Cloning, functional expression, and mRNA tissue distribution of the rat 5-hydroxytryptamine1A receptor gene.

G protein-coupled receptors comprise a family of genes that share significant sequence similarity. We have screened a rat genomic library under low stringency hybridization conditions with the coding portion of the hamster beta 2-adrenergic receptor gene to isolate new members of this gene family. We show that one of these clones, clone D, codes for a 5-hydroxytryptamine1A (5-HT1A) binding site since: 1) it possesses an intronless open reading frame encoding a protein with seven putative transmembrane domains and 89% amino acid identity with the human 5-HT1A receptor (G21); 2) when transfected into Ltk- cells, it expresses a ligand-binding site with the pharmacology of the 5-HT1A receptor subtype, including 5-HT- and spiroxatrine-displaceable binding of 8-hydroxy-(2-(N,N-di[2,3-3H]propylamino)-1,2,3,4-tetrahydronaphthalene (KH = 0.8 nM). We further show that clone D encodes a functional receptor because its binding site interacts with G proteins and because it mediates agonist-induced inhibition of basal and stimulated cAMP accumulation in transfected GH4C1 pituitary cells. Finally, we have analyzed the tissue distribution of 5-HT1A receptor mRNA in rat brain and have found that 5-HT1A mRNA is present with the expected distribution of the 5-HT1A receptor (highest in septum and hippocampus) but is present as three RNA species (3.9, 3.6, and 3.3 kilobases). These studies represent the first characterization of receptor function and brain distribution of the cloned rat 5-HT1A receptor.

G protein-coupled receptors comprise a family of genes that share significant sequence similarity. We have screened a rat genomic library under low stringency hybridization conditions with the coding portion of the hamster &adrenergic receptor gene to isolate new members of this gene family. We show that one of these clones, clone D, codes for a B-hydroxytrypta-minelA (B-HTIA) binding site since: 1) it possesses an intronless open reading frame encoding a protein with seven putative transmembrane domains and 69% amino acid identity with the human S-HTIA receptor (G21); 2) when transfected into Ltk-cells, it expresses a ligand-binding site with the pharmacology of the 5-HTIA receptor subtype, including 5-HT-and spiroxatrine-displaceable binding of 8-hydroxy-(2-(N,Ndi[2,3-3H])propylamino)-1,2,3,4-tetrahydronaphthalene (KH = 0.8 nM) . We further show that clone D encodes a functional receptor because its binding site interacts with G proteins and because it mediates agonist-induced inhibition of basal and stimulated CAMP accumulation in transfected GH& pituitary cells. Finally, we have analyzed the tissue distribution of 5-HTIA receptor mRNA in rat brain and have found that 5-HTIA mRNA is present with the expected distribution of the S-HTIA receptor (highest in septum and hippocampus) but is present as three RNA species (3.9, 3.6, and 3.3 kilobases).
These studies represent the first characterization of receptor function and brain distribution of the cloned rat 5-HTIA receptor.
The cloning of membrane-bound neurotransmitter, peptide, and hormone receptors has been complicated because of the difficulties inherent in the solubilization and purification of these proteins. The traditional cloning approach of protein purification, isolation of peptide sequence, and construction of oligonucleotides for probing genomic or cDNA libraries have been used successfully to clone the hamster &adrenergic receptor (Dixon et al., 1986) and the porcine Ml (Kubo et ai, 1986) and M2 (Peralta et al., 1987)  larities between these G protein-coupled receptors for small cationic agonists (noradrenaline and acetylcholine) at both the nucleotide and amino acid levels (Dohlman et al., 1987). Based on the similarity between these receptors, we hypothesized that G protein-coupled receptors for other important neurotransmitters, especially small cationic ligands such as noradrenaline, dopamine, or serotonin (&HT),' may have a similar structure. We have used the hamster &adrenergic receptor coding sequence to probe a rat genomic library under low stringency conditions and isolate new clones with presumed nucleotide similarity to the &adrenergic receptor. Using this homology probing approach, we have reported previously the cloning of a rat dopamine-DZ receptor cDNA (Bunzow et al., 1988), and others have reported the cloning of G21 (the human 5-HT1* receptor gene Fargin et al., 1988)) and the human fll-adrenergic receptor gene . We report here the cloning, expression, and functional characterization of the rat 5-HTlA receptor gene. We present new results showing the brain distribution of multiple mRNA transcripts of this receptor, and we provide evidence that activation of the cloned 5-HTIA receptor causes inhibition of CAMP accumulation.

EXPERIMENTAL PROCEDURES
Isolation of Clone D-A rat genomic library in EMBL-3 was probed as described (Bunzow et al., 1988)  as above and then diluted 3-fold by the addition of 500 ~1 of ice-cold binding buffer containing unlabeled DPAT (final concentration of 1 FM). After various times at 4 "C, the samples were filtered as above and counted (Koch and Schonbrunn, 1984). Membrane proteins were solubilized overnight in 0.1 M NaOH and quantified by the method of Bradford (1976). For CAMP assay, GH, cells were plated in six-well 35-mm dishes 3-7 days prior to experimentation.
Cells were preincubated in 2 ml/ well warm F-10 + 0.1% (-)-ascorbic acid + 20 mM Tris (pH 7.2) (FAT) for 5-10 min followed by addition of 1 ml/well of FAT + 100 pM 3-isobutyl-1-methylxanthine + experimental compounds and incubated for 30 min at 37 "C. Experimental compounds were diluted lOOO-fold from stock solutions made immediately prior to assay. The final ethanol concentration never exceeded O.l%, a concentration without effect on basal or VIP-enhanced CAMP or prolactin levels in GH,ZR7 cells. Media were collected, and the cells were lysed immediately in 1 ml of boiling water. Cell lysates and media were centrifuged (2000 x g, 10 min, 4 "C) and the supernatants collected for assay as cell extracts.
Cell extracts and media samples were frozen at -20 "C until assay, if not assayed immediately.
CAMP was assayed by a specific radioimmunoassay (ICN) as described (Dorflinger and Schonbrunn, 1983) with antibody used at 1:500 dilution. After 16 h of incubation at 4 "C, 20 ~1 of 10% bovine serum albumin and 1 ml of 95% ethanol were added consecutively to precipitate the antibodyantigen complex.
Standard curves showed It&, of 0.5 f 0.2 pmol using CAMP as standard.
For pretreatment with pertussis toxin (Sigma), cells were treated with aseptic toxin dissolved in sterile water (100 rg/ml) and maintained in growth medium for 16-20 h. The final concentration of toxin used (100 rig/ml) has been shown to elicit maximal blockade of inhibition of adenylyl cyclase by somatostatin in GHICl cells (Koch et al., 1985). The plates were rinsed immediately prior to experiments in Hepes-buffered medium and processed as described above. RNA Isolation and Northern Blot Analysis-Brain or peripheral tissues were dissected (Gispen et al., 1972) and extracted with guanidinium thiocyanate (Fluka), centrifuged (33,000 rpm, 16 h) through a 1.7-g/ml CsCl pad, and the pellets extracted with phenol/chloroform and ethanol precipitated (Chirgwin et al., 1979). RNA was resuspended and quantitated by UV absorbance at OD 260 nm. For Northern blots, RNA was denatured in glyoxal/dimethyl sulfoxide (1 h, 50 "C) and run on a 1% agarose gel in 10 mM sodium phosphate. RNA was blotted overnight onto nylon membrane (N-bond, Amersham) and immobilized by baking at 80 "C for 2 h. The Northern blots were prehybridized overnight at 42 "C in 50% formamide, 0.2% polyvinylpyrrolidone (Mr 40,000), 0.2% Ficoll (M, 400,000), 0.2% bovine serum albumin, 50 mM Tris (DH 7.5). 1 M NaCl. 0.1% sodium pyrophosphate, 1% SDS, 100 pg/ml sonicated and denatured salmon sperm DNA. For hybridization, 5 X 10" dpm/ml random primed 32Plabeled 0.9-kb BalI/PuuII fragment of clone D (l-2 x lo9 dpm/pg) was added to fresh prehybridization solution.
The blots were hybridized for 28 h at 42 "C in this solution.
The blots were then washed twice in 2 x SSC, 1% SDS at 55 "C for 10 min; twice in 0.2 X SSC, 1% SDS at 65 "C for 15 min; and once in 0.1 X SSC, 1% SDS at 65 "C for 15 min and exposed to x-ray film overnight at -80 "C, with intensifying screen.

AND DISCUSSION
Isolation of a Rat G2l Homologue-A rat genomic DNA library in XEMBL-3 was probed under low stringency conditions with the 32P-labeled 1.3-kb Hind111 fragment of the hamster &-adrenergic receptor gene, which contains most of the coding sequence of the gene (Dixon et al., 1986). Since several known G protein-coupled receptors lack introns, we attempted to isolate clones containing complete coding sequences for homologous receptors. Three of the clones that hybridized most strongly to the &adrenergic receptor gene were isolated, purified, subjected to Southern blot analysis, and found to be derived from the same gene. Partial sequencing of one of these clones (X clone D) revealed moderate nucleotide and amino acid similarities to hydrophobic domains I and II of the &adrenergic receptor, and this clone was studied in detail. A 2.3-kb BamHI/PstI fragment was found to hybridize strongly with the &-adrenergic receptor gene. The restriction map and full sequence of this fragment were determined (Fig. 1 1C) possesses standard characteristics of the G protein-coupled receptor gene family (Dohlman et al., 1987) including seven hydrophobic domains; several N-linked glycosylation motifs in the N-terminal domain; a potential phosphorylation site in the third cytoplasmic loop; and most of the amino acid motifs found to be conserved among G protein-coupled receptors, e.g. conserved sequences in hydrophobic domain II (Ala/ Val-Ile-Ala-X-Asp-Arg-Tyr-X-Ala-Ile).
Like receptors thought to couple to inhibitory G proteins (e.g. the dopamine-Dz receptor (Bunzow et al., 1988)), clone D has a long (128 amino acids) third cytoplasmic loop and a short (16 amino acids) C-terminal cytoplasmic tail. The hybridization between the hamster &-adrenergic clone and clone D was due presumably to the nucleotide similarity (90% identical) of a contiguous stretch of 19 out of 20 identical amino acids in predicted transmembrane domain VI. The predicted amino acid sequence of the open reading frame of rat genomic clone D was most closely related to the PI-adrenergic  and &adrenergic receptors (50 and 43% identity between hydrophobic domains, respectively) and somewhat homologous (40-41% identity between hydrophobic domains) to the rat 5-HTlc (Julius et al., 1988), 5-HTz (Pritchett et al., 1988, and dopamine-Dz (Bunzow et al., 1988) receptor clones. By contrast, the sequence of clone D is 89% identical to and only 1 amino acid longer than G21 Fargin et al., 1988), the human 5-HT1* receptor gene (Fig. 1C). The degree of amino acid identity between this rat clone (clone D) and the human 5-HTIA receptor is similar to the identity between rodent muscarinic cholinergic (Bonner et al., 1988) or adrenergic receptor (Dohlman et al., 1987) subtypes and the respect.ive human counterparts (i.e. from 89 to 98% amino acid identity).
As observed for muscarinic cholinergic and adrenergic receptor homologues, there was nearly complete (98.6%) amino acid conservation between clone D and the human 5-HTlA receptor in hydrophobic domains. The third cytoplasmic loop, a region of low conservation between clone D and the human 5-HTiA receptor, retained proline-rich segments and a putative phosphorylation site in both clones. Positively charged regions bordering the fifth and sixth hydrophobic domains, which are thought to confer G protein selectivity (Strader et al., 1987;Kobilka et al., 1988;O'Dowd et al., 1988), were also present in the third cytoplasmic loop of both clones. Based on this homology, it appeared likely that clone D encodes the rat homologue of G21, i.e. the rat 5-HTl~ receptor.
Binding Characteristics of Expressed Clone D-In order to test directly the hypothesis that clone D encodes the rat 5-HT,* receptor, the BumHI/PstI fragment of clone D was placed in the pZEM-3 expression vector, 3' to the mouse metallothionein promotor and 5' to the hGH polyadenylation signal. This plasmid, named pZEM-Dsp, and a plasmid containing the neo-resistance gene (pRSV-neo) were co-transfected into mouse Ltk-fibroblast-like cells (for binding assay) or GH,CI pituitary cells (for CAMP assay) which had been shown not to express clone D corresponding mRNA. Stable transfectant colonies were selected by growth in G418-containing media, and total RNA from lo-12 individual clones was prepared for Northern blot analysis. Two cell lines (LZD7 and GH4ZDlo) expressing the highest levels of clone D mRNA were studied further. The basal expression of clone D mRNA was much greater in the LZDT cells than in GH4ZDlo cells or rat hippocampus (data not shown). In GH,ZDlo cells, pretreatment with 100 ELM Zn2+ enhanced markedly the levels of clone D mRNA, consistent with transcriptional regulation of the clone DsP insert by the zinc-sensitive mouse metallothionein promotor of the transfected pZEM-Dsp plasmid (data not shown).
Membranes from LZDT ( Fig. 2A)  whereas wild-type Ltk-and GHICl cell membranes possessed no detectable specific [3H]DPAT binding (data not shown). On average, the binding capacity for LZD7 membranes was 1.89 f 0.08 pmol/mg of protein (n = 3). Scatchard analysis of the binding of [3H]DPAT to LZD7 membranes was fit by a model assuming the presence of two affinity states of the receptor ( Fig. 2A, inset), where on average, 32 + 9% of the sites were in the high affinity form (Ku = 0.78 + 0.28 nM) and the remaining sites in a low affinity form (average & = 8.7 f 1.8 nM). Membranes from GH,ZD,, cells pretreated with Zn" also expressed specific [3H]DPAT binding with a similar K. (3.4 + 0.77 nM) but about half the receptor density (B,,, = 1.1 + 0.3 pmol/mg of protein) of LZD7 membranes, consistent with the lower levels of clone D mRNA in GH4ZDla cells (data not shown). Thus, transfection with a plasmid vector directing the synthesis of clone D mRNA led to the presence of [3H]DPAT-binding sites in two different 5-HTIA receptor-negative cell lines, indicating that the expressed mRNA encodes a 5HT1*-binding site.
To obtain further evidence that the binding site expressed in LZD7 cells had the pharmacologic profile of the 5-HTlA subtype, competition experiments were performed using a variety of agonists and antagonists, shown in Fig. 2B. Both DPAT and 5-HT displaced [3H]DPAT binding at nM concentrations, whereas dopamine was effective only at PM concentrations. Agonists (such as DPAT and 5HT) displayed shallow displacement curves that were best fit by assuming the presence of two binding affinity components (Hall et al., 1985;Fargin et al., 1988). Antagonists displayed steeper displacement curves with Hill coefficients closer to 1, and were fit by assuming a single component of binding. Average KD or Kr values calculated from 50% inhibitory concentration (IC& values of three of four independent trials for each compound are listed in Table I. For agonists, both high and low affinities were determined. The calculated affinities for DPAT (i.e. KH = 0.4 nM and KL = 16 nM) and the proportion of high affinity sites (54 k 10%) were similar to binding affinities of [3H] DPAT (above) obtained from saturation analysis ( Fig. 2A Fig. 4B were converted to K, values as described (Cheng and Prusoff, 1973 (Fargin et al., 1988)) and to affinities obtained in rat and human brain membrane preparations (KH = 3 nM (Gozlan et al., 1983;Hall et al., 1985;Hoyer et al., 1986)). However, in membranes prepared from GH4ZDlo cells pretreated for 16 h with 50 rig/ml pertussis toxin, competition with unlabeled DPAT indicated the sole presence of the low affinity binding site (K, = 14 nM). This result suggests that high affinity binding of DPAT is generated by interaction with a G protein that can be dissociated by ADP-ribosylation of the a-subunit with pertussis toxin to yield the low affinity state (Dolphin, 1987). This is an initial indication that the cloned receptor is coupled functionally to G proteins. Spiroxatrine, a potent antagonist at the 5-HTIA receptor (Nelson and Taylor, 1986), antagonized [3H]DPAT to LZD-, membranes with a Kr of 25 nM, slightly greater than that reported in rat brain membranes (K, = 4 nM). The least potent antagonist tested, 5-HTz-selective ketanserin (Leysen et al., 1982), displaced [3H]DPAT binding only at very high concentrations.
Specific binding of DPAT and its inhibition by spiroxatrine argue against identification of the receptor as the 5-HTls subtype, since neither of these compounds binds to 5-HT1~ sites (Hoyer et al., 1986). Surprisingly, the partially selective 5-HTIB receptor agonist CGS12066B (Neale et al., 1987) with a lo-fold higher potency than expected at the 5-HT1* receptor. Small discrepancies, as above, may arise from differences in membrane environments or post-translational processing between nonneuronal cultured cells and neurons. Nevertheless, the specific binding of 5-HTIA receptor agonist [3H]DPAT and its competition by antagonist spiroxatrine favor classification of the binding site expressed from pZEM-Dsp as the 5-HT1* subtype. Other antagonists that are less potent than spiroxatrine inhibited binding of [3H]DPAT to LZDT membranes at the concentrations known to be effective in brain membrane preparations.
Spiperone and propranolol, which are known to act more potently at dopamine-Dz and ,&adrenergic receptors, respectively, inhibited [3H]DPAT binding with the expected affinity of the 5-HT1* receptor (Hoyer et al., 1986;Nelson and Taylor, 1986;Alexander and Wood, 1987). The ability of these antagonists to function at the 5-HT1* receptor correlates with structural similarities between the 5-HTlA receptor and /3-adrenergic or dopamine-Dz receptors. By using chimeric receptor constructs (Kobilka et al., 1988;Frielle et al., 1988), it has been shown that transmembrane domains VI and VII are the most important determinants of antagonist specificity in adrenergic receptors. Clone D and the human 5-HTiA receptor  possess highest amino acid sequence similarity with &adrenergic and dopamine-D2 receptors in precisely these regions, and this homology may account for the observed overlap of antagonist specificity between these receptors. Taken together, the affinities and rank order of potency of the compounds tested are strikingly similar to that presented for the expressed human 5-HTIA receptor (Fargin et al., 1988) and for the native rat 5-HTIA receptor. We conclude that the open reading frame contained within the BamHI/PstI fragment of clone D encodes a rat 5 HTIA-binding site.

Modulation of r3H]DPAT
Binding by Guanine Nucleotides-The existence of two agonist-binding affinity states and modulation of DPAT affinity by pertussis toxin pretreatment suggested that the gene product of clone D may interact with G proteins, which are known to modulate agonist binding to this family of receptors (Dohlman et al., 1987). Membranes prepared from LZD7 cells were incubated with ["HIDPAT in the presence of increasing concentrations of guanine nucleotides, and specific binding was assessed (Fig. 3A) respectively, whereas ATP (10 PM) had no effect (not shown). GTP hydrolysis upon binding to the G protein (Dolphin, 1987) may explain why the EC& value for GTP is higher than for GTPrS, which is not hydrolyzed. Guanine nucleotides are thought to act by binding to G proteins associated with the 5-HT1,, receptor (Dolphin, 1987), causing dissociation of the G protein and leaving the receptor in a low affinity agonist state (Stratford et at, 1988). Scatchard analysis of [3H]DPAT binding to GH4ZDlo membranes in the presence of 30 nM GTPyS indicates the presence of a single binding site with KD = 8.5 + 1.5 nM (n = 2), the value obtained for low affinity binding of DPAT.
In order to differentiate between a competitive inhibition of binding (where agonist dissociation rate is unaltered) and a GTP-induced decrease in agonist affinity, the dissociation of t3H]DPAT was measured in LZDi cell membranes in the absence and presence of GTPyS (Fig. 3B). observed between high and low affinity states of [3H]DPAT binding. These data indicate that the 5-HTIA receptor interacts with G proteins (Hall et al., 1985;Stratford et al., 1988;Fargin et al., 1988) such that agonist affinity is decreased in the presence of guanine nucleotides due to an enhanced dissociation rate, as observed for other G protein-coupled receptors (Dolphin, 1987). Inhibition of CAMP Accumulation-To test whether the cloned rat 5-HTIA-binding site is a functional receptor able to transduce biological actions, the GH,ZDlo transfectant was used to study the actions of 5-HT on CAMP accumulation. The recipient GH& cells are a pituitary-derived strain (Tashjian, 1979) that provides a highly responsive system in which to study receptor signaling by a variety of transduction mechanisms, including activation or inhibition of adenylate cyclase (Dorflinger and Schonbrunn, 1983). Accumulation of CAMP was measured in the presence of 100 pM 3-isobutyl-lmethylxanthine to inhibit CAMP degradation by phosphodiesterases (Fig. 4). Thus, the observed changes in CAMP accumulation are due primarily to changes in CAMP synthesis. Incubation of intact GH,ZDlo cells with a maximal concentration of vasoactive intestinal peptide (VIP) (200 nM) induced a 2-fold increase in intracellular CAMP levels and a 3-fold increase in extracellular CAMP levels, similar to changes observed in wild-type GH& cells. Addition of 200 nM 5-HT reduced basal CAMP accumulation by 30-60% and reduced intra-and extracellular CAMP accumulation in the presence of VIP to nearly basal levels. In wild-type GH& cells, 5-HT had no effect on VIP-induced enhancement of CAMP accumulation.
The inhibitory actions of 5-HT were reversed by the 5-HT,;\ receptor-selective antagonist, spiroxatrine (2 PM). These data indicate that the cloned rat 5-HTiA receptor is negatively coupled to adenylate cyclase to reduce basal and stimulated CAMP accumulat,ion in GH& cells. The native 5-HT,, receptor is a member of a class of receptors (Nicoll, 1988) coupled by pertussis toxin-sensitive G proteins to inhibition adenylate cyclase (DeVivo and Maay ani, 1986;Dumuis et al., 1987) as well as membrane hyperpolarization by opening of potassium channels (Andrade and Nicoll, 1987;Colino and Halliwell, 1987;Innis et al., 1988). Inhibition of both basal and VIP-enhanced accumulation of CAMP in media by activation of t.he cloned rat 5-HTiA receptor in GH.,ZD,,, cells was almost completely inhibited by pertussis toxin pretreatment (Table II). There was a consistent lowering of basal CAMP accumulation by toxin pretreatment, an effect that has not been observed in several other GH,C, subclones (Koch et al., 1985;Albert et al., 1990). However, enhancement by VIP (5-fold uersus over 6-fold after toxin) was not inhibited by toxin treatment. These studies show that the gene product of clone D is equivalent functionally to the 5-HT].., receptor, inhibiting adenylyl cyclase by coupling to pertussis toxin-sensitive G protein(s). Studies in membranes from rat or guinea pig brain have suggested that in addition to inhibiting stimulated adenylate cyclase activity, activation of 5-HT,.\ receptors stimulates basal adenylate cyclase activity (Markstein et al., 1986;Shenker et al., 1987). We observed no stimulation of CAMP accumulation by the cloned 5-HTIA receptor in GH,ZDlo cells, cells in which VIP enhanced CAMP accumulation.
In other experiments, DPAT inhibited CAMP accumulation as efficaciously as 5-HT. Thus, with respect to inhibition of adenylate cyclase, DPAT is a full agonist in GH4ZDlo cells. We find no evidence for stimulation of adenylate cyclase by activation of the cloned rat 5-HT,, receptor. Note that while this paper was under revision, Fargin et al. (1989) have shown that the cloned human 5-HTIA receptor expressed in HeLa or COS-7 cells also inhibits CAMP formation by a pertussis toxinsensitive pathway. It remains possible that other 5-HT receptor subtypes exist (Peroutka, 1988;Dumuis et al., 1988) which mediate stimulation of adenylate cyclase activity by serotonin. mRNA Distribution in Brain-To analyze which rat tissues express the 5-HTIA gene, Northern blot analyses of mRNA isolated from brain and pituitary regions were conducted (Fig.  5). The BalI/PuuII fragment of clone D, containing the sequence between the second and sixth predicted transmembrane segments and including the unique third cytoplasmic region, was labeled by random priming and used as a probe to detect complementary mRNA sequences under high stringency hybridization conditions. Three species of clone D mRNA (3.9, 3.6, and 3.3 kb) were identified consistently in Actions ofpertussis toxin pretreatment on CAMP accumulation in GH4ZD,o cells revealed a single species when probed for other G proteincoupled receptors (data not shown). To eliminate the possibility that additional RNA forms were due to hybridization of contaminants from other segments of pGEM-Dsp, the BalII PvuII fragment was subcloned and reisolated for use as a probe in Fig. 5. The mRNA species of clone D has similar relative abundance in all brain regions examined, with a conserved rank order of 3.9-> 3.6->> 3.3-kb form. Thus, there appears to be no markedly region-specific expression of one or other species of mRNA but rather a homogeneous rank order of expression of all species in each brain region. Southern analysis of rat genomic DNA digested with BamHI/PstI has revealed a single 2.3-kb species, indicating the absence of other closely related genes (data not shown). Since we have shown that the rat 5-HT1~ gene is intronless, the different species of clone D mRNA could represent results from different transcriptional start sites or different polyadenylation sites. The equivalent distribution of the different mRNA forms in all brain regions studied suggests transcriptional regulation by a common tissue-specific promotor. Another possibility is that the other mRNA species encode different receptor subtypes (e.g. the 5-HTis receptor) that have an mRNA distribution similar to that of the 5-HTlA receptor. This seems unlikely since at the stringency of hybridization used, only products containing sequences very closely related to the clone D probe should be observed. Southern analysis of rat genomic DNA digested with BamHI/PstI has revealed a single 2.3-kb species, indicating the absence of other closely related genes (data not shown). It is interesting that in man, where only one species of 5-HTiA receptor mRNA has been described, 5-HT1a receptors are absent (Hoyer et al., 1986). Further DNA sequencing of the gene (and possibly the cDNAs) is necessary to determine the origin and nature of the different species of clone D mRNA. However, as described above, the tissue distribution of RNA hybridizing to clone D is consistent with the known distribution of 5-HT1~ receptors.

CONCLUSION
Rat genomic clone D, isolated on the basis of its nucleotide similarity to the hamster &-adrenergic receptor, has been shown to encode a functional 5-HTIA receptor based on the following criteria. The nucleotide sequence of clone D codes for a protein similar in structure to other G protein-coupled receptors and nearly identical in amino acid sequence to the human 5-HTIA receptor. Transfection of clone D into two different eucaryotic cell lines leads to expression of a binding site with the pharmacological profile of a 5-HTiA receptor. As for other receptors that interact with G proteins, the binding of agonists to the expressed clone is shifted to a low affinity state by guanine nucleotides and by pretreatment with pertussis toxin. We find that the affinity shift is due to a guanine nucleotide-induced enhancement of agonist dissociation rate. The cloned rat 5-HT1A receptor is coupled to inhibition of CAMP accumulation and is thus functional. This action of the receptor is blocked by pertussis toxin pretreatment, indicating mediation by coupling to one or more pertussis toxinsensitive G proteins.
We present the first studies of the brain distribution of 5-HTIA receptor mRNA, which is present as three species in limbic structures and medulla. 5-HTiA receptor mRNA is also present at high levels in the thalamus and mesencephalon.
The observed mRNA distribution is analogous to the known distribution of native rat 5-HT1*-binding sites. The presence of three mRNA species has not been observed in human peripheral tissues, and the functional significance of the phenomenon remains to be elucidated.