It is now clearly established that cerebral ischemia overactivates PARP-1 in several cell types of the ischemic region and significantly contributes to the extension of ischemic damage. PARP-1 activation occurs not only in injured neurons but also in endothelial and glial cells as well as in pericytes and infiltrating leukocytes. Consistent with PARP-1 involvement in ischemic brain damage, enzymatic activity of PARP-1 in gerbils subjected to transient (5 min) ischemia increases in the injured tissue 4.3- and 1.7-fold at 1 and 24 h of reperfusion, respectively (Nagayama et al. 2000). Prolonging duration of ischemia (10 min) leads to significant increases of PARP-1 activity up to the seventh day of reperfusion (Strosznajder et al. 2003). Accordingly, various studies report PAR accumulation in the ischemic brain tissue. In a focal and transient model of brain ischemia in mice, PAR formation is highly increased in the nuclei of cells of cerebral cortex compared to those present in the contralateral one. Of note, polymer formation is drastically decreased both in the ischemic and contralateral cortex of PARP-1-/- mice subjected to MCAo (Eliasson et al. 1997). A parallel study in mice reports that following 2 h MCAo, increased PAR formation in the ischemic cortex occurs as early as 5 min after reperfusion in cells showing swelling and nuclear disruption (Endres et al. 1997). PAR accumulation, however, is not evident at later times (3-6 h) of reperfusion or after milder ischemic insult (1 h MCAo) (Endres et al. 1997). PAR formation has also been investigated in a permanent model of MCAo in rats. In this study, it is reported that PAR immunoreactivity increases in the ischemic core and penumbra 2-8 h after ischemia, returning to basal levels 16 h post-ischemia. Importantly, PAR immuno-reactive cells of the ischemic cortex show the classical morphology of pyramidal neurons (Tokime et al. 1998). Similarly, the ischemic cortex of rats subjected to distal MCAo and transient bilateral common carotid artery occlusion have a threefold increase of PAR immunoreactive cells 10 min after 1.5 h ischemia. Notably, inhibition of PARP-1 completely abrogates increase of PAR immunoreactivity (Takahashi et al. 1999). Remarkably, PAR formation also occurs in neural cells of the infarcted human brain at 18-24 h post-insult and rapidly declines, thereafter. A second wave of poly(ADP-ribosyl)ation is due to PAR-positive macrophages, which start infiltrating the ischemic human brain 3 days after the ischemic injury (Love et al. 2000). In a subsequent study in humans, it was reported that brain ischemia causes accumulation of PAR in the ischemic core and penumbra mostly during the first 2 days after cardiac arrest. Importantly, double immunostaining for PAR and the neuronal marker MAP2 indicates that the majority of PAR-positive cells are neurons. It has also been reported that brain ischemia alters PARP-1 expression levels. For example, Love and colleagues show that expression of PARP-1 increases in the infarcted tissue of the human brain 18-24 h post-insult and, similarly to PAR, declines thereafter (Love et al. 2000). Accordingly, PARP-1 protein as well as its mRNA levels increase in the nucleus of cultured cerebellar granular cells upon exposure to neurotoxic concentrations of glutamate (Cosi et al. 1994). Results obtained in a model of transient global ischemia in gerbils are contradictory. Indeed, the group of Sharp reports an increase of PARP-1 mRNA in the dentate gyrus of gerbil brains 4 h after 10 min of global ischemia and return to basal levels 8 h after ischemia (Liu et al. 2000). Conversely, Nagayama and associates (Nagayama et al. 2000) as well as Strosznajder and colleagues (Strosznajder et al. 2003) show that PARP-1 mRNA and protein do not increase in the gerbil hippocampus in the same ischemia model. Nevertheless, these results taken together unambiguously establish a central role of cerebral ischemia in altering the homeo-stasis of poly(ADP-ribosyl)ation in neurons of different species including humans.
Was this article helpful?