Adenosine is a ubiquitous nucleoside that, in addition to its established role within metabolism, has widespread neuromodulatory actions in the CNS through an interaction with heterogeneous P1 purinoreceptors (A1, A2A, A2B, A3). The concept of ‘homeostatic neuromodulation’ by adenosine has been introduced to account for the co-existence of a predominant synaptic inhibition by A1 and facilitation by A2 receptors (Cunha, 2001). A1 receptors localized to the substantia gelatinosa of the rat spinal cord exhibit antinociceptive actions that are attributed to presynaptic inhibition of glutamatergic synaptic transmission (Lao et al. 2001). Pro-nociceptive actions have been attributed to spinal A2 receptors (Burnstock & Wood, 1996). Given this complexity of action, up- or down-regulation of the adenosine concentration within the synaptic environ may impact on neurotransmission that is modulated by P1 receptors. In this respect, the equilibrative (ENT) and concentrative (CNT) nucleoside transporters that move adenosine into and out of cells could provide a means to achieve this. In this study, we have used immunohistochemical techniques to assess the cellular and regional distribution of the adenosine transporter ENT1 within the spinal dorsal horn. In addition, we have tested the functional consequences of reduced adenosine uptake via the ENT1 transporter on nociceptive synaptic transmission in rat substantia gelatinosa neurons in vitro. In the light of data showing a morphine-induced release of spinal adenosine via a putative action on nucleoside transport systems (Sweeney et al. 1993), we have evaluated whether µ opioid receptor-induced synaptic inhibition in substantia gelatinosa could be partly mediated by endogenous adenosine acting through presynaptic A1 receptors.

Immunoblotting of dorsal horn samples with ENT1 antisera revealed a band at ~50 kDa, consistent with the size that is predicted from the established sequence of rat ENT1. Immunohistochemical analysis of spinal cord sections using selective ENT1 antisera revealed dense staining for ENT1 within laminae I and II and moderate to low staining in deeper laminae. This distribution of ENT1 overlaps with expression of the A1 receptor, which is especially dense within substantia gelatinosa. Ultrastructural analysis of ENT1 expression within the substantia gelatinosa with electron microscopy and immuno-gold labelling revealed localization of ENT1 to presynaptic terminals and postsynaptic neuronal elements. Significant A1 receptor immunoreactivity was also found predominately within presynaptic terminals. In patch clamp recordings of rat substantia gelatinosa neurons in vitro, monosynaptic EPSCs were significantly attenuated by either adenosine (1 µM) or the selective A1 receptor agonist 2-chloro-N6-cyclopentyl adenosine (CCPA, 1 µM). This synaptic inhibition was mimicked by a reduction of adenosine uptake with the selective ENT1 inhibitor nitrobenzylthioinosine (NBMPR, 1-1000 nM). These data suggest inhibition of synaptic transmission as a consequence of an accumulation of endogenous extracellular adenosine and activation of A1 receptors. Further evidence for this mechanism of action is the finding that facilitated breakdown of endogenous adenosine by inclusion of adenosine deaminase (30 mg ml^-1) in the superfusate partially reversed NBMPR-induced attenuation of the EPSC amplitude. The selective A1 receptor antagonist 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) offset EPSC inhibition by both adenosine and NBMPR. Analysis of the paired-pulse ratio for evoked EPSCs and spontaneous miniature EPSC amplitude and frequency revealed a presynaptic locus of action and a reduced probability of transmitter release after NBMPR. An AMPA-evoked inward current in single cells was unaltered by NBMPR indicating that postsynaptic modulation cannot account for these effects. The µ opioid receptor agonist D-ala(2), N-Me-Phe(4), Gly-ol(5) enkephalin (DAMGO, 1 µM) significantly reduced the evoked EPSC amplitude and this action was partially offset by DPCPX. These data suggest a contributory role of adenosine and presynaptic A1 receptors to opioid-induced antinociception.

In interpreting these data, we propose that adenosine transporters, including the ENT1 subtype, may play a role in adenosine homeostasis within the vicinity of synapses expressing adenosine receptors. A corollary of this is that altered activity of these nucleoside transporters within substantia gelatinosa will indirectly modify nociceptive synaptic transmission as a secondary consequence to manipulation of extracellular adenosine concentrations. Furthermore, a closer evaluation of the proposal that opioid-induced antinociception may be enhanced by inhibition of adenosine reuptake (Keil & Delander, 1995) may be justified.

This work is financed by the MRC. Technical provision was from J. Daniel and J. Ingham. R.J.G. was supported by a University of Leeds studentship. A1R antisera were kindly provided by Dr M. Yates, University of Leeds. ENT1 antibody was synthesized by Z. Berry, University of Leeds. With thanks to J. Deuchars for assistance with EM studies.