Adenosine receptors are coupled to GTP-binding proteins. When bound with adenosine, these receptors activate G proteins to produce a variety of intracellular effects. These G protein-coupled receptor responses, however, are intrinsically slower and longer lasting than those of ligand-gated ion channels. For example, the light-flash response of turtle cone photoreceptors (coupled to G proteins) starts within 20ms, reaches a peak in 100ms, and is finished in 300ms (Baylor, et al., 1974), whereas most fast chemical reactions occur within 0.1-2ms. Therefore, adenosine could not be released and produce its effect in a 10ms interval of stimulation.

However, receptors for purines other than adenosine also include a ligand-gated ion channel which mediate much more rapid effects of neurotransmitter release. Recently, evidence has shown that transmission of ATP at synapses in both the peripheral and central nervous systems is mediated by the P2 purinoceptors which may be subdivided into P2x- and P2y-purinoceptors (Burnstock, G and Kennedy, C, 1985; Dubyak, G and El-Moatassim, C, 1993). The P2x-purinoceptors appear to be largely ionotropic and are present in the hippocampus (Abbraechio, MP and Burnstock G, 1994; Balcar, VJ, et al., 1995; Bo, X and Burnstock, G, 1994). Furthermore, P2 purinoceptor antagonists modulate glutamatergic transmission in the hippocampus (Motin, L and Bennett, MR, 1995).

Many drugs which bind to adenosine receptors and act as antagonists also affect the levels of cAMP, an important intracellular signaling molecule. These drugs—which include methylxanthines—inhibit the metabolic enzyme responsible for converting cAMP into AMP (Hall, 1992). Consequently, the levels of cAMP are elevated in the cell.

In this study, we have determined that an adenosine antagonist has a much stronger effect on axons near the point of stimulation as opposed to those farther away—suggesting a profound yet novel effect by adenosine. We have investigated the effects of 8-cyclopentyl-1,3-dimethylxanthine (CPT)—an adenosine A₁ receptor antagonist—in the CA1 region of the rat hippocampus. To further investigate this novel effect of adenosine on glutamate transmission in CA1, we have compared the electrophysiological effects of A₁ receptor agonists and antagonists to other purinoceptor antagonists and modulators of cAMP levels on synaptic transmission in the hippocampus.

We have found that the effects of adenosine, caffeine (a methylxanthine), forskolin (cAMP elevator), and suramin (P2x ionotropic receptor antagonist) shared similar results in both the near and far pathways. Further studies would therefore be necessary to further elucidate the mechanism of CPT's novel effect.

ELECTROPHYSIOLOGICAL RECORDING AND DATA ACQUISITION
Extracellular field recordings were taken from the stratum radiatum in the CA1 region with recording electrodes filled with 2M NaCl. Two stimulating electrodes were colinearly inserted in the stratum radiatum of CA1 approximately 4mm apart (Figure 1). A recording electrode was inserted immediately adjacent to one of the stimulating electrodes, thereby creating a near-far recording setup. The strength of the near stimulation ranged from 30-60mA, and the far ranged from 14-45mA. Excitatory postsynaptic potentials (EPSP) were evoked at 10 second intervals, and the slope of the EPSP's was measured by AxoBasic software and stored in an IBM computer.

Figure 1: Extracellular field recordings were made from the stratum radiatum in the CA1 region of the hippocampus. Two stimulating electrodes were collinearly inserted in the middle of the stratum radiatum at about 4mm apart, and a recording electrode was inserted immediately adjacent to one of the stimulating electrodes, thereby creating a near-far recording setup.

CHEMICALS
The adenosine antagonist, CPT (8-cyclopentyl-1,3-dimethylxanthine), was prepared as a fresh 1mM stock solution in 0.2 N NaOH. The A₁ adenosine receptor agonist, N⁶-phenyladenosine, was also prepared as a fresh stock solution in 0.2N NaOH. Caffeine (1,3,7-trimethylxanthine), a phosphodiesterase inhibitor, was prepared as a fresh stock solution in distilled water. Forskolin, was prepared as a stock solution in anhydrous DMSO (Dimethyl sulfoxide) and stored at -20ºC. Suramin was prepared as a fresh stock solution in distilled water.

Chemical compounds were purchased from Research Biochemicals International (Natick, MA, U.S.A.). After establishing 10 minutes of a stable baseline recording, the chemicals were added directly to the artificial cerebrospinal fluid for 15 minutes.

Results

In extracellular field EPSP recordings from the CA1 area of the hippocampus, addition of 1mM CPT to the ACF resulted in a larger increase in the EPSP's stimulated in the near pathway than in the EPSP's stimulated in the far pathway (Figure 2). The response to CPT in the near pathway was 120% greater than that in the far pathway (n=16), demonstrating a much greater effect of CPT near the point of stimulation.

Figure 2: CPT, the A1 adenosine antagonist, produces a greater increased effect near the point of stimulation opposed to a distance further away. In fact, the near pathway is 120% greater than that of the far pathway. (Average of 16 experiments.)

Figure 3: N6-phenyladenosine, the potent A1 adenosine receptor agonist, has identical effects at pathways both near and far from the point of stimulation. After four experiments, no significant difference appeared between the two pathways.

Figure 4: Caffeine (1,3,7-Trimethylxanthine), the phosphodiesterase inhibitor, has a slightly greater increased effect on the EPSP of the far pathways opposed to that of near pathways of stimulation. Certainly the results do not mirror the differential effects of CPT.

Figure 5: Forskolin, the adenylate cyclase activator, shows no effect on the EPSP's stimulated at either the near or far pathways (n=5).

Figure 6: Suramin, a purinoceptor antagonist, shows no effect on EPSP's at either the near or far pathways (n=5).

To explain this effect, we first hypothesized that adenosine is being released at sites near the recording site but not farther away. Perhaps, two sets of fibers are present in the slices: short ones with more adenosine receptors and longer ones with fewer receptors. Our second hypothesis was that another purine, such as ATP, is being released at ligand-gated ion channels near the point of stimulation. Perhaps, this effect is modulated by adenosine at A₁ receptors. A third hypothesis explored the effects of CPT and other methylxanthines in increasing the levels of cAMP in the cell. Perhaps, more cAMP is being increased at points near the spot of stimulation as opposed to distances further away.

To test the first hypothesis, we added N⁶-phenyladenosine, an A₁ receptor agonist, to the slices. We observed no significant difference in the EPSP's between the two pathways. This suggests that CPT acts by a mechanism other than on A₁ adenosine receptors.

To test the second hypothesis about ligand-gated ion channels, we added suramin, a purinoceptor antagonist, to the slices. No observable change occurred in the EPSP's of either pathway. This suggests that activation of an ionotropic purinoceptor does not mirror the effects of CPT, therefore CPT is not acting to modulate a P2x ligand-gated channel.

To test the final hypothesis about cAMP levels, caffeine (a PDE inhibitor) was added to the slices. We identified a slightly greater response in the far pathway than in the near pathway. Yet, since caffeine is also an A₁ and A₂ receptor antagonist, the results do not solely reflect phosphodiesterase's potential role. To further investigate the role of cAMP, forskolin—an adenylate cyclase agonist—was added to the slices. No observable changes in synaptic response were evident in either pathway. This demonstrates that a rise in cAMP does not mirror the effects of CPT's, and suggests that the mechanism of CPT does not act through increased levels of cAMP.

None of these hypotheses appear to account for the novel effects of CPT on the CA1 region of the hippocampus. Further studies are needed explore the possibilities of novel fibers running throughout the hippocampus. Perhaps, longitudinal or oblique slicing would reveal fibers that run in a non-parallel direction. Other studies would investigate the effects of additional ligand-gated ion channels such as serotonin or nicotine. Still other studies could explore the differential effects of cGMP as opposed to cAMP.