Transient forebrain ischemia in rodents induces a

Transient forebrain ischemia in rodents induces a reduction in expression of mRNA encoding the AMPA receptor subunit GluR265, 66, 67, 68, 69 (Fig. 3). gluR2 mRNA expression is markedly reduced in the CA1 region (the Angiotensin Fragment 1-7 acetate region most vulnerable to ischemia-induced damage), but not in the CA3 region or the dentate gyrus. Analysis of emulsion-dipped sections indicates that the reduction is specific to CA1 pyramidal cells, and reveals that it clearly precedes any histological sign of cell damage. The decline in gluR2 mRNA in the CA1 region is first detectable at about 12h after ischemia, and is about 30% of control at 24h. GluR3 expression is also reduced, but less so (to about 50% of control), whereas gluR1 and nr1 mRNA expression are not significantly changed. The (GluR1 + GluR3)/ GluR2 ratio (see above) is markedly increased in the CA1 region at 24h, but not in CA3 neurons or the dentate gyrus. Because the GluR2 subunit limits the Ca2+ permeability of AMPA channels, these findings predict the increased formation of Ca2+-permeable AMPA receptors in CA1 neurons after transient global ischemia at times preceding neurodegeneration. Moreover, timing of the switch is consistent with a causal role for AMPA receptor-mediated Ca2+ influx in post-ischemic damage. Reduction in RNA editing at the Q/R site of GluR2 could also increase Ca2+ permeability of AMPA receptors, but this mechanism is unchanged after global ischemia70, 71, 72. Electrophysiological and optical imaging studies69, 73, 74 demonstrate increased AMPA receptor-mediated Ca2+ influx in CA1 pyramidal cells post-ischemia but prior to cell death (see Appendix A). These findings provide evidence that AMPA receptors lacking the GluR2 subunit become permeable to Ca2+ in post-ischemic brain, and that this modification contributes to delayed CA1 pyramidal cell death. Recently, in dissociated rat hippocampal neurons, sublethal oxygen-glucose deprivation induced an increase in gluR4 mRNA expression relative to other AMPA receptor subunits (H.S. Ying, J. Weishaupt, M. Grabb, L.M.T. Canzoniero, S.L. Sensi, C.T. Sheline, H. Monyer and D.W. Choi, unpublished observations). Conditioned neurons also exhibited an increase in kainate-induced, Joro toxin-sensitive Ca2+ influx, suggesting that the increase in GluR4 was associated with decreased expression of GluR2. In addition, susceptibility to kainate-induced cell death was enhanced. The molecular mechanisms underlying regulation of GluR2 expression post-ischemia are as yet unknown. A number of diverse strategies and drug treatments that afford neuroprotection in animal models of global ischemia prevent downregulation of GluR2 expression in the CA1 region. These include preconditioning with a sublethal ischemic episode and administration of hyperpolarizing agents (that is, activators of ATP- sensitive K+ channels or adenosine A1 receptor agonists) in rats[68], or i.c.v. injection of the endonuclease inhibitor aurintricarboxylic acid in gerbils (E. Aronica, J.A. Gorter, M.V.L. Bennett, R.S. Zukin and D.M. Rosenbaum, unpublished observations). In contrast, the AMPA receptor antagonist NBQX, given either at the time of the ischemic episode or later, does not prevent the reduction in GluR2[67], suggesting that NBQX affords neuroprotection by direct block of modified AMPA channels (that is, NBQX is likely to remain in the system long enough to act on the newly formed GluR2-less, Ca2+-permeable receptors). The NMDA antagonist MK-801 is not neuroprotective, given early or late, and does not prevent the downregulation of GluR2[67]. These data suggest that the GluR2 reduction post-ischemia is not triggered by activation of ionotropic glutamate receptors either during the early rise in extracellular glutamate or later after ischemia (but see [75], which proposes a role for metabotropic glutamate receptors in triggering delayed neurodegeneration). The findings with hyperpolarizing agents suggest that increased extracellular K+ concentration during ischemia[76] is an early mechanism that may trigger downregulation of GluR2. Cultured cerebellar granule cells grown in medium containing increasing concentrations of KCl exhibit decreased gluR2 mRNA and protein expression as well as increased AMPA-evoked 45Ca2+ influx[77].