJ5-HT, nM

FIG. 2. Determination of equilibrium dissociation constant (Kd)of [3H]5-HT for cloned human 5-HT11< (A) and 5-HTDwp (B) recep-tors. Membranes harvested from stable transfectants were incubatedwith 9-10 concentrations of [3H]5-HT (0.25-50 nM) in absence orpresence of 10 ,uM 5-HT for 30 min at 37°C in the dark. Each datapoint is the mean of triplicate determinations; standard deviationsaveraged <5%. Kd and B. values were determined by nonlinearregression analysis, and are illustrated in the form ofa Scatchard plot(Insets). Calculated Kd andB. values are 4.5 nM and 5.8 pmol/mgfor clone 5-HT1,D and 5.0 nM and 2.0 pmol/mg for clone 5-HT1Dp.

the relationship between closely related subtypes. Further-more, the pharmacological binding properties and second-messenger coupling of this clone are very similar to theproperties of the 5-HT1D<, clone (see below). We have there-fore named this second receptor 5-HTlDs. A comparison ofthe human 5-HT1,D and 5-HTDp receptor sequences to allother known serotonin receptors is shown in Fig. 1. Allserotonin receptors contain certain structural features thatare invariant, including the aspartic residues of transmem-brane regions II and III, the Asp-Arg-Tyr sequence at the endof transmembrane region III, and the conserved prolineresidues of transmembrane regions IV-VII (20).

Clonal cell lines stably expressing the 5-HTID1 or 5-HTDpreceptor gene were generated from murine LM(tk-) fibro-blasts. Membranes harvested from cells transfected witheither the 5-HTD1 or the 5-HTDp gene displayed an appar-ently homogeneous population of high-affinity, saturable[3H]5-HT binding sites (Fig. 2). The equilibrium dissociationconstants (Kd) of [3H]5-HT for the 5-HT1,D (Kd = 5.1 ± 0.4nM; n = 5) and 5-HTlDp (Kd = 4.3 ± 0.5 nM; n = 5) receptorswere similar. The density (Bmt) of binding sites was 2-foldhigher in the 5-HTD1 receptor cell line (B., = 4.0 + 0.5pmol/mg) than in the 5-HTlDg3 receptor cell line (B. = 2.3± 0.4). Specific binding was >85% of total [3H]5-HT bindingat the Kd concentration of radioligand, for both clones.The pharmacological profiles of both clones were deter-

mined in parallel experiments from analysis ofcompetition of[3H]5-HT binding (Table 1). The rank order of potency ofthese ligands to compete for the [3H]5-HT binding site isconsistent with a5-HTID receptor pharmacological profile forboth clones: 5-carboxamidotryptamine > 5-HT > yohimbine> 8-hydroxy-2-(di-n-propylamino)tetralin > spiperone > za-copride. Compounds exhibiting high affinity for both the5-HTD. and 5-HTDs clones and the pharmacologically de-fined 5-HT1D receptor in native brain tissue includedtryptamine derivatives (5-carboxamidotryptamine, 5-HT,

Table 1. Apparent dissociation constants (Kj) and Hill coefficients (nH) of serotonergic ligands for the cloned human5-HT1D, and 5-HTDs receptors in comparison to the pharmacologically defined 5-HT1D receptor from native brainmembrane preparations 4

Drug* Human 5-HTD1 receptor Human 5-HTDp receptor Bovinemembranest

No. Name Ki, nM nH Ki, nM nH Ki, nM1 Lysergol 0.63 ± 0.10 1.04 + 0.12 1.19 ± 0.13 0.87 + 0.04 ND2 5-CT 0.70 + 0.06 0.% ± 0.03 1.6 ± 0.2 0.81 + 0.03 2.53 CGS-12066B 2.9 + 0.3 0.91 ± 0.04 2.1 + 0.3 0.91 ± 0.07 324 Sumatriptan 3.4 ± 0.3 0.80 ± 0.04 7.7 ± 0.5 0.88 ± 0.07 29S Methysergide 3.6 + 0.6 1.00 ± 0.02 25 ± 5 0.91 ± 0.03 46 5-HT 3.9 ± 0.3 0.93 ± 0.02 4.3 ± 0.9 0.94 ± 0.01 47 5-MeOT 4.8 ± 0.3 1.02 ± 0.02 34 + 6 0.95 ± 0.01 48 1-NP 7.4 ± 0.6 1.05 ± 0.04 12 ± 2 1.05 ± 0.03 159 DP-5-CT 13 + 1 0.97 ± 0.09 42 ± 3 0.91 ± 0.02 6310 Rauwolscine 16 ± 2 0.90 + 0.05 40 ± 7 0.87 ± 0.06 2011 Yohimbine 22 ± 2 0.91 ± 0.04 27 ± 3 0.97 ± 0.04 7912 TFMPP 64 ± 5 0.95 + 0.03 114 ± 18 1.00 + 0.02 28213 Tryptamine 86 + 6 0.95 ± 0.02 521 + 82 0.91 + 0.05 4014 DPAT 120 ± 11 0.90 ± 0.05 260 ± 32 0.91 ± 0.07 126015 mCPP 216 + 20 0.93 ± 0.05 361 ± 31 0.93 ± 0.01 155016 2-Methyl-5-HT 915 ± 68 0.90 ± 0.02 860 ± 109 0.95 ± 0.09 39817 Spiperone 995 ± % 0.86 ± 0.03 >10,000 ND 500018 Quipazine >1000 ND >1000 ND 138019 Pindolol 4100 ± 500 0.91 ± 0.04 4900 ± 300 0.91 ± 0.02 630020 Zacopride >10,000 ND >10,000 ND NDK; values and Hill coefficients are means + SEM from three to seven determinations. ND, not determined.

*Numbers correspond to those in Fig. 3. 5-CT, 5-carboxamidotryptamine; 5-MeOT, 5-methoxytryptamine; 1-NP,1-naphthylpiperazine; DP-5-CT, N, N-dipropyl-5-carboxamidotryptamine; TFMPP, 1-[3-(trifluoromethyl)phenyllpipera-zine; DPAT, 8-hydroxy-2-(di-n-propylamino)tetralin; mCPP, 1-(3-chlorophenyl)piperazine.tData from ref. 9.

Proc. Natl. Acad Sci. USA 89 (1992)

Dow

nloa

ded

by g

uest

on

Aug

ust 1

, 202

0

Proc. NatL. Acad. Sci. USA 89 (1992) 3633

5-methoxytryptamine), ergoline analogs (lysergol and meth-ysergide), and a2-adrenergic antagonists (rauwolscine andyohimbine). Sumatriptan was one of the most potent com-pounds for both clones (K1 = 3.4 nM for 5-HT1TE; K1 = 7.7nM for 5-HTlD3). With the exception of CGS-12066B, allcompounds tested exhibited higher affinity for the 5-HT1,<receptor than for the 5-HTlDp receptor (Table 1). This con-sistent difference in affinity values was maintained when theexpression densities were reversed (i.e., BX, for 5-HT1Dc =1 pmol/mg of protein; Bm. for 5-HTDs = 2.1 pmol/mg ofprotein). Structurally diverse compounds that discriminated(>7-fold) between these closely related subtypes includedmethysergide, 5-methoxytryptamine, tryptamine, and spip-erone. High correlations were obtained between the apparentaffinities of compounds at the pharmacologically defined5-HT1D receptor from native brain membranes and each ofthe cloned receptor subtypes (Fig. 3 A and B). In addition, avery high correlation coefficient (r = 0.96) was obtained inthe comparison of the cloned 5-HT1,D, and 5-HTDs subtypes(Fig. 3C).The functional coupling of these cloned 5-HTID receptor

subtypes to second-messenger responses was determined inintact cells stably expressing each one ofthese genes (Fig. 4).Forskolin (10,4M) produced a 3- to 5-fold increase in cAMPabove basal values. 5-HT (1 ,tM) inhibited forskolin-stimulated cAMP accumulation by 70 ± 4% in transfectedcells expressing the 5-HT1,D receptor and by 92 ± 1% intransfected cells expressing the 5-HTlDs receptor. The EC50for 5-HT was 2.49 ± 0.65 nM for 5-HT1TE and 1.8 ± 1.2 nMfor 5-HTlDp (Fig. 4). Sumatriptan was found to produceagonist responses (EC50 = 3.2 ± 0.67 nM for 5-HTlDa and 5.2± 1.2 nM for 5-HTD,8). The responses to 5-HT or sumatriptan

III

10IAv

9 r G 7>9

I7a- 110100 0

160 1

°175 6 7 8 9 l1

plO, Clone 5-HT1Da

pKi, Clone 5-HTlDp101= C

0 9 32IV- r aO.96 6032Aa ~~1184

S 12 9O7

0Y13

5 6 7 8 9 10

pKi, Clone 5-HTiDa

FIG. 3. Correlations between affinity constants (pK; values) ofserotonergic ligands for the cloned human 5-HT1,D (A) or 5-HT1Dp(B) receptor and the pharmacologically defined 5-HT1D receptor.Correlation between pK, values of the same ligands for the clonedhuman 5-HTlD0 and 5-HT1Dq receptors is also shown (C). Numberscorrespond to compounds listed in Table 1. The correlation coeffi-cient (r) is listed in each panel.

were completely blocked by the nonselective antagonistmethiothepin (10 juM) in both transfected cell lines (data notshown). In the absence of forskolin, 5-HT did not stimulateadenylate cyclase in these transfected cells. In addition, 5-HT(1 ;LM) was unable to produce any alterations in phosphoti-dylinositol metabolism. In untransfected cells, 5-HT did notstimulate or inhibit cAMP accumulation.

DISCUSSIONIn the current study, two separate human 5-HTlD receptorgenes (5-HTjEr and 5-HTDp) have been isolated and char-acterized. The dog homolog of the 5-HT1Dc, receptor, knownas the RDC4 sequence (11), was the first 5-HTE1, receptorgene to be isolated, although its identity as a 5-HTlD receptorwas only recently recognized (12). The second human 5-HT1Dreceptor to be cloned and characterized, the 5-HTDp recep-tor, exhibits pharmacological and functional properties thatare strikingly similar to those ofthe 5-HTlDa receptor. The rathomolog ofthe human 5-HTDp receptor was recently cloned,characterized, and shown to encode a 5-HT1l receptor site(21, 22). This result confirmed the predicted relationship ofthe rat 5-HT1l and human 5-HTlD sites (23). The fact thatsufficient pharmacological differences were present to haveoriginally led to the naming of the 5-HTlD and 5-HT1l sitesas separate receptor subtypes is a reflection of the unusualdegree of species variation shown by the 5-HTlDp receptor.The binding properties of these two 5-HT1D receptor

subtypes are so similar that a linear correlation coefficient (r)of 0.96 was calculated in the comparison of log Ki values of19 compounds at these two receptor sites. It is interesting thatthe 5-HT1Ea receptor displayed higher affinity for all struc-tural classes of compounds tested than did the 5-HT1D8receptor. Although the 5-HT1,D cell line had a higher expres-sion level than did the 5-HTDp cell line, this difference doesnot account for the differences in binding affinity. Specifi-cally, when the receptor densities of two cell lines werereversed and studied in parallel for a selection of compoundsshowing the largest differences in affinity values, the 5-HTDcreceptor still showed higher affinity for these compounds

'D

.E >*0

=cL

0

_xT Is

-0-

0-0I0

0" 10 9 8-log [drug] (M)

FIG. 4. Concentration-response curve for the inhibition of for-skolin-stimulated cAMP production by 5-HT (o) and sumatriptan (o)in clonal cell lines stably expressing the 5-HTD1 (A) or 5-HT1D3 (B)receptor gene. Curves shown were fitted to a four-parameter logisticequation to obtain response parameters. Calculated EC50 values for5-HT and sumatriptan in the experiments shown were 1.7 nM and 2.4nM in the 5-HTEa cell line and 1.3 nM and 3.3 nM in the 5-HT1Dscell line. Each data point is the mean of triplicate determinations;standard deviations averaged <5% of the mean. The point precedingthe break in the curve represents the mean of triplicate measure-ments of forskolin-stimulated cAMP accumulation in the absence ofagonist. Basal cAMP levels were subtracted from all cAMP valuesobtained by using forskolin and were normalized to 100%o relative toforskolin-stimulated cAMP levels in the absence of agonist. Theexperiment was replicated twice with similar results.

Pharmacology: Weinshank et al.

0

Dow

nloa

ded

by g

uest

on

Aug

ust 1

, 202

0

3634 Pharmacology: Weitishank et al.

than did the 5-HTlD, receptor. The strong similarity inproperties of the 5-HT1,D and 5-HTIDfl receptors makes itdifficult to determine whether any of the multiple physiolog-ically or pharmacologically defined 5-HTID subtypes matchesthe properties of these genetically defined receptors. Thatthese two cloneos are not sufficient to explain the known rangeof 5-HT10 subtypes means that additional 5-HTID-like recep-tor subtypes may remain to be cloned.

Similar to the 5-HTID receptor in native tissues, bothcloned human 5-HT1D receptors were found to couple toadenylate cyclase inhibition. Neither produced a phosphati-dylinositol response; The EC"0 values obtained for com-pounds at each receptor.subtype closely matched the affinityconstants for binding. These results contrast with the similarcomparison of agonist affinities for binding and function ofthe 5-HTID receptor in the calf substantia nigra (9). It is likelythat the high degree of agonist activity in these heterologousexpression systems relative to those reported in the literaturecan be attributed to differences in receptor reserve. Since the5-HT1A receptor has been shown to activate potassiumchannels in hippocampal neurons through a pertussis toxin-sensitive G-protein (24), it will be interesting to determinewhether the 5-HTlD receptor subfamily shares this property.The presence in the human genome oftwo different 5-HTlD

receptor genes that encode receptors which are remarkablysimilar in their pharmacological selectivity, second-messenger coupling, and amino acid sequence cannot beattributed to the trivial explanation that only one gene isexpressed. The 5-HTID receptor was isolated from bothhuman (present study) and dog (11) cDNA libraries. The5-HTlDp- receptor was expressed in the human brain, sincemRNA encoding this receptor was detected by PCR ampli-fication of cDNA derived from human cortex (data notshown). Thy possibility that these two 5-HT1D receptors areexpressed in different brain regions or in different cells withinthe same region needs to be investigated. Other possibleexplanations include different rates of desensitization, cou-pling to different G proteins, or different ratios of high- andlow-affinity agonist states in native tissues.The 5-HTID receptor has been implicated in a variety of

important clinical problems, but only one partially selectiveligand`for the 5-HTID receptor site, sumatriptan, is available.This new anti-migraite cnompound has been shown to interactwith 5-HT1D (and 5-HTiB) sites in animal preparations, withapparent dissociation constants ranging from 17 to 251 nM (8,25, 26). The two cloned human 5-HT11, receptor subtypesboth exhibit high-affinity dissociation constants (3.4 nM and7.7 nM) for sumatriptan, and therefore both the 5-HT1, and5-HTlDs receptors are likely targets for the therapeuticactions and side effects of this drug.

We acknowledge the excellent technical assistance provided byMr. Marcelino Dizon, Ms. Barbara Dowling, Ms. Anastasia Kokki-nakis, Mr. Harvey Lichtblau, and Ms. Michelle Smith. We thankDrs. Nika Adham, Jonathan Bard, and Lee Schechter for experi-mental support. We thank Mr. Ernest Lilley and Mr. GeorgeMoralishvili for producing the illustrations. This research was funded

in part by Grant 1R43NS27789 from the National Institutes of Healthand in part by Eli Lilly and Company.

1. Heuring, R. E. & Peroutka, S. J. (1987) J. Neurosci. 7, 894-903.

2. Humphrey, P. P. A., Feniuk, W., Perrin, M. J., Connor,H. E., Oxford, A. W., Coates, I. H. & Butina, D. (1988) Br. J.Pharmacol. 94, 1123-1132.

3. Bond, R. A., Craig, D. A., Charlton, K. G., Ornstein, A. G. &Clarke, D. E. (1989) J. Auto. Pharmacol. 9, 201-210.

4. Molderings, G. J., Werner, K., Likunga, J. & Gothert, M.(1990) Naunyn-Schmiedeberg's Arch. Pharmacol. 342, 371-377.

5. Schlicker, E., Fink, K., Gothert, M., Hoyer, D., Molderings,G., Roschke, I. & Schoeffier, P. (1989) Naunyn-Schmiede-berg's Arch. Pharmacol. 340, 45-51.

6. Hoyer, D. & Schoeffier, P. (1988) Eur. J. Pharmacol. 147,145-147.

7. Herrick-Davis, K., Titeler, M., Leonhardt, S., Struble, R. &Price, D. (1988) J. Neurochem. 51, 1906-1912.

8. Peroutka, S. J. & McCarthy, B. G. (1989) Eur. J. Pharmacol.163, 133-136.

9. Waeber, C., Schoeffier, P., Hoyer, D. & Palacios, J. M. (1990)Neurochem. Res. 15, 567-582.

10. Humphrey, P. P. A., Feniuk, W., Motevalian, M., Parsons,A. A. & Whalley, E. T. (1991) in Serotonin: MolecularBiology,Receptors, and Functional Effects, eds. Saxena, P. & Fozard,J. R. (Birkhauser, Basel), pp. 421-429.

11. Libert, F., Parmentier, M., Lefort, A., Dumont, J. E. &Vassart, G. (1989) Science 244, 569-572.

12. Zgombick, J. M., Weinshank, R. L., Macchi, M., Schechter,L. E., Branchek, T. A. & Hartig, P. R. (1991) Mol. Pharmacol.40, 1036-1042.

13. Branchek, T., Zgombick, J., Macchi, M., Hartig, P. & Wein-shank, R. (1991) in Serotonin: Molecular Biology, Receptors,and Functional Effects, eds. Saxena, P. & Fozard, J. R.(Birkhauser, Basel), pp. 21-32.

14. Hamblin, M. W. & Metcalf, M. A. (1991) Mol. Pharmacol. 40,143-148.

15. Feinberg, A. P. & Vogelstein, B. (1983) Anal. Biochem. 132,6-13.

16. Weinshank, R. L., Zgombick, J. M., Macchi, M., Adham, N.,Lichtblau, H., Branchek, T. A. & Hartig, P. R. (1990) Mol.Pharmacol. 38, 681-688.

17. Libert, F., Parmentier, M., Lefort, A., Dumont, J. E. &Vassart, G. (1990) Nucleic Acids Res. 18, 1916.

18. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl.Acad. Sci. USA 74, 5463-5467.

19. Cheng, Y. C. & Prusoff, W. H. (1973) Biochem. Pharmacol.22, 3099-3108.

20. flartig, P. R., Kao, H.-T., Macchi, M., Adham, N., Zgombick,J., Weinshank, R. & Branchek, T. (1990) Neuropsychophar-macology 3, 335-347.

21. Voigt, M.M., Laurie, D.J., Seeburg, P.H. & Bach, A. (1991)EMBO J. 10, 4017-4023.

22. Adham, N., Romanienko, P., Hartig, P., Weinshank, R. L. &Branchek, T. (1992) Mol. Pharmacol. 41, 1-7.

23. Hoyer, D. & Middlemiss, D. M. (1989) Trends Pharmacol. Sci.10, 130-132.

24. Andrade, R., Malenka, R. & Nicholl, R. A. (1986) Science 234,1261-1265.

25. Schoeffter, P. & Hoyer, D. (1989) Naunyn-Schmiedeberg'sArch. Pharmacol. 339, 675-683.

26. Sumner, M. J. & Humphrey, P. P. A. (1989) Br. J. Pharmacol.98, 29-31.

Proc. Nad. Acad. Sci. USA 89 (1992)

Dow

nloa

ded

by g

uest

on

Aug

ust 1

, 202

0