Distribution of &, Adrenoceptors in Rat Brain Revealed by in Situ

The Journal of Neuroscience, July 1994, 14(7): 4252-4268

Distribution of &, Adrenoceptors in Rat Brain Revealed by in situ Hybridization Experiments Utilizing Subtype-Specific Probes

Vincent A. Pieribone, Anthony P. Nicholas, Ake Dagerlind, and Tomas Hiikfelt Department of Neuroscience, Karolinska Institutet, S-17177 Stockholm, Sweden

The distribution of neurons in the rat CNS that synthesize to any published (Y, receptor distribution pattern, indicating mRNA for the (Y,~,~ and CY,~adrenoceptors was revealed by that this mRNA likely encodes for a novel adrenoceptor. The the in situ hybridization method. Forty-eight-mer DNA probes present findings further expand the heterogeneity of adren- were synthesized to two different and unique regions of both oceptor mRNAs presented in two accompanying studies the (Y,~,~ and (Y,~ mRNAs. Tissue sections from all levels of (Nicholas et al., 1993a,b). This differential distribution of ad- the CNS and some peripheral ganglia were incubated in a renoceptors subtypes provides a framework for the func- hybridization cocktail containing one of these four probes. tional diversity to the apparently widespread, diffuse, and The two mRNAs were expressed in a discrete and often rather homogeneous noradrenergic innervation of the CNS. complementary manner to each other, and identical hybrid- [Key words: adrenaline, noradrenaline, epinephrine, nor- ization patterns were seen for the probes directed against epinephrine, adrenergic, locus coeruleus, cardiovascular] the same mRNA. The (Y,A,D probes hybridized heavily with neurons in the Central noradrenergic(Foote et al., 1983) and adrenergic(Hok- internal granular and internal plexiform layers of the olfactory felt et al., 1984) neurotransmissionis important in a variety of bulb, in layers II-V of most areas of the cerebral cortex, and different neuronal functions including arousal,sleep, mood, car- in the lateral aspect of the lateral amygdaloid nucleus, with diovascular regulation, feeding, pain, and motor output. Phar- pyramidal neurons of CAl-CA4 regions, hilar and granular macological manipulations of these two brain catecholamine neurons of the dentate gyrus, and neurons in the reticular systemsare usedin the clinical treatment of a variety of diseases thalamic nucleus, cranial and spinal motor nuclei, and the including depressionand cardiovascular disorders. Noradren- inferior olivary nucleus. Light labeling was seen in a variety aline-containing fibers are found throughout the entire neuraxis of other regions in the brain and spinal cord. innervating areas of the brain involved in all the above types The (Y,~ probes hybridized heavily with neurons in the mid of CNS functions (Dahlstrom and Fuxe, 1964; Fuxe, 1965a,b; layers of cerebral cortex and with virtually all neurons in the Ungerstedt, 1971; Swanson and Hartman, 1975; Moore and thalamus, except the reticular and habenular nuclei. In ad- Card, 1985) whereasthe adrenergic system has a more limited dition, labeling was seen in the lateral and central amyg- distribution (Hokfelt et al., 1984).The majority ofnoradrenergic daloid nuclei, in brainstem and spinal motor nuclei, over most fibers in the CNS arise from a single source in the pons, the neurons of the dorsal and medullary raphe nuclei and neu- locuscoeruleus (LC; A6 accordingto Dahlstrom and Fuxe, 1964). rons of the intermediolateral cell column in the spinal cord. However, neuronal activity in LC is highly homogeneous;that Light labeling was seen in the septal nucleus, the horizontal is, theseneurons respond in unison to a variety ofafferent inputs limb of the diagonal band, the paraventricular and lateral (Foote et al., 1980, 1983; Aston-Jones and Bloom, 198 1). There- hypothalamic nuclei, the pontine and medullary reticular for- fore, the multitude of effects resulting from noradrenergic ac- mation, and in most laminae in the spinal cord. tivation is likely a result of different postsynaptic responsesto The patterns of labeling obtained with the alB probes re- noradrenaline. In fact, it is possiblethat the various functional semble the labeling seen in previous autoradiographic ligand modalities attributed to noradrenaline and adrenaline are a re- binding studies utilizing “general” or, ligands, while the la- sult of a differential distribution of adrenoceptors. beling patterns seen with the alAiD probes do not correspond NA mediatesits effectsin the CNS via several different G-protein-linked receptors broadly classifiedas a,, 01*,and /3 adrenoceptors (Ahlquist, 1948; Berthelsen and Pettinger, 1977; Ray- Received Aug. 2, 1993; revised Dec. 28, 1993; accepted Jan. 13, 1994. mond et al., 1991). The present study will focus on the CNS We acknowledge the superb technical assistance of Ms. K. Friberg, Ms. S. Nilsson, Ms. K. .&man, and W. Hiort. V.A.P. was supported by a grant from the distribution of only one classof thesereceptors, the 01,receptors. National Science Foundation (INT-8908720) and the Fogarty International Center The o(, adrenoceptorshave been subdivided into allA, ailB,a,,, (1 F20 TWO 1586-O I). A.P.N. was supported by grants from The Jeane B. Kemp- and possibly01,~) receptors, basedupon their differential affinities ner Foundation of the University of Texas Medical Branch at Galveston and by the National Heart, Lung and Blood Institute of the NIH (F32HL08494-Olal). to a variety of agonistsand antagonists,their differential sus- A.D. was supported by grants from Svenska SLllskapet fdr Medicinsk Forskning, ceptibility to covalent inactivation by chlorethylclonidine (CEC), Stiftelsen Sigurd och Elsa Goljes Minne, Anders Otto Swlrds Stiftelse, Svenska and their activation of different second messengercascades LBkareslllskapet and Karolinska Institutets Fonder. Grants from the Swedish Medical Research Council (04X-2887) and the NIMH (MH 43230) are also grate- (McGrath, 1982; Morrow et al., 1985; Hieble et al., 1986; Mor- fullv acknowledged. row and Creese,1986; Han et al., 1987a; Johnson and Minne- dorrespondence should be addressed to Vincent A. Pieribone, Laboratory of Molecular and Cellular Neuroscience, The Rockefeller University, 1230 York man, 1987; McGrath and Wilson, 1988; Minneman, 1988; Ray- Avenue, New York, NY 1002 l-6399. mond et al., 1990; Schwinn et al., 1990; Han and Minneman, Copyright 0 1994 Society for Neuroscience 0270-6474/94/144252-17$05.00/O 1991). The Journal of Neuroscience, July 1994, 14(7) 4253

Numerousstudies employing the tissuesection ligand binding Alpha, N ,, receptor mRNA autoradiography technique developed by Young and Kuhar (1979a) have examined the distribution of 0, adrenoceptorsin ’ 500 1000 1500 2000 2500 2947 ! ! I the CNS (Young and Kuhar, 1979b, 1980; Unnerstall et al., Translated region 1982; Dashwood, 1983; Rainbow and Biegon, 1983; Jones et al., 1985; Palacioset al., 1987; Unnerstall, 1987; Chamba et al., 480 2159 1991). Three ligandshave beenpreferentially usedin thesestud- . 1 ies, 3H-WB4 101, 3H-prazosin, and lz51-HEAT. Recent studies Probe #2 Probe #l have utilized predominately prazosin and HEAT, because 241-267 1167-1214 WB4 101 was found not be specific for LY,adrenoceptors (Hoff- man and Lefkowitz, 1980; Lyon and Randall, 1980; Rehavi et Alpha, B receptor mRNA al., 1980). While the binding patterns of prazosin and HEAT have generally given comparableresults, HEAT is preferred for a variety of reasons(see Unnerstall, 1987). However, neither of thesecompounds binds to only one subtype of the a), adrenoceptor family. Furthermore, the low cellular resolution of the autoradiographic ligand binding technique does not allow ac- Probe #l curate identification ofthe cellsin the CNS that expressa specific Probe #2 170-217 1420-1467 (Y, adrenoceptor subtype. In addition, the autoradiographic labeling seenwith HEAT hasoften contradicted biochemicalstud- Figure I. A schematicdiagram of the alAD and alBreceptor mRNAs ieson the localization ofa, receptors. For example, such studies indicatingthe locationsof the sequences from which the four oligo- have indicated that (Y, receptorsare presentin the hippocampus, nucleotideprobes were derived. Two 4%merreverse and complementary probeswere made to eachof the mRNAs.For eachof the mRNAs, while many autoradiographic studiesutilizing HEAT and pra- one probe was madeto the regionof the mRNA that wastranslated zosin binding have consistently failed to identify substantial into the receptorprotein while anotherwas madeagainst the 5’ un- binding in the hippocampus (Unnerstall et al., 1982, 1985; translatedregion ofthe mRNA. Eachofthe probeswas used on alternate Dashwood, 1983;Rainbow and Biegon, 1983; Joneset al., 1985; sectionsin someexperiments and the labelingobtained compared (Fig. Palacios et al., 1987; Unnerstall, 1987; Marks et al., 1990; 12). Chamba et al., 1991; but seeZilles et al., 1991). Utilizinga biochemicalmethod that combinesthe preferential the properties of the (Y,,, adrenoceptor. Additionally, a cDNA covalent modification of (Y,* adrenoceptorsby CEC with com- clone was isolated coding for a receptor with the pharmacolog- petitive ligand binding, Wilson and Minneman (1989) have ical properties of the classic N,” adrenoceptor (Voight et al., reported regional differencesin the distribution of CY,*and (Y,~ 1990). The present study reveals a unique and widespreaddis- adrenoceptors.These investigators have found that (Y,~receptors tribution pattern for Q’,&.~and cy,” adrenoceptor mRNAs. predominate in the hippocampus, pons-medulla, and spinal cord, while aYIBreceptors predominate in the thalamus, hypo- Materials and Methods thalamus, and cerebellum. After the first cloning of an adrenoreceptorby Letkowitz and Preparation qf the oligonucleotide probes. In efforts to developprobes specificfor either of the two subtypesof the rat (Y, adrenoceptorsse- collaborators (Cotecchia et al., 1988), the genetic sequencesof quencedso far, nonhomologousregions of the mRNA wereused in a number of adrenoceptorshave been determined (Buckland et generating the probes. Two probes directed against each of the receptors al., 1990; Machida et al., 1990; Schwinn et al., 1990; Shimomura were constructed, two of which were complementary to regions in the and Terada, 1990; Voigt et al., 1990, 1991; Zeng et al., 1990; translatedhydrophilic domain (third intracellularloop) of eachof the mRNAsand two complementaryto the 5’ untranslatedregion of each Flordellis et al., 1991; Lanier et al., 199 1; Lomasneyet al., 1991). of the mRNAs(Fig. I). UsingMACVECTOR software(IBI, New Haven, This has made it possible to utilize the in situ hybridization CT), 48-nucleotide segments were selected based on a determined op- technique (seeYoung, 1990) to examine the distribution of cells timum ratio for total G/C content (60-65%)and minimal homology in the CNS that synthesizemRNA to thesevarious adrenocep- (notgreater than 80%) with GenBank-entered eukaryotic gene sequences tars (Nicholas et al., 199l), revealing distinctly differential dis- (as of August 1991)and specificallynonhomologous with the other adrenergicreceptors mRNA (a?,BT and @, ?) sequences (not greater than tribution patterns for the (Y*-and p-adrenergicreceptor mRNAs 65% homology; Cotecchia et al., 1988; Buckland et al., 1990; Machida (Nicholas et al., 1993a,b). In addition to specificity, thesetypes et al., 1990; Schwinnet al., 1990;Shimomura and Terada,1990; Zeng of studiesoffer a cellular resolution difficult to obtain with tra- et al., 1990;Voigt et al., 1991; Lanieret al., 1991). Reversedand com- ditional biochemical or autoradiographic techniques. On the plementaryDNA probeswere made (Scandinavian Gene Synthesis AB, Kiiping, Sweden)to the followingsequences: 24 I-287 and I 167-I 2 14 other hand, the in situ hybridization technique only reveals the of rat (Y,~,, receptor (Lomasney et al., 199 I) and 170-2 I7 and 1420- cell bodies in which the receptorsare synthesized and not the 1467of rat a,, receptor(Voigt et al., 1990)mRNAs. Eighty nanograms distribution of the receptor protein in the dendritic and axonal (5 pmol)of oligonucleotideprobe were labeled at the 3’-endusing ter- ramifications. Thus, information from the two techniques is minal deoxynucleotidyltransferase(Amersham) and %S-dATP(New complementary and together provides important clues to the EnglandNuclear) to a specificactivity of 1-2 x lob cpm/ng. Preparation qftissue and in situ hybridzzation. Male Sprague-Dawley functional role of receptors. rats(n = 8; 250-350gm) weredecapitated, and the brain, spinalcord A receptor initially termed the CY,~adrenoceptor was recently and superiorcervical, dorsal root (LS),and nodose ganglia were rapidly cloned from a rat brain cDNA library (Lomasneyet al., 1991). removed and frozen with gas produced from liquid carbon dioxide. However, this mRNA doesnot seemto code for the classicCY,~ Serial coronaland sagittalcryostat sections (14 pm) weremelted onto FisherProbe-On slides (Fisher Scientific, Pittsburgh, PA). In srtuhy- adrenoceptor but rather a heretofore unidentified “alD” (Perez bridizationexperiments were performed as previously described (Schall- et al., 199 1) or “ffIwVD” adrenoceptor (Schwinn and Lomasney, ing et al., 1988;Schalling, 1990; Young, 1990;Dagerlind et al., 1992). 1992). This receptor (here termed CY,~~)has some but not all Briefly, adjacentunfixed tissue sections (Dagerlind et al., 1992) were air 4254 Pieribone et al. - 01. Receptor mRNA Distribution in the CNS

Figure 2. A-H, Schematic diagrams indicating the brain regions containing cells that hybridized with the alk/D and qB receptor mRNA probes. The sections on the right indicate the areas of the brain to which the aIA/D (right side, left hemisection) and alB (right side, right hemisection) probes hybridized. The hemisections on the left side indicate the cytoarchitectonic boundaries and nomenclature. Two patterns are used to indicate cells exhibiting heavy hybridization (dark shading) versus light hybridization or lightly hybridized cells (light shading) with either of the probes. In addition, areas that contained low number of isolated labeled cells are indicated by large dots. The boundaries and nomenclature are derived mainly from Paxinos and Watson (199 1). AH, anterior hypothalamic nucleus; Amb, nucleus ambiguus; AON, anterior olfactory nucleus; AP, area postrema; &?rV, bed nucleus of the stria terminalis (ventralis); Cu, cuneate nucleus; DR, dorsal raphe; FrCx, frontal cortex; GR, gracile nucleus; HDB, horizontal limb of the diagonal band; Hip, hippocampus; ZGR, internal granular cell layer of the olfactory bulb; IO, inferior olivary complex; ZPI, internal plexiform layer of the olfactory bulb; LaDL, lateral amygdaloid nucleus, dorsolateral part; LRt, lateral reticular nucleus; LSV, lateral septal nucleus (ventralis); MeRet, median reticular formation; MeV, medial vestibular nucleus; MGD, medial geniculate body; Pi, pineal gland; Pir, piriform cortex; PO, preoptic nucleus; PVN, paraventricular nucleus; RA rhinal fissure; RMg, raphe magnus; Rob, raphe obscurus; RPu, raphe pallidus; Rt, reticular thalamic nucleus; SHY, septohypothalamic nucleus; Sp5, spinal trigeminal nucleus; VII, facial nucleus; XII, hypoglossal nucleus. dried and incubated at 42°C for 16-18 hr with 0.5-l .O ng of the labeled were allowed to expose 10-17 weeks in light-free metal containers at probe per 100 ~1 of hybridization cocktail consisting of 50% deionized -2o”C, and then developed in Kodak D 19 (3 min), fixed in Kodak 3000 formamide, 4 x SSC (1 x SSC is 0.15 M NaCl, 0.0 15 M sodium citrate), A and B (6-7 min), and coverslipped in glycerol. Some tissue sections 0.02% bovine serum albumin, 0.02% Ficoll, 0.02% polyvinylpyrroli- were counterstained with 0.1% toluidine blue with 0.5% sodium borate done, 1% sarkosyl, 0.02 M sodium phosphate (pH 7.0), 10% dextran and coverslipped with Entellan (Merck). All sections were examined sulfate, 200 mM dithiothreitol, and 500 &ml ofheat-denatured salmon under light- or dark-field illumination with a Nikon Microphot-FX testis DNA. For competitive binding control experiments, radiolabeled microscope. Photomicrographs were taken with Kodak T-Max 100 black- probes were mixed with an excess (100 x) of the respective nonradiol- and-white film. In some cases photographic prints were made directly abeled probe. Sections were then washed in 1 x SSC buffer (4 x 15 min) from the autoradiographic films. at 55”C, slowly cooled to room temperature over 30 min while in the Evaluation of results. Labeled structures were evaluated in sections final rinse, briefly transferred through distilled water, and dehydrated that had been counterstained with toluidine blue. The distribution of in 60% and then 90% ethanol. Following this procedure, sections were the two mRNA is shown in several coronal planes and’ represents a either opposed to autoradiographic film (Hyperfilm-pmax; Amersham) subjective integration of cell density and intensity of labeling. or dipped in Kodak NTB2 autoradiography emulsion (50% in water). Northern blot analysis. Approximately 500 mg of frozen tissue (ce- Following 4-17 weeks exposure at -2o”C, films were developed with rebral cortex or thalamus/hypothalamus) was placed in 4 M guanidine LX 24 (Kodak) for 2 min, rinsed in distilled water (30 set), and fixed isothiocyanate, 0.1 M P-mercaptoethanol, 0.025 M sodium citrate (pH in AL 4 (Kodak) for 15 min. Sections dipped in photographic emulsion 7.0) and immediately homogenized with a Polytrone. Each tissue ho- The Journal of Neuroscience, July 1994, 14(7) 4255

Figure 3. Distribution of aIA,D mRNA hybridization in the forebrain. A-C, Heavy hybridization is seen in the olfactory bulb predominantly in the internal granule cell (IGL) and internal plexiform (ZPL) layers of the olfactory bulb. Levels are just at background in the external plexiform (EPL) and glomerular (GL) layers. Occasionally, it appears that labeling in glomerular layer was slightly above background (A, B). A and B are dark-field photomicrographs- of emulsion-dipped sections, while C shows the same region as in B under bright-field illumination to reveal the Nissl staining. D, A dark-field photomontage of the parietotemporal cortex showing the widespread distribution of cells expressing (Y,~,~ adrenoceptor mRNA. Labeline is found in lavers II-V. while laver I is unlabeled. A few weaklv positive cells are seen in laver VI. The ventral border of the labeling was generally the rhinal fissure (;f). CPU, nucleus caudatus putamen. Scale bars, 100 pm. mogenate was layered over a 4 ml cushion of 5.7 M CsCl in 0.025 M sample and 4 pg RNA size marker (Promega, Madison, WI) were sep- sodium citrate (pH 5.5) and centrifuged at 20°C in a Beckman SW41 arated on a 1% agarose gel containing 0.7% formaldehyde, blotted onto rotor at 35,000 rpm for 18 hr. The amount of total RNA was quantified Hybond-N membranes (Amersham), and cross-linked by UV illumi- spectrophotometrically at 260 nm and 15 fig of total RNA from each nation. Membranes were prehybridized at 42°C overnight in a solution 4256 Pieribone et al. - a, Receptor mRNA Distribution in the CNS

Figure 4. A-E, Distribution of CQ,,, mRNA hybridization in cortex (A-D) and hippocampus (E). A-D, Numerous cells, mainly in layers II-V, are labeled, with low numbers of weakly positive cells in layer VI. At a higher magnification it can be seen that many cells in outer layer V (arrows) The Journal of Neuroscience, July 1994, 14(7) 4257 containing 50% formamide (J. T. Baker BV, Deventer, The Nether- strongly labeled(Fig. 6A), while the remainder of the amygdala lands), 5 x SSPE (1 x SSPE = 0.15 M sodium chloride, 0.0 1 M sodium was unlabeled. Numerous lightly labeled neurons were seenin dihydrogen phosphate monohydrate, 1 mM EDTA) pH 7.7, 0.02 M the medial septum/diagonal band complex, and scattered neu- sodium Dhosohate buffer DH 7.0. 0.1% sodium dodecvl sulfate (SDS). 1 x Denhardi’s (0.02% each of pdlyvinylpyrrolidone, $icoll, bovme se: rons were found in the vertical and horizontal limbs of the rum albumin; Sigma), and 100 &ml of sheared and heat denatured diagonal band and the magnocellularpreoptic nucleus(Fig. 2C). salmon testis DNA (Sigma). Eight nanograms of 32P-labeled (l-2 x 1O6 The basal ganglia including the caudate putamen, the globus cpm/ng) kD or ollB oligonucleotide probe were added per milliliter of pallidus, and the substantia nigra were all unlabeled. hybridization solution, and the incubation was continued for an additional 18-20 hr at 42°C. Membranes were washed in 2 x SSPE and 0.1% Diencephalon. The entire hypothalamus and thalamus were SDS (2 x 10 min at 55°C) 1 x SSPE and 0.1% SDS (1 x 20 min at unlabeled, with the exception of the reticular thalamic nucleus, 55°C) and 0.1 x SSPE and 0.1% SDS (2 x 10 min at room temperature which was heavily labeledthroughout its entire extent (Figs. 5A, = 25°C) and exposed to film (New England Nuclear Reflection) with 6B). intensifying screens (New England Nuclear Reflection) at 4°C for 3 d. Mesencephalonand rhombencephalon.Virtually all of the cranial motor nuclei were labeled (Fig. 2G,H), including the ocu- Results lomotor, facial, hypoglossal, ambiguus, and motor trigeminal The alAIDand LY,~receptor mRNAs had a wide distribution in nuclei. The dorsal motor nucleus of the vagus was, however, the rat CNS (Figs. 2-l 1). Nissl-counterstained tissue sections unlabeled. A small collection of lightly labeled cells was con- revealed that the labeling of both mRNAs was accumulated sistently seenin the area of the lateral dorsal tegmental nucleus over neuron-like somata, although glial labeling can not be ex- in the pontine central gray matter. In addition, numerous scat- cluded (Stone and Ariano, 1989; Aoki, 1992). However, hy- tered labeledcells were seenthroughout the reticular formation bridization with the probe for the untranslated region of the o(,~ of the pons and medulla. adrenoceptor mRNA resulted, in addition, in a diffuse signal The most intensely labeled structure in the brainstem was the over white matter (seeFig. 12D,E). Therefore, the probes di- inferior olivary complex (Fig. 6C). Virtually all neurons in all rected against the translated regions of each mRNA were used subdivision of the inferior olive were heavily labeled (Fig. 60). for the mapping study. Generally very long exposure times (8- The labeling extended from the most rostra1to the most caudal 12 weeks) were required to visualize labeling of both mRNAs. extent of the structure. Another heavily labeledstructure in the The distribution patterns are demonstrated in schematicdraw- caudal medulla was the cuneate nucleus (Fig. 2H), while the ings (Fig. 2A-H) reproduced from the atlas of Paxinos and Wat- external cuneate nucleuswas unlabeled. The gracile nucleuswas son (199 1) and in micrographs of bright- and dark-field auto- lightly labeled (Fig. 2H). Scattered positive cells could also be radiograms of emulsion-dipped, toluidine blue-counterstained seenin the medial vestibular nucleus, the area postrema and sections(Figs. 3-l 1). the lateral reticular nucleus(Fig. 2H). Light but distinct labeling was seenin the superficial layers of the trigeminal spinal nucleus a,, ,) receptor mRNA distribution (Fig. 2H). Telencephalon. Pronounced labeling was seenin the internal Spinal cord. The most prominently labeledcells in the spinal granular and plexiform cell layers of the olfactory bulb (Fig. 3A- cord were the ventral horn motor neurons at all levels. In the C). This labeling was found uniformly throughout the bulb but remaining laminae of the spinal cord a large number of scattered did not extend into the external plexiform layer or the mitral neurons was seenin the dorsal aspectsof the ventral horn. cell layer (Fig. 3B,C). Light labeling was seenin the glomerular Peripheral tissues.The pineal gland was unlabeled with the layer over the periglomerular neurons(Fig. 3B,C). No labeling LY,~,~probes, as were the superior cervical ganglia, lumbar dorsal was seenin the anterior olfactory nucleus. root ganglia and the nodoseganglia. Neurons throughout the cerebral cortex labeled extensively and heavily with the CY,~,~probes (Figs. 30, 4A-D). The most cylBreceptor mRNA distribution pronounced labeling was seenin the prefrontal, cingulate, and Telencephalon.The olfactory bulb was virtually unlabeledwith occipital cortex, while less labeling was seenin the temporal the o(,~probe. A very light labeling was consistently observed cortex. The piriform cortex was generally unlabeled. Ventrally, only in the internal plexiform layer. Distinct though light la- the labeling generally ended at the level of the rhinal fissure beling was seenwithin the ventral division of the anterior ol- (Figs. 30, 6A). Labeled neurons were seenin all layers of the factory nucleusand within the rostra1piriform cortex (Fig. 2B). cortex except layer I, and only a few labeled neurons were ob- The cerebral cortex was strongly labeled with the aIB probe served in layer VI. Some superficial pyramidal neuronsin layer but not to the extent seenwith the CY,~,~)probes (Figs. 7, SA,B). V were often heavily labeled, whereas others had only low or The strongestlabeling was observed in layers III-V, while the no detectable labeling (Fig. 4A-D). In Nissl-stainedsections of superficial layers were unlabeled. Single labeled cells formed a the cingulate cortex most neurons in layer II-V were labeled. thin band in deep layer VI overlying the dorsal aspectsof the In the hippocampal formation the pyramidal and granular corpus callosum (Fig. 2C,D). As with the OI,~,”probe the large cell layers were heavily labeledwith the (Y,~,,,probes (Figs. 4E, pyramidal neuronsof layer IV were distinctly labeled with the 5A; seealso Fig. 12A,B). The dorsal and the ventral hippocam- cylBprobe. Although the piriform cortex within the accessory pus were equally labeled. olfactory bulb was lightly labeled, the remainder of it was un- The dorsolateral aspectof the lateral amygdaloid nucleuswas labeled.Generally the cortical labelingwith the cylUprobes ended t are strongly labeled and some of them are pyramidal cells. Other pyramidal cells are either very weakly labeled (double arrowheads)or not labeled at all (arrowheads). B and D show the same region as in A and C, respectively, under bright-field illumination and after Nissl staining. E, The pyramidal cell layers of CAl-CA4 and the granular cell layers of the dentate gyrus (DG) are heavily labeled. Scale bars, 100 Km. 4258 Pieribone et al. * ol, Receptor mRNA Distribution in the CNS

Figure 5. A and B, Distribution of rzIA,,, (A) and (Y,* (R) mRNA labeling in the dorsal thalamus and dorsal lateral hippocampus (Hi). A and B show adjacent sections. Strong (Y~+,,~labeling is seen in the thalamic reticular nucleus (Rt) but not in the remaining thalamus, for example, in the adjacent ventral posterior nucleus (VP) (A). Instead, the remaining thalamus is rich in cylsmRNA-expressing cells; however, they are not found in the reticular nucleus (B). Note strong aIA,D labeling in the pyramidal cells in the hippocampus (A) while no qB labeling is seen in the hippocampus in the adjacent section (B). Scale bar, 100 pm.

just ventral to the rhinal fissure (Fig. 2). Area 1 of the parietal 9C). In striking contrast to the labeling pattern seenwith the cortex was lesslabeled than the surrounding cortex. As with the LX,~,~probe, no certain cylBlabeling was seenin the reticular o(,~,,,probe, the dorsolateral division of the lateral amygdaloid thalamic nucleus (Fig. 5B). nucleus was heavily labeled. However, unlike the CY,~,,,probe, The lateral hypothalamic nucleusand the paraventricular nu- the central amygdaloid nucleus was lightly labeled with the o(,~ cleusboth exhibited very light but significant labeling (Fig. 20). probe, as was the bed nucleus of the stria terminalis and the Mesencephalonand rhombencephalon.The most pronounced ventral septalarea (Fig. 2CJ). The vertical and horizontal limbs labeling in the brainstem was seenin the dorsal raphe complex of the diagonal band were labeled(Fig. 2C). There was no dis- (Fig. lOA). Virtually all neurons in the dorsal raphe (including cernible signal in any of the basalganglia structures. the lateral wings) were labeled. This labeling extended ventrally Diencephalon.One of the most striking and distinct labeling to include the central superior nucleus and caudally to the me- patterns was seenin the thalamus (Figs. 5B, 9A-C). With the dulla oblongata including the raphe magnus,pallidus, and ob- exceptionsof the reticular and habenular nuclei, the entire thal- scurusgroups. Prominent labeling was seenover cranial motor amus was heavily labeled. All of the medial and lateral tier of neurons (Fig. 8B) and in the lateral reticular nucleus. Motor thalamic nuclei (except the reticular nucleus)(Fig. 5B), including neuronsin the facial (Fig. SB), hypoglossal,ambiguus, and mo- the geniculate bodies (Fig. 9,4-C), was heavily labeled. In Nissl- tor trigeminal nuclei were all well labeled. However, the dorsal stained sectionsof the thalamus it appearedthat virtually every motor nucleus of the vagus was not labeled. The lateral reticular neuron in the labeled nuclei expressedcylB receptor mRNA (Fig. nucleus was also heavily labeled throughout its rostrocaudal The Journal of Neuroscience, July 1994, 14(7) 4259

Figure 6. A-D, qND mRNA labeling of the amygdala and cortex (A), thalamus (B), and inferior olivary complex (C and D). A, In the amygdala only the lateral aspect of the lateral amygdaloid nucleus is labeled, and the labeling is very strong. B, The reticular thalamic nucleus was the only labeled structure in the thalamus. This nucleus was heavily labeled throughout its rostrocaudal extent. C and D, The entire inferior olivary complex with all of its subdivisions was heavily labeled (C). Upon Nissl counterstaining it appeared that virtually all neurons were labeled (D). Arrowheads in C represent isolated cells labeled in the medullary reticular formation. Scale bars: A-C, 200 pm; D, 50 pm. extent (Fig. 2G,H). In addition, the external cuneate nucleus Peripheral tissues. The pineal gland was very heavily labeled was lightly labeled, as were neurons in the spinal trigeminal with the (Y,~probe (Fig. 8C), while the peripheral gangliashowed nucleus. A large number of scattered, positive cells were seen no labeling. in the pontine and medullary reticular formation, including a well-defined group in the paragigantocellularisnucleus region Control experiments immediately adjacent to the inferior olive (Fig. 1B). A very faint Three types of control experiments were performed. First, to signal was detected in the cerebellar cortex upon prolonged ex- control for the nonspecificbinding of the probe to cellular com- posure.The labeling wasgenerally over the granule and Purkinje ponents, incubations of alternate sectionswith probes against cell layers. the translated regions were done in the presenceof an excess Spinal cord. Spinal motor neurons were heavily labeled (Fig. (100 x ) of the respective unlabeled probe. In all cases,sections lOC,D), and scattered labeled cells were seenthroughout the incubated with an excessof unlabeled probe showed no hy- various laminae in the remainder of the grey matter of the spinal bridization (not shown). Second, to exclude “nonspecific” hy- cord. Presumably preganglionic neurons in the intermediola- bridization to unknown related RNA, additional probes were teral cell columns of the thoracic spinal cord were also labeled. generatedto the 5’ untranslated regionsof either the (Y,~~ or +, 4260 Pieribone et al. - 01, Receptor mRNA Distribution in the CNS

Figure 7. Cortical labeling with the ollB mRNA probe shown in a dark- field photomontage of the lateral frontal lobe. Intense O(!~ labeling is seen in the intermediate layers of the cortex, while the deep (VI) and FIgwe 8. A-C, alB mRNA labeling of cortex (A) and pineal gland (C). superficial layers (I and II) are very weakly labeled. Scale bar, 100 pm. A and B, The comparison of the dark-field (A) and bright-field (B) micrographs show that most labeled cells are seen in layers IV and V, with more weakly labeled cells in layer VI. C, A strong signal is seen in the pineal gland. Scale bar, 100 pm. The Journal of Neuroscience, July 1994, 14(7) 4261 mRNAs, that is, regions that normally contain very low sequence homologies between even closely related mRNAs. These probes were used on adjacent sections to those hybridized with the original probes, and the patterns of labeling were compared. Probes made against the untranslated or the translated region gave identical labeling patterns (Fig. 12A-E). However, the CX,~ probe against the untranslated mRNA region in addition showed a signal over the white matter (Fig. 12DJ). Finally, Northern blot (Fig. 13) of total RNA was used to ensure that a single major RNA species was being detected with the in situ probes and that the size of the transcript compared to published reports (Lomasney et al., 199 1; McCune and Voigt, 199 1). Both probes labeled only a single band in each lane, a 3.0 kb transcript for the CY,A,D probe and a 2.4 kb transcript for the a,* probe. A greater amount of OI,~,~ mRNA was seen in the cerebral cortex than in the thalamus, while the ailB probe revealed an opposite pattern, with a greater amount of aIB mRNA transcript seen in the thalamus and less transcript seen in the cerebral cortex. The Northern blot data therefore agree with the in situ hybridization results in the thalamus and cerebral cortex and reveal a single transcript of correct size for each of the probes. Discussion Using in situ hybridization histochemistry the cellular localization of (Y,~,~ and ollB mRNA has been analyzed. Results show both complementary and overlapping distribution patterns for the two receptors mRNAs. Since earlier autoradiographic ligand binding studies have mainly identified alB receptors (for references, see below), the present results on the o(,~,, receptors are of particular interest. However, the cDNA encoding the true LY,~ adrenoceptor has likely not yet been cloned. Conversely, it does not appear that the cylc adrenoceptor is present in the CNS. However, it is possible that other cy,-like adrenoceptors will be cloned in the future. The term “(Y,~,~ adrenoceptor” is used here, because the biochemical classification of the receptor has been equivocal. In the initial cloning report (Lomasney et al., 199 l), the properties of this receptor in COS-7 cells mostly resembled those of the classical LY,~adrenoceptor. In another study (Perez et al., 199 l), apparently the same cDNA sequence (with two base pair differences that were likely due to an initial sequencing mistake) was isolated, cloned, and also expressed in COS-7 cells. The resulting receptors properties appeared to represent an, as yet unidentified, o(, adrenoceptor, which was termed the O(,~adrenoceptor and which, only in some respects, resembled the CY,~ adrenoceptor (Perez et al., 199 1). Subsequently, authors of the first report suggested that, in fact, the receptor they had cloned was likely a novel a, adrenoreceptor and termed it the “a,*,,, adrenoceptor” (Schwinn and Lomasney, 1992). Our results in- Fgure 9. A-C, LX,,,mRNA labeling of the medial geniculate body. A, Intense labeling is seen in the thalamic nuclei including the medial dicate that this mRNA likely encodes for a novel adrenoceptor geniculate bodies (MC). B, In a bright-field view of the same region it (and not the aIA adrenoceptor). is evident that the labeling is restricted to the medial geniculate body and does not include the neighboring hippocampus (HI) and midbrain Specificity of the in situ method gray matter. C, At higher power, the geniculate body labeling can be It is well known that members of the G-protein-linked receptor seen over most large cells, presumably representing neurons (largear- row) but many small cells are unlabeled (smaN arrow).It is not possible family in general share a great homology (Raymond et al., 1990), to state whether these cells represent neurons or glia. Scale bars: A and making it difficult to distinguish between different types of ad- B, 100 pm; C, 50 pm. renergic receptors using hybridization techniques. We have therefore used several probes directed against different parts of each individual transcript of the (Y,~,~ and LY,~receptor mRNAs Furthermore, Northern blot analysis using the same oligonu- and compared their labeling patterns. Our results show that the cleotide probe as in the in situ hybridization experimentsshowed two probes raised against 01,~~ and (Y,~ mRNA, respectively, a single mRNA specimen correlating well with the previously show identical labeling, thus supporting specificity of the results. reported sizesfor the individual mRNAs (Lomasneyet al., 1991; 4262 Pieribone et al. l 01~ Receptor mRNA Distribution in the CNS

Figure 10. Raphe nucleus and motor neuron labeling with the aIB mRNA probe. A, The most heavily labeled structure in the brainstem is the raphe complex. In A both the dorsal (DR) and median (MnR) raphe nuclei are intensely labeled. Virtually all neurons in the dorsal raphe are labeled, both in the central and “lateral wing” zones. B-D, Motor neurons in the facial (B) and the spinal ventral horn (C) are labeled, as is evident also after Nissl staining (D). aq, midbrain aqueduct; cc, central canal; DR, dorsal raphe; mlf, medial longitudinal fasciculus; MnR, median raphe nucleus; PN, pontine nuclei; py, pyramidal tracts; VH, ventral horn; VII, facial nucleus. Scale bars: A-C, 100 wrn; D, 50 pm. The Journal of Neuroscience, July 1994, 74(7) 4263

Fzgure I I. A and B, Labeling with olle mRNA probe in the ventral gigantocellular reticular nucleus at the rostra1 level of the inferior olive (IO). A small group of cells immediately dorsolateral to the inferior olive is strongly labeled (arrowheads) as can be seen, when comparing the dark-field (A) and bright-field (B) micrographs of the same tissue section. Scale bar, 100 pm.

McCune and Voigt, 1991; but seePieribone et al., 1992).Finally, CNS, but could not confirm the McCune and Voigt (199 1) find- after incubation with an excessof unlabeled probe, the hybrid- ings of alB receptor mRNA in the hippocampus, and we only ization patterns were abolished, ruling out unspecific hybrid- find a weak signal in the cerebellum. ization to cellular components other than mRNA. Schwinn et al. (1990) have cloned and expressedan cu,-like adrenoceptor from a bovine brain cDNA library. This receptor, Localization of receptor mRNA versusreceptor protein termed 01,c,although pharmacologically similar to the CY,~re- Several caveats must be made when interpreting in situ hybrid- ceptor, is more sensitive to CEC inactivation (68% vs 15% in- ization data. First, the lack of detectable mRNA levels in a activation of each receptor, respectively). However, (~,c is in- particular area of the CNS does not exclude presence of the activated by CEC to a lesserdegree than in the alBreceptor (68% mRNA of interest, since the levels may be below the detection vs 95%). These authors also state that Southern blot analysis limit of our method. It is therefore possible that (Y, receptor- showsthat rats have a gene analog to the bovine (Y,~receptor synthesizing neurons have a more widespreaddistribution than and different from the aYIA,,,or ol,a receptors, but they found no shown here. Second, the technique doesnot prove translation expressionof a bovine a,,-like receptor mRNA in rat cerebral into a functional receptor protein. Finally, the presenceof re- cortex, hippocampus,or brainstem or in bovine cerebral cortex ceptor mRNA in the neuronal somadoes not indicate where on (Schwinn et al., 1990). the cell’s surface the receptor protein is deployed. A cell may preferentially express a particular receptor on its presynaptic In situ versusautoradiographic ligand localization of a, terminal that would only be visible in the target area using the adrenoceptors autoradiographic receptor ligand binding technique (Young and Numerous autoradiographic studies have examined the distri- Kuhar, 1979a)or an immunohistochemical procedure utilizing bution of LY,adrenoceptors in the CNS. Early studies (Young antibodies directed againstthe receptor (Aoki et al., 1987). and Kuhar, 1979b, 1980) utilizing 3H-WB4101 as a ligand showedvery limited and weak binding in the granule cell layer Comparison with studiesusing Northern blot analysis of CNS of the hippocampus and dentate gyrus as well as within the o(, adrenoceptor mRNAs periaqueductal gray. These early studies have been difficult to Several studieshave examined CNS concentrations of the var- interpret (Unnerstall, 1987) since 3H-WB4101 is not specific ious a, adrenoceptor mRNAs by Northern blot analysis of for o(, adrenoceptorsbut also binds to 5-HT receptors(see below mRNA extracted from various brain regions (Schwinn et al., and Hoffman and Lefkowitz, 1980; Lyon and Randall, 1980; 1990; Lomasneyet al., 1991; McCune and Voigt, 1991). McCune Rehavi et al., 1980; Morrow and Creese,1986; Unnerstall, 1987). and Voigt (199 1) and Lomasney et al. (199 1) have presented The ligands now in most common use are lZ51-HEAT (IBE conflicting resultsabout the distribution of ollBreceptor mRNA 2254), ‘H-WB4101, and ‘H-prazosin (Unnerstall et al., 1982, in the rat CNS. McCune and Voigt (199 1) report that CX,~receptor 1985; Dashwood, 1983; Rainbow and Biegon, 1983; Jones et mRNA is present in comparableamounts in the cerebral cortex, al., 1985; Palacios et al., 1987; Unnerstall, 1987; Marks et al., cerebellum, hippocampus, and hypothalamus, while 20% less 1990; Chamba et al., 1991). While HEAT and prazosin are was seenin the midbrain and brainstem. In contrast, Lomasney reported to bind to both alA and aIB sitesin homogenates(Han et al. (1991) report a),a receptor mRNA only in the brainstem et al., 1987b; Morrow and Creese, 1986; Minneman, 1988) and cerebral cortex and not in the hippocampus or cerebellum. autoradiographic studiesutilizing theseligands have shown very Our results are in agreementwith the findings of Lomasney et little labeling in the hippocampusor reticular thalamic nucleus. al. (1991) on cylBreceptor mRNA distribution patterns in the These areascontain exclusively LY,~receptors, according to pre- 4264 Pie&one et al. l ur Receptor mRNA Distribution in the CNS

Figure 12. A-E, Comparison of hybridization obtained with the four different probes directed against the qMD and Q, mRNAs. A and B, Labeling is seen in two adjacent sections of the ventral hippocampus hybridized with the oligonucleotide probes, one (A) complementary to a region in the translated domain of the alA,D mRNA (probe 1 in Fig. 1, top) and the other one (B) to a region in the 3’ untranslated region of the mRNA (probe 2 in Fig. 1, top). Note that in both regions the probes give an identical pattern of labeling. C-E, Direct prints from film autoradiograms of three adjacent tissue sections including sections from several brain regions hybridized with the oligonucleotide probe complementary to the translated region of the qB mRNA (C, probe 1 in Fig. 1, bottom), the untranslated 5’ region of the mRNA (E; probe 2 in Fig. 1, bottom) or a combination of both probes (D). Note that the medial geniculate body (long arrows), the thalamus (open arrows), the dorsal raphe nucleus (short arrows), and the cortex (large arrowheads) are all labeled with both the translated and untranslated probes. However, the probe to the untranslated 3’ region also gives a signal over white matter (small arrowheads) (E), which is accentuated after hybridization with the mixture of the two probes (0). Scale bar: 100 pm for A and B, 1000 wrn for C-E. vious studies(Morrow et al., 1985; Morrow and Creese,1986; Recently, WB4 101 has been “rediscovered” as a potential qA Wilson and Minneman, 1989; however, see also Zilles et al. ligand (Morrow and Crease, 1986; Blendy et al., 1990, 1991; 1991). It appearsthat while theseligands are useful in biochem- Grimm et al., 1992). When 3H-WB4101 is combined with un- ical assaysto identify CY,~receptors, they are not as suitable for labeled 5-HT (to avoid its binding to SHT,, sites),it is reported this purpose in autoradiographic studies. to label only LY,~sites (Morrow and Crease,1986; Blendy et al., The Journal of Neuroscience, July 1994, M(7) 4265

1990, 1991; Grimm et al., 1992). Oddly, when one compares al AID al B the “aIA-specific” 3H-WB4101 binding with the “cu,,-specific” ‘H-prazosin binding (preabsorbtion of tissue with cold WB4 10 l), TCCT the patterns look rather similar (Blendy et al., 1990, 199 1; Grimm et al., 1992). These studies point to further problems using this ligand. For example, the medial thalamus shows greater aYIA labeling than either the hippocampus or the reticular thalamic nucleus, which does not agree with previous biochemical studies showing low alA binding in the thalamus and higher binding in the hippocampus (Wilson and Minneman, 1989). Wilson and Minneman (1989) used CEC inactivation combined with the competitive binding of WB4 10 1 to distinguish between CY,~and (Y,* receptors. The alA receptor, which is insensitive to inactivation by CEC, is found to predominate in the hippocampus, Fzgure13. A, Northern blot oftotal RNA extracted from rat thalamus/ brainstem, and spinal cord, while CEC-sensitive sites (ailB) are hypothalamus (7) and cerebral cortex (C) probed with the oligonucle- found in the thalamus, hypothalamus, and cerebellum. While otide probes for the OI,~,~(I eji szd e ) an d ollB (right szde)mRNA. The size ‘H-WE4101 appears to label some CY,~~ sites it may not ade- of the c~,~,n mRNA is approximately 3.0 kb, whereas the size of ollB quately reflect the total LY,* receptor distribution. In some ob- mRNA is approximately 2.4 kb. In agreement with the in situ hybridization results, there is higher expression of qA,n mRNA in cerebral vious cases, however, HEAT may label OI,~~ receptors. For cortex than there is in the thalamus/hypothalamus, while the reverse is example, the inferior olivary complex and the olfactory bulb, true for the LY,~mRNA. The migration of three RNA size markers (4.0, which both contain a large amount of a,A,D mRNA (and no LY,~ 2.8, and 1.9 kb, respectively; Promega) is indicated to the right. mRNA), were also heavily labeled with 1251-HEAT (Jones et al., 1985; Unnerstall et al., 1985; Palacios et al., 1987; Unnerstall, 1987). Previously it has been shown that the locus coeruleus contains son and Hartman, 1975; Jones and Moore, 1977; Hijkfelt et al., a high density of (Y, receptors as determined by receptor auto- 1984; Moore and Card, 1984; Jonesand Yang, 1985). However, radiography (Jones et al., 1985; Palacios et al., 1987; Unnerstall, in retrospect it may be possibleto determine what the physio- 1987; Chamba et al., 199 1). However, physiological experi- logic effects of stimulating the 01,~~adrenoceptor are, since pre- ments both in vivo and in vitro have failed to identify any o(,- vious studies have examined the physiological actions of LY, like responses in adult locus coeruleus neurons (Cederbaum and adrenoceptor agonistsin areasshown in the present study to be Aghajanian, 1977; Egan et al., 1983). In the present study, nei- enriched in (Y,A/D adrenoceptors (Baraban and Aghajanian, 1980; ther LY,*,~ nor alB mRNA was found in neurons of the locus Bradshaw et al., 1981; Fung and Barnes, 1981; Aghajanian, coeruleus. It is possible that 01, binding seen in the locus coe- 1985; Freedman and Aghajanian, 1987; McCormick and Prince, ruleus represents presynaptic receptors or perhaps alA receptors 1988; McCormick and Wang, 1988; McCormick, 1992). (Schwinn et al., 1990). Basedon biochemical studiesof brain homogenates,the two In conclusion, it is obvious that the available ligands used to original a), receptor subtypes, CX,~and o(,~adrenoceptors, differ study the autoradiographic distribution of 01, adrenoceptors are from each other with regard to the secondmessenger cascades not specific for either the LY,~or a,B adrenoceptors and definitely that they activate (Johnson and Minneman, 1986, 1987; Han do not recognize the receptor encoded by the recently identified et al., 1987a; Minneman, 1988; Raymond et al., 1990). The LY,~ o(,*,~ mRNA. Complementary in situ hybridization studies may receptor is linked via a G-protein directly to a dihydropyri- therefore help to localize the many types of closely related ad- iine-sensitive Ca2+channel. The CX,,,receptor, on the other hand, renoceptors that are expressed in the CNS. activates phospholipaseC to hydrolyze phosphoinositol into inositol triphosphate and diacylglycerol to activate Ca2+influx Functional roles for dlflerent o(~receptors (Minneman and Johnson, 1984; Schoeppet al., 1984; Johnson In many CNS areas only one of the two 01, subtype receptors and Minneman, 1986, 1987; Han et al., 1987a; Minneman, was found (Table 1). The most striking examples were the re- 1988; Michel et al., 1990). While theseare the effectsobserved ticular thalamic nucleus, on the one hand, and the remaining in biochemical studies,they are not the major effectsseen phys- thalamus, the raphe nuclei, the inferior olivary nucleus, the iologically in the CNS by activation of LY,receptors. In most pineal gland, and the hippocampus, on the other. In each of studied areasof the brain, including the cerebral cortex, medial these cases,only one of the probes heavily labeled neurons in and lateral geniculate nuclei, the reticular thalamic nucleus,dor- eachrespective area.Conversely, in someother areasboth probes salraphe and spinal motor neurons,(Y, receptor activation causes showed hybridization, for example, in the midlayers of the cor- strictly excitatory responsesboth in vivo and in vitro (Baraban tex, the dorsal division of the lateral amygdaloid nucleus, and and Aghajanian, 1980; Bradshaw et al., 1981, 1985; Fung and in the brainstem and spinal motor neurons. These differences Barnes, 1981; Aghajanian, 1985; Freedman and Aghajanian, in distribution of receptor mRNAs do not seemsimply to reflect 1987; McCormick and Prince, 1988; McCormick and Wang, the differential input from the endogenousadrenoceptor ligands, 1991; McCormick, 1992; for review, see Szabadi, 1979; Foote adrenaline and noradrenaline. Moreover, these two catechola- et al., 1983; Szabadi and Bradshaw, 1987). However, it appears minesare not likely to reflect the regional differencesseen, since that this activation is due largely to a decreasein resting K+ as noted above the noradrenaline-producing locus coeruleusis conductance and not to a robust increasein Ca2+ conductance the major afferent source of the endogenousligand for both (Baraban and Aghajanian, 1980; Aghajanian, 1985; McCormick receptor subtypesevaluated in the presentstudy, in which adren- and Prince, 1988; McCormick and Wang, 1991; McCormick, aline has a very restricted distribution (Ungerstedt, 197 1; Swan- 1992). In all thesecases, this K+ conductance reduction is not 4266 Pieribone et al. * 01, Receptor mRNA Distribution in the CNS

Table 1. Differential distribution of (Y,,~,~ and (Y,~ receptor mRNAs and a, ligand binding

ffl ~lAIU cflL3 ligand Structure mRNA mRNA binding Predominately aIAL1 receptor mRNA Internal granule cell layer of the olfactory bulb +++ - + Hippocampus +++ - + Reticulated thalamic nucleus +++ - + Inferior olivary complex +++ - ++ Cuneate nucleus ++ - Predominately N,~ receptor mRNA Paraventricular hypothalamic nucleus - + ++ Thalamus (except ret. thal. nut.) - +++ +++ Raphe nuclei +++ ++ External cuneate nucleus - ++ ++ Intermediolateral column in the spinal cord + + Pineal gland - +++ ++ High levels of ai, receptor binding but little mRNA External plexiform layer of the olfactory bulb - - +++ Locus coeruleus - - +++ Overlapping distribution of o(,~,~ and ailB receptor mRNAs and CI, ligand binding Cerebral cortex (layers II-V) +++ +++ +++ Dorsolateral lateral amygdala +++ +++ +++ Cranial and spinal motor neurons +++ +++ +++ Sensory trigeminal nucleus + + + Pontine and medullary reticular formation + + +

This table shows the differential distribution of a,4 ,, and c(,” mRNA in selected regions of the CNS as compared to previous ligand binding studies (Unnerstall et al., 1982, 1985; Dashwood, 1983; Rainbow and Biegon, 1983; Jones et al., 1985; Palacios et al., 1987; Unnerstall, 1987; Marks et al., 1990; Cbamba et al., 1991). This is meant for comparison only and is relative and subjective. Single pluses indicate low, two pluses indicate moderate, and three pluses indicate high intensity of the hybridization signal found over neurons in the in situ experiments or the autoradiographic grain density in the binding studies.

dependent on extracellular Ca*+ entering the cell. In CNS areas Conclusions such as the reticular thalamic nucleus, that apparently contain only o(,~,~ receptors, which should directly activate a Ca*+ cur- In conclusion, we have found that two 01, receptor mRNAs are rent, no such activation has been reported (McCormick and expressed partly in different and partly in the same regions of Wang, 199 1; McCormick, 1992). In the case of raphe neurons, the rat brain, suggesting involvement of these receptors in many shown in the present study to contain a large amount of aIB CNS functions. These findings imply that pharmacological agents receptor mRNA, 01,agonists do, in addition to reducing a resting specific for a,*,,, and N,~ receptors are likely to have different A current, activate a slow Ca *+-dependent K+ current (Baraban effects related to their regional distribution. Depressive disor- and Aghajanian, 1980; Aghajanian, 1985). However, this sec- ders (Blendy et al., 1990, 199 1) or depression-mimicking con- ondary effect seen in the raphe neurons is opposite to the effect ditions (Grimm et al., 1992) seem to be linked mainly to ollD seen by ailB receptor activation in the medial thalamic nuclei, receptor, and revealing the distribution patterns of the cells including the geniculate bodies (McCormick and Wang, 1991; synthesizing this receptor in the human brain may in the future McCormick, 1992). In the geniculate bodies, shown here to also give further clues to the substrate for these mental disorders. be rich in o(,~ but not in alA,,, receptors, N, activation, in addition References to reducing a resting A current, causes a reduction in a slow Aghajanian GK (1985) Modulation of a transient outward current in Ca*+-dependent K+ current (McCormick and Prince, 1988). In serotonergic neurones by u,-adrenoceptors. Nature 315:501-503. these two different areas, which are enriched in only one of the Ahlquist RP (1948) A study of adrenotropic receptors. Am J Physiol two different o(, receptor subtypes, cy, activation leads to the 153:586-600. same “primary” effect on resting K+ conductance but causes Aoki C (1992) P-Adrenergic receptors: astrocytic localization in the opposing “secondary” effects on the afterhyperpolarization cur- adult visual cortex and their relation to catecholamine axon terminals as revealed by electron microscopic immunocytochemistry. J Neu- rent. Taken together, it appears that the most robust effect of rosci 12:781-792. either LY,~,~or aIB activation is a reduction in resting K+ con- Aoki C, Joh TH, Pickel V (1987) Ultrastructural immunocytochemical ductance, while the secondary effects can include a reduction or evidence for presynaptic localization of P-adrenergic receptors in the an enhancement of a Ca2+ and/or Ca*+-dependent K+ current. striatum and cerebral cortex of rat brains. Ann NY Acad Sci 604: Thus, although the (Y,~,,, 582-585. and cylBreceptors often have a different Aston-Jones G, Bloom FE (1981) Activity of norepinephrine con- localization, the effects of activation of these two subtypes by a taining locus coeruleus neurons in behaving rats anticipates fluctua- catecholamine seem to be similar. tions in the sleep-waking cycle. J Neurosci 1:876-886. The Journal of Neuroscience, July 1994, 74(7) 4267

Baraban JM, Aghajanian GK (1980) Suppression of firing activity of ergic receptors revealed by chlorethylclonidine. Mol Pharmacol 32: 5-HT neurons in the dorsal raphe by fl-adrenoceptor antagonists. 505-5 10. Neuropharmacology 19:355-363. Hieble HP, DeMarinis RM, Matthews WD (1986) Evidence for and Berthelsen S, Pettinger WA (1977) A functional basis for classification against heterogeneity of alpha- I adrenoceptors. Life Sci 38: 1339- of a-adrenergic receptors. Life Sci 2 1:595-606. 1350. Blendy JA, Grimm LJ, Perry DC, West-Johnsrud L, Kellar KJ (1990) Hoffman BB, Letkowitz RJ (1980) [‘HI WB410 l-caution about its Electroconvulsive shock differentially increases binding to alpha- I role as an a-adrenergic subtype selective radioligand. Biochem Phar- adrenergic receptor subtypes in discrete regions of rat brain. J Neu- macol 29:1537-1541. rosci 10:2580-2586. Hijkfelt T, Johansson 0, Goldstein M (1984) Central catecholamine Blendy JA, Perry DC, Lellar KJ (I 99 1) Electroconvulsive shock in- neurons as revealed by immunohistochemistry with special reference creases a,-adrenoceptor binding sites but not norepinephrine stim- to adrenaline neurons. In: Handbook of chemical neuroanatomy, Vol ulated phosphoinositide hydrolysis in rat brain. Eur J Pharmacol 156: 2, Classical transmitters in the CNS, Pt I (Bjorklund A, Hokfelt T, 267-270. eds), pp 157-276. Amsterdam: Elsevier. Bradshaw CM, Pun RYK, Slater NT, Szabadi E (1981) Comparison Johnson RD, Minneman KP (I 986) Characterization of a,-adreno- of the effects of methoxamine with those of noradrenaline and phen- receptors which increase cyclic AMP accumulation in rat cerebral ylephrine on single cortical neurones. Br J Pharmacol 41:462-479. cortex. Eur J Pharmacol 129:293-305. Bradshaw CM, Sheridan RD, Szabadi E (1985) The pharmacology of Johnson RD, Minneman KP (I 987) Differentiation of a, -adrenergic the neuronal cu-adrenoceptor in the cerebral cortex. In: Pharmacology receptors linked to phosphatidylinositol turnover and cyclic AMP of adrenoceptors (Szabadi E, Bradshaw CM, Nahorski SR, eds). Ba- accumulation in rat brain. Mol Pharmacol 3 1:239-246. singstoke: Macmillian. Jones BE, Moore RY (1977) Ascending projections of the locus coe- Buckland PR, Hill RM, Tidmarsh SF, McGuffin P (1990) Primary ruleus in the rat. II. Autoradiographic study. Brain Res 127:25-53. structure of the rat beta-2 receptor gene. Nucleic Acids Res 18:682. Jones BE, Yang Y (1985) The efferent projections from the reticular Cederbaum J, Aghajanian GK (1977) Catecholamine receptors of lo- formation and the locus coeruleus studied by anterograde and ret- cus coeruleus neurons: pharmacological characterization. Eur J Phar- rograde axonal transport in the rat. J Comp Neurol 242:56-92. macol 44:375-386. Jones LS, Gauger LL, Davis JN (1985) Anatomy ofbrain a,-adrenergic Chamba G, Weissmann D, Rousset C, Renaud B, Pujol JF (1991) receptors: in vitro autoradiography with [1251]-HEAT. J Comp Neurol Distribution of alpha-l and alpha-2 binding sites in the rat locus 231:190-208. coeruleus. Brain Res Bull 26: 185-193. Lanier SM, Downing S, Duzic E, Homey CJ (1991) Isolation of rat Cotecchia S, Schwinn DA, Randall RR, Lefkowitz RJ, Caron MG, genomic clones encoding subtypes of the a,,-adrenergic receptor. Kobilka BK (1988) Molecular cloning and expression of the cDNA Identification of a unique receptor subtype. J Biol Chem 266: 10470- for the hamster cu,-adrenoceptor. Proc Nat1 Acad Sci USA 85:7 l59- 10478. 7163. Lomasney JW, Cotecchia S, Lorenz W, Leung W-Y, Schwinn DA, Yeng- Dagerlind A, Friberg K, Bean AJ, Hiikfelt T (1992) Sensitive mRNA Feng TL, Brownstein M, Letkowitz RJ, Caron MG (199 I) Molecular detection using unfixed tissue: combined radioactive and nop-radio- cloning and expression of the cDNA for the a,,-adrenergic receptor. active in situ hybridization histochemistry. Histochemistry 98:39-49. J Biol Chem 266:6365-6369. Dahlstrom A, Fuxe K (1964) Evidence of the existence of monoamine Lyon TF, Randall WC (1980) Multiple central WB4 10 1 binding sites containing neurons in the central nervous system. I. Demonstration and the selectivitv of urazosin. Life Sci 26:1121-l 129. of monoamines in the cell bodies of brain stem neurons. Acta Physiol Machida CA, Bunzow JR, Searles RP, Van To1 H, Tester B, Neve KA, Stand 62[Suppl 232]:1-55. Teal P, Nipper V, Civelli 0 (1990) Molecular cloning and expression Dashwood MR (1983) Central and peripheral prazosin binding: an in of the &-adrenergic receptor gene. J Biol Chem 265: 12960-l 2965. situ autoradiographic study in the rat. Eur J Pharmacol 78:121-123. Marks SA, Stein RD, Dashwood MR, Gilbey MP (1990) [3H]prazosin Egan TM, Hederson G, North RA, Williams JG (1983) Noradrenaline- binding in the intermediolateral cell column and the effects of ion- mediated synaptic inhibition in rat locus coeruleus neurons. J Physiol tophoresed methoxamine on sympathetic preganglionic neuronal ac- (Lond) 345:477-488. tivity in the anesthetized cat and rat. Brain Res 530:321-324. Flordellis CS, Handy DE, Bresnahan MR, Zannis VI, Gavras H (199 1) McCormick DA (1992) Cellular mechanisms underlying cholinergic Cloning and expression of a rat brain oc],-adrenergic receptor. Proc and noradrenergic modulation of neuronal firing mode in the cat and Nat1 Acad Sci USA 88:1019-1023. guinea pig dorsal lateral geniculate nucleus. J Neurosci 12:278-289. Foote SL, Aston-Jones G, Bloom FE (1980) Impulse activity of locus McCormick DA, Prince DA (1988) Noradrenergic modulation of fir- coeruleus neurons in awake rats and monkeys is a function of sensory ing pattern in guinea pig and cat thalamic neurons, in vitro. J Neu- stimulation and arousal. Proc Nat1 Acad Sci USA 77:3033-3037. rophysiol 591978-996. Foote SL. Aston-Jones G. Bloom FE (1983) Nucleus locus ceruleus: McCormick DA, Wang Z (199 I) Serotonin and noradrenaline excite new evidence of anatomical and physiological specificity. Physiol Rev GABAergic neurones of the guinea-pig and cat nucleus reticularis 63:844-914. thalami. J Physiol (Lond) 442:235-255. Freedman JE, Aghajanian GK (1987) Role of phosphoinositide me- McCune SK, Voigt MM (I 99 1) Regional brain distribution and tissue tabolites in the prolongation of afterhyperpolarizations by a,-adren- ontogenic expression ofa family ofalpha-adrenergic receptor mRNAs oreceptors in rat dorsal raphe neuro&J Neurosci 7:3897-3906. in the rat. J Mol Neurosci 3:29-37. Funrr SJ. Barnes CD (198 1) Evidence of facilitatorv coerulosninal ac- t&t in lumbar motoneurbnes of cats. Brain Res 216:299-3I 1. McGrath JC (1982) Evidence for more than one type of postjunctional Fuxe K (1965a) Evidence for the existence of monoamines neurons alpha-adrenoceptor. Biochem Pharmacol 3 1:467-484. in the central nervous system. III. The monoamine nerve terminal. McGrath J, Wilson V (1988) a-Adrenoreceptor subclassification by Z Zellforsch Mikrosk Anat 65:573-596. classical and response-related methods: same questions different an- Fuxe K (1965b) Evidence for the existence of monoamines neurons swers. Trends Pharmacol Sci 9: 162-l 65. in the central nervous system. IV. The distribution of monoamine Michel MC, Hanft G, Gross G (1990) a,, but not ollv adrenoceptors nerve terminals in the central nervous system. Acta Physiol Stand mediate inositol phosphate generation. Naunyn Schmiedebergs Arch [Suppl] 247:39-85. Pharmacol 341:385-387. Grimm LJ, Blendy JA, Kellar KJ, Perry DC (1992) Chronic reserpine Minneman KP (1988) ai,-Adrenergic receptor subtypes, inositol phos- administration selectively up-regulates p,- and a,,-adrenergic recep- phates, and sources of cell Ca’+. Pharmacol Rev 40:87-l 19. tors in rat brain: an autoradiographic study. Neuroscience 47:77-86. Minneman KP, Johnson RD (1984) Characterization of a, -adrenergic Han CD, Minneman KP (199 I) Interaction of subtype-selective an- receptors linked to (‘H) inositol metabolism in rat cerebral cortex. J tagonists with cr, adrenergic receptor binding sites in rat tissues. Mol Pharmacol Exp Ther 230:3 17-323. Pharmacol 40:531-538. Minneman KP, Han C, Abel PW (1988) Comparison of o(,-adrenergic Han C, Abel PW, Minneman KP (I 987a) a,-Adrenoreceptor subtypes receptor subtypes distinguished by chlorethylclonidine and WB4 IO 1. linked to different mechanisms for increasing intracellular Ca2+ in Mol Pharmacol 33:509-5 14. smooth muscle. Nature 329:333-335. Moore RY, Card P (1984) Noradrenaline-containing neuron systems. Han C, Abel PW, Minneman KP (1987b) Heterogeneity of a,-adren- In: Handbook of chemical neuroanatomy, Vol 2, Classical transmit- 4266 Pieribone et al. * 01, Receptor mRNA Distribution in the CNS

ters in the CNS, Pt I (BjGrklund A, Hijkfelt T, eds), pp 123-156. Stone EA, Ariano MA (1989) Are glial cells targets of the central Amsterdam: Elsevier. noradrenergic system? A review of the evidence. Brain Res Rev 14: Morrow AL, Creese I (1986) Characterization of cu,-adrenergic recep- 297-309. tor subtypes in rat brain: reevaluation of [)H] WB4104 and [‘HI Surprenant A, Horstman DA, Akbarali H, Limbird L (1992) A point prazosin binding. Mol Pharmacol 29:321-330. mutation of the az-adrenoceptor that blocks coupling to potassium Morrow AL, Battaglia G, Norman AB, Creese I (1985) Identification but not calcium currents. Science 257:977-980. of subtypes of [‘HI prazosin-labelled (Yereceptor binding sites in rat Swanson LW, Hartman BK (1975) The central adrenergic system. An brain. Eur J Pharmacol 109:285-287. immunofluorescence study of the location of cell bodies and__- their ______Nicholas AP, Pieribone VA, Elde R, HGkfelt T (1991) Initial obser- efferent connections in the rat utilizing dopamine /3-hydroxylase as a vations on the localization of mRNA for (Yand p adrenergic receptors marker. J Como Neurol 163:467-506. in brain and peripheral tissues of rat using in situ hybridization. Mol Szabadi E (1979) Adrenoreceptors on central neurones: microelectro- Cell Neurosci 2:344-350. phoretic studies. Neuropharmacology 18:83 l-843. Nicholas AP, Pieribone V, Hakfelt T (1993a) Distributions of mRNAs Szabadi E, Bradshaw CM (1987) Alpha- 1 adrenergic receptors in the for alpha-2 adrenergic receptor subtypes in rat brain: an in situ hy- central nervous system. In: The a-1 adrenergic receptors (Ruffolo RR, bridization study. J Comp Neurol 328:585-604. ed), pp 405-453. Clifton, NJ: Humana. Nicholas AP, Pieribone V, Hiikfelt T (1993b) Cellular localization of Terman BI, Riek RP, Grodski A, Hess H-J, Grahm RM (1990) Iden- mRNA for beta-l and beta-2 adrenergic receptors in rat brain: and tification and structural characterization of ai,-adrenergic receptor in situ hybridization study. Neuroscience 56:1023-1039. subtypes. Mol Pharmacol 37:526-534. Palacios JM, Hoyer D, CortCs R (1987) 01,-Adrenergic receptors in the Ungerstedt U (197 1) Stereotaxic mapping ofthe monoamine pathways mammalian brain: similar pharmacology but different distribution in in the rat brain. Acta Phvsiol Stand 821su~~l 3671:1-48. rodents and primates. Brain Res 4 19:65-75. Unnerstall JR (1987) Localization of &I adrenergic receptor in the Paxinos G. Watson G ( I99 1) The rat brain in stereotaxic coordinates, central nervous system. In: The LY- 1 adrenergic receptors (Ruffolo RR, 2d ed. Gew York: Adade&. ed), pp 71-109. Clifton, NJ: Humana. Perez DM, Piascik MT, Graham RM (1991) Solution-phase library Unnerstall JR, Orensanz LM, Femandez I, Kuhar MJ (1982) On the screening for the identification of rare clones: isolation of an a,,- selectivity of WB4 104 as an a, ligand: an autoradiographic study. adrenernic receptor cDNA. Mol Pharmacol 40:876-883. Neurosci Lett [Suppl] lO:S494. PieriboneVA, Biodin L, Friberg K, Dahlstrand J, SGderberg C, Lar- Unnerstall JR, Femandez I, Orensanz LM (1985) The o(,- adrenergic hammar D. Hakfelt T (1992) Differential expression of mRNAs for receptor: radiohistochemical analysis of functional characteristics and neuropeptide-Y related peptides in rat nervous tissues: possible evo- biochemical differences. Pharmacol Biochem Behav 22:859-874. lutionary conservation. J Neurosci 12:336 l-337 1. UPrichard DC, Chamess ME, Robertson D, Snyder SH (1978) Pra- Rainbow TC, Biegon A (1983) Quantitative autoradiography of [‘HI zosin: differential affinities for two populations of oc-noradrenergic prazosin binding sites in the rat forebrain. Neurosci Lctt 40:221-226. recentor bindine sites. Eur J Pharmacol 50:87-89. Raymond JR, Hnatowich M, Lefkowitz RJ, Caron M (1990) Adren- Voigt MM, Kispert J, Chin HM (1990) Sequence of a rat brain cDNA ergic receptors. Models for regulation of signal transduction proFesses. encoding an LY,,?adrenergic receptor. Nucleic Acids Res 18: 1053 Hypertension 15: 119-I 3 1. Voigt MM, McCune SK, Kanterman RY, Felder CC (199 1) The rat Rehavi M, Yavetz B, Ramot 0, Sokolovsky M (1980) Regional het- az-C4 adrenergic receptor gene encodes a novel pharmacological sub- erogeneity oftwo high affinity binding sites for [JH] WB4 10 1 in mouse type. FEBS Lett 278:45-50. brain. Life Sci 26:6 15-62 1. Wiiion KM, Minneman KP (1989) Regional variations in oc,-adren- Schalling M (1990) In situ hybridization studies on regulatory mole- ergic receptor subtypes in rat brain. J Neurochem 53: 1782-1786. cules in neural and endocrine tissue with special reference to expres- Young WS III (1990) In situ hybridization histochemistry. In: Hand- sion of coexisting peptides. Thesis ISBN 9 1 6280036 11, Karolinska book of chemical neuroanatomy: analysis of neuronal microcircuits Institute, Stockholm, Sweden. and synaptic interactions (BjBrklund A, HGkfelt T, Wouterlood FG, Schalling M, Dagerlind A, BrenC S, Hallman H, Djurfeldt M, Persson van den Pol AN, eds), pp 48 l-5 12. Amsterdam: Elsevier. H, Terenius L, Goldstein M, Schlesinger D, Hijkfelt T (1988) Co- Young WS III, Kuhar MJ (1979a) A new method for receptor auto- existence and gene expression of phenylethanolamine N-methyltrans- radiography. ‘H-opioid receptor labeling in mounted tissue sections. ferase, tyrosine hydroxylase, and neuropeptide tyrosine in the rat and Brain Res 179:255-270. bovine adrenal glands: effects of reserpine. Proc Nat1 Acad Sci USA Young WS III, Kuhar MJ (1979b) Noradrenergic a, and a, receptors: 85:8306-8310. autoradioeraohic visualization. Eur J Pharmacol 59:3 17-3 1o ,. Schoepp DD, Knepper SM, Rutledge CO (1984) Norepinephrine stim- Young WS fiI,‘Kuhar MJ (1980) Noradrenergic a, and ~1~receptors: ulation of phosphoinositide hydrolysis in rat cerebral cortex is as- light microscopic autoradiographic localization. Proc Nat1 Acad Sci sociated with the a,-adrenoceptor. J Neurochem 43: 1758-l 76 1. USA 77: 1696-1700. Schwinn KA, Lomasney JW (1992) Pharmacological characterization Zeng D, Harrison JK, D’Angelo D, Barber CM, Tucker AL, Lu Z, Lynch of cloned a,-adrenoceptor subtypes: selective antagonists suggest the KR (1990) Molecular characterization of a rat a,,-adrenergic re- existence of a fourth subtype. Eur J Pharmacol Mol 227:433-436. ceptor. Proc Nat1 Acad Sci USA 87:3 102-3 106. Schwinn KA, Lomasney JW, Lorenz W, Szklut PJ, Fremeau RT, Yang- Zilles K, Gross G, Schleicher A, Schildgen S, Bauer A, Bahro M, Schwen- Feng TL, Caron MC, Lefkowitz RJ, Cotecchia S (1990) Molecular demann G, Zech K, Kolassa N (199 1) Regional and laminar dis- cloning and expression ofthe cDNA for a novel a,-adrenergic receptor tributions of a,-adrenoceptors and their subtypes in human and rat subtype. J Biol Chem 265:8 183-8 189. hippocampus. Neuroscience 40:307-320. Shimomura H, Terada A (1990) Primary structure of the rat beta-l adrenergic receptor gene. Nucleic Acids Res 18:459 1.