Localization of α2A- and α2B-Adrenergic Receptor Subtypes in Brain
Abstract Recent studies have shown that all three subtypes of α2-adrenergic receptor (α2-AR) are found in brain. The purpose of this study was to map the subtype localization of the α2A- and α2B-ARs in brain structures. RNase protection shows that both the α2A- and α2B-ARs are detectable in cortex, cerebellum, pons-medulla, and hypothalamus. We tested probes derived from the α2A- and α2B-AR cDNAs on cell lines that express each of the α2-AR subtypes to establish the subtype specificity of these probes for in situ hybridization. Then we used the α2A- and α2B-AR probes for in situ hybridization on sagittal and coronal sections of rat brain. Both α2A and α2B mRNA were detected throughout the brain. Overall, there appears to be a greater expression of message for α2A- than α2B-AR in most brain areas, with the exception of the thalamus. Developing these probes for in situ hybridization is an important step for further studies on the exact role of the α2-AR subtypes in neurons that modulate cardiovascular function.
The α2-adrenergic receptors (α2-ARs) are believed to have an important role in the central regulation of autonomic cardiovascular function.1 The lower brain stem structures contain a high concentration of α2-ARs2 and are responsible for maintaining vasomotor sympathetic tone.3 Moreover, the α2-ARs within the brain stem are the target of the central hypotensive actions of clonidine.4
The α2-ARs are a heterogeneous group of receptors that mediate their effects through the Gi class of G proteins and bind to the naturally occurring ligands epinephrine and norepinephrine. As in other G protein–coupled receptors, α2-AR subtypes have been identified that differ in their structure, pharmacological profile, and tissue distribution.5 6 7 8 9 Pharmacological and molecular studies of α2-ARs have suggested at least three distinct genes in both human and rat that code for three distinct subtypes, designated α2A, α2B, and α2C.5 Recent studies show that all three subtypes are found in brain and kidney.6 7 10
Localization of α2-ARs in the rat brain has been accomplished with radiolabeled ligands,11 and localization of α2-AR transcripts has been done with RNAse protection assays or Northern blots.7 10 However, these data cannot provide unequivocal information on the cellular distribution of the various α2-AR subtypes because the radioligands have broad specificities for the α2 class of receptors. Subtype-specific probes have been used in protection analysis or Northern blots, but these methods rely on the isolation of RNA from larger dissected tissue sections. Recent reports have made progress in determining the localization of these receptors in rat brain cells by using immunohistochemistry or in situ hybridization12 13 14 15 16 17 ; however, there are inconsistencies among the several studies that used in situ hybridization.12 13 14 15 16 17
To address these problems, we developed a unique set of riboprobes, and in this article we describe the localization of α2A- and α2B-ARs in the normotensive rat brain using in situ hybridization with subtype-specific RNA probes.
We subcloned regions of the rat α2A- and α2B-AR genomic DNA and cDNA, respectively, into an SP6 promoter containing vector, psP65 (Promega), to allow the synthesis of antisense RNA probes that are specific for transcripts from either the α2A- or α2B-AR genes. To detect the α2A mRNA, we subcloned a 600-bp SstI/XbaI fragment that encompasses part of the 5′ untranslated region of rat α2A transcript into psP65 at the Sst I/Xba I sites. A 269-bp Pst I fragment of RNGα2 cDNA was cloned into the Pst I site of psP65 to create a probe specific for the transcripts that encode the α2B-AR. RNGα2 cDNA18 was a kind gift of Dr Kevin R. Lynch, University of Virginia School of Medicine, Charlottesville, Va.
Tissue Preparation and In Situ Hybridization
Male Wistar rats (250 to 300 g) under anesthesia (sodium pentobarbital, 50 mg/kg IP) were perfused with cold (4°C) phosphate-buffered saline for 3 to 5 minutes and decapitated. The brain was rapidly removed and frozen in isopentane at −70°C. Serial coronal cryostat sections (15 μm) of this material were thaw-mounted onto silanized slides, rapidly dried, and stored at −20°C until used (3 to 12 weeks). In situ hybridization experiments were performed as previously described.19 Briefly, unfixed, adjacent tissue sections were thawed rapidly, air dried, and incubated in 35 000 cpm/μL of one of the 35S-labeled RNA probes diluted in a hybridization solution consisting of 50% deionized formamide, 0.3 mol/L NaCl, 20 mmol/L Tris-HCl (pH 7.4), 5 mmol/L EDTA, 10 mmol/L NaH2PO4·H2O (pH 8.0), 10% dextran sulfate, 1× Denhardt’s, and 0.5 mg/mL total yeast RNA. Sections were incubated in humidified boxes at 52°C for 17 hours. Sections were washed extensively in 0.75 mol/L NaCl, 0.075 mol/L sodium citrate, and 10 mmol/L dithiothreitol for 30 minutes at 50°C and 0.30 mol/L NaCl, 0.030 mol/L sodium citrate, 50% formamide, and 0.1 mol/L dithiothreitol for 20 minutes at 65°C before digestion with 0.02 mg/mL RNase A for 60 minutes at 37°C. After dehydration in a series of ethanols from 30% to 100%, the slides were air-dried, dipped in Kodak NTB2 autoradiography emulsion, and exposed for 6-7 days in a light-free black plastic box at 4°C. Finally, these slides were developed in Kodak D19 for 4 minutes at 14°C and fixed in Kodak Rapid Fix A (solution A) for 3 minutes. Sections were rinsed in distilled water, stained through the emulsion with 0.05% thionin, and dehydrated in graded ethanols, and glass coverslips were affixed with Cytoseal 60 mounting medium. All the sections were examined under bright- and dark-field illumination with a microscope. Photomicrographs were taken with Polaroid Type 55 or Kodak Contrast Process Ortho 4×5 films. The identification and nomenclature of rat brain structures were based on the atlas of Paxinos and Watson.20
In addition to this microscopic analysis, we also used a computerized video densitometry system (Inquiry, Loats Associates) to perform an overall analysis of the labeling obtained with these two methods. In this case, one brain was sagittally sectioned, and adjacent sections were processed with both sense and antisense probes for both the α2A- and α2B-ARs. These sections were then exposed on a single piece of Hyperfilm (Amersham) for 5 days, after which the film was developed in D19 and digitized on the Inquiry densitometry system. Since the overall labeling of the α2A-AR was much greater than for the α2B-AR, the background level of labeling of the pseudocolor map was adjusted so that the range of densities for the two probes was normalized. In this way, areas that differed from the global mean could be visually revealed and used to confirm the differential distribution and labeling of these two probes as well as to guide further analysis.
In situ hybridization of cultured cell lines was performed after cells were grown on chamber slides. These cells were not stained. Permanently transfected CHO cells that express the rat α2A-AR (CHO-A), α2B-AR (CHO-B), and α2C-AR (CHO-C)7 were maintained in the alpha modification of minimum essential medium without ribonucleosides and deoxyribonucleosides supplemented with 10% fetal calf serum. For CHO-A and CHO-C cells, we added 0.005 and 0.02 μmol/L methotrexate to the media, respectively. Untransfected CHO cells (dhfr−) (CHO-d) were grown in alpha minimum essential medium with ribonucleosides and deoxyribonucleosides supplemented with 10% fetal calf serum. RINm5F cells, a rat insulin–producing cell line that naturally expresses the α2A-AR, were maintained in RPMI-1640 medium supplemented with 10% fetal calf serum.21 All media were from GIBCO/BRL.
RNase Protection Assay
RNA from Wistar rat brains was isolated through a CsCl gradient as described.22 RNase protection was performed as described.23 Briefly, antisense probes were linearized before RNA synthesis, and antisense RNA was synthesized in the presence of [α-32P]CTP (Amersham) with SP6 polymerase. RNA duplexes were formed between the radiolabeled probe and RNA samples by overnight hybridization at 45°C in 80% formamide, 0.4 mol/L NaCl, 1 mmol/L EDTA, and 40 mmol/L piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES) (pH 6.7). Single-stranded RNA was digested by a mixture of RNase A and T1 at 37°C for 1 hour. Protected bands were visualized on a denaturing 6 mol/L urea, 6% polyacrylamide gel. Labeled pBR322 HinfI fragments were used as size markers.
In Situ Hybridization in Transfected CHO and RINm5F Cells
To verify the specificity and sensitivity of our probes, we performed in situ hybridization on permanently transfected CHO cells expressing α2A-, α2B-, and α2C-ARs and RINm5F cells. Fig 1⇓ shows the common structure of α2-AR and relative positions of rat α2A- and α2B-AR subtype probes.
The α2B antisense probe hybridizes strongly with the transcripts in CHO-B cells (Fig 2A⇓) but does not hybridize to transcripts in the CHO-A (Fig 2C⇓), CHO-C (Fig 2D⇓), or nontransfected CHO-d cells (Fig 2E⇓), showing a high specificity of this probe for the rat α2B mRNA. The α2A antisense probe strongly hybridizes to all the RINm5F cells (Fig 3A⇓). Furthermore, the α2A probe did not react positively against the CHO-B, CHO-C, or nontransfected CHO-d cells (data not shown). Figs 2B⇓ and 3B⇓ also show that α2B and α2A sense probes, respectively, produce very little background labeling.
Dissection of brain into hypothalamus, pons-medulla, cerebellum, and cortex shows that the α2A-AR transcripts are most concentrated in hypothalamus and pons-medulla, followed by cortex, and are much lower in cerebellum (Fig 4⇓). The α2B-AR probe shows a different distribution pattern, with the strongest signal in hypothalamus, weaker bands in cerebellum and cortex, and a faint but visible band in medulla.
In Situ Hybridization in Rat Brain
To begin the analysis of α2-AR subtypes in brain, we studied the distribution of α2A- and α2B-AR mRNAs throughout the brain. Specificity of hybridization was indicated by comparison between the low, generally uniform levels of background labeling resulting from hybridization with each labeled sense-strand RNA probe and the strikingly different labeling pattern, the clustering of silver grains over neurons characteristic of specific hybridization between RNA transcripts and the antisense RNA probe. For example, Fig 5⇓ shows a typical comparison of the labeling patterns of sense and antisense probes. Clusters of silver grains are present over neurons in the cerebellum when either the α2A or α2B antisense probe is used (Fig 5A⇓ and 5B⇓, respectively), whereas few silver grains are apparent when either the α2A or α2B sense probe is used (Fig 5C⇓ and 5D⇓, respectively). Comparison of sense and antisense signals showed message for both the α2A- and α2B-ARs to be expressed throughout the rat brain. The Table⇓ shows the relative distribution of these receptor subtypes in various brain structures. Fig 6⇓ represents a digitized scan of autoradiographic images showing the localization of α2A and α2B mRNAs throughout the brain. Overall, there appeared to be a greater expression of α2A- than α2B-AR in the brain, with the exception of some thalamic structures.
The most intense hybridization signal in the medulla was seen with the α2A probe in the lateral reticular nucleus, hypoglossal nucleus, inferior olive, and ambiguous nucleus, followed by the gigantocellular reticular nucleus and dorsal motor nucleus of vagus (Table⇑). The spinal nucleus of the trigeminal nucleus of the solitary tract, and spinal vestibular nucleus were also labeled by the α2A probe. As shown in Fig 7G⇓, 7H⇓, and 7I⇓, the α2A probe showed higher density of α2A message in the nucleus of XII than in the nucleus of the solitary tract. The reverse distribution was found for the α2B message.
Although α2B transcripts were never detected at the same overall intensity as α2A transcripts in medulla, relatively dense α2B labeling was seen in the lateral reticular nucleus and nucleus of the solitary tract. The gigantocellular reticular nucleus, dorsal motor nucleus of vagus, hypoglossal nucleus, and inferior olive were labeled with medium intensity, and the spinal nucleus of the trigeminal, ambiguous nucleus, and spinal vestibular nucleus were weakly labeled by the α2B probe.
Neurons of both the molecular and granule cell layers of the cerebellum (see Fig 5⇑) were labeled with both the α2A and α2B probes, although the α2B signal was less pronounced. Purkinje cells also showed more signal with the α2A probe, whereas α2B-specific labeling was less pronounced (Table⇑ and Fig 7D⇑, 7E, 7F).
Pons and Midbrain
Both α2A- and α2B-AR transcripts were present at relatively high levels in the facial nucleus, gigantocellular reticular nucleus, pontine nucleus, and motor trigeminal nucleus (Table⇑). In contrast, locus coeruleus showed the highest intensity signal with the α2A probe, and much less α2B signal was detected (Fig 7D⇑, 7E, 7F). The central gray was one of the few regions where the α2B-AR appeared to be the dominant adrenoceptor. Dorsal and ventral cochlear nuclei, parvicellular reticular nucleus, and dorsal and ventral parabrachial nuclei showed similar, relatively intense signals of α2A- and α2B-ARs.
The structures of the hippocampus and cerebral cortex showed similar, high-intensity signals for both α2A- and α2B-ARs, although α2A was higher than α2B in the hippocampus and the superficial layers of the neocortex (Fig 7A⇑, 7B, 7C). In contrast, thalamus and hypothalamus were most strongly labeled by the α2B probe.
We have mapped the localization of α2A- and α2B-AR mRNAs in brain by in situ hybridization and RNase protection. These data suggest the presence of both of these messages throughout the brain, with different intensity patterns. The regions that appeared to have the most mRNA by protection assay (Fig 4⇑) did not correlate with the regions that had the strongest signal, as judged by intensity imaging (Fig 6⇑) of the in situ data. However, the protection assay was performed per amount of total RNA, whereas the intensity imaging was determined over an area of brain tissue. The in situ hybridization signal was not affected by the amount of total RNA in a particular structure, only the amount of the specific RNA targeted with the probe. On the other hand, RNase protection measured the amount of a particular α2-AR mRNA present in a mixture of RNAs. In this case, the signal detected would be affected by the total amount of RNA present.
Our results with the α2A probe largely agree with the results of studies using immunohistochemical methods12 13 as well as an in situ study by Nicholas et al.15 However, the current study has several differences compared with other in situ studies.16 17 For example, these other studies did not find α2A-AR expression in thalamus or the trigeminal nucleus, whereas we found α2A mRNA in both of these regions and in especially high levels in the trigeminal nucleus. In addition, we found α2A-AR mRNA in both the granule cell and molecular layers of the cerebellum, in contrast to Scheinin et al,17 who reported α2A mRNA only in the granule cell layer.
Unlike Nicholas et al,15 we detected α2B mRNA throughout the brain, whereas their report finds the α2B-AR only in the thalamic structures. This is in contrast to an earlier article by this same group that reported α2B expression throughout the brain.14
Most of the other in situ studies14 15 16 used end-labeled oligonucleotide probes, whereas we used single-stranded RNA probes that are synthesized in the presence of a radiolabeled ribonucleotide, producing a probe that is labeled throughout the synthesized RNA. Because of these differences in probes, the investigators with oligonucleotide probes used 4 to 6 weeks of exposure time. Similarly, another report17 used full-length cDNA to produce riboprobes; however, these investigators apparently used much fewer counts per microliter in their probe hybridizations and 8 weeks of exposure time. In contrast, we used only 6-7 days of exposure time. This suggests that our probes are much more sensitive than the oligonucleotide probes, and by altering probes and counts, we can detect much lower levels of mRNA with riboprobes than with oligonucleotide probes.
One could argue that cross-hybridization of our probes could account for our results. However, our probes have been extensively tested against cell lines that express only one α2-AR subtype, with no evidence for cross-reactivity. In addition, there are structures of the brain where the relative intensities of α2B and α2A differ; for example, in the medulla, the α2B message is more abundant in the nucleus of the solitary tract than the nucleus of XII, and the α2A message is more abundant in the nucleus of XII than the nucleus of the solitary tract. If the α2B cross-hybridizes to α2A mRNA, then one would expect to find α2B signal greatest in the regions where α2A has the greatest expression. This indicates that choosing RNA probes gives a higher sensitivity and specificity for visualization of various brain structures. It also suggests that cross-hybridization should be tested in known cell structures or cell lines that express a specific mRNA.
Our results have many implications for understanding the role of α2-AR in controlling cardiovascular regulation. We have found α2A- and α2B-AR mRNA in neurons of the central pathways involved in baroreflex, ie, within regions such as the nucleus of the solitary tract, the dorsal motor nucleus of the vagus, the ambiguous nucleus, the inferior olive, and the reticular nucleus. Both α2A- and α2B-AR mRNAs are also detected in other areas involved with blood pressure regulation and the control of sympathetic outflow, such as the locus coeruleus or hypothalamus, although α2A predominates in the locus coeruleus. Our studies cannot determine whether both receptors are present in the same neurons or whether both receptors contribute to the control of arterial and pulse pressures. Analysis of subtype expression patterns in hypertensive versus normotensive animal models may provide insights into which subtype controls blood pressure. Previous studies have described differences in central α2-AR levels in normotensive and hypertensive animals.24 25 26 These studies have shown specific alterations in α2-AR in the medulla24 and alterations of α2-AR mRNA in the pons-medulla26 regions of spontaneously hypertensive versus Wistar-Kyoto rats. However, these studies did not precisely localize these changes. It is possible that in situ analysis will show whether these differences in expression occur in a region of the pons-medulla involved in blood pressure control. However, it is likely that more direct methods that can block or stimulate only one α2-AR subtype may be necessary to clarify the role of each α2-AR subtype in cardiovascular function. For example, antisense therapies or knockout animal models may be used to block specific subtype expression, or specific drugs can be used to inhibit or stimulate a single α2-AR subtype.
In terms of applying any findings in animal models to treatment strategies for human hypertension, the human brain has many more levels of complexity. In addition, although there are three α2-AR subtypes in human and rodent species, previous work has shown that differences in the pharmacology of analogous α2A-ARs in rodents and humans arise from a few specific nucleotide changes that specify changes in key amino acid residues and result in distinct ligand binding properties of highly related proteins.27 Likewise, genetic differences in regulatory regions of human and rodent α2-AR genes could contribute to differences in tissue-specific expression patterns between species.
In conclusion, using RNA antisense probes, we have found a significantly different distribution of mRNAs for α2-AR subtypes than other researchers using DNA probes. In addition to the increased sensitivity, we have found the RNA probes to be less time-consuming to use, with no evidence for cross-hybridization. Developing these probes for in situ hybridization is an important step for further studies on the exact role of the α2-AR subtypes in cardiovascular function modulated by neurons expressing these receptors.
This study was supported by National Institutes of Health grants P01 AG00001, P01 NS31649, HL-48181, and HL-46693. Agostinho Tavares was supported in part by Brazilian grants 201579/90.1 (CNPq) and 93/4414-9 (FAPESP). We acknowledge the excellent technical assistance of Andrew Doolittle, Claudia Fitzgerald, and Ramy Rizkalla in the Department of Anatomy and Neurobiology.
- Received September 12, 1995.
- Revision received October 23, 1995.
- Accepted November 6, 1995.
Hausler G. Cardiovascular regulation by central adrenergic mechanism and characterization by hypertensive drugs. Circ Res. 1975;36/37(suppl I):I-232-I-233.
Reis JD, Golanov EV, Ruggiero DA, Sun MK. Sympatho-excitatory neurons of the rostral ventrolateral medulla are oxygen sensors and essential elements in the tonic and reflex control of the systemic and cerebral circulations. J Hypertens. 1994;12:S159-S180.
Bylund DB. Subtypes of α1 and α2-adrenergic receptors. FASEB J. 1992;6:832-839.
Lorenz W, Lomasney JW, Collins S, Regan JW, Caron MG, Lefkowitz RJ. Expression of three α2-adrenergic receptor subtypes in rat tissues: implications for receptor classification. Mol Pharmacol. 1990;38:599-603.
Coupry I, Duzic E, Lanier SM. Factors determining the specificity of signal transduction by guanine nucleotide-binding protein-coupled receptors, II: preferential coupling of the α2C-adrenergic receptor to the guanine nucleotide binding protein, Go. J Biol Chem. 1992;267:9852-9857.
Flordellis CS, Berguerand M, Gouache P, Barbu V, Gavras H, Handy DE, Bereziat G, Masliah J. α2-adrenergic receptor subtypes expressed in Chinese hamster ovary cells activate differentially mitogen-activated protein kinase by a p21ras independent pathway. J Biol Chem. 1995;270:3491-3494.
Boyajian CL, Loughlin SE, Leslie FM. Anatomical evidence for alpha-2 adrenoceptor heterogeneity: differential autoradiographic distribution of [3H]rauwolcine and [3H]idaxozan in rat brain. J Pharmacol Exp Ther. 1987;241:1079-1091.
Zeng D, Harrison JK, D’Angelo DD, Barber CM, Tucker AL, Lu Z, Lynch KR. Molecular characterization of a rat alpha2B-adrenergic receptor. Proc Natl Acad Sci U S A. 1990;87:3102-3106.
Paxinos G, Watson C. The Rat Brain in Stereotaxic Coordinates. New York, NY: Academic Press; 1982.
Praz GA, Halban PA, Wollheim CB, Blondel B, Strauss AJ, Renold AE. Regulation of immunoreactive-insulin release from a rat cell line (RINm5F). Biochem J. 1983;210:345-352.
Flordellis C, Handy DE, Bresnahan MR, Zannis VI, Gavras H. Cloning and expression of a rat brain alpha2B-adrenergic receptor. Proc Natl Acad Sci U S A. 1991;88:1019-1023.
Link R, Daunt D, Barsh G, Chruscinski A, Kobilka B. Cloning of two mouse genes encoding α2-adrenergic receptor subtypes and identification of a single amino acid in the mouse α2-C10 homolog responsible for an interspecies variation in antagonist binding. Mol Pharmacol. 1992;42:16-27.