Heat shock protein overexpression - effect on experimental stroke

Abstract

Heat shock proteins (HSPs) have been reported to increase cell survival in response to a wide range of cellular challenges. In order to investigate the effect of these proteins in vivo, transgenic (tg) mice overexpressing HSP27 or HSP70 have been compared to wild-type (WT) mice in a middle cerebral artery occlusion (MCAO) model of permanent cerebral ischaemia. To assess the protective effect of HPSs following permanent ischaemia, the lesion size was estimated after 24 hours of ischaemia using a multi-slice T2-weighted MRI scan. 4 groups of mice were used: 1. HSP27 tg (n=5), 2. HSP27 WT (n=6), 3. HSP70 tg (n=5) and 4. HSP70 WT (n=6). The MCA was permanently occluded by advancing a 180-mm-diameter filament into the internal carotid artery past the MCA junction. Multi-slice coronal images were obtained 24 hours after the onset of stroke. A 2.35 T horizontal bore SMIS MR scanner was used with the following imaging parameters: T2-weighted SE sequence with FOV 20 mm, 1 mm slice thickness, 9 slices and 128 64 pixels. The relative infarcted area per slice was defined as the ratio of the lesion area to the whole brain size. All data are presented as mean SD. The relative infarcted area was larger in the WT mice than in the HSP27-overexpressing animals for every slice (P < 0.05 using degrees of freedom adjusted repeated measures ANOVA). The same is true for HSP70-overexpressing mice. Overall, the lesion is 30 3 % and 27 4 % of the whole brain volume in the HSP27 WT and HSP70 WT mice respectively, compared to 20 7 % and 19 6 % in the HSP27tg and HSP70tg animals. In-situ hybridisation showed high expression levels of HSP27 and HSP70 throughout the brain in HSP27tg and HSP70tg mice respectively. With this particular model, the lesion is significantly smaller for both HSP27 and HSP70 overexpressing mice compared to WT animals. Interestingly, similar work using HSP27 or HSP70 gene delivery via a viral vector in a rat model of transient MCAO showed protection for HSP27, but not for HSP70 1. The difference between these experiments is probably partly due to the amounts of HSP overexpression and their spatial distribution, which are not the same in the transgenic mouse model and the viral delivery rat model. Furthermore, the different MCAO models (permanent versus transient) are known to influence the amount of apoptosis expected in infracted tissue, and thus the differences between the two models may reflect the different inhibitory actions of HSP27 and HSP70 in the apoptotic cascade.