Sigma (sigma) receptors have generated a great deal of interest on the basis of their possible role in psychosis, neuroprotection and various other behaviors including learning processes. The existence of at least two classes of sigma receptor binding sites (sigma(1) and sigma(2)) is now well established. The recent cloning of the mouse, guinea pig and human sigma(1) receptors has allowed the study of the discrete distribution of the sigma(1) receptor mRNA in rodent and human brain tissues using in situ hybridization. Overall, the sites of expression of specific sigma(1) receptor mRNA signals were in accordance to the anatomical distribution of sigma(1) receptor protein first established by quantitative receptor autoradiography. Specific sigma(1) receptor hybridization signals were found to be widely, but discretely distributed, in mouse and guinea pig brain tissues. The highest levels of transcripts were seen in various cranial nerve nuclei. Lower, but still high hybridization signals were observed in mesencephalic structures such as the red nucleus, periaqueductal gray matter and substantia nigra, as well as in some diencephalic structures including such as the habenula and the arcuate, paraventricular and ventromedial hypothalamic nuclei. Superficial (I-II) and deeper (IV-VI) cortical laminae were moderately labeled in the mouse brain. Moderate levels of sigma(1) receptor mRNA were also found in the pyramidal cell layer and the dentate gyrus of the hippocampal formation. Other structures such as the thalamus and amygdaloid body also expressed the sigma(1) receptor mRNA although to a lesser extent. In murine peripheral tissues, strong hybridization signals were observed in the liver, white pulp of the spleen and the adrenal gland. In the postmortem human brain, moderate levels of sigma(1) receptor mRNA, distributed in a laminar fashion, were detected in the temporal cortex with the deeper laminae (IV-VI) being particularly enriched. In the hippocampal formation, the strongest hybridization signals were observed in the dentate gyrus while all other subfields of the human hippocampal formation expressed lower levels of the sigma(1) receptor mRNA. Antisense oligodeoxynucleotides against the purported sigma(1) receptor were used next to investigate the possible role of this receptor in dizocilpine (MK-801)/NMDA receptor blockade-induced amnesia. Following a continuous intracerebroventricular infusion of a specific sigma(1) receptor antisense into the third ventricle (0.4 nmol/h for 5 days), sigma(1)/[3H](+)pentazocine binding was significantly reduced in mouse brain membrane homogenates while a scrambled antisense control was without effect. Moreover, the sigma(1) receptor antisense treatments (5 nmol/injection, every 12 hx3 or 0.4 nmol/h for 5 days) attenuated (+)MK-801/NMDA receptor blockade-induced cognitive deficits in the treated mice while a scrambled antisense control had no effect. Taken together, these results demonstrate the widespread, but discrete, distribution of the sigma(1) receptor mRNA in the mammalian central nervous system. Moreover, antisense treatments against the purported sigma(1) receptor gene reduced specific sigma(1)/[3H](+)pentazocine binding and modulated cognitive behaviors associated with NMDA receptor blockade providing further evidence for the functional relevance of the cloned gene.