The role of Bcl-2 homology domain (BH) 3-only proteins in stress-induced cell death of mouse neocortical neurons associated with ischemic brain injury

In this project, we set out to characterise the involvement of the Bcl-2 homology domain 3 only (BH3-only) proteins in excitotoxic signalling and cellular stress associated with ischemic stroke. Cerebral ischemia occurs when there is a restriction in the blood supply to tissues of the brain, the extent and duration of which determines the degree of neural damage and infarct volume.

The BH3-only proteins are a pro-apoptotic group of proteins which belong to the b cell lymphoma 2 (Bcl-2) family and are responsible for the regulation of the intrinsic apoptotic pathway. During apoptosis, the BH3-only proteins translocate to the mitochondria, induce permeabilisation of the mitochondria1 outer membrane, and promote release of integral proteins into the cytosol such as cytochrome c (cyt-c), second mitochondria1 activator of caspases (Smac/DIABLO) and apoptosis inducing factor (AIF). Previous work has implicated the BH3-only proteins in ischemic and seizureinduced neuronal injury. However, it remains unclear which of these family members are induced during stress conditions, and the significance of their induction in contributing to neuronal apoptosis.

Excitotoxicity is a phenomenon where over-excitation of the glutamate receptors results in neuronal death. In this study, we established an excitotoxic model of neuronal injury in murine neocortical neurons by N-Methyl-D-Aspartate (NMDA) and used this to investigate the signalling pathways induced during excitotoxicity. We identified the BH3-only protein, Bcl-2 interacting mediator (Bim) as an important moderator of excitotoxic injury and highlighted roles for two protein kinases, c-Jun NH2- terminal kinase (JNK) and AMP-activated protein kinase (AMPK). We demonstrated that JNK is activated in our model and that inhibition of JNK protects neurons from cell death mediated via Bim. Activation of AMPK in our model was shown to correlate to reduced levels of adenosine tri-phosphate (ATP). Overexpression of a constitutively active ampk or AMPK activation using a pharmacological inducer of AMPK, 5-amino,4-imidazolecarboxarnide riboside (AICAR), resulted in increased neuronal death associated with JNK activation and Bim induction, whereas inhibition of ampk by siRNA resulted in neuronal protection from excitotoxic injury. We propose that, in response to acute excitotoxicity, AMPK and JNK led to the activation of the BH3-only protein Bim thereby promoting neuronal apoptosis.

In addition, we investigated the signalling pathways induced following proteasomal stress, a cellular dysfunction also associated with ischemia. We induced proteasomal stress in cortical neurons using a pharmacological inhibitor of the proteasome, epoxomicin, and identified the BH3-only protein, p53 upregulated modulator of apoptosis (PUMA) as being an important mediator of proteasomal stress-induced apoptosis. PUMA dependent apoptosis was associated with the release of cyt-c, caspase activation and nuclear condensation. However, we also demonstrate that cell death can proceed in the absence of puma following prolonged proteasornal stress. Using a fluorescent resonance energy transfer (FRET) probe and live cell confocal imaging, we examined the dynamics and kinetics of caspase 3-like activity on a single cell level. We propose that three main cell death pathways may exist or be recruited in specific cell types during acute proteasomal stress in cortical neurons, and provide evidence of caspase-dependent and caspase-independent cell death occurring in the absence of puma expression.

In an alternative approach to identify novel regulators of BH3 only proteins, we employed protein array technology to seek novel binding partners for PUMA. We used a peptide mimicking the BH3 domain of PUMA to identify an interaction between Ski oncogene (c-Ski) and our PUMA peptide. This interaction was validated in vitro by immunoprecipitation, where we pulled down a PUMA/c-Ski protein complex following overexpression of puma. In addition, we were able to demonstrate that c-Ski is transcriptionally induced in response to PUMA dependent cell death by epoxomicin in cortical neurons. Finally, we established that expression of c-Ski has the ability to attenuate PUMA-mediated cell death in response to epoxomicin, and other stimuli in various cell types, highlighting a physiologically relevant role for c-Ski in PUMA-dependent apoptosis.