INABIS Y2K, Poster #93, Matsuoka et al. Discussion

DISCUSSION
Although the CA1 was not damaged by the injection of KA alone, the addition of an A1 adenosine receptor antagonist, CPT, exacerbated neuronal damage in the hippocampus, i.e., neuronal cell loss also occurred in the CA1. Microglia have thin and longer processes that were typically ramified in both non-treated and vehicle-injected rat brain. However, microglia were activated and showed a typical ameboid type morphology after the injection of KA or KA/CPT. In addition, MHC class II antigen was detected in the region of neuronal cell loss, but was not detected in the vehicle-injected rat brain. Several recent papers have reported that KA neurotoxicity and ischemia induce MHC class II-immunoreactive ameboid microglia (Akiyama et al., 1988; Finsen et al., 1993; Matsuoka et al., 1998b). In addition, reactive microglia that express MHC class II have been observed phagocytosing degenerated neuronal elements in Alzheimer's disease, Parkinson's disease, acquired immunodeficiency syndrome, and other neuronal degenerative disorders of humans (Dickson et al., 1993; McGeer et al., 1993). Therefore, the induction of MHC class II antigen suggests the strong activation of glial cells. The regions in which glial activation occurred correlated well with those that showed KA- or KA/CPT-induced neurodegeneration.

The expression of c-Jun participates in neuronal cell death (Bossy-Wetzel et al., 1997). After KA-injection, numerous pyramidal neurons in the CA3 showed apoptosis markers (Pollard et al., 1994). In this study, c-Jun phosphorylation may have preceded neurodegeneration. c-Jun phosphorylation occurred in the same region as where neurons underwent apoptotic cell death. Recently, it has been shown that the phosphorylation of c-Jun by the activation of c-Jun N-terminal kinase (JNK) is important for neuronal apoptosis using cultured cells in vitro (Bossy-Wetzel et al., 1997; Eilers et al., 1998; Herdegen et al., 1998; Schwarzschild et al., 1997; Watson et al., 1998). These reports suggest that phosphorylation of c-Jun by the activation of JNK is closely associated with KA-induced neuronal apoptosis in the rat hippocampus in vivo. In this study, we showed that treatment with CPT induces neuronal and glial changes, such as neuronal cell loss, neuronal apoptosis, and the induction of MHC class II antigen, in the CA1 as well as in the CA3. These results clearly indicate that adenosine has neuroprotective effects through similar events in the CA1 following treatment with KA.

Adenosine increased membrane potassium conductance and opened specific chloride channels through the activation of A1 receptor (Trussell and Jackson, 1985; Mager et al., 1990). Both act together in generating hyperpolarizing currents that stabilized the neuronal resting membrane potential and antagonized depolarization, and then the synaptically evoked neuronal calcium influx through voltage-sensitive ion channels is reduced (Schubert, 1988). An A1 adenosine receptor can be activated by nanomolar concentrations of extracellular adenosine present under physiological conditions in the normal rat brain (Ballarin et al., 1991). An increase in the extracellular adenosine concentration up to the micromolar range, as observed in the ischemic brain (Hagberg et al., 1987), will presumably further strengthen the A1 receptor activation and raise the threshold effect for a toxic calcium influx to cause neuronal damage. Such a protective threshold effect of endogenous adenosine against neuronal damage is supported by in vivo experiments (Donaghy and Scholfield, 1994). The massive calcium influx, through a series of as yet unclear steps that may involve phospholipase activity and the production of free radicals from the arachidonic acid pathway, eventually leads to lipid peroxidation and cell death. Based upon the concept that adenosine has neuroprotective effects (Schubert et al., 1997), the effects of adenosine receptor agonists were examined. Previously, administrated (not endogenous) adenosine and its derivatives have been shown to have neuroprotective effects against ischemia and KA-induced neurodegeneration (Heron et al., 1994; MacGregor and Stone, 1993, respectively). In this study, we found that antagonism of the A1 adenosine receptor exacerbated neurodegeneration. Furthermore, this exacerbated neurodegeneration was attenuated by the coadministration of an A1 adenosine receptor agonist. These results strongly suggest that endogenous adenosine has neuroprotective effects in neurons.

The selective vulnerability of CA3 pyramidal neurons to KA stimulation can be explained by the predominant distribution of KA receptors in the CA3 (Monaghan and Cotman, 1982; Werner et al., 1991). This hyperexcitability in the CA3 appears to be limited to within the CA3, although CA1 neurons receive synaptic inputs directly from CA3 neurons via the Schaffer collaterals. One possible pathway for CA1 neuronal death observed in this study was that CPT antagonized the protective effects of endogenous adenosine released in CA1. The synaptic transmission of the Schaffer collateral/commissural afferent in CA1 is suppressed by extracellular adenosine released during a tetanic stimulation (Sekino and Koyama, 1992; Mitchell et al., 1993; Manzoni et al., 1994). Thus, the hyperexcitation of CA3 neurons induced by the KA treatment can release adenosine that might prevents neurons in CA1 from glutamate excitotoxicity. However, the inhibitory effect of CPT against NMDA- and hypoxia-induced suppression of synaptic responses evoked by CA1 stimulus was limited to partial (Manzoni et al., 1994; Arlinghaus and Lee, 1996). These reports suggest that another additional factor can be involved in the widespread of neuronal death in CA1. Recently, a novel intrahippocampal pathway, the CA3-CA2-CA1 circuit, was identified by a physiological technique (Sekino et al., 1997). In the transverse hippocampal slices, neuronal activity in CA3 propagates to CA1 only when the slices were treated by CPT and activation of the CA3-CA2-CA1 circuit is supposed to be involved (Sekino and Obata, 1995). As seen in the present study, CPT seems to open a gate for synaptic transmission between CA3 and CA1. This study also suggests that endogenous adenosine plays a role in regulating the signal flow from CA3 to CA1. Immunohistochemical studies using an anti-A1 adenosine receptor antiserum raised against purified rat brain A1 adenosine receptor (Nakata, 1993) have shown that the adenosine A1 receptor thought to be functioning postsynaptically was distributed predominantly in the CA3a/CA2 subfield (Ochiishi et al., 1999). Removing the tonic inhibition of this field can result in enhancement of neuronal activity in CA1. The present result also supports the hypothesis that the CA2 field acts as a gate for signal flow from the CA3 to the CA1. Based on these observations, we consider that the administration of A1 adenosine receptor antagonist exacerbates neurodegeneration by i) negating the neuroprotective effects of adenosine, and reducing the threshold for bursting; and ii) reducing the threshold of CA2 neurons which contribute to the CA3-CA2-CA1 circuit.