Kindling-induced neuroprotection

While attempting to produce brain damage in hippocampal kindled rats, using systemic kainic-induced (KA) SE as the induction agent, we discovered that previous kindling remarkably protected the brain against damage in several vulnerable structures, even though the seizures in the kindled rats were as severe or more so than naïve rats (Kelly & McIntyre, 1994).

We have been studying this fascinating phenomenon for several years now and now know a good deal more about it. Dorsal hippocampal kindling protects the brain bilaterally, and includes structures like the piriform cortex, substantia nigra and the hippocampus that normally would be badly damaged; this bilateral outcome is consistent with what we know about the functional anatomy of hippocampal kindling.

More recently, we have explored the anatomical specificity of the kindling neuroprotection against KA and are showing that it is due to regional activation and not systemic release of some neuroprotective agent. For example, amygdala kindling mostly protects the ipsilateral amygdala/piriform cortex from damage but not the contralateral area, and it provides virtually no protection for the dorsal hippocampus. Similarly, partial kindling of the dorsal hippocampus (15 stimulations without generalization) provides complete protection for that hippocampus, less protection for the contralateral hippocampus and no protection for the amygdala/piriform cortex. Thus, the protective effect in brain of stimulation is very regional and dependent upon previous high levels of induced activity. However, there is no kindling pretreatment that provides protection for some structures like the midline thalamus and parts of the endopiriform nucleus (Kelly and McIntyre, in preparation).

The mechanisms of this protection need to be examined. Recently we have observed that there is significant and protracted upregulation of neuronal apoptotic inhibitory proteins (NAIP), as well as another endogenous protective agent, kynurenic acid, in protected structures but not in unprotected structures. In addition, we have seen that the message for BDNF (which is often neuroprotective) upregulates in the presence of KA much faster in the protected structures of kindled rats (as does c-fos) than it does in controls, suggesting a facilitated network of biological 'good things' associated with the kindling process. Further, we have reported that the brain's immune system is dramatically and regionally activated by a kindled seizure (Plata-Salamán et al., 2000), and this also might be part of the kindling protection effect. Lastly, and most importantly, the kindling neuroprotection extends beyond seizure-based challenges, since we have observed that previous hippocampal kindling dramatically reduces the penumbra by ~70% following MCAO. Interestingly, many of the neuroprotective observations associated with kindling have been confirmed in preliminary observations with human epileptics (Nancy Walton, personal communication).

In future experiments, we will explore regional brain differences in activity and biochemistry during kindling-induced neuroprotection against challenges such as KA SE and ischemia. We will compare these outcomes to those induced by our ambulatory model of SE, where the kindled focus is stimulated to trigger the SE, and it is the kindled amygdala/piriform area that is damaged, while the contralateral area is not. Of course, the opposite profile occurs in amygdala kindled rats exposed to KA SE. Comparisons between these two basic challenges (electrical triggered SE and KA SE) should define many potential contributors to the neuroprotection, contributors that regionally increase the hardiness of the brain, and assist the brain in its periodic but intense encounters with a hostile ambient environment.