• cGMP;
  • homeostatic feedback;
  • NMDA;
  • phosphodiesterase;
  • plasticity;
  • zaprinast


Adenosine is an important regulator of neuronal excitability. Zaprinast is a cyclic nucleotide phosphodiesterase inhibitor, and has been shown in the hippocampal slice to suppress excitation. This action can be blocked by an adenosine receptor antagonist, and therefore is presumably due to adenosine release stimulated by exposure to zaprinast. To explore the mechanism of this phenomenon further, we examined the effect of zaprinast on adenosine release itself in cultured rat forebrain neurons. Zaprinast significantly stimulated extracellular adenosine accumulation. The effect of zaprinast on adenosine appeared to be mediated by increasing intracellular cyclic adenosine monophosphate (cAMP) and activation of protein kinase A (PKA): (i) zaprinast stimulated intracellular cAMP accumulation; (ii) a cAMP antagonist (Rp-8-Br-cAMP) significantly reduced the zaprinast effect on adenosine; (iii) an inhibitor of phosphodiesterase (PDE)1 (vinpocetine) and an activator of adenylate cyclase (forskolin) mimicked the effect of zaprinast on adenosine. We also found that zaprinast had no effect on adenosine in astrocyte cultures, and tetrodotoxin completely blocked zaprinast-evoked adenosine accumulation in neuronal cultures, suggesting that neuronal activity was likely to be involved. Consistent with a dependence on neuronal activity, NMDA receptor antagonists (MK-801 and D-APV) and removal of extracellular glutamate by glutamate–pyruvate transaminase blocked the effect of zaprinast. In addition, zaprinast was shown to stimulate glutamate release. Thus, our data suggest that zaprinast-evoked adenosine accumulation is likely to be mediated by stimulation of glutamate release by a cAMP- and PKA-dependent mechanism, most likely by inhibition of PDE1 in neurons. Furthermore, regulation of cAMP, either by inhibiting cAMP–PDE activity or by stimulating adenylate cyclase activity, may play an important role in modulating neuronal excitability. These data suggest the existence of a homeostatic negative feedback loop in which increases in neuronal activity are damped by release of adenosine following activation of glutamate receptors.