Increased protein SUMOylation following focal cerebral ischemia

Abstract

Stroke is a major cause of death and disability, which involves excessive glutamate receptor activation leading to excitotoxic cell death. We recently reported that SUMOylation can regulate kainate receptor (KAR) function. Here we investigated changes in protein SUMOylation and levels of KAR and AMPA receptor subunits in two different animal stroke models: a rat model of focal ischemia with reperfusion and a mouse model without reperfusion. In rats, transient middle cerebral artery occlusion (MCAO) resulted in a striatal and cortical infarct. A dramatic increase in SUMOylation by both SUMO-1 and SUMO-2/3 was observed at 6 h and 24 h in the striatal infarct area and by SUMO-2/3 at 24 h in the hippocampus, which was not directly subjected to ischemia. In mice, permanent MCAO resulted in a selective cortical infarct. No changes in SUMOylation occurred at 6 h but there was increased SUMO-1 conjugation in the cortical infarct and non-ischemic hippocampus at 24 h after MCAO. Interestingly, SUMOylation by SUMO-2/3 occurred only outside the infarct area. In both rat and mouse levels of KARs were only decreased in the infarct regions whereas AMPARs were decreased in the infarct and in other brain areas. These results suggest that posttranslational modification by SUMO and down-regulation of AMPARs and KARs may play important roles in the pathophysiological response to ischemia.

Introduction

Cerebral ischemia has profound effects on multiple cellular processes including posttranslational modification of proteins (Dorval and Fraser, 2007, Durukan and Tatlisumak, 2007). One such posttranslational modification is SUMOylation that occurs when Small Ubiquitin like MOdifier (SUMO), a 97-residue protein, covalently binds to lysine residues on target proteins. Protein SUMOylation has been most extensively studied in the nucleus where it can alter the subcellular localization, protein partnering, and DNA-binding and/or transactivation functions of transcription factors (Hilgarth et al., 2004).

There are four SUMO paralogues (SUMO-1–4) in vertebrates. Of these SUMO-4 is mainly localized to the kidney (Bohren et al., 2004) while the other forms are present in brain. SUMO-2 and SUMO-3 differ by only four N-terminal amino acids and no difference in their functional roles has yet been reported (Hay, 2005). Together SUMO-2/3 shares only ∼50% sequence identity to SUMO-1.

It is becoming increasingly apparent that protein SUMOylation is involved in diverse cellular processes including transcriptional regulation, nuclear transport, maintenance of genome integrity, cell signalling, plasma membrane depolarization and signal transduction (Johnson, 2004, Wilson and Rosas-Acosta, 2005, Kerscher et al., 2006, Scheschonka et al., 2007). The functional consequences of SUMO attachment vary greatly from substrate to substrate, and in many cases are not well understood. SUMO-1 and SUMO-2/3 also differ in their conjugation dynamics and responses to cellular stress. Under resting conditions, very little SUMO-1 is present in an unconjugated form, yet there is a large free pool of SUMO-2/3. In response to cellular stresses, such as oxidative stress, osmotic stress or heat shock, an increase in SUMO-2/3 conjugation occurs suggesting that SUMO-2/3 could act as a cellular SUMO reserve to allow an efficient response to stress (Saitoh and Hinchey, 2000).

Protein SUMOylation is dramatically increased in the brain of ground squirrels during hibernation and it has been proposed that this may provide a mechanism to protect cells from otherwise lethally low levels of oxygen and glucose due to reduced blood flow (Lee et al., 2007). Furthermore, recently it has been reported that transient global cerebral ischemia induces a marked increase in protein SUMOylation in the hippocampus and cerebral cortex by SUMO-2/3 in mice (Yang et al., in press).

There are multiple SUMOylated proteins at synapses (Martin et al., 2007) consistent with bioinformatics analysis data indicating that a diverse range of synaptic proteins are potential SUMO targets (Yang et al., 2006). More specifically, the KAR subunit GluR6 is a SUMO target. GluR6 exhibits low levels of SUMOylation under resting conditions but is rapidly SUMOylated in response to agonist activation. Reducing GluR6 SUMOylation using the specific de-SUMOylation enzyme SENP-1 prevented the agonist-evoked KAR endocytosis. KAR-mediated excitatory postsynaptic currents (EPSCs) in hippocampal slices are decreased by SUMOylation and enhanced by de-SUMOylation (Martin et al., 2007). This is of interest because GluR6 has been implicated in neuronal cell death following ischemic insult via a JNK activation pathway (Pei et al., 2006, Zhang et al., 2006) and KARs are also involved in ischemia-induced white matter injury (Tekkok et al., 2007). Therefore, we formulated the hypothesis that one role of increased SUMOylation could be to down-regulate the surface expression of KARs, which in turn, would reduce excitotoxicity and subsequent cell death.

Here we have investigated changes in the pattern of protein SUMOylation and in levels of AMPAR and kainate receptor subunits in MCAO-induced focal ischemia in rat and mouse, which is a model for human stroke. In mice permanent middle cerebral artery occlusion results in a cortical infarct whereas in rat transient MCAO results in a predominantly striatal infarct. At 6 h and 24 h post-MCAO, ischemic tissue from cortex and striatum was dissected and processed to analyse protein SUMOylation. In addition, non-ischemic hippocampal tissue was analysed from both MCAO models. Thus we studied protein SUMOylation in brain tissue from the infarcted area as well as outside the ischemic regions from two different species with or without reperfusion. Our data show that ischemia evokes dramatic changes in the pattern of protein SUMOylation and causes decreases in both AMPAR and KAR levels.