ADENOSINE A 1 AND A 2A RECEPTORS ARE CO-EXPRESSED IN PYRAMIDAL NEURONS AND CO-LOCALIZED IN GLUTAMATERGIC NERVE TERMINALS OF THE RAT HIPPOCAMPUS N. REBOLA, a R. J. RODRIGUES, a L. V. LOPES, a P. J. RICHARDSON, b C. R. OLIVEIRA a AND R. A. CUNHA a * a Center for Neurosciences of Coimbra, Institute of Biochemistry, Fac- ulty of Medicine, University of Coimbra, 3004-504 Coimbra, Portugal b Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD, UK Abstract—Adenosine is a neuromodulator that controls neu- rotransmitter release through inhibitory A 1 and facilitatory A 2A receptors. Although both adenosine receptor-mediated inhibition and facilitation of glutamate release have been observed, it is not clear whether both A 1 and A 2A receptors are located in the same glutamatergic nerve terminal or whether they are located on different populations of these terminals. Thus, we have tested if single pyramidal glutama- tergic neurons from the hippocampus simultaneously ex- pressed A 1 and A 2A receptor mRNA and if A 1 and A 2A recep- tors were co-localized in hippocampal glutamatergic nerve terminals. Single cell PCR analysis of visually identified py- ramidal neurons revealed the simultaneous presence of A 1 and A 2A receptor mRNA in four out 16 pyramidal cells pos- sessing glutamatergic markers but not GABAergic or astro- cytic markers. Also, A 1 and A 2A receptor immunoreactivities were co-localized in 264% of nerve terminals labeled with antibodies against vesicular glutamate transporters type 1 or 2, i.e. glutamatergic nerve terminals. This indicates that glu- tamatergic neurons in the hippocampus co-express A 1 and A 2A receptors and that these two receptors are co-localized in a subset of glutamatergic nerve terminals. © 2005 IBRO. Published by Elsevier Ltd. All rights reserved. Key words: co-expression, co-localization, glutamate, synap- tosomes. Adenosine is a neuromodulator which inhibits the release of excitatory neurotransmitters and decreases neuronal excitability through the activation of inhibitory A 1 receptors, which are the most abundant of the four known adenosine receptors in the brain (reviewed in Dunwiddie and Masino, 2001). Recent studies have revealed that the A 2A receptor has opposite effects, i.e. facilitating neurotransmitter re- lease (reviewed in Cunha, 2001a). These A 2A receptors are considerably less abundant than A 1 receptors in the hippocampus (Lopes et al., 2004) but play a major role in neuroprotection (reviewed in Cunha, 2005), possibly due to their ability to control glutamate release (Marcoli et al., 2003; Popoli et al., 2003). Thus, the control of glutamate release in the hippocampus is under the dual control of inhibitory A 1 and facilitatory A 2A receptors (Lopes et al., 2002), but it is still debatable whether the same nerve terminal is simultaneously equipped with these two aden- osine receptors exhibiting opposite effects on neurotrans- mitter release. We have previously provided functional evidence suggesting that both A 1 and A 2A receptors might be present in the same glutamatergic nerve terminals (Lopes et al., 2002), but this has never received direct confirmation. Thus, this study was designed to test if the same glutamatergic pyramidal neuron could express both A 1 and A 2A receptors and if these two adenosine receptors could be co-located in the same glutamatergic nerve ter- minal in the rat hippocampus. EXPERIMENTAL PROCEDURES All experiments were conducted according to EU guidelines on the ethical use of experimental animals (86/609/EEC), with particular care to minimize the number of animals used and their suffering. Single cell PCR of laser-dissected neurons Coronal sections (6 and 10 m thick) were obtained (between -4.52 and -2.80 mm from bregma) from frozen brains of male Sprague– Dawley rats (6 – 8 weeks old). These sections were dehydrated in 100% ethanol then re-hydrated gradually (100 –50% ethanol), Nissl- stained and dehydrated again with xylene. The labeled cell bodies of hippocampal CA1 or CA3 pyramidal neurons were micro-dissected using a laser-equipped microscope Arcturus LCM 210 PixCell II with a spot size of 7.5 m, according to manufacturer’s protocol (see Lopes et al., 2003). Total RNA was then extracted according to StrataPrep Total RNA microprep kit instructions from Stratagene (London, UK) and subjected to reverse transcription (RT) and cDNA amplification. Each laser-captured pyramidal cell body was evalu- ated individually, i.e. there was no pooling of cells or their cDNA thus affording a single neuron analysis. The 3= end amplification PCR protocol used in this study was carried out as previously described (Richardson et al., 2000). Briefly, samples of the pre-amplified cDNA were subjected to 45 rounds of gene-specific PCR in 20 l of 45 mM Tris–HCl (pH 8.1), 12.5% (w/v) sucrose, 12 mM (NH 4 ) 2 SO 4 , 3.5 mM MgCl 2 , and 0.5 mM deoxynucleotide triphosphates, with 100 ng of the forward and reverse primer. The following primers were used: ribosomal protein L11 (Rpl11; forward primer 215-234: GAAAATCGCTGT- TCACTGCA; reverse primer 401-383: CAGGCCATAGATGC- CAATG; PCR product length of 187 base pairs, bp); intronic marker (region chromosome 3; forward primer: GCCTGCAT- TCATCTTCATCTGC; reverse primer: AAAGGTGGAACTCGC- CCGTTT; PCR product length of 149 bp); adenosine A 1 receptor (Adora1; forward primer 1225-1242: TTCCGATGCCAGC- CTAAG; reverse primer 1392-1373: CAGCTGGGAAAACT- GAGGAG; PCR product length of 168 bp); adenosine A 2A recep- *Corresponding author. Tel: +351-239-820190; fax: +351-239-822776. E-mail address: racunha@clix.pt (R. A. Cunha). Abbreviations: bp, base pairs; BSA, bovine serum albumin; GFAP, glial fibrillary acidic protein; PBS, phosphate-buffered saline; RT, re- verse transcription; vGluT, vesicular glutamate transporter. Neuroscience 133 (2005) 79–83 0306-4522/05$30.00+0.00 © 2005 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2005.01.054 79