Molecular cloning and expression of the cDNA for a novel alpha 1-adrenergic receptor subtype.

A novel alpha 1-adrenergic receptor subtype has been cloned from a bovine brain cDNA library. The deduced amino acid sequence is that of a 466-residue polypeptide. The structure is similar to that of the other adrenergic receptors as well as the larger family of G protein-coupled receptors that have a presumed seven-membrane-spanning domain topography. The greatest sequence identity of this receptor protein is with the previously cloned hamster alpha 1B-adrenergic receptor being approximately 72% within the presumed membrane-spanning domains. Localization on different human chromosomes provides evidence that the bovine cDNA is distinct from the hamster alpha 1B-adrenergic receptor. The bovine cDNA clone expressed in COS7 cells revealed 10-fold higher affinity for the alpha 1-adrenergic antagonists WB4101 and phentolamine and the agonist oxymetazoline as compared with the alpha 1B receptor, results similar to pharmacologic binding properties described for the alpha 1A receptor. Despite these similarities in pharmacological profiles, the bovine alpha 1-adrenergic receptor is sensitive to inhibition by the alkylating agent chloroethylclonidine unlike the alpha 1A-adrenergic receptor subtype. In addition, a lack of expression in tissues where the alpha 1A subtype exists suggests that this receptor may actually represent a novel alpha 1-adrenergic receptor subtype not previously appreciated by pharmacological criteria.

A novel al-adrenergic receptor subtype has been cloned from a bovine brain cDNA library.
The deduced amino acid sequence is that of a 466-residue polypeptide. The structure is similar to that of the other adrenergic receptors as well as the larger family of G protein-coupled receptors that have a presumed sevenmembrane-spanning domain topography.
The greatest sequence identity of this receptor protein is with the previously cloned hamster alB-adrenergic receptor being -72% within the presumed membrane-spanning domains.
Localization on different human chromosomes provides evidence that the bovine cDNA is distinct from the hamster alB-adrenergic receptor. The bovine cDNA clone expressed in COS7 cells revealed lo-fold higher affinity for the al-adrenergic antagonists WB4101 and phentolamine and the agonist oxymetazoline as compared with the (YlB receptor, results similar to pharmacologic binding properties described for the alA receptor.
Despite these similarities in pharmacological profiles, the bovine al-adrenergic receptor is sensitive to inhibition by the alkylating agent chloroethylclonidine unlike the alA-adrenergic receptor subtype.
In addition, a lack of expression in tissues where the a1A subtype exists suggests that this receptor may actually represent a novel al-adrenergic receptor subtype not previously appreciated by pharmacological criteria.
Epinephrine and norepinephrine mediate their effects via binding to adrenergic receptors (aI, alp, &, p2) (1,2). These receptors are encoded by different genes and are members of a much larger family of guanine nucleotide regulatory protein (G protein)-coupled receptors (3). Molecular cloning studies have revealed a growing heterogeneity of receptor subtypes. For example five different muscarinic cholinergic (4)(5)(6)(7)(8), three serotonergic (g-12), and two a*-adrenergic receptor subtypes (13,14) have been recently identified by molecular cloning. Heterogeneity of al-adrenergic receptors (LY~* and alB sub-* This work was supported in part by the American Society of Anesthesiologists/Association of University Anesthesiologists fellowship award, the Deutsche Forschungsgemeinschaft fellowship award, and by Grant 5R37-HL-16037 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This  and antagonists (e.g. WB4101 and phentolamine) as well as on different requirements of al-adrenergic receptor-induced responses for extracellular calcium (20). Recently we reported the cloning of the cDNA which encodes the hamster cul-adrenergic receptor purified from DDT1-MF, cells (21). The pharmacological properties of this CQadrenergic receptor resemble those described for the @Breceptor subtype. Here we present the cloning, sequencing, and expression of another oc,-adrenergic receptor subtype from bovine brain. This receptor shows the pharmacological properties proposed for the cYIA-adrenergic receptor subtype, but on the basis of a lack of expression in tissues where the alA subtype exists and its sensitivity to CEC', it may actually represent a novel a!,-adrenergic receptor subtype not previously appreciated by pharmacological criteria.

RESULTS
Screening of a human leukocyte genomic library with the labeled oligonucleotide probe constructed from the hamster cu,B-adrenergic receptor cDNA (nucleotides 1028-1094) identified one clone under high stringency conditions (0.2 x SSC at 60 "C). This clone contained a 0.86-kb EcoRI/HindIII restriction fragment with an open reading frame between nucleotides 459 and 855. The open reading frame encoded a peptide where the first 43 residues were 67% identical with the hamster oci-adrenergic receptor putative seventh transmembrane domain and the beginning of the carboxyl terminus (nucleotides 951-1077 of the hamster ai-adrenergic receptor) (21). Thereafter, the open reading frame diverged completely from the hamster ol,-adrenergic receptor sequence. In order to obtain a full-length clone, a size-selected (2.0-4.4 kb) bovine brain cDNA library was screened using as a probe a 0.32-kb PuuII/HindIII restriction fragment containing most of the open reading frame of the human genomic clone previously obtained. A single clone with a 3.1-kb insert was isolated under high stringency conditions (0.2 x SSC at 60 "C). This bovine clone contained the same PuuII/HindIII restriction fragment as the human genomic clone. An initiator methionine at nucleotide 741 of the bovine cDNA clone started a 1398 bp open reading frame.  (967 bp)). The open reading frame encodes a polypeptide of 466 amino acids with a calculated molecular mass of 51 kDa. Comparison of the deduced amino acid sequence of the bovine clone with that of previously cloned G protein-coupled receptors revealed striking amino acid identity in the putative transmembrane domains of the various adrenergic receptors but most strikingly with the hamster cui-adrenergic receptor. The percentage identities in the putative transmembrane domains with each adrenergic receptor are the following: hamster (Yin, 72.1%; human (r&4, 43.2%; human (~~-C10, 41.5%; human /&, 43.2%; human pZ, 42.1% (13,14,21,25,34,35). The level of amino acid identity within the putative transmembrane domains between the bovine cDNA clone and the hamster cylB-adrenergic receptor (72.1%) is similar to the level of amino acid identity within the transmembrane domains between the two a*-adrenergic receptor subtypes (75%) or between the ,6,-and &adrenergic receptor subtypes (75%). Comparison of the entire bovine cDNA-deduced amino acid sequence and the hamster (YlBadrenergic receptor (Fig. 2, solid circles are identical amino acids) reveals that the NH* terminus (27% amino acid identity), the COOH terminus (12% amino acid identity), and the third cytoplasmic loop (50% amino acid identity) represent the most divergent domains. These regions are also ones that differ the most in length and amino acid composition among other adrenergic receptors and G protein-coupled receptors. Since amino acid identity between hamster and human (YlB' adrenergic receptors is 99%,' this suggests that the bovine clone does not encode the bovine homolog of the hamster alB' adrenergic receptor but rather a different oc,-adrenergic receptor subtype. Three potential sites for N-linked glycosylation are present in the NH2 terminus (asparagine residues 7, 13, and 22). Several threonines and serines are present in the second and third cytoplasmic loops of this cDNA clone, representing potential sites for protein kinase C phosphorylation. A con-sensus sequence for protein kinase A phosphorylation (amino acid residues 211-215) is present in the bovine a,-adrenergic receptor subtype cDNA in an analogous position to the consensus site seen in the hamster arls-adrenergic receptor cDNA. Phosphorylation has been observed as a mechanism of regulation for the ells-adrenergic receptor (37,38). The presence of these consensus sequences suggests that this new cyl-adrenergic receptor may also be regulated by phosphorylation.
To further confirm that the bovine cDNA obtained represented a different gene product than the hamster LY~ -adrenergic receptor, human somatic cell hybridization analysis was performed using as probes the 0.32-kb PuuII/HindIII fragment of the bovine cDNA and the 0.7-kb XhoI/BanHI fragment of the hamster cY1-adrenergic receptor. These studies showed that while the gene corresponding to the hamster aIadrenergic receptor is located on human chromosome 5, the gene corresponding to the bovine cDNA is located on human chromosome K3 To further assess the functional identity of the cDNA isolated, a 2.4-kb fragment representing the coding region and the entire 3'-untranslated region was inserted into the expression vector pBC12BI and used to transfect COS-7 cells. The COS-7 cells transfected with the vector containing the bovine cDNA were able to bind the al-adrenergic antagonist [lz51] HEAT with high specific activity (15 pmol/mg of protein) and with an affinity (60-70 PM) similar to that of the cloned hamster al=-adrenergic receptor. No binding activity was detected in untransfected COS-7 cells. Analysis of adrenergic agonist and antagonist competition curves showed the appropriate pharmacology for al-adrenergic receptor binding. To compare ligand binding characteristics between the newly cloned bovine ol,-adrenergic receptor and the hamster (Ylsadrenergic receptor, competition curve analysis of agonists and antagonists for [lz51]HEAT binding was performed on membranes prepared from COS-7 cells transfected with the cDNA for either the bovine cY1-adrenergic receptor subtype or the hamster LulB-adrenergic receptor subtype. Comparison of the K, values for different compounds (Table I) reveals striking differences between these two crl-adrenergic receptors. Specifically, the bovine al-adrenergic receptor subtype has more than lo-fold higher affinity for the (Y antagonists WB4101, phentolamine, corynanthine, and indoramin, as well as for the agonists oxymetazoline and methoxamine when compared with the hamster al-adrenergic receptor (Fig. 3).
The LYE selective antagonist prazosin shows the same affinity for both cY,-adrenergic receptor subtypes, while epinephrine and norepinephrine are slightly more potent at the hamster culB-adrenergic receptor. Recently the existence of two (Yeadrenergic receptor subtypes ((Y~* and (Y1s) has been suggested on the basis of differences in the affinities of the antagonists WB4101 and phentolamine and the agonist oxymetazoline for the two putative receptor subtypes (15)(16)(17)(18)(19)(20), with the (YlAadrenergic receptor subtype having lo-fold higher affinity for WB4101 than the (YIB. The lo-fold higher affinities of WB4101, phentolamine, and oxymetazoline for the bovine versus the hamster al-adrenergic receptor subtypes are in close agreement with those described for the n1A and alBadrenergic receptor subtypes, respectively. Ahother approach used to pharmacologically discriminate different al-adrenergic receptor subtypes has been the irreversible inactivation of aI-adrenergic receptor binding by the alkylating derivative of clonidine, CEC (16,17,19). Minneman and others (17,20) have shown that treatment with CEC inactivates 50-60% of the al-adrenergic receptors in rat cerebral cortex. Since the population of receptors left after CEC Solid circles indicate amino acids common to the corresponding position in the hamster ala-adrenergic receptor (21). Transmembrane domains are defined based on hydropathicity analysis (36). Potential N-linked glycosylation sites are shown as crosses.
inactivation shows only high affinity for WB4101, the (YlAadrenergic receptor subtype has been suggested to be insensitive to CEC. To assess the effect of CEC on the bovine and hamster cY1-adrenergic receptors, transfected COS cell membranes were treated with 100 WM CEC for 20 min at 37 "C, and al-receptor ligand binding was measured with a saturating concentration of [lZ51]HEAT. CEC treatment inactivated 95 + 1 and 68 f 3% (mean + S.E., n = 3, p < 0.02) of the CQ binding in COS cells individually expressing the hamster (Y~Band bovine al-adrenergic receptor, respectively. In rat cortex membranes, the same treatment with CEC inactivated 72 f 7.5% of al-adrenergic receptors, in agreement with the values previously reported (17,19,20). These results indicate that the bovine al-adrenergic receptor subtype is less sensitive to CEC inactivation than the hamster oils-adrenergic receptor subtype but more sensitive than has been suggested for the oclA-adrenergic receptor subtype (16,17,19).
In order to explore the tissue expression of the bovine aladrenergic receptor and compare it with that of the hamster alB subtype, we performed Northern blot analysis on poly(A)+-selected mRNA of various rat and bovine tissues as well as utilized the DNA PCR with cDNA prepared from RNA of various bovine tissues. Northern blot analysis of rat mRNA has confirmed that the hamster al-adrenergic receptor previously cloned corresponds to the subtype described as the alB-receptor as indicated by strong hybridization to rat liver and cerebral cortex mRNA (15,17,19). However, no signal was observed with the bovine (Ye subtype in any of the rat tissues (cerebral cortex, pituitary, hippocampus, brainstem, liver, heart, lung, kidney, spleen, aorta, adipose tissue, skeletal muscle, vas deferens) or bovine tissues (cerebral cortex, liver, heart, kidney, lung, adrenal) examined by Northern blot analysis or by PCR (data not shown). In situ hybridization of human dentate gyrus using antisense RNA probes derived from the human homolog of the bovine al-adrenergic receptor subtype revealed a restricted pattern of expression only in the granular cell layer (Fig. 4). These data suggest that the expression of this receptor subtype is either very low or restricted to specific tissues or cellular populations. DISCUSSION We have cloned a cDNA encoding a novel al-adrenergic receptor subtype from bovine brain. Sequence homology with the previously cloned hamster alB-adrenergk? receptor (72.1% identity in the putative transmembrane domains) suggests the bovine cDNA is an al-adrenergic receptor. High affinity for the al-adrenergic antagonist [lz51]HEAT and overall rank order of potencies of agonists and antagonists confirms that the cDNA encodes an Lu,-adrenergic receptor. However, human chromosome analysis provides evidence that the bovine oll-adrenergic receptor (localized to chromosome 8) is distinct from the hamster alB-adrenergic receptor (localized to chromosome 5). In addition, the homology between the bovine aladrenergic receptor and the hamster ollB-adrenergic receptor subtype in the putative transmembrane domains of the receptor is similar to that between az-adrenergic subtypes (75%) and @-adrenergic receptor subtypes (75%), suggesting that these receptors represent distinct cYl-adrenergic receptor subtypes. Ligand binding studies confirm that the bovine (Y,adrenergic receptor subtype and the hamster ale-adrenergic receptor subtype have clear pharmacological differences. Specifically, the a-antagonists WB4101, phentolamine, corynanthine, and indoramin, as well as agonists oxymetazoline and methoxamine, have IO-fold higher affinity for the bovine oiadrenergic receptor subtype compared with the hamster CQ~adrenergic receptor subtype. These data are in close agreement with ligand binding properties of the 0(iA-and LYE,adrenergic receptors described in the literature, indicating that the bovine cDNA encodes a receptor with pharmacologic properties similar to the cYiA-adrenergic receptor subtype.
In contrast to the clear agreement between ligand binding properties reported for the cYiA-adrenergic receptor in the literature and those described for the bovine a,-adrenergic receptor subtype, CEC inactivation studies show some differences. (Y~B-and alA-adrenergic receptors have been described as being sensitive or insensitive to the alkylating agent CEC,respectively (17,19,20). This classification originated from experiments in various tissues where CEC inactivation correlated with a loss of low affinity sites for WB4lOl (corresponding to alB-adrenergic receptors), leaving only high affinity sites (corresponding to alA-adrenergic receptors). Cur data with transfected COS cell membranes treated with 100 PM CEC for 20 min at 37 "C show that the previously cloned (Y~Badrenergic receptor subtype is in fact totally (95%) inactivated by CEC while the bovine Lu,-adrenergic receptor subtype is only partially (68%) inactivated. It is very difficult to conclude based on ligand binding studies with membranes derived from tissue homogenates containing mixtures of al-receptor subtypes whether the alA-receptor is completely insensitive to CEC or whether it is merely less sensitive than the (Yis. Although it seems clear that the arl-receptors remaining after CEC treatment in such preparations have ligand binding properties identical to those of the cloned bovine crl-adrenergic receptor (e.g. high affinity for WB4101), it is also possible that some unspecified proportion of the al,+-receptor in the tissue homogenate was inactivated. Moreover, it is possible that another subtype of cq-adrenergic receptor exists with  A, x-ray autoradiographs of coronal section of the human dentate gyrus hybridized with the ?3labeled antisense strand probe specific for the human homolog of the bovine cu,-adrenergic receptor. Specific labeling occurs over the granular cell layer. B, x-ray autoradiographs of serial sections hybridized with the 35Slabeled sense strand control probe resulted in only background labeling. C, emulsion autoradiographs (Kodak NTB2, hematoxylin and eosin) of the section shown in panel A magnified x 250. Specific hybridization to the granular cell layer (dense band of cells running through picture) is seen; pyramidal cells were not labeled. D, emulsion autoradiographs of the section shown in panel B magnified x 250 result in background labeling only. similar pharmacological properties which may be completely insensitive to CEC inactivation.
In agreement with this hypothesis is the fact that while the pharmacological properties of the bovine cui-adrenergic receptor subtype suggest that it might represent the (YlA subtype, we have been unable to establish this identification by observing mRNA species in tissues where the oclA-adrenergic receptor subtype has been previously described such as rat vas deferens and hippocampus. These observations strongly suggest that the bovine or,-adrenergic receptor represents a novel cYi-adrenergic receptor subtype and that expression of this receptor is quite specialized in either tissue distribution or developmental stage. In fact, even if at present we have not identified the bovine or rat tissues where the bovine CQAR subtype is expressed, in situ hybridization of human dentate gyrus slices reveals the presence of the human homolog of the bovine LQAR specifically in the granular cell layer. Recently the human and rat genes of a fifth muscarinic receptor have been cloned, but the expression of this receptor could not be detected in several rat tissues (8). These observations indicate that receptor heterogeneity might be more complex than predicted by pharmacological studies, and the isolation of different receptor subtypes by molecular cloning provides the most direct tool for the attribution of distinct receptor sub-types to specific tissues. They also indicate the sensitivity and power of techniques such as PCR and in situ hybridization studies in determining the localization of new receptor subtypes. Since we have evidence for the presence of the gene analogous to the bovine alAR subtype in both the human and rat genome, a more extensive investigation by in situ hybridization of human and rat tissues will be required to elucidate the tissue distribution as well as the level of expression of this novel alAR subtype.
The structural features of the new Lu,-receptor subtype are consistent with those of the other adrenergic receptors which have been cloned as well as with the wider family of Gprotein-coupled receptors. The remarkable conservation of sequence within the presumed membrane-spanning domains between the two cri-receptor subtypes is consistent with their very similar ligand binding properties. This sequence conservation extends to those regions of the third cytoplasmic loop and carboxyl-terminal cytoplasmic tail which are presumed to lie closest to the plasma membrane.
Inasmuch as these regions of the receptor molecule have been suggested to be those responsible for coupling to G-proteins, this might suggest that the effector function of these receptor subtypes might also be similar. However it has been speculated that al-receptor subtypes might mediate distinct functions (19).